In Vitro Activity of Quinolones Against S

.lr...o_,:11 J.,,-!J �I U$.:I

..::.,L..,.�I !""l..Li:l fr-' -,I°J ,;yWI �I tJLJ.J..... _,:!.I J_r!J �)'I �I erJ.;..,.; Ji �)I �I �I 1,1° ..!,Ll.) _r),J -!.i�)il �l::iJ 4:,1\..i_,ll ..::.,\J.,J.IJ �l.il.lJ .-l}i/1 J�\;:JJ ,t.. �JjUIJ �I ..::.,l.,�I ij oJ.t..l..:1,.1 ..::.,IJ�W.IJ �I ..::.il.JSJIJ ,�I .:,t--11 .,1..,;a&.i JS'J! 4f>.'Y' -,/°J .J....._,:).1 Jr-�� lf.... � L., ,_1,:.J ,..::.it..._.,L.ll .;r �I �l_,J)'IJ ;_JWI L-.,.:Ji � C;' ..;Jl.dl _pl)IJ ,�I yp1 .r), ..::.,L..,li.;l\ i.1$'J ,�I .u.WI J\....J

,-1>.-Jb:-J �)'I ij �4

EASTERN MEDITERRANEAN HEALTH JOURNAL IS the official health journal published by the Eastern Mediterranean Regional Office of the World Health Organization. It is a forum for the presentation and promotion of new policies and initiatives in health services; and for the exchange of ideas, concepts, epidemiological data, research findings and other information, with special reference to the Eastern Mediterranean Region. It addresses all members of the health profession, medical and other health educational institutes, interested NGOs, WHO Collaborating Centres and individuals within and outside the Region.

LARE VUE DE SANTE DE LA MEDITERRANEE ORIENTALE EST une revue de sante officiellepubliee par le Bureau regional de \'Organisation mondiale de la Sante pour la Mediterranee orientale. Elle offre une tribune pour la presentation et la promotion de nouvelles politiques et initiatives dans le domaine des services de sante ainsi qu'a l'echange d'idees, de concepts, de donnees epidemiologiques, de resultats de recherches et d'autres informations, se rapportant plus particulierement a la Region de la Mediterranee orientale. Elle s'adresse a tous les professionnels de la sante, aux membres des instituts medicaux et autres instituts de formation medico-sanitaire, aux ONG, Centres collaborateurs de \'OMS et personnes concernes au sein et hors de la Region.

The Eastern Mediterranean Health Journal is abstracted/indexed in the Index Medicus and MEDLIN£ (Medical Literature Analysis and Retrieval Systems on Line) and the ExtraMed-Full text on CD-ROM, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), CAB International, Lexis Nexis™ and the Index Medicus for the WHO Eastern Mediterranean Region (IMEMR). EMHJ is also available on the World Wide Web: http://wwwlemro.who.int/emhj.htm

ALL ARTICLES ARE PEER REVIEWED

ISSN 1020-3397

© WORLD HEALTH ORGANIZATION, 2005 Publications of the World Health Organization enjoy copyright protection in accordance with the provisions of Protocol 2 of the Universal Copyright Convention. All rights reserved. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitations of its frontiers or boundaries. The mention of specific companies or of certain manufacturers' products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. The authors alone are responsible for the views expressed, which do not necessarily reflect the opinion of the World Health Organization or of its Member States.

COVER: Remote sensing image of the area of the world that includes the Eastern Mediterranean Region. Cover image produced by the Canada Centre for Remote Sensing, Ottawa, Canada. (NOAA AVHRR Composite). Reproduced with permission.

Cover designed by Ahmed Hassanein Printed on acid-free paper Printed by League of Arab States Printshop

6 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Cholera epidemic in Baghdad during 1999: clinical and bacteriological profile of hospitalized cases A.M. Al-Abbassi,1 S. Ahmed 2 and T. Al-Hadithi 2

ABSTRACT An epidemic of cholera in Iraq was anticipated for the year 1999 and a plan of notification and treatment of cases of diarrhoea was made. This paper documents the clinical and bacteriological profile of cholera cases admitted to the 6 hospitals in Baghdad during the epidemic. The number of stool culture- positive cases was 874. The peak incidence of cases (June 1999) was earlier than previous epidemics with no cases registered in November/December 1999. All age groups were affected and the case fatality rate was 1.3%. Strains isolated were Vibrio cholerae El-Tor O1, serotypes Ogawa (79.6%) and Inaba (12.1%), V. parahaemolyticus (2.0%) and non-agglutinable vibrios (6.1%). V. cholerae O139 was isolated from 2 cases (0.2%) for the first time in Iraq. Antibiotic resistance was noted, especially to tetracycline.

Épidémie de choléra à Bagdad en 1999 : profil clinique et bactériologique des cas hospitalisés RÉSUMÉ Une épidémie de choléra en Iraq a été anticipée pour l’année 1999 et un plan a été établi pour la notification et le traitement des cas de diarrhée. Cet article documente le profil clinique et bactériologique des cas de choléra admis dans les six hôpitaux de Bagdad pendant l’épidémie. Le nombre de cas dont les cultures de selles étaient positives s’élevait à 874. Le pic d’incidence des cas (juin 1999) est survenu plus tôt que lors des épidémies précédentes, aucun cas n’ayant été enregistré en novembre/décembre 1999. Tous les groupes d’âge étaient touchés et le taux de létalité était de 1,3 %. Les souches isolées étaient Vibrio cholerae El Tor O1, sérotypes Ogawa (79,6 %) et Inaba (12,1 %), V. parahaemolyticus (2,0 %) et des vibrions non agglutinables (6,1 %). V. cholerae O139 a été isolé chez 2 cas (0,2 %) pour la première fois en Iraq. Une résistance aux antibiotiques a été constatée, en particulier à la tétracycline.

1Department of Medicine; 2Department of Community Medicine, College of Medicine, University of Baghdad, Baghdad, Iraq (Correspondence to A.M. Al-Abbassi: [email protected]). Received: 26/06/02; accepted: 15/09/04

02 Cholera epidemic in Baghdad.pmd 6 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 7

Introduction of deaths and many sectors of the city were completely depopulated [14]. The Cholera is a specific infectious disease of disease spread to Baghdad, with similar humans caused by Vibrio cholerae sero- consequences. After that, cholera contin- group O1and O139. It is an enzymatic pro- ued to appear in several epidemic forms cess causing a brief, acute onset diarrhoeal during the years 1871, 1889, 1894, 1899 illness, with recurrent vomiting and stools and 1917 [15], after which the disease that resemble rice water. This leads to completely disappeared from Iraq to reap- acute rapid dehydration with fluid and elec- pear again in August 1966 as a part of the trolyte loss, ending with acidosis and hypo- 7th pandemic spread [11,16]. After subsid- volaemic shock, which is usually fatal if ence of the 7th pandemic in Iraq, occasion- untreated [1–7]. Cholera spreads mainly al outbreaks of cholera continued in Iraq; via drinking water supplies contaminated Figure 1 shows the registered cholera cas- by human excreta, especially from sub- es in Iraq from 1991–99. However, there clinical carriers and mild cases. Humans was no official announcement of another are the reservoir of the disease, despite iso- epidemic until the year 1999. During 1991, lation of V. cholerae O1 from blue crabs a third of the total number of reported cas- [8] and from the faeces of some aquatic es were in Baghdad city (602) whereas in birds [9]. the inter-epidemic period 1990–98, Bagh- Cholera differs from other enteric dis- dad has a smaller proportion of the total eases in its clinical course (a very short in- (197). cubation period) and epidemic pattern After the escalation in the incidence of (rapidly spreading to different countries cholera cases in Iraq during 1998, and due and disappearing rapidly when outbreaks to serious shortages of safe water supplies subside) [10]. Cholera is endemic in Asia; for human use, the epidemic of cholera in from 1817 to 1923 there were 6 pandemic the year 1999 was anticipated. A plan of waves which moved through Asia then notification for cases of diarrhoea and their through the Americas and Africa [11–13]. treatment in general hospitals or isolation When the 3rd pandemic begun in 1852, it wards was arranged for a prospective extended to Persia, Mesopotamia and Eu- study. This paper documents the clinical rope as a whole. England was severely af- and bacteriological data of all cholera cases fected, allowing John Snow to complete admitted to the 6 hospitals in Baghdad dur- his studies of the relationship of the disease ing the epidemic of 1999 in Iraq. to water supplies at the Broad Street pump [11]. Throughout history, cholera has en- gendered fearful reactions; nowadays, out- Methods breaks often lead to inappropriate The 6 Baghdad hospitals chosen for the responses including trade embargoes, tour- management of diarrhoea cases during the ism restrictions, quarantine or excessive 1999 epidemic were the Medical City isolation and mass chemoprophylaxis. Teaching Hospital, Al-Mansour Paediatric Iraq is at risk of epidemics spreading Hospital, Ibn El-Khateeb Hospital for Infec- from neighbouring countries because it lies tions Diseases, the Central Paediatric Hos- on the routes of pilgrimage to Mecca and pital, Al-Kindy General Hospital and contains a number of holy shrines. During Al-Kadisia General Hospital. the epidemic of 1820, when cholera first spread to Basra, there were a great number

02 Cholera epidemic in Baghdad.pmd 7 10/25/2005, 1:20 AM 8 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Figure 1 Annual number of reported cholera cases in Iraq, 1991–99 (Source: Department of Community Disease Control, Ministry of Health, Iraq)

Information about the presenting symp- was done in peptone water and TCBS toms, including the state of dehydration, (thiosulfate–citrate–bile salts–sucrose was recorded for all patients attending the agar) medium. Stool cultures were isolated hospitals. Stool samples were collected for and confirmed using specific sub-strains analysis. Blood urea, serum creatinine and Ogawa and Inaba antisera. Patients with serum electrolytes were tested for severe negative stool culture results were omitted cases of dehydration. Haematocrit values from the study. Antimicrobial resistance and other investigations were not routinely testing was carried out routinely as part of done. The clinical course and the outcome the general stool examination according to of cases were recorded. Vandepitte et al. [17], although for time rea- The intravenous fluids available for re- sons the choice of antibiotics for treatment suscitation of severe cases were Ringer’s was not dependent on the results of sensi- lactate, normal saline (0.9%), sodium bi- tivity tests. carbonate (1.4%) and molar (M/6) sodium The number of cases of cholera admit- lactate. The oral fluid used for rehydration ted to the Baghdad hospitals during each of mild cases was the World Health Organi- month of 1999 was charted. zation’s recommended dextrolyte solution

containing 3.5 g NaCl, 2.5 g NaHCO3, 1.5 g KCl and glucose 20 mg/L (or sucrose 40 g/ Results L). Because of concerns about resistant stains, different antibiotics, singly or in A total of 874 patients diagnosed with chol- combination, were used for 3–5 days. era were admitted to the 6 hospitals in The stool samples were transported in Baghdad governorate (Table 1). They were saline containers to the Reference Labora- 424 males and 450 females with a tory at the Central Public Health Laborato- male:female ratio of 0.94:1. Table 2 shows ry, where further processing of the stool the age and sex distribution of the patients;

02 Cholera epidemic in Baghdad.pmd 8 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 9

Table 1 Distribution of cholera patients dehydration (98.9%), vomiting (98.1%), admitted to 6 Baghdad hospitals in 1999 and abdominal colic (13.0%). Dehydration was graded severe for 58.6%, moderate Admitting hospital No. of % for 29.3% and mild for 11.0%; 10 cases patients were not dehydrated. Fever (raised temper- Central Paediatric Hospital 233 26.7 ature between 37.5–38.0 °C) was present Al-Kadisia Hospital 220 25.2 in 24.0% of patients, mainly children. The stool was reported as white (rice water) for Al-Kindy Hospital 141 16.1 83.8% of the patients; the rest were yellow Baghdad Teaching Hospital 106 12.1 (13.1%), bloody (1.3%), green (1.1%) or Al-Mansour Hospital 105 12.0 brown (0.7%). Ibn-El-Khateeb Hospital 69 7.9 The strains isolated from the stool sam- Total 874 100.0 ples of these hospitalized cases were mostly V. cholerae biotype El-Tor, serogroup O1, with serotypes Ogawa comprising 79.6% and Inaba 12.1%. V. parahaemolyt- those below puberty were 61.4% of the to- icus was found in 2.0% of cases and non- tal number of cholera cases admitted. Most agglutinable vibrios (poly-negative isolates) of the cases were from the centre of Bagh- in 6.1%. The strain V. cholerae O139 was dad city and from Al-Thawra district in the isolated from 2 cases (0.2%). ≤ eastern side of Baghdad, which are poor The period of stay in the hospital was districts suffering from shortages in the 5 days for 72.9% of the admitted patients water supply with unsafe drinking water except those with complications who need- most of the time. ed a further few days in hospital (6–8 days The main presenting symptoms and 20.9%, 9+ days 6.1%). The outcome was signs were diarrhoea (100% of patients), recovery without complications for 92.1% of cholera cases, recovery with complications for 6.6% and death for 1.3%. Of the 11 patients who died, 5 were males and 6 Table 2 Distribution of cholera patients females from all age groups (Table 2). The admitted to Baghdad hospitals in 1999 by cause of death was mainly due to severe age and sex dehydration, uraemia and acute heart failure among severely dehydrated patients Age Admitted Died presenting late. Of the 58 cases developing (years) Male Female Total (n = 11) complications most of these were due to (n = 424) (n = 450) (n = 874) uraemia (43 cases), the rest due to convul- % % No. % No. sions, intrauterine death, haematemesis or < 1 9.4 7.1 72 8.2 1 other causes (15 cases). 1–5 38.2 28.9 292 33.4 1 A variety of antibiotics, singly or in 6–15 20.8 18.9 173 19.8 1 combination, were used to treat cases; the choice of antibiotics was according to their 16–25 7.3 21.6 128 14.6 3 availability in the hospitals. First-line treat- 26–35 5.7 10.0 69 7.9 2 ment was with tetracycline, ampicillin or 36–46 4.7 5.6 45 5.1 0 cotrimoxazole; second-line treatment was 50+ 13.9 8.0 95 10.9 3 erythromycin, doxycycline, chlorampheni-

02 Cholera epidemic in Baghdad.pmd 9 10/25/2005, 1:20 AM 10 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

col, gentamicin or metronidazole. Single observed in Iraq in 1998 (Figure 1), which drug use was for 34.3% of patients and led to the expectation of more cases during combinations for 65.7%. Resistance of up 1999. to 30% was recorded for tetracycline and In the Iraqi epidemic of 1991, and the cotrimoxazole and lower rates for other an- inter-epidemic period, the number of cases tibiotics, but with a good sensitivity of of clinical disease typically peaked during more than 90% to ampicillin. the autumn months, especially September The time trend of reported cholera cas- and October. The earlier peak of the 1999 es for the year 1999 is shown in Figure 2. epidemic in Baghdad (June) was perhaps This shows no case registration during the related to the great shortage of drinking winter and spring times, while the peak ad- water supply for human use in addition to mission was during June, which was near- the lowered standards of hygiene and con- ly halved during July then rapidly decreased tinuous contamination of the water sources to zero during November and December by human faeces at this time. The attack 1999. rate among previously uninfected populations living in crowded unsanitary conditions can be very high, even though most Discussion infections are mild or sub-clinical. The epidemic ended in November 1999 where The 7th pandemic of El-Tor cholera that there were no clinical cases registered and started in 1961 was still active in 1998, no culture-positive results for V. cholerae when a marked increase was recorded in were found from patients presenting with the number cases in all countries affected, diarrhoea. with a total of 293 121 cases and 10 586 All ages were affected, but the majority deaths reported to the World Health Organi- (61.4%) were 15 years or under. These zation [18]. This amounts to almost double findings are similar to those reported from the number of reported cases in 1997 other cholera-endemic areas, where the (147 425 cases and 6274 deaths) [18–20]. highest incidence of clinical cholera is usu- This was the same upsurge of the disease

Figure 2 Number of cholera patients admitted to Baghdad hospitals in 1999 by month (data missing for 36 patients)

02 Cholera epidemic in Baghdad.pmd 10 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 11

ally observed among toddlers, pre-school the lower cross-river of Nigeria [30]. In children and women of childbearing age agreement with our study, they reported [2,3,7,10,21,22]. the non-O1 cholera infection was associat- The clinical presentation of diarrhoea, ed with watery diarrhoea, mild to moderate vomiting, rice-water stool and severe dehy- in severity, and the diarrhoea was some- dration observed in our patients is the clas- times bloody, occasionally accompanied by sic symptomatology of cholera [1,2,10]. vomiting, abdominal cramps and low-grade The case fatality rate among our hospital fever. patients was 1.3%, less than the global case Due to the emergence of multiple-resis- fatality rate of cholera of 4.3% during 1997 tant strains of V. cholerae [24–26], we and 3.6% during 1998 [18]. A high mortal- used a variety of antibiotics singly or in ity rate in cholera epidemics could be due to combination. However, antibiotic resis- inadequate use of oral rehydration, use of tance, especially to tetracycline, was noted. inappropriate intravenous fluids and inade- Mahon et al. found in a study of cholera O1 quate experience of health workers in man- in the United States that the proportion of agement of severe cholera [23–28]. The isolates resistant to at least 1 antibiotic rose higher global case fatality rate might be be- from 3% in 1992 to 93% in 1994 [27]. A cause nearly 72% of world cholera cases single dose of ciprofloxacin is currently were in African countries where poverty more effective than a single dose of doxy- and lack of resources are a more serious cycline and may be particularly valuable problem. Furthermore, our anticipation of where resistant strains to tetracycline are the 1999 epidemic meant that hospitals and common [24]. wards had made preparations to receive the clinical cases, ensuring the availability of proper fluids for rehydration and training in Conclusions and their correct use; this is also likely to great- recommendations ly reduce the case fatality rate. V. cholerae O139 was isolated from 2 The study has shown V. cholerae sero- cases who had severe dehydration diar- strain O139 isolated in Iraq for the first rhoea indistinguishable from the clinical time. Early planning to contain the disease signs of cholera due to V. cholerae O1. A may have reduced the case fatality rate, es- resurgence has been seen of this Bengal pecially through proper regimens of rehy- strain in many countries since 1992 [29] dration. Resistance to different antibiotics and now in Iraq. Non-O1 vibrios strains was noted. Careful and regular stool cultur- have been implicated in outbreaks of food- ing from diarrhoea cases suspected of borne disease, isolated from as many as cholera throughout the year is recommend- 13% of patients with cholera-like disease ed for detection of index cases before epi- during cholera epidemics. Eko et al. report- demics start. ed the spectrum of Vibrio spp. diarrhoea in

References

1. Cholera. In: Christie AB, ed. Infectious practice, 4th ed. Volume 1. Edinburgh, diseases: epidemiology and clinical Churchill Livingstone, 1987:193–222.

02 Cholera epidemic in Baghdad.pmd 11 10/25/2005, 1:20 AM 12 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

2. Cholera. In: Benson AS, ed. Control of 12. Cholera 1996. Weekly epidemiological communicable diseases manual, 16th record, 1997, 72(31):229–35. ed. Washington DC, American Public 13. Singh J et al. Endemic cholera in Delhi, Health Association, 1995:94–101. 1995: analysis of data from a sentinel 3. Bhattacharya SK et al. Cholera in young centre. Journal of diarrhoeal diseases children in an endemic area. Lancet, research, 1998, 16(2):66–73. 1992, 340:1549. 14. Cholera epidemic. In: Al-Wardi A. 4. Glass RI et al. Cholera in Africa: lessons Lamahat Ijtima‘iyya min Ta’rikh al-‘Iraq on transmission and control for Latin al-Hadith [Social aspects of Iraqi modern America. Lancet, 1991 338:791–5. history]. Baghdad, Matba‘a al-Adib al- Baghdadiyya, 1969:244–5. 5. Greenough WBV. Cholerae and cholera. In: Mandell GL, Douglas RG, Bennett JE, 15. Longrigg SH. Iraqi government at the eds. Bennett’s principles and practices of end of the 19th century. In: Al-Khayatt J infectious diseases, 4th ed. New York, (transl). Four centuries of Iraq’s modern Churchill Livingstone, 1995:1934–43. history, 4th ed. Baghdad, Iraq, Al- Arabeyah Library, 1968:380. 6. Gunnlaugsson G et al. Funerals during the 1994 cholera epidemic in Guinea- 16. Al-Awqati QS, Mekkiya M, Thamer M. Es- Bissau. Epidemiology and infection, tablishment of a cholera treatment unit 1998, 120(1):7–15. under epidemic conditions in a developing country. Lancet, 1969, 1:252–3. 7. Vibrio cholerae and Asiatic cholera. In: Todar K, ed. Todar’s online textbook of 17. Vandepitte J et al. Guidelines for anti- bacteriology. Madison, University of Wis- microbial resistance surveillance. Alex- consin, Department of Bacteriology, andria, World Health Organization, Re- 2002 (http://textbookofbacteriology.net/ gional Office for the Eastern cholera.html, accessed 12 December Mediterranean, 1996. 2004). 18. Cholera 1998. Weekly epidemiological 8. Huq A et al. Colonization of the gut of the record, 1999, 74(31):257–64. blue crab (Callinectes sapidus) by Vibrio 19. Cholera 1999. Weekly epidemiological cholerae. Applied and environmental record, 2000, 75(31):249–26. microbiology, 1986, 52(3):658–8. 20. Cholera 2000. Weekly epidemiological 9. Ogg JE, Ryder RA, Smith HL Jr. Isolation record, 2001, 76(31):233–40. of Vibrio cholerae from aquatic birds in Colorado and Utah. Applied and envi- 21. Franco AA et al. Cholera in Lima, Peru, ronmental microbiology, 1989, 55(1): correlates with prior isolation of Vibrio 95–9. cholerae from the environment. Ameri- can journal of epidemiology, 1997, 10. Mintz ED, Tauxe RV, Levine MM. The glo- 146(12):1067–75. bal resurgence of cholera. In: Noah N, O’Mahony M, eds. Communicable dis- 22. Fournier JM, Villeneuve S. Actualité du ease epidemiology and control. London, choléra et problématique vaccinale Wiley, 1998:63–94. [Cholera update and vaccination problems]. Médecine tropicale: revue du 11. Benson AS. Cholera. In: Evans AS, Corps de santé colonial, 1998, 58(suppl. Brachman PS, eds. Bacterial infection of 2):32–5. humans: epidemiology and control, 2nd ed. New York, Plenum, 1991:207–25.

02 Cholera epidemic in Baghdad.pmd 12 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 13

23. Siddique AK et al. Why treatment centres miology. Journal of the American Medi- failed to prevent deaths among cal Association, 1996, 27:307–12. Rwandan refugees in Goma. Lancet, 28. Khan WA et al. Randomised controlled 1995, 345:359–61. comparison of single-dose ciprofloxacin 24. Dubon JM et al. Emergence of multiple and doxycycline for cholera caused by drug-resistant Vibrio cholera 01 in San Vibrio cholerae O1 or O139. Lancet, Pedro Sula, Honduras. Lancet, 1997, 1996, 348:296–300. 349:924. 29. Mahalanabis D et al. An epidemic of 25. Duval P et al. Cholera in Madagascar. cholera due to V. cholerae O139 in Lancet, 1999, 353:2068. Dhaka, Bangladesh: clinical and epidemiological features. Epidemiology and 26. Nizami SO, Farooqi BJ. Cholera in chil- infection, 1994, 112(3):463–71. dren in Karachi from 1990 through 1995: a study of cases admitted to a tertiary 30. Eko FO, Udo SM, Antia-Obong OE. Epi- hospital. Journal of the Pakistan Medical demiology and spectrum of vibrio diar- Association, 1998, 48(6):171–3. rheas in the lower cross river basin of Nigeria. Central European journal of 27. Mahon BE et al. Reported cholerae in public health, 1994, 2(1):37–41. the United States, 1992–1994: a reflection of global changes in cholera epide-

International travel and health 2005. Situation as on 1 January 2005 International travel is undertaken by large, and ever increasing, numbers of people for professional, social, recreational and hu- manitarian purposes. Travellers are exposed to a variety of health risks in unfamiliar environments. Most such risks, however, can be minimized by suitable precautions taken before, during and after travel, and it is the purpose of this publication to provide guidance on measures to prevent or reduce any adverse consequences for travellers’ health. The book is addressed primarily to medical and public health professionals who provide health advice to travellers, but it is also intended to provide guidance to travel agents and or- ganizers, airlines and shipping companies. As far as possible, the information is presented in a form readily accessible to interested travellers and non-medical readers. The printed edition is now published only every second year but an Internet version (http:// www.who.int/ith) allows on-going updating and provides easy links to other information, such as news of current disease outbreaks of international importance. Further information on this publication can be obtained from WHO Press: http://www.who.int/bookorders/ anglais/home1.jsp?sesslan=1

02 Cholera epidemic in Baghdad.pmd 13 10/25/2005, 1:20 AM 14 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Épidémiologie des méningites à Haemophilus influenzae en Tunisie M.S. Soltani,1 A. Bchir, 1 F. Amri, 3 N. Gueddiche, 1 T. Sfar, 2 S. Sahloul 4 et M. Garbouj, 5

RÉSUMÉ Nous avons étudié l’incidence des méningites à Haemophilus influenzae (Hib) chez les enfants de moins de 5 ans en Tunisie. Pour ce faire, un système de surveillance a été mis en place pendant 14 mois à partir de juin 2000, comprenant une surveillance populationnelle dans 3 gouvernorats et une surveillance sentinelle dans 2 gouvernorats. Chaque enfant de moins de 5 ans suspect de méningite a subi une ponction lombaire. Pour tout prélèvement de liquide céphalorachidien, on a pratiqué un examen macroscopique, une numération cellulaire, une analyse chimique et une culture. Au terme de 14 mois de suivi, 80 cas de méningite bactérienne ont été colligés. Les données de la surveillance populationnelle montrent une prédominance des cas chez les enfants de moins de 1 an (73,6 %) et chez les garçons (64 %). H. influenzae a été isolé dans 38 % des cas, le pneumocoque dans 13 % des cas et le méningocoque dans 7 % des cas. L’incidence des méningites à Hib confirmées est de 14,4 pour 100 000. L’évaluation des coûts de prise en charge des méningites à Hib et de leurs complications montre que ceux-ci sont plus élevés que le coût d’introduction du vaccin. Ainsi, sur la base de cette étude, le Ministère de la Santé a décidé d’introduire la vaccination anti-Hib. Epidemiology of Haemophilus influenzae b meningitis in Tunisia ABSTRACT The incidence of Haemophilus influenzae b meningitis (Hib) in children < 5 years in Tunisia was studied through a surveillance system set up in June 2000 and followed for 14 months. Population-based surveillance began in 3 governorates and sentinel surveillance in 2. Children < 5 years suspected of meningitis had lumbar puncture, macroscopic exam, blood count, chemical analysis and culture carried out. In the 14 months, 80 cases of meningitis were recorded. From the population-based surveillance most cases were children < 1 year (73.6%) and boys (64%). H. influenzae was isolated in 38% of cases, pneumococci in 13% and meningococci in 7%. The incidence of confirmed Hib was 14.4/100 000 children. The estimated cost of identifying and treating Hib meningitis and its complications was greater than the cost of vaccine introduction. Based this study, the Ministry of Health has decided to introduce Hib vaccination.

1C.H.U. de Monastir ; 2C.H.U. de Mahdia ; 3C.H.U. de Kairouan ; 4C.H.U. de Sousse, (Tunisie) (Correspondance à adresser à M.S. Soltani : [email protected]). 5Direction des Soins de Santé de Base du Ministère de la Santé publique, Tunis (Tunisie). Reçu : 31/12/02 ; accepté : 08/04/03

03 EpidÈmiologie des mÈningites.pmd 14 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 15

Introduction l’épidémiologie et le poids économique des méningites à Hib. À terme, ce système de Haemophilus influenzae de sérotype b surveillance fournira aux décideurs de la (Hib) est une cause majeure d’infections santé publique les éléments nécessaires bactériennes sévères chez le nourrisson et pour une prise de décision éclairée en ma- l’enfant. Il est notamment le premier agent tière d’introduction du vaccin contre le Hib. bactérien responsable de méningite purulente chez le nourrisson [1-3]. Un vaccin contre le Hib est actuellement Description du système de disponible et a permis une diminution rapide surveillance des pathologies graves à Hib dans de nom- Objectifs de la surveillance breux pays industrialisés où les nourrissons Ce système de surveillance a pour objectifs sont habituellement vaccinés. L’introduc- principaux : tion du vaccin contre le Hib dans le pro- d’estimer l’incidence de la méningite à gramme de vaccination doit considérer • Hib dans les régions retenues pour la l’épidémiologie de l’infection, les priorités surveillance et dresser le profil épidé- nationales et les capacités économiques de miologique de la méningite à Hib ; chaque pays. L’utilisation de ce vaccin dans les pays • de déterminer le taux de létalité des en développement a été limitée, en partie à méningites à Hib ; cause du manque de données sur • de déterminer la proportion des ménin- l’importance des pathologies graves à Hib gites bactériennes causées par Hib ; [4] et à cause du coût élevé du vaccin. • d’estimer la fréquence des séquelles En Tunisie, l’analyse des données pro- liées aux méningites à Hib ; venant d’études hospitalières indique que le • de calculer les coûts médicaux occa- Hib est la cause principale des méningites sionnés par la maladie. bactériennes des enfants avec un taux d’incidence de méningites à Hib proche de Modalités de la surveillance ceux rapportés par d’autres pays de la ré- Le système de surveillance comprend deux gion, soit 16 à 25 cas pour 100 000 enfants volets : une surveillance sur une base popu- âgés de moins de 5 ans. Le taux de létalité lationnelle et une surveillance à partir de atteint 15 % selon les études [5,6]. Il a été sites sentinelles. estimé par ailleurs que le germe est responsable de 750 à 3000 épisodes de pneumo- La surveillance populationnelle pathies par an [7]. Il s’agit d’une surveillance sur une base Ces données ont été jugées insuffisantes populationnelle qui permet de recenser tous pour décider de la pertinence de les cas de méningites bactériennes et de dé- l’introduction du vaccin dans le pro- terminer ceux dus à Hib. Afin d’estimer le gramme national de vaccination. Face à taux d’incidence de la méningite à Hib de l’insuffisance de données populationnelles manière fiable, il faut donc choisir des permettant d’estimer la fréquence des in- zones géographiques pour lesquelles un re- fections à Hib, le Ministère de la Santé pu- censement exhaustif des cas est assuré. blique a décidé de mettre en place un Le choix a été porté sur les 4 gouver- système de surveillance des méningites. Ce norats de Sousse, Monastir, Mahdia et système devrait permettre de déterminer

03 EpidÈmiologie des mÈningites.pmd 15 10/25/2005, 1:20 AM 16 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Kairouan. Ce choix a été dicté par les con- hôpitaux de la ville de Sfax. Ces hôpitaux sidérations suivantes : les 4 gouvernorats ont d’importants services de pédiatrie et comptent une population de plus d’un mil- bénéficient de l’appui de laboratoires per- lion d’habitants. Le nombre de structures formants. de pédiatrie prenant en charge les ménin- gites est limité. L’accès aux soins est facile. Méthodes de collecte des données Des laboratoires performants existent sur Tout enfant de moins de 5 ans qui présente place. L’infrastructure privée d’hospitali- une suspicion de méningite et qui consulte sation n’est pas très importante. Collective- dans l’un des sites de surveillance (struc- ment, les structures sanitaires prennent en ture publique ou privée) subit une ponction charge l’ensemble des cas de méningite lombaire et le liquide céphalorachidien dans les 4 régions. Et, finalement, plusieurs (LCR) est envoyé immédiatement au labo- équipes ont déjà travaillé sur le sujet. ratoire. Pour tout prélèvement de LCR [8], La surveillance cible les enfants âgés de on pratique l’examen macroscopique du moins de 5 ans résidant dans ces 4 gouver- liquide, une numération cellulaire, une ana- norats. Tous les enfants atteints de ménin- lyse chimique (glycorachie et protéinora- gite bactérienne sont concernés par la chie) et une culture à la recherche de surveillance, qu’ils soient hospitalisés dans germes pathogènes. le secteur public ou privé. Un registre est tenu au laboratoire et La surveillance des méningites à Hib tous les prélèvements de LCR des enfants devrait durer une année, ce qui permettra suspects de méningite sont notés. Ce regis- d’identifier une trentaine de cas (sous tre est tenu conjointement par le chef de l’hypothèse d’un taux d’incidence annuelle laboratoire et le pédiatre responsable du de 20 pour 100 000 enfants âgés de moins système de surveillance. Il comporte les de 5 ans). Ce nombre attendu de cas de renseignements suivants : nom et prénom méningite à Hib est toutefois faible pour dé- du malade, âge, date du prélèvement, as- terminer le taux de létalité, le coût de la pect du liquide, numération cellulaire, gly- prise en charge et la fréquence des compli- corachie, protéinorachie, coloration de cations. Ainsi il est important d’associer à Gram, résultat de la culture et test de latex. ce système de surveillance populationnelle Ce registre est rempli quotidiennement et un autre système s’appuyant sur des sites une copie est adressée à la Direction des sentinelles. Soins de Santé de Base (Ministère de la Santé publique) mensuellement. La surveillance à partir des sites Suite aux résultats des examens sentinelles biologiques, les malades sont classés dans Les sites sentinelles serviront à identifier un l’une des catégories suivantes : nombre plus important de méningites à Hib, • méningite bactérienne confirmée à Hib ; ce qui permettra de compléter les données • méningite bactérienne confirmée à autre de la surveillance populationnelle afin de germe (ou méningite purulente à culture calculer la fréquence des complications et positive) ; le coût de la prise en charge. • méningite bactérienne probable (ou Les deux sites sentinelles choisis sont méningite purulente à culture négative). l’hôpital d’enfants de Tunis et les deux

03 EpidÈmiologie des mÈningites.pmd 16 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 17

Les cas confirmés de méningite bacté- Description de la mise en rienne sont définis par : œuvre du système de • une culture positive du LCR (utilisant le surveillance latex au besoin) ; La phase préparatoire • un germe mis en évidence à l’examen Une commission nationale a été créée ; elle direct. est chargée de discuter la stratégie pro- Les cas probables de méningite bacté- posée et de suivre la mise en œuvre du rienne ont été définis par un LCR qui système. Après l’accord sur la stratégie de présente l’une des caractéristiques sui- surveillance des méningites, il y a eu identi- vantes : fication des structures de pédiatrie (pu- • LCR trouble ; bliques et privées) engagées dans la prise en • 100 globules blancs quelle que soit la charge des méningites et des laboratoires formule ; (publics et privés) et évaluation de leur ca- pacité de diagnostic de Hib. • 10 à 99 globules blancs avec une Par la suite, la commission nationale a protéinorachie > 1 g/L et une défini les modalités de mise à niveau des glycorachie < 2,2 mmol/L. différents laboratoires : utilisation d’un Pour être classés probables, tous ces manuel de procédures, standardisation de la cas doivent avoir en plus une évolution cli- procédure du diagnostic, contrôle de la nique compatible avec le diagnostic de qualité. Nous avons également décidé de méningite bactérienne. Ainsi le diagnostic tester les différents outils au cours d’une de méningite virale est retenu si, au cours période d’essai. de l’évolution clinique, l’enfant ne présente plus de fièvre, ni de signes inflammatoires La période d’essai et que son état de santé au cours de la visite Elle s’est déroulée au cours du mois de mai de contrôle est normal. 2000 et a été suivie d’une réunion Pour chaque cas de méningite con- d’évaluation au cours de laquelle nous firmée ou probable, une fiche de déclara- avons passé en revue les difficultés de mise tion est envoyée à la Direction des Soins de en œuvre, en particulier l’implication du Santé de Base à la fin de l’hospitalisation. privé et des pédiatres de Tunis et la qualité Cette fiche comporte des informations sur des informations collectées. Suite à quoi, le lieu de la résidence, les résultats des exa- nous avons décidé que : mens biologiques et le coût de la prise en • les médecins coordinateurs régionaux charge dans la structure sanitaire. devraient relancer les pédiatres privés ; Les techniques de laboratoire • dans chaque service de pédiatrie, un L’isolement de Haemophilus influenzae a responsable serait chargé du dossier été réalisé après un ensemencement en pour assurer le bon suivi du système ; stries sur une gélose chocolat complé- • pour Tunis, il faut charger un coordina- mentée en facteurs X et V et incubée à 35- teur par service. 37 °C sous une atmosphère contenant 5 à Nous avons finalement décidé de débu- ter officiellement la surveillance en juin 10 % de CO2 pendant au moins 4 jours. Le diagnostic de Haemophilus influenzae a été 2000. également réalisé par le test d’agglutination sur particules de latex.

03 EpidÈmiologie des mÈningites.pmd 17 10/25/2005, 1:20 AM 18 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Évaluation du processus • relancer les pédiatres coordinateurs pour s’assurer de la marche du système Le suivi a été assuré par des contacts télé- dans leur région (fiche de déclaration, phoniques, par des courriers pour les dif- registre de laboratoire, données man- férentes régions et des visites de quantes, exhaustivité, acheminement supervision. des rapports à la Direction des Soins de L’évaluation à mi-parcours en janvier Santé de Base à temps) et s’assurer de 2001 a consisté en une visite des différents l’existence ou non de cas de méningite services de pédiatrie avec vérification des dans le secteur privé ; différents supports du système de surveil- • vérifier systématiquement l’exhaustivité lance ainsi qu’une réunion le 13 janvier des cas sur le registre d’admission des 2001. Le constat de cette évaluation est services ; que : • compléter les données manquantes au • le système fonctionne correctement niveau des fiches de déclaration pour dans les gouvernorats de Mahdia, Mo- toute la période ; nastir et Kairouan et à Sfax, hormis quelques données manquantes au • lister tous les cas probables biologiques niveau des déclarations et des rapports et prendre, sur ces derniers, les mensuels et le retard de l’acheminement renseignements cliniques d’évolution. de quelques rapports. Durant la 2e période du suivi, nous • Le système n’a pas très bien fonctionné avons veillé à l’application de ces recom- à Sousse : le registre de laboratoire n’est mandations. C’est ainsi que : pas tenu à jour ; manque de coordina- • 5 réunions de suivi avec tous les inter- tion entre les 2 laboratoires de biochimie venants ont été organisées ; et de bactériologie, manque de coordi- • les fiches de déclaration comportant des nation entre le pédiatre et les labora- données manquantes ont été adressées toires. aux pédiatres, complétées et re-saisies ; • Le fonctionnement du système dans le • l’exhaustivité des déclarations a été site sentinelle de Tunis est très défi- vérifiée dans tous les services en les cient : les médecins-chefs de service ne confrontant aux données de laboratoire semblaient pas adhérer au système. et du registre d’admission ; Cette évaluation afin d’assurer le bon • les registres du laboratoire de Sousse fonctionnement du système nous a permis ont été complétés ; d’adopter les décisions suivantes : • le listage des cas probables biologiques • exclure le site de Tunis et considérer a été élaboré pour chaque service. Pour Sousse comme un site sentinelle à cause chaque cas, les renseignements d’évo- des difficultés rencontrées dans la mise lution clinique ont été revus sur le dos- en œuvre du système, reconnaître sier clinique et une concertation avec le l’importance du secteur privé ; pédiatre sur ces cas a été assurée afin de • organiser et multiplier des visites de su- statuer sur ces cas. pervision au niveau du terrain ; • assurer des réunions de suivi men- suelles pour détecter les problèmes à temps et y remédier ;

03 EpidÈmiologie des mÈningites.pmd 18 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 19

Évaluation des résultats Les 53 cas de méningites bactériennes originaires de 3 gouvernorats de la zone de Analyse globale surveillance populationnelle se composent Cas déclarés de méningites bactériennes de 20 cas de Hib, de 13 cas de méningites Nous avons colligé au cours des 14 mois de bactériennes confirmées et de 20 cas de surveillance 80 cas de méningites bacté- méningites bactériennes probables. riennes chez les enfants âgés de 1 à 59 mois, qui ont été déclarés au niveau des Caractéristiques des méningites sites de surveillance populationnelle et des bactériennes sites sentinelles de Sfax et Sousse. Le La répartition selon l’âge des méningites tableau 1 décrit la répartition de l’ensemble bactériennes déclarées montre que 63 % des cas de méningites selon leur lieu ont été recensées chez des enfants âgés de d’hospitalisation. 1 à 12 mois (Tableau 3). Sur l’ensemble des La répartition des cas de méningites 26 méningites à Hib, 11 (43 %) sont surve- selon le gouvernorat d’origine (Tableau 2) nues chez des enfants âgés entre 7 et montre que 53 cas sont originaires de la 12 mois. zone de surveillance populationnelle et Les méningites bactériennes déclarées 27 cas sont originaires des gouvernorats sont plus fréquentes chez les enfants de des sites sentinelles ou d’autres gouverno- sexe masculin que chez ceux de sexe fémi- rats limitrophes (Sidi Bouzid : 2 cas ; Za- nin (68 % de garçons contre 32 % de filles) ghouan : 2 cas ; Kasserine : 2 cas ; Gaf- (Tableau 3). Parmi les 26 cas de méningites sa :1 cas). bactériennes à Hib, 15 cas ont été déclarés Les méningites bactériennes à Hib chez des garçons, soit 57 % des cas. représentent 32,5 % de l’ensemble des Les principaux germes responsables méningites déclarées (26 cas). Les ménin- des méningites bactériennes sont le Hib gites bactériennes confirmées à autres ger- dans 1/3 des cas, le pneumocoque dans 1/6 mes sont au nombre de 21 cas, soit 26 %. des cas et le méningocoque dans 7,5 % des Celles probables sont au nombre de 33 cas, cas (Tableau 4). soit 41,5 %.

Tableau 1 Répartition des cas de méningites bactériennes selon le lieu d’hospitalisation

Sites de Méningites Méningites Méningites Total surveillance à Hib bactériennes à bactériennes autres germes probables

Kairouan 5 4 16 25 Mahdia 8 6 2 16 Monastir 6 3 2 11 Sfax 4 4 5 13 Sousse 3 4 8 15 Total 26 21 33 80

03 EpidÈmiologie des mÈningites.pmd 19 10/25/2005, 1:20 AM 20 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Tableau 2 Déclaration des méningites Étude des méningites bactériennes par gouvernorat d’origine bactériennes déclarées dans la zone de surveillance Gouvernorat Méningite bactérienne populationnelle d’origine Hib Autres Probable Total Dans les trois gouvernorats de la surveil- germes lance populationnelle, nous avons colligé Kairouan 6 3 16 25 durant les 14 mois de surveillance 53 cas de Mahdia 8 5 2 15 méningites bactériennes dont 20 cas sont à Monastir 6 5 2 13 Hib (six cas proviennent de Kairouan, 8 cas proviennent de Mahdia et 6 cas proviennent Sfax 4 3 4 11 de Monastir). Sousse 0 3 6 9 Régions Caractéristiques des méningites limitrophes 2 2 3 7 bactériennes de la zone de surveillance Total 26 21 33 80 populationnelle L’étude des caractéristiques des méningites bactériennes recensées dans la zone de surveillance populationnelle montre une pré- Quarante pour cent (40 %) des cas de dominance chez les enfants âgés de moins méningites bactériennes ont été recensés en de 1 an (73,6 % des cas) et chez le sexe hiver et en automne ; 70 % des cas de mé- masculin (64 %) (Tableaux 3 et 5). Elles ningites à Hib ont été déclarés pendant ces surviennent surtout en hiver et en automne. deux saisons avec une prédominance au Le germe responsable est le Hib dans mois de septembre (19 % des cas). 38 % des cas, le pneumocoque dans 13 %

Tableau 3 Répartition des méningites bactériennes déclarées selon l’âge et le sexe

Variable Méningites Autres Méningites Total à Hib méningites bactériennes bactériennes probables

Âge (mois) 1 à 6 4 9 14 27 7 à 12 11 4 9 24 13 à 24 8 6 6 20 > 24 3 2 4 9 Sexe Masculin 15 15 25 55 Féminin 11 6 8 25 Total 26 21 33 80

03 EpidÈmiologie des mÈningites.pmd 20 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 21

Tableau 4 Types de germes des méningites Pour avoir une estimation plus correcte bactériennes déclarées du taux d’incidence des méningites à Hib, il faut inclure dans le numérateur le nombre Type de germe Nombre % de méningites dues probablement à Hib. Ce Haemophilus 26 32,5 nombre est estimé en multipliant le nombre de méningites bactériennes probables par la Pneumocoque 12 15,0 proportion de méningites à Hib parmi les Méningocoque 6 7,5 méningites bactériennes confirmées ; ainsi Pyocyanique 1 1,2 par exemple, pour Kairouan, la proportion Enterobacter 1 1,2 des méningites à Hib représente les 2/3 des Salmonella 1 1,2 méningites bactériennes confirmées ; donc Germe non identifié 33 41,2 les cas de méningites à Hib parmi les mé- Total 80 100,0 ningites probables (n = 16) seraient de 2/3 de ce nombre, soit 16 × 2/3 = 10,7 (Tableau 6). Ainsi le taux d’incidence des méningites des cas et le méningocoque dans 7 % des à Hib confirmées et probables dans les cas. 3 gouvernorats est de 23,7 pour 100 000 enfants âgés de moins de 5 ans (Tableaux 6 Incidence des méningites bactériennes et 7). à Hib Les études antérieures estimaient que le Le taux d’incidence des méningites bacté- taux d’incidence des méningites bacté- riennes à Hib confirmées dans l’ensemble riennes à Hib se situait entre 16 et 25 pour de ces 3 gouvernorats, sachant que la durée 100 000 enfants âgés de 1 à 59 mois. Notre globale de la surveillance est de 14 mois, système de surveillance confirme cette hy- est de : 20 × 12 × 100 000 ÷ 119 160 × 14 pothèse. = 14,4 pour 100 000 enfants âgés de 1 à Nous pensons que ce taux d’incidence 59 mois. reflète de façon correcte la situation épidé- miologique au cours de la période de l’étude car tous les cas hospitalisés ont été recensés et surtout la disponibilité des don- Tableau 5 Répartition des méningites nées cliniques a permis d’estimer de façon bactériennes selon l’âge et le sexe fiable le nombre de cas probables. En effet, Variable Hib Autres probables Total habituellement les cas sont classés proba- germes bles seulement sur la base de critères biologiques, alors que dans notre système Âge (mois) tous les cas biologiquement probables ont 1-5 4 6 10 20 6-11 11 2 6 19 été revus et reclassés tenant compte de 12-24 4 4 3 11 l’évolution clinique (température, signes in- > 24 1 1 1 1 flammatoires, état de l’enfant à la sortie et Sexe au cours du suivi ultérieur). Ceci a permis Masculin 11 8 15 34 de réduire au minimum le nombre de cas de Féminin 9 5 5 19 méningites probables. De plus, l’approche Total 20 13 20 53 prospective adoptée pour l’estimation du taux d’incidence garantit une meilleure

03 EpidÈmiologie des mÈningites.pmd 21 10/25/2005, 1:20 AM 22 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Tableau 6 Estimation des cas de Hib parmi les cas probables

Région Population Nombre Nombre total % des Nombre Nombre d’enfants de cas de de cas de cas de de cas de cas de âgés de 1 méningites méningites Hib probables méningites à 59 mois à Hib bactériennes à Hib parmi confirmées les cas probables

Kairouan 50 239 6 9 66,7 16 10,7 Mahdia 32 472 8 13 61,6 2 1,2 Monastir 36 449 6 11 54,5 2 1,1 Les 3 gouvernorats 119 160 20 33 60,6 20 12,1

qualité et validité de l’information que cours de l’automne 2001 dépasse de loin ce l’approche rétrospective souvent utilisée qui était rapporté au cours de la même pé- dans les autres études. riode durant l’année de l’étude. Il faut néanmoins souligner que le taux Le taux d’incidence retrouvé à travers d’incidence peut varier d’une année à notre système de surveillance est proche l’autre et que les pédiatres ont remarqué des taux rapportés par d’autres pays de la que la fréquence des méningites à Hib était région durant les années 90 tels qu’Israël, particulièrement basse au cours de l’année Koweït et Qatar où il varie entre 16 et 31 de l’étude. Cette hypothèse a été rapide- pour 100 0000. Il est par contre plus bas ment vérifiée puisque dans certains gouver- que ceux rapportés dans certains pays de norats (Monastir en particulier) le nombre l’Afrique subsaharienne où celui-ci se situe de méningites à Hib diagnostiquées au entre 40 et 57 pour 100 000 [8].

Tableau 7 Taux d’incidence des méningites à Hib chez les enfants âgés de 1 à 59 mois

Région Incidence des cas Incidence des cas confirmés à Hib pour confirmés et des cas 100 000 enfants probables de méningites de 1 à 59 mois à Hib pour 100 000 enfants de 1 à 59 mois

Kairouan 10,2 28,5 Mahdia 21,08 24,3 Monastir 14,14 16,7 Les 3 gouvernorats 14,4 23,7

03 EpidÈmiologie des mÈningites.pmd 22 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 23

Détermination de la gravité de la maladie taires qui ont eu lieu après l’hospitalisation. Nous avons recensé deux décès parmi les Par ailleurs, il faut noter que le coût ainsi 20 cas de méningites bactériennes à Hib. Le calculé est vraisemblablement en dessous taux de létalité est donc de 10 %. La cause de la réalité car certains examens complé- de décès dans les deux cas a été un état de mentaires ou traitements ont pu ne pas être choc septique. notifiés dans les dossiers. Il s’agit donc Dans certains pays développés, on a es- d’une estimation des coûts directs sup- timé que 10 % des enfants atteints de mé- portés par le secteur hospitalier pour la ningites à Hib décèdent et 10 à 35 % prise en charge de la phase aiguë de la ma- survivent avec des séquelles invalidantes ladie et non de ses séquelles. Les autres pa- telles que la surdité ou le retard mental [9]. thologies causées par Hib ne sont pas Dans les pays en développement, le taux de considérées dans cette estimation du coût. létalité peut atteindre 40 % [9]. L’estimation des coûts de ces com- Parmi les 20 cas de méningites à Hib, posantes a été réalisé comme suit et résumé 12 enfants, soit 60 % des cas, ont présenté dans le tableau 8. au moins une complication. L’analyse de Pour le calcul du coût de séjour : la l’ensemble de ces complications montre durée moyenne du séjour est de 18 ± qu’elles sont dominées par les convulsions 15 jours. Ce séjour varie entre 1 jour (cas (4 cas), le coma (2 cas), les arthrites (2 décédé) et un maximum de 56 jours. Le cas), l’état de choc septique (2 cas) et coût de séjour moyen est de TND 606 ± l’hémorragie digestive (2 cas). 386. Quant aux séquelles, six patients (30 % Pour celui des examens radiologiques et des cas) ont gardé au moins une séquelle, et les actes : sont comptabilisées dans cette ce seulement au cours de la période initiale rubrique toutes les explorations autres que d’hospitalisation et de suivi immédiat. Les biologiques, essentiellement les examens principales séquelles sont à type d’atrophie radiologiques. Ces examens coûtent en corticale (2 cas), épilepsie, hydrocéphalie moyenne TND 197 ± 247. (2 cas) et hypotonie axiale (2 cas). Concernant le coût des explorations biologiques : tous les examens biologiques Étude du coût de la prise en charge des pratiqués chez ces enfants et inscrits sur méningites à Hib les déclarations ont été comptabilisés. Nous allons considérer ici le coût de la prise en charge des méningites à Hib dé- clarées dans la zone de surveillance popula- Tableau 8 Coûts moyens de la prise en charge tionnelle. Ces coûts de prise en charge des méningites bactériennes à Hib exprimés englobent le coût du séjour hospitalier en dinars tunisiens (TND) (dans un service de pédiatrie, le coût d’une journée d’hospitalisation est de 30 dinars), Prise en charge Coût moyen Min/Max le coût des examens complémentaires ra- (TND) diologiques et biologiques et le coût des Séjour 606 ± 386 30/1290 médicaments. Il est à noter que cette esti- Examens radiologiques 197 ± 247 13,5/1101 mation des coûts ne comprend pas les Examens biologiques 145 ± 64 73/320 dépenses encourues par la famille avant l’hospitalisation, le coût des consultations Médicaments 136 ± 125 8/457 ambulatoires et des examens complémen- Total 1086 ± 740 156/3096

03 EpidÈmiologie des mÈningites.pmd 23 10/25/2005, 1:20 AM 24 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Leurs coûts s’évaluent en moyenne à • 122 à 195 cas de méningites bacté- TND 154 ± 64. riennes à Hib, en appliquant le taux Finalement pour le coût des médica- d’incidence annuelle des méningites à ments : la méningite est une maladie grave ; Hib confirmées (14,4 pour 100 000) et en général, les malades ont bénéficié d’une le taux des méningites confirmées et association d’antibiotiques à base d’une probables (23,7 pour 100 000) ; céphalosporine de 3e génération et d’un • le nombre de décès à déplorer par an au aminoside. Il est à noter que certaines cours de l’hospitalisation des cas de présentations d’antibiotiques injectables ne méningites à Hib varierait de 12 à 19 sont utilisées qu’en partie (dosage élevé (taux de létalité : 10 %) selon que l’on pour les besoins pédiatriques). Le coût cal- considère l’un ou l’autre des taux culé ne tient cependant compte que de la d’incidence susmentionnés ; quantité prescrite, ce qui sous-estime le • 37 à 58 cas de séquelles de méningites à coût réel. Le coût moyen du traitement est Hib considérant une fréquence de évalué à TND 136 ± 125. séquelles de 30 % ; Par ailleurs, l’étude des coûts totaux par • 610 à 975 épisodes de pneumonie à Hib rubrique montre que les coûts du séjour avec 30 à 49 cas de décès (considérant hospitalier sont les plus importants, suivis les deux taux d’incidence susmention- par ceux des actes et des examens ra- nés et que, selon l’OMS, il y a un cas de diologiques. méningite pour 5 cas de pneumonie) (Tableau 9). Estimations nationales de la morbidité et du poids économique des pathologies liées Pour l’estimation du coût de la prise en à Hib charge hospitalière des méningites à Hib at- L’extrapolation des données de la surveil- tendues dans la population tunisienne lance populationnelle à l’ensemble de la d’enfants âgés de 1 à 59 mois, nous avons population tunisienne d’enfants âgés de 1 à considéré le coût moyen de toutes les mé- 59 mois montre que le poids de la morbidité ningites confirmées à Hib et déclarées par annuelle par Hib attendue serait de : l’ensemble des structures participant à la

Tableau 9 Projection nationale de la morbidité annuelle par Hib attendue chez les enfants de 59 mois en Tunisie, basée sur les données de la surveillance populationnelle

Méningites à Hib Méningites confirmées confirmées par la culture et probables

Cas attendus de méningites à Hib 848 966 × 14,4 ÷ 100 000 = 122 848 966 × 23,7 ÷ 100 000 = 195 Cas de décès attendus par méningites à Hib (10 %) 12 19 Cas de séquelles attendues par méningites à Hib (30 %) 37 58

La population d’enfants tunisiens âgés de 1-59 mois était de 848 966 en 1999 selon les estimations de l’INS (9 % de la population totale).

03 EpidÈmiologie des mÈningites.pmd 24 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 25

surveillance (sites sentinelles et sites de la • coût direct et indirect supporté par les surveillance populationnelle), soit 26 cas. familles ; Ce coût serait de TND 211 770 (195 • coût de la prise en charge médicale pour cas × TND 1086) considérant le taux le suivi et l’exploration des séquelles. d’incidence maximum (cas confirmés et A titre indicatif, signalons que l’OMS probables). En plus des réserves susmen- estime à TND 15 le coût journalier de la tionnées concernant la validité de prise en charge d’un enfant handicapé, soit l’estimation du coût, il faut aussi mention- TND 5460 par an. Considérant que 58 en- ner que ce calcul a supposé que le coût fants par an présentent des séquelles de moyen estimé à partir des services de pé- méningite nécessitant une prise en charge diatrie hospitalo-universitaires peut être ex- spécialisée, et qu’un enfant est suivi au trapolé aux autres structures hospitalières moins pendant 10 ans, nous estimons que pouvant prendre en charge ce genre de pa- le coût annuel de leur prise en charge est de thologie. Ceci n’est pas nécessairement TND 3 166 800 (TND 5460 × 58 enfants × vrai. En effet, dans les hôpitaux régionaux, 10 ans). les facilités de laboratoire et de traitement Ainsi le coût global de prise en charge ne sont pas les mêmes que celles des hôpi- des méningites à Hib d’une année et de taux universitaires. De ce fait, la durée leurs séquelles sur une période seulement d’hospitalisation peut être plus longue, aug- de 10 ans serait de TND 3 378 570, sans mentant ainsi le coût de la prise en charge. compter la prise en charge des autres Concernant l’estimation du coût de la problèmes dus à Haemophilus. prise en charge hospitalière des autres pathologies liées au Hib, il faut préciser que le Estimation du coût de système mis en place ne comportait pas la l’introduction éventuelle du vaccin surveillance et donc l’estimation des coûts anti-Hib des pneumonies à Hib ni des autres patholo- Pour le calcul du coût de l’introduction du gies. Une revue de la littérature nationale en vaccin contre le Hib dans le calendrier tu- matière de pneumopathie de l’enfant [10] a nisien de vaccination, nous avons con- été faite et a montré que le séjour moyen sidéré les éléments suivants : hospitalier par pneumopathie est de 7,3 ± • 3 doses de vaccin par enfant, sans rap- 7,5 j, soit 40 % de celui d’une méningite pel (3, 4 et 5 mois ) ; bactérienne à Hib ; il est toutefois difficile d’utiliser ces données pour estimer le coût • les prix du vaccin sont ceux proposés de la prise en charge des pneumonies à Hib. par l’Institut Pasteur de Tunis ; En effet, le protocole thérapeutique et la • le taux de perte avec un vaccin multi- fréquence des complications sont dif- doses est de 30 % selon la Direction des férents de ceux liés aux méningites à Hib. Soins de Santé de Base ; Afin de disposer d’une estimation réelle • les coûts relatifs à la formation des du coût médico-social des pathologies liées agents de santé, la mobilisation sociale à Hib, il faudrait pouvoir comptabiliser et et au changement des formulaires ad- ajouter les coûts suivants : ministratifs ne sont pas comptabilisés. • coût de la prise en charge sociale des La Tunisie compte 160 000 naissances enfants handicapés à cause du Hib ; par an ; ainsi 480 000 doses seront néces- • coût des vies perdues ; saires par an en cas d’utilisation d’un vac-

03 EpidÈmiologie des mÈningites.pmd 25 10/25/2005, 1:20 AM 26 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Tableau 10 Coûts des vaccins en dinars tunisiens (TND)

Variable Vaccin combiné Hib-DTC Vaccin monovalent Hib Vaccin combiné (10 doses/flacon) (unidose) Hib-DTC (unidose)

Prix de la dose TND 5,507 TND 5,906 TND 7,263

Besoins 686 000 doses 480 000 doses 480 000 doses

Coût du vaccin combiné TND 3 777 802 TND 2 834 880 TND 3 486 240

Coût de la composante TND 176 720 - TND 176 720 DTC Coût de la composante Hib TND 3 601 082 - TND 3 309 520

cin monodose et 686 600 en cas d’utilisa- et leur poids économique sur le système de tion d’un vaccin multidose. soins qui est important. La décision L’estimation du coût des différentes d’introduire le vaccin dans le calendrier de présentations de vaccin Hib est indiquée au vaccination doit tenir compte des autres tableau 10. priorités nationales et des possibilités finan- Il ressort de la comparaison des coûts cières du pays. reliés à la prise en charge des méningites à Finalement, sur la base de cette étude, Hib et de ceux de l’achat du vaccin que l’introduction du vaccin contre Haemophi- l’achat du vaccin revient moins cher pour lus influenzae a été décidée en Tunisie étant le Ministère. Notons au passage que les donné que le coût du vaccin est moins im- coûts occasionnés par les autres patholo- portant que le coût de la prise en charge des gies liées à Haemophilus (pneumonie, etc.) méningites à Hib et de leurs séquelles. n’ont pas été comptabilisés. Une partie du système devrait être main- tenue au cours des années à venir surtout si le vaccin est introduit ; en effet, ceci per- Conclusion mettrait de déterminer l’impact de cette action sur l’épidémiologie des méningites à Le système de surveillance a permis Hib. d’estimer la fréquence des méningites à Hib

Références

1. Pilly E. Infections à Haemophilus influ- 2. Levine OS, Schwartz B. The rationale for enzae dans : Les maladies infectieuses. population-based surveillance for Édition 2M2 1993 par l’Association des Haemophilus influenzae type b meningi- professeurs de pathologie infectieuse et tis. Pediatric infectious diseases journal, tropicale (APPIT). 1998, 17 (9 suppl.):5195–8.

03 EpidÈmiologie des mÈningites.pmd 26 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 27

3. Schuchat A, Wenger JD. Épidémiologie in Tunisia 10–19 September 1999. OMS : des méningites bactériennes. Annales article publié. Nestlé, 1997, 55:87–100. 8. Levine OS et al. Protocole générique de 4. Dumonceaux A et al. Impact de la vacci- surveillance des populations pour Hae- nation anti-haemophilus influenzae sur mophilus influenzae type b. Genève, l’incidence des infections invasives à Organisation mondiale de la Santé, Haemophilus influenzae de type b dans 1998. la région Nord-Pas-de-Calais. Archives 9. Introduction of Hib vaccine into immuni- de pédiatrie, 1999, 6:617–24. zation programmes: management 5. Chaabouni M et al. Les méningites guidelines. Geneva, World Health Orga- purulentes du nourrisson : à propos de nization, 2000. 52 cas. Maghreb médical, 1999, 337:8– 10. Bouzaiene Y. Morbidité hospitalière chez 14. les enfants d’âge préscolaire dans le 6. Boukadida J et al. Aspects bactério- gouvernorat de Monastir [Thèse de logiques des méningites purulentes: Médecine]. Monastir, Faculté de étude de 732 bactéries caractérisées Médecine de Monastir, 1998. dans le CSF. Revue maghrébine de 11. Peltola H et al. Perspective: A five-coun- pédiatrie, 1998, 8(5):227–33. try analysis of the impact of four different 7. Levine OS. Consultation of surveillance Haemophilus influenzae type b conju- for Haemophilus influenzae type B (Hib) gates and vaccination strategies in disease and evaluation of the need for Scandinavia. Journal of infectious dis- routine infant immunization against Hib eases, 1999, 179:223–9.

Activités de vaccination de l’OMS Dans le domaine de la vaccination, l’OMS travaille avec des partenaires tels que des gouvernements, des institutions des Na- tions Unies et d’autres organisations internationales, des organismes publics bilatéraux de santé et de développement, des organisations non gouvernementales, des associations profes- sionnelles et le secteur privé. Les responsabilités propres à l’OMS sont les suivantes : • Soutenir et faciliter la recherche et le développement ; • Assurer la qualité et l’innocuité des vaccins ; • Élaborer des politiques et des stratégies pour accroître au maximum l’utilisation des vaccins ; • Réduire les obstacles financiers et techniques à l’adoption des vaccins et des technologies ; et • Aider les pays à acquérir les compétences et les infrastructures nécessaires pour combattre et éradiquer les maladies. Source : Aide-mémoire N° 288, 2005

03 EpidÈmiologie des mÈningites.pmd 27 10/25/2005, 1:20 AM 28 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Kawasaki disease in East Mazandaran, Islamic Republic of Iran, 1997–2002 M.J. Saffar1 and F. Reshidighader 1

ABSTRACT We conducted a retrospective review of all cases of Kawasaki disease admitted to the major referral centres in Sari, East Mazandaran from 1 November 1997 to 30 October 2002. Of 29 probable cases, 25 were confirmed, giving an average annual incidence rate of 7.3 per 100 000 in children 0–5 years. The mean age of all cases was 38 months (range: 3.5–80 months). There was a male predominance. The mean time between the onset of illness and diagnosis was 8.2 ± 3.1 days. Five of the 25 patients had cardiovascular abnormalities before treatment. All patients were treated with intravenous immunoglobulin and 22 (88%) responded to a single dose of treatment and high doses of acetylsalicylic acid; 2 patients needed a second dose of IVIG, and 1 patient required a third. Complete recovery was noted in all patients except for 1, who is under follow-up and treatment.

Maladie de Kawasaki dans la partie orientale de la province de Mazandaran (République islamique d’Iran), 1997-2002 RÉSUMÉ Nous avons réalisé une étude rétrospective de tous les cas de maladie de Kawasaki admis dans les principaux centres d’orientation-recours de Sari (partie orientale de la province de Mazandaran) du 1er novembre 1997 au 30 octobre 2002. Vingt-cinq (25) des 29 cas probables ont été confirmés, ce qui donne un taux d’incidence annuel de 7,3 pour 100 000 chez les enfants de 0-5 ans. L’âge moyen de tous les cas était de 38 mois (extrêmes 3,5-80 mois). Il y avait une prédominance masculine. Le temps moyen entre le début de la maladie et le diagnostic était de 8,2 ± 3,1 jours. Cinq des 25 patients présentaient des anomalies cardio- vasculaires avant le traitement. Tous les patients ont été traités par immunoglobuline intraveineuse et 22 (88 %) ont répondu à une seule dose de traitement et à de fortes doses d’acide acétylsalicylique ; 2 patients ont eu besoin d’une seconde dose d’immunoglobuline intraveineuse et un patient a eu besoin d’une troisième dose. Une guérison complète a été constatée chez tous les patients sauf un, qui fait l’objet d’un suivi et d’un traitement.

1Department of Paediatrics, Boali Hospital, Mazandaran University of Medical Sciences, Sari, Islamic Republic of Iran (Correspondence to M.J. Saffar: [email protected]). Received: 05/06/03; accepted: 25/02/04

04 Kawasaki disease.pmd 28 10/25/2005, 1:20 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 29

Introduction findings, management and outcomes of Kawasaki disease in Iranian children in East Kawasaki disease is an acute vasculitis of Mazandaran (northern Islamic Republic of infants and children which has been report- Iran) during the 5-year period between ed worldwide. High fever, bilateral non- 1997 and 2002. purulent conjunctivitis, cervical lymphadenopathy, skin rash, and oral mucosal and extremity changes characterize the disease. Methods The diagnosis of Kawasaki disease is clinical and based on the presence of fever and East Mazandaran is located in the north of at least 4 of the above-mentioned clinical the Islamic Republic of Iran. It has an esti- findings, after exclusion of other childhood mated population of 1.5 million and a tem- conditions that may give rise to similar perate climate. Six hospitals in East signs and symptoms [1,2]. Kawasaki dis- Mazandaran admit and care for paediatric ease affects the coronary arteries in 20%– patients (1 university teaching hospital and 25% of untreated patients [3,4]. Admi- 5 others with limited paediatric beds). We nistration of intravenous immunoglobulin undertook a retrospective review of the (IVIG) within the first 10 days of illness medical records of all cases with Kawasaki decreases the prevalence of coronary ar- disease referred, treated or discharged tery abnormalities (CAA) to 2%–5% [5–7]. from these hospitals in the 5-year period Several infectious agents, environmental from November 1997 to October 2002. We and bacterial toxins as well as unusual re- confirmed diagnosis of Kawasaki disease sponse to a variety of different superanti- based on criteria outlined by the American gens have been suggested to cause Heart Association [1,2]. The diagnosis of Kawasaki disease, however the actual CAA by echocardiography (2D ECHO) cause is unknown [8,9]. was based on the criteria established by the Epidemiological studies indicate that Japanese Ministry of Health [25]. Informa- host factors, including age, sex and race, tion concerning demographic, clinical and influence the occurrence of the disease [9]. laboratory data, and management and out- Kawasaki disease occurs worldwide [10– come of each case was obtained from pa- 18] with the highest incidence reported in tients records. Parametric data were Japan [10,19]. This disease is now the expressed as mean and standard deviation most common cause of acquired heart dis- (SD) and range, incidence of the disease ease in North American and Japanese chil- was calculated by dividing the number of dren [11,19,20]. positive cases by the population of the re- Kawasaki disease was first recognized gion obtained from the census office of in Iranian children in the early 1980s [21]. Sari (53 000 for 0–5 years for each year; However, except for a few case reports 73 000 for 5–9 years for each year). [22–24], little information is available on the epidemiology of the disease in our Results country [21]. This kind of information is particularly lacking in the northern part of Of 29 probable cases in the 5-year period, the country. Thus we aimed to provide 25 were confirmed as Kawasaki disease. more detailed data on the epidemiological Yearly distribution of cases was: 2, 9, 5, 5, picture, clinical presentation, laboratory 4 for the years 1997–2002 respectively.

04 Kawasaki disease.pmd 29 10/25/2005, 1:21 AM 30 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

The average annual incidence rate of the seen by an outside physician and had re- disease was 7.3 per 100 000 in the 0–5 year ceived antibiotics prior to their admission. age group. Figure 1 shows the age distribu- Mild anaemia [mean haemoglobin level of tion of Kawasaki disease. As observed in 9.5 (SD 0.92) g/dL (range: 7.5–11.5 g/ this study, 60% of cases were younger dL)], and leukocytosis [mean white blood than 3 years, with a mean age of 38 (SD cell count of 15 700 (SD 6721)/mm3 22.3) months (range 3.5–80 months). (range: 8200–37 100/mm3)] were observed There was a male predominance of 56% in most patients (Table 2). Sequential plate- (1.27/1). let counts showed thrombocytosis [mean As shown in Figure 2 the majority of pretreatment platelet of 350 000 (SD Kawasaki disease cases occurred in au- 152 000)/mm3 (range: 110 000–598 000) tumn (10 of 25, 40%), and 32%, 16% and /mm3)]. However, 1–2 weeks post-treat- 12% occurred in the spring, summer and ment mean platelet counts were 655 000 winter respectively. The mean time interval (SD 130 000)/mm3 (range: 379 000– between the occurrence of the first sign 914 000/mm3). Most patients had evidence and diagnosis of the disease was 8.2 (3.1) of an acute phase response with a rise in days (range 3–17 days). their erythrocyte sedimentation rate (ESR) Fever for longer than 5 days was the and/or C-reactive protein concentrations, most common finding in all patients, fol- the mean ESR being 86 (SD 39.5) mm/h lowed by red lips and oral changes (92%), (range: 17–152 mm/h). bilateral conjunctival congestion (92%), All patients received 1 course of IVIG changes in peripheral extremities (84%) (23 patients, single infusion: 2 g/kg body and polymorphous skin rash (84%). Cervi- weight, 2 patients 1 g/kg), and high dose cal lymph node enlargement was the least acetylsalicylic acid (60–100 mg/kg per frequently reported cardinal sign (72%), day). Seventy-two (72) hours after com- and 84% of patients had some other signif- pletion of IVIG infusion, 22 (88%) of the icant signs and symptoms, such as irritabil- 25 patients responded and became afebrile. ity, diarrohea, vomiting and joint pain The 3 (12%) patients with persistent fever (Table 1). Most of the patients had been were re-treated with a second infusion of

Age group (months)

Figure 1 Age distribution of the 25 patients with Kawasaki disease in East Mazadaran by month

04 Kawasaki disease.pmd 30 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 31

Figure 2 Monthly and seasonal distribution of Kawasaki disease in children in East Mazandaran

IVIG, 2 of whom responded to treatment. disease on ECG and mild pericardial effu- Because of progressive CAA and continued sion on ECHO, and 4 patients with CAA on fever in 1 patient (resistant case), a third ECHO. During the 3–12 months of follow dose of IVIG (1 g/kg) was given and the up, CVA and CAA of 4 of these patients im- patient subsequently improved. proved. One patient continued to have mild ECG and ECHO detected cardiovascu- CAA. Patients’ characteristics and haema- lar abnormalities (CVA)-CAA in 5 out of 25 tological findings of patients with or with- patients (20%) in the first 48 hours of diag- out cardiac complications are shown in nosis and/or treatment: one 3.5-month-old Table 3. CVA-CAA were significantly more patient with evidence of ischaemic heart common among those who had a longer time to diagnosis and treatment [11.67 (SD 4.6) days before treatment versus 7.19 (SD Table 1 Patients fulfilling the diagnostic 2.6) days before treatment, P = 0.024] and criteria of Kawasaki disease older patients [60.75 (SD 20.22) months Criteria No of patient % versus 32.8 (SD 21.45) months, P = 0.03]. (n = 25) Fever > 5 days 25 100 Discussion Oral mucosal changes 23 92 Conjunctivitis 23 92 The patients described in this study had characteristics and results of laboratory in- Rash 21 84 vestigations similar to those reported in se- Peripheral extremity changes 21 84 ries from other countries. The disease was Cervical lymphadenopathy 18 72 seen more often in boys than girls (1.27/1) Othersa 21 84 as reported in other studies [3,10–19]. The estimated annual incidence rate of Kawasa- aOthers: associated signs and symptoms such as nausea, vomiting, abdominal pain, irritability, ki disease in this study was 7.3/100 000 drowsiness, headache, arthralgia, arthritis, dysuria, children under 5 years of age, which is frequency, diarrhoea, hepatomegaly, tachycardia, lower than that reported from Japan (90– sweating and aseptic meningitis.

04 Kawasaki disease.pmd 31 10/25/2005, 1:21 AM 32 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 2 Haematological laboratory findings in the patients with Kawasaki disease

Variable Mean SD Range

Haemoglobin (g/dL) 9.5 0.92 7.5–11.5 White blood cell counts (/mm³) 15 700 6 721 8 200–37 100 Neutrophil counts (/mm³) 11 564 2 605 4 900–29 680 Platelet count (/mm3): Before treatment 350 000 152 000 110 000–598 000 2–3 weeks after treatment 655 000 130 000 379 000–916 000 Erythrocyte sedimentation rate (mm/h) 86 39.5 17–152

SD = standard deviation.

112/100 000 children) [10,19,21] and China finding that 84% of cases were children (18.2–30.6/100 000 children) [16], similar under 5 years is consistent with 80% of to some populations in the United States of age-specific clustering of illness in Japan America (USA) (4.3–47.1/100 000 chil- and the USA [1–13,19]. However, our inci- dren) [2,11–13,26] and more than that re- dence rate is higher than previously report- ported from England and Australia ed (60%) from southern Islamic Republic (3.4–3.7/100 000 children) [14,15] and Ja- of Iran [21]. The mean age of presentation maica (2.7/100 000 children) [17]. The (38 months) was higher than that in Japan

Table 3 Characteristics of patients with Kawasaki disease according to presence or absence of cardiac complications

Variable Cardiac complication No cardiac complication P-value (n = 5) (n =20) Mean SD Mean SD

Sex ratio: male/female 4/1 1/1 Age (months) 60.75 20.22 32.84 21.45 0.03 Duration of fever before treatment (days) 11.67 4.60 7.19 2.60 0.02 Haemoglobin (g/dL) 9.32 1.08 9.53 0.80 NS White blood cell count (/mm³) 16180 3105 20922 19066 NS Neutrophil count (%) 75.2 6.76 74.55 9.46 NS Platelets before treatment (/mm³) 255.20 181.73 359.22 140.96 NS Peak platelet count (/mm³) 674.60 84.55 659.83 84.55 NS Erythrocyte sedimentation rate (mm/h) 75.80 46.51 85.55 35.84 NS

SD = standard deviation. NS = not significant.

04 Kawasaki disease.pmd 32 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 33

(9–10 months) [19], China (median 2.3 Male gender, older age and delayed di- years) [16] and the USA (2.3 years) agnosis were identified as the risk factors [12,13,26] but was similar to that in Thai for development of CVA-CAA in patients children (37 months) [18] and lower than with Kawasaki disease which concurs with reported from southern Islamic Republic of some studies [9,28–30] but not with others Iran (53 months) [21]. [27,30,31]. Haematological laboratory The number of children with the disease findings of patients with and without cardi- peaked in the autumn and spring in our ac involvement were not to significantly study, whereas the number of patients indifferent from each other. creased in winter in Japan [19], in winter A large dosage of IVIG in conjunction and spring in the USA [12,13,26] and with high dose of acetylsalicylic acid (60– southern Islamic Republic of Iran of Iran 100 mg/kg per day) initiated within 10 days [21], and in spring and summer in China of the onset of the fever decreases the oc- [16]. These findings suggest that suscepti- currence of CVA-CAA [1,2,5,6,27]. After a bility to Kawasaki disease may be related to single infusion of IVIG (2g/kg) plus acetyl- climate and season. salicylic acid (60–100 mg/kg/day), 88% of The clinical manifestations of Kawasaki our patients (22 of 25) responded to the disease in the children in our study were treatment within 48–72 hours; they became similar to those reported previously and remained afebrile and progression of [1,2,9,21]. Of the 5 cardinal clinical fea- cardiac disease was stopped and improve- tures, lymphadenopathy was the least fre- ments noted. These results are similar to quently seen. This concurs with 70% of other reports that indicated more than 85% reports from Japan [19], USA [11,12] and of patients with Kawasaki disease respond- China [16]. Conjunctivitis, exanthem, ed to 1 course of IVIG [1,5,6,7,27]. Re- oropharyngeal and changes in peripheral treatment with IVIG or administration of extremities were seen in more than 80% of pulse steroids or cytotoxic drugs in chil- the patients and were consistent with the dren with persistent fever or worsening findings of other studies [1,2,9,16–18] as CAA is advocated by some clinicians well as one report from southern Islamic [27,31–36]. In our study, 3 patients who Republic of Iran of Iran [21]. failed to respond to the initial IVIG and ace- Cardiovascular complications are the tylsalicylic acid treatment were treated with most common cause of both short- and a second dose of IVIG, 2 of them respond- long-term morbidity and mortality in pa- ed well, but the third patient did not; his fe- tients with Kawasaki disease, as noted by ver persisted and CAA worsened. He finally previous reports [3,7,9,11]. In our study, responded after a third dose of IVIG and the frequency of CVA-CAA in the patients, became afebrile and CAA progression before or shortly after the start of treatment stopped. was 20% (5 out of 25 patients). This con- In conclusion, our study provides the trasts with 2 multi-centre studies of Ka- first representative information on the epi- wasaki disease which reported 6 of 158 demiological characteristics of Kawasaki (3.8%) patients and 17 of 523 (3.2%) pa- disease among children in East Mazanda- tients with CVA-CAA [5,7]. However, our ran, Islamic Republic of Iran. The annual frequencies are within the range reported in incidence of the disease in our sample is other reports—55 of 377 patients (14.6%) similar to that reported by most studies un- [27] and 6 of 28 patients (21%) [28]. dertaken in industrialized countries and

04 Kawasaki disease.pmd 33 10/25/2005, 1:21 AM 34 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

most of the patients responded to the rec- considered in any child with persistent fe- ommended treatment regimen. Because of ver, signs of mucocutaneous inflammation the danger of the cardiac involvement, the and laboratory data associated with sys- diagnosis of Kawasaki disease should be temic inflammation.

References 1. Dajani AS et al. Diagnosis and therapy of pediatric rheumatology, 3rd ed. Philadel- Kawasaki disease in children. Circula- phia, WB Saunders, 2001:580–94. tion, 1993, 87:1776–80. 10. Yanagawa H et al. Incidence survey of 2. American Heart Association Committee Kawasaki disease in 1997 and 1998 in on Rheumatic Fever, Endocarditis and Japan. Pediatrics, 2001, 107(3):E33. Kawasaki Disease. Diagnostic guide- 11. Taubert KA, Rowley AH, Shulman ST. lines for Kawasaki disease. American Seven year national survey of Kawasaki journal of diseases of children, 1990, disease and acute rheumatic fever. Pedi- 144:1218–9. atric infectious disease journal, 1994, 3. Rowley AH, Shulaman ST. Kawasaki 13(8):704–8. syndrome. Pediatric clinics of North 12. Holman RC et al. Kawasaki syndrome America, 1999, 46:313–29. among American Indian and Alaska Na- 4. Kato H et al. Fate of coronary aneurysm tive children, 1980 through 1995. Pediat- in Kawasaki disease: Serial coronary ric infectious disease journal, 1999, angiography and long-term follow-up 18(5):451–5. study. American journal of cardiology, 13. Bronstein DE, Besser RE, Burns TC. Pas- 1982, 49:1758–66. sive surveillance for Kawasaki disease 5. Newburger JW et al. The treatment of in San Diego County. Pediatric infectious Kawasaki syndrome with intravenous disease journal, 1997, 16(11):1015–8. gamma-globulin. New England journal 14. Dhillon R et al. Management of Kawa- of medicine, 1986, 315:341–7. saki disease in the British Isles. Archives 6. Furusho K et al. High-dose intravenous of disease in childhood, 1993, 69:631– gammaglobulin for Kawasaki disease. 6. Lancet, 1984, 2(8411):1055–8. 15. Royle JA et al. Kawasaki disease in Aus- 7. Newburger JW et al. A single intrave- tralia 1993–1995. Archives of disease in nous infusion of gamma globulin as childhood, 1998, 78:33–9. compared with four infusions in the treat- 16. Du ZD et al. Epidemiologic picture of ment of acute Kawasaki syndrome. New Kawasaki disease in Beijing from 1995 England journal of medicine, 1991, 324: through 1999. Pediatric infectious dis- 1633–9. ease journal, 2002, 22(2):103–7. 8. Meissner HC, Leung DYM. Superanti- 17. Pierre R, Sue-Ho R, Watson D. Kawasaki gens, conventional antigens and the eti- syndrome in Jamaica. Pediatric infec- ology of Kawasaki syndrome. Pediatric tious disease journal, 2000, 19(6):539– infectious disease journal, 2000, 19(2): 43. 91–4. 18. Thisyakorn C, Thisyakorn U. Kawasaki 9. Petty RE, Cassidy JT. Kawasaki disease disease in Thai children. Pediatric infec- In: Petty RE, Cassidy JT, eds. Textbook of

04 Kawasaki disease.pmd 34 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 35

tious disease journal, 1995, 14(4):324– 27. Burns JC et al. Intravenous gammaglo- 5. bulin treatment and retreat-ment in Kawasaki disease. Pediatric infectious 19. Yanagawa H et al. Epidemiologic pic- disease journal, 1998, 17(12): 1144–8. tures of Kawasaki disease in Japan from the nationwide incidence survey in 1991 28. Stockheim JA, Innocentini N, Shulman and 1992. Pediatrics, 1995, 95:475–9. ST. Kawasaki disease in older children and adolescents. Journal of pediatrics, 20. Taubert KA, Rowley AH, Shulman ST. 2000, 137(2):250–2. Nationwide survey of Kawasaki disease and acute rheumatic fever. Journal of 29. McCrindie BW et al. Summary and ab- pediatrics, 1991, 119:279–82. stract of the sixth international Kawasaki disease symposium. Pediatric research, 21. Sadeghi E, Amin R, Agamee GH. 2000, 47:544–70. Kawasaki syndrome: the Iranian experience. Eastern Mediterranean health 30. Koren G et al. Kawazaki diseases: review journal, 2000, 7(1/2):16–25. of risk factors for coronary aneurisms. Journal of pediatrics, 1986, 108:388–92. 22. Sarram A. Kawasaki disease. New concepts and 15 case reports. Paper pre- 31. Fukunishi M et al. Prediction of non- sented at the Annual Congress of the responsiveness to intravenous high- Iranian Society of Pediatrics and the dose gammaglobolin therapy in patients 23rd Memorial Congress of Professor with Kawasaki disease at onset. Journal Mohammad Gharib, 4–8 May 2002, of pediatrics, 2000, 137:172–6. Thran, Iran [in Persian]. 32. Hashino K et al. Re-treatment for im- 23. Kordivarian R. Kawasaki disease. Re- mune globulin-resistant Kawasaki dis- port of 13 cases from Alzahra Hospital, ease: a comparative study of additional Isfahan. Isfahan medical sciences jour- immune globulin and steroid pulse nal, 1337 (1998), 16(3):77–6 [in Per- therapy. Pediatrics internationalto : offi- sian]. cial journal of the Japan Pediatric Soci- ety, 2001, 43(3):211–7. 24. Soleimani GhR, Siadati A. Etiologic evaluation of Kawasaki disease in pa- 33. Onouchi Z, Kawasaki T. Overview of tients referred to a pediatric medical cen- pharmacological treatment of Kawasaki ter (Markaz Tebbi). Paper presented at disease. Drugs, 1999, 58(5):813–22. the 14th International Congress of Pedi- 34. Raman V et al. Response of refractory atrics, 12–17 October, 2002, Tehran, Kawasaki disease to pulse steroid and Iran. cyclosporin A therapy. Pediatric infec- 25. Research committee on Kawasaki dis- tious disease journal, 2001, 20(6):635– ease. Report of subcommittee on stan- 7. dardization of diagnostic criteria and 35. Dale RC et al. Treatment of severe com- reporting of coronary artery lesion in plicated Kawasaki disease with oral Kawasaki disorder. Tokyo, Ministry of prednisolone and aspirin. Journal of pe- Health and Welfare, 1984. diatrics, 2000, 137:723–6. 26. Belay ED et al. The incidence of 36. Wallace CA et al. Initial intravenous Kawasaki syndrome in West Coast gammaglobulin treatment failure in health maintenance organizations. Pedi- Kawasaki disease. Pediatrics, 2000, atric infectious disease journal, 2000, 105(6):E78. 19(9):826–32.

04 Kawasaki disease.pmd 35 10/25/2005, 1:21 AM 36 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 In vitro activity of quinolones against S. pneumoniae, H. influenzae and M. catarrhalis in Saudi Arabia H.H. Balkhy,1,3 Z.A. Memish,2,3 A. Shibl,4 A. Elbashier 5 and A. Osoba6

ABSTRACT Susceptibility of 88 clinical Streptococcus pneumoniae isolates, 116 Haemophilus influenzae isolates and 80 Moraxella catarrhalis isolates to 6 fluoroquinolones—ciprofloxacin, ofloxacin, levofloxacin, trovafloxacin, grepafloxacin and gemifloxacin—were determined. Isolates were from patients with invasive disease at 4 hospitals in Saudi Arabia between 1996 and 1998. S. pneumoniae isolates were fully susceptible to trovafloxacin, grepafloxacin and gemifloxacin; susceptibility to ofloxacin and levofloxacin was 97.7% and 98.9% respectively. H. influenzae isolates were susceptible to all agents, except for trovafloxacin (99.1%). M. catarrhalis strains were fully sensitive to all agents except ofloxacin (97.5%). No isolates were resistant to gemifloxacin or grepafloxacin. Activité in vitro des quinolones vis-à-vis de S. pneumoniae, H. influenzae et M. catarrhalis en Arabie saoudite RÉSUMÉ La sensibilité de 88 isolats cliniques de Streptococcus pneumoniae, de 116 isolats d’Haemophilus influenzae et de 80 isolats de Moraxella catarrhalis à six fluoroquinolones – ciprofloxacine, ofloxacine, lévo- floxacine, trovafloxacine, grépafloxacine et gémifloxacine – a été déterminée. Les isolats provenaient de patients présentant une maladie invasive dans quatre hôpitaux d’Arabie saoudite entre 1996 et 1998. Les isolats de S. pneumoniae étaient très sensibles à la trovafloxacine, à la grépafloxacine et à la gémifloxacine ; la sensibilité à l’ofloxacine et à la lévofloxacine était de 97,7 % et 98,9 % respectivement. Les isolats de H. influenzae étaient sensibles à tous les agents sauf à la trovafloxacine (99,1 %). Les souches de M. catarrhalis étaient très sensibles à tous les agents sauf à l’ofloxacine (97,5 %). Aucun isolat n’était résistant à la gémiflox- acine ou à la grépafloxacine.

1Department of Paediatrics; 2Department of Internal Medicine; 3Department of Infection Prevention and Control, King Abdulaziz Medical City, King Fahad National Guard Hospital, Riyadh, Saudi Arabia (Correspondence to Z.A. Memish: [email protected]). 4Department of Microbiology, King Saud University, Riyadh, Saudi Arabia. 5Department of Microbiology, Qatif Central Hospital, Qatif, Saudi Arabia. 6Division of Microbiology, King Abdulaziz Medical City, King Khalid National Guard Hospital, Jeddah, Saudi Arabia. Received: 09/03/03; accepted: 02/12/03

05 In vitro activity.pmd 36 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 37

Introduction Methods Respiratory quinolones, as they are some- Between 1996 and 1998, 284 isolates of S. times referred to, have become a class of pneumoniae, H. influenzae and M. catarrh- their own among the quinolones because of alis were collected from patients with inva- their broad spectrum activity against sive diseases including pneumonia, Gram-positive, Gram-negative and atypical bacteraemia and meningitis in the western, pathogens [1]. These newer fluoroquinolo- central and eastern provinces of Saudi Ara- nes like levofloxacin, grepafloxacin, trova- bia. Patients were seen at King Fahad Na- floxacin and gemifloxacin, even though tional Guard Hospital in Riyadh, King Saud they are not as effective as ciprofloxacin University Hospital in Riyadh, King Khalid against Pseudomonas spp., have improved National Guard Hospital in Jeddah or Qatif activity against Gram-positive organisms, Central Hospital in Qatif. No patients had especially Streptococcus pneumoniae, thus taken antimicrobials for the 2 weeks pre- making them more efficacious in the treat- ceding the study. ment of community-acquired pneumonia This was a multicentre survey in which [1–4]. In Saudi Arabia, the only approved non-repeat isolates were collected by each fluoroquinolones are ciprofloxacin, ofloxa- centre. Organisms were included if: cin and norfloxacin. Only very recently did • They were obtained from fresh clinical the Ministry of Health approve moxifloxa- material received in the laboratory. cin for general prescription. • They were obtained from appropriate At present, there are only a few reports samples from patients with clinical indi- globally on resistance of common respira- cations of respiratory tract infections. tory pathogens to the newer quinolones; none of these reports are from Saudi Arabia • They were isolated using routine meth- [5,6]. However, as there is no effective an- ods. tibiotic control by the government in our • They were identified using conventional country, it would not be surprising to find methods (Optochin susceptibility for S. that these drugs may soon be overused. pneumoniae and X and V factor re- Even though the isolation and detection of quirement for H. influenzae. M. ca- fluoroquinolone-resistant Haemophilus in- tarrhalis was identified using oxidase fluenzae strains have been rare, the respira- spot test and butyrate esterase disc). tory quinolones need to be evaluated to • Strains were stored within 48 hours of determine their comparative clinical poten- isolation at –70 °C by inoculation into cies and in vitro efficacy against H. influ- routine storage medium. enzae, Moraxella catarrhalis and S. Organisms were excluded if they were pneumoniae in each community. isolated from throat swabs from healthy The purpose of our study was to deter- carriers or duplicate isolates from the same mine in vitro susceptibility patterns to patient. ciprofloxacin and some of the newer qui- Isolates were inoculated into Mueller– nolones among Saudi Arabian isolates of S. Hinton (MH) broth, colonies were removed pneumoniae, H. influenzae and M. catarrh- from 24-hour chocolate agar and the sus- alis and to ascertain if resistant strains have pension was adjusted to 0.5 McFarland tur- emerged. This will provide supportive data bidity standard. A sterile cotton swab was for the more effective use of all quinolone dipped into the adjusted suspension and antibiotics, especially the new ones.

05 In vitro activity.pmd 37 10/25/2005, 1:21 AM 38 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

used to inoculate the Haemophilus spp. test ism against each antimicrobial agent. Tests medium for H. influenzae strains. S. pneu- of MIC values were controlled with organ- moniae strains were plated on MH agar isms of known susceptibility. For S. pneu- supplemented with 5% sheep cells and M. moniae testing, Escherichia coli ATCC catarrhalis strains were plated on MH agar. 235218, S. pneumoniae ATCC 49619 and Using a template, we applied antibiotic epsi- Staphylococcus aureus ATCC 29213 were lometer (Etest) strips of the antibiotics to used with each batch on MH agar plus 5% the agar surface (AB Biodisk, Solna, Swe- sheep blood. For H. influenzae testing, E. den). After 18–24 hours incubation at coli ATCC 35218 and H. influenzae ATCC

35 °C in 5% CO2 for H. influenzae strains 49247 and ATCC 49766 were used with and in air for S. pneumoniae and M. ca- each batch on haemophilus test medium tarrhalis, the minimum inhibitory concen- (HTM) agar. Results were collated from tration (MIC) values were read. We the combined data from Riyadh, Jeddah interpreted the elliptical zones of inhibition and Qatif. as per manufacturer’s instructions, i.e. reading the value on the Etest strip where the thin end of the ellipse, or teardrop, in- Results tersected the strip. National Committee of S. pneumoniae Clinical and Laboratory Standards (NC- Table 1 shows the MIC and MIC values CLS) criteria for breakpoint values were 50 90 for S. pneumoniae against each agent. Of used to determine the susceptibility of iso- the 6 agents, gemifloxacin exhibited the lates to the various antibiotics [3,7]. Sus- lowest MIC and MIC against S. pneumo- ceptibility rates were determined as: 50 90 niae, closely followed by grepafloxacin and ciprofloxacin ≤ 1.0 µg/mL, ofloxacin ≤ 2.0 trovafloxacin. The MIC of ofloxacin and µg/mL, levofloxacin ≤ 2.0 µg/mL, trova- 90 ciprofloxacin were nearly the same at 2.0 floxacin ≤ 1.0 µg mL and grepafloxacin ≤ mg/L and exceeded the breakpoint recom- 0.5 µg/mL. For gemifloxacin the break- mended by the NCCLS. The most active point was ≤ 0.25 µg/mL, which corre- agent was gemifloxacin with an MIC of sponded to a zone diameter of 20–39 mm 90 0.06 mg/L. [3]. The MIC50 and MIC90 were based on the cumulative MIC values of each organ-

Table 1 Minimum inhibitory concentration (MIC) values for 88 isolates of Streptococcus pneumoniae

Antimicrobial MIC (mg/L) Susceptibility No. of

MIC50 MIC90 MIC range (%) isolates Gemifloxacin 0.03 0.06 0.008–0.25 100 88 Ciprofloxacin 1 2 0.03–4.0 98.9 87 Grepafloxacin 0.06 0.12 0.047–0.75 100 88 Levofloxacin 1 1 0.12–4.0 98.9 87 Ofloxacin 2 2 0.12–4.0 97.7 86 Trovafloxacin 0.06 0.12 0.015–0.5 100 88

05 In vitro activity.pmd 38 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 39

Table 1 also shows the percentage of S. 6 agents, gemifloxacin and trovafloxacin

pneumoniae isolates susceptible to each had the lowest MIC90 values at 0.03 mg/L antimicrobial. Among the strains of S. each, closely followed by ciprofloxacin, pneumoniae, the prevalence of fluoroqui- grepafloxacin and levofloxacin. Of the 80 nolone resistance is low, but has been M. catarrhalis strains tested, only 2 were emerging to less potent agents such as resistant to ofloxacin. Therefore, of all ciprofloxacin, levofloxacin and ofloxacin. agents tested, gemifloxacin had the lowest We identified 1 strain resistant to ciproflox- MIC values for all 3 pathogens and had the acin, 1 strain resistant to levofloxacin and 2 lowest overall MIC values against S. pneu- strains resistant to ofloxacin. moniae.

H. influenzae

Table 2 shows MIC50 and MIC90 values for Discussion test antimicrobials against H. influenzae. Each quinolone was active against H. influ- The quinolones have become an important enzae. All 116 strains of H. influenzae dem- group of antibiotics over the past 2 decades onstrated 100% susceptibility to all tested since they were introduced into clinical use agents except trovafloxacin, which had a for the treatment of serious infections susceptibility rate of 99.1%. Gemifloxacin, [8,9]. The older antimicrobials belonging to ciprofloxacin, grepafloxacin and levofloxa- this group such as ciprofloxacin, ofloxacin and norfloxacin have potent in vitro activity cin each had MIC90 of 0.03 mg/L against H. influenzae. Of these, gemifloxacin had the against many Gram-negative organisms. However, their efficacy against some im- lowest MIC50 with a value of 0.008 mg/L. Trovafloxacin was the most active agent portant Gram-positive organisms and atypical agents was below clinical expectations with MIC50 and MIC90 of 0.015 mg/L. Only 1 H. influenzae strain was resistant to tro- [10]. The new fluoroquinolones appear to vafloxacin. have several advantages over their older prototypes, including a broader spectrum M. catarrhalis of activity as well as improved bio- availability and safety. In addition, they Table 3 shows the MIC50 and MIC90 values of the quinolones for M. catarrhalis. Of all may offer shorter duration of therapy lead-

Table 2 Mean inhibitory concentration (MIC) values for 116 isolates of Haemophilus influenzae

Antimicrobial MIC (mg/L) Susceptibility No. of

MIC50 MIC90 MIC range (%) isolates Gemifloxacin 0.008 0.03 0.001–0.25 100 116 Ciprofloxacin 0.015 0.03 0.015–1.0 100 116 Grepafloxacin 0.015 0.03 <0.015–0.25 100 116 Levofloxacin 0.03 0.03 <0.015–1.0 100 116 Ofloxacin 0.06 0.06 <0.06–2.0 100 116 Trovafloxacin 0.015 0.015 <0.015–4.0 99.1 115

05 In vitro activity.pmd 39 10/25/2005, 1:21 AM 40 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 3 Mean inhibitory concentration (MIC) values for 80 isolates of Moraxella catarrhalis

Antimicrobial MIC (mg/L) Susceptibility No. of

MIC50 MIC90 MIC Range (%) isolates Gemifloxacin 0.015 0.03 0.004–0.12 100 80 Ciprofloxacin 0.03 0.06 <0.015–0.12 100 80 Grepafloxacin 0.015 0.06 0.015–0.25 100 80 Levofloxacin 0.06 0.06 0.03–0.25 100 80 Ofloxacin 0.012 0.25 <0.06–4.0 97.5 78 Trovafloxacin 0.015 0.03 <0.015–0.06 100 80

ing to improved compliance and cost sav- pneumococcal isolates with reduced sus- ings [1,11]. In 1998, the Infectious Disease ceptibility to fluoroquinolones increased Society of America guidelines included the from 0.5% in 1993–1994 to 2.0% in 1997– fluoroquinolones as an option in empirical 1998. In Canada, more resistant isolates management of community-acquired pneu- were more common among people aged 65 monia due to their activity against interme- years and older and among people from diately and highly resistant strains of S. Ontario. This was associated with a 5-fold pneumoniae [8,12]. Concerns regarding increase in the number of fluoroquinolones the liberal use of these agents and the rapid prescribed during 1993–1998 [22]. More emergence of resistance, especially among reports have indicated the increasing resis- S. pneumoniae, however, have been raised tance of S. pneumoniae isolates to the qui- [13,14]. nolones, including to the newer agents Many reports and surveillances have such as levofloxacin [23,24]. documented the global emergence of resis- The resistance of H. influenzae to tance to penicillins, cephalosporins and ampicillin appears to be steadily rising in macrolides among common agents causing Saudi Arabia from 4% in 1982 to 17% in community-acquired pneumonia [15–17]. 1994 [25–28]. The prevalence rate of S. The Alexander Project, which was begun in pneumoniae resistance to penicillin and 1992 and included isolates from many ampicillin has varied considerably by city, countries including Saudi Arabia, found but in general it has dropped from 100% that the resistance of H. influenzae and S. sensitivity in the early 1980s to below 40% pneumoniae isolates to ofloxacin and in the last few years [29–35]. Sensitivity ciprofloxacin had not exceeded 1% and testing to penicillin has shown, however, identified no M. catarrhalis quinolone- that most resistant strains are intermediate- resistant strains [18]. Unfortunately more ly resistant, while no more than 10% is recent data are not as encouraging. The highly resistant S. pneumoniae [18,29,33]. prevalence of S. pneumoniae with de- There is strong evidence that a number of creased susceptibility to fluoroquinolones strains resistant to agents commonly used is increasing [19–21]. When over 7000 S. to treat community-acquired pneumonia pneumoniae isolates were examined be- have emerged in our country and continue tween 1988 and 1998, the percentage of to become more common [33].

05 In vitro activity.pmd 40 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 41

Since new quinolones provide a valid al- The literature on M. catarrhalis and its ternative antibacterial therapy, especially in susceptibility to antibiotics in Saudi Arabia areas where the prevalence of penicillin- is scant. In Riyadh, 27 of 32 strains (84%) resistant and macrolide-resistant organisms isolated from clinical specimens produced dominate, and since they have become li- beta-lactamases and were resistant to censed for the first time in the Kingdom of erythromycin and clindamycin phosphate. Saudi Arabia, we examined the local sus- All isolates, however, were susceptible to ceptibility pattern of the most common ciprofloxacin, tetracycline and trimethop- agents of community-acquired pneumonia rim–sulfamethoxazole [39]. In our study, to some of them. Among the 284 clinical the MIC90 was well below the breakpoint of isolates of H. influenzae, S. pneumoniae the various quinolones tested (Table 3). and M. catarrhalis, only 7 isolates were re- Although clinical studies are needed to sistant (4 S. pneumoniae, 1 H. influenzae confirm our in vitro results, resistance and 2 M. catarrhalis). This indicated excel- rates to each newer quinolone were too low lent sensitivity among our local isolates and to be of clinical concern. In a study from could be easily explained by the rare use of Canada, a high sensitivity rate to levofloxa- the newer quinolones since only ciproflox- cin was documented, while other quinolo- acin is registered on the formulary of most nes were not included [40]. Similarly, in the hospitals in the country. Alexander project, low rates of resistance One S. pneumoniae strain was sensitive to ciprofloxacin and cefuroxime were doc- to levofloxacin but resistant to ciprofloxa- umented from 266 isolates collected be- cin. Perhaps an active efflux mechanism in tween 1996 and 1997 [18]. this particular strain resulted in lower resis- Our findings suggest that gemifloxacin, tance. In some strains of S. pneumoniae, grepafloxacin, trovafloxacin and levofloxa- an efflux protein, PmrA, mediates this cin would be effective agents in respiratory mechanism [36,37]. In 2000, after the tract infections caused by the common completion of our study, ciprofloxacin was pathogens in our country. Nevertheless, in widely used for the first time during the pil- order to maintain the efficacy of this group grimage event as a decolonizing agent for of antimicrobials, we should enforce their returning pilgrims to protect their family judicious use and implement continued sur- against meningitis [38]. For this reason, it veillance of susceptibility of the different is expected that resistance to ciprofloxacin organisms to this class of fluoroquinolo- will gradually increase if the practice is nes. continued or abused. References

1. Blondeau JM. A review of the compara- Report from the SENTRY Antimicrobial tive in vitro activities of 12 antimicrobial Surveillance Program (2000) in Europe, agents, with a focus on five new respira- Canada and the United States. Diagnos- tory quinolones. Journal of antimicrobi- tic microbiology and infectious disease, als and chemotherapy, 1999, 43 (suppl. 2001, 39(4):245–50. B):1–11. 3. Biedenbach DJ, Jones RN. Evaluation of 2. Biedenbach DJ, Jones RN, Pfaller MA. in vitro susceptibility testing criteria for Activity of BMS 284756 against 2681 re- gemifloxacin when tested against cent clinical isolates of Haemophilus Haemophilus influenzae strains with re- influenzae and Moraxella catarrhalis: duced susceptibility to ciprofloxacin and

05 In vitro activity.pmd 41 10/25/2005, 1:21 AM 42 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

ofloxacin. Diagnostic microbiology and 12. Marre R, Trautmann M. Ambulant infectious disease, 2002, 43(4):323–6. erworbene Atemwegsinfektionen. Ak- tuelle Daten zur Wirksamkeit und 4. Rittenhouse S et al. In vitro antibacterial Resistenzsituation verschiedener Anti- activity of gemifloxacin and comparator biotikagruppen gegenuber den wichti- compounds against common respiratory gsten bakteriellen Spezies. [Community pathogens. Journal of antimicrobials acquired respiratory tract infections. Cur- and chemotherapy, 2000, 45(suppl. 1): rent data on the efficacy of various 23–7. classes of antibiotics and antibiotic re- 5. Gould IM, Forbes KJ, Gordon GS. sistance of the main prevalent bacteria Quinolone resistant Haemophilus species.] Medizinische Klinik, 1999, influenzae. Journal of antimicrobials and 94(11):609–13. chemotherapy, 1994, 33(1):187–8. 13. Heffelfinger JD et al. Management of 6. Biedenbach DJ, Jones RN. Fluoro- community-acquired pneumonia in the quinolone-resistant Haemophilus in- era of pneumococcal resistance: A report fluenzae: frequency of occurrence and from the Drug-Resistant Streptococcus analysis of confirmed strains in the SEN- pneumoniae Therapeutic Working TRY antimicrobial surveillance program Group. Archives of internal medicine, (North and Latin America). Diagnostic 2000, 160(10):1399–408. microbiology and infectious disease, 14. Hooper DC. Flouroquinolone resistance 2000, 36(4):255–9. among gram-positive cocci. Lancet in- 7. National Committee for Clinical Labora- fectious diseases, 2002, 2(9):530–8. tory Standards. Performance Standards 15. Doern GV et al. Haemophilus influenzae for Antimicrobial Susceptibility Testing, and Moraxella catarrhalis from patients Twelfth Informational Supplement, with community-acquired respiratory M100–S12, vol. 22, no. 1. Wayne, Penn- tract infections: antimicrobial susceptibil- sylvania, National Committee for Clini- ity patterns from the SENTRY antimicro- cal Laboratory Standards, 2002. bial surveillance program (United States 8. Bartlett JG et al. Community-acquired of America and Canada, 1997). Antimi- pneumonia in adults: guidelines for man- crobial agents and chemotherapy, 1999, agement. The Infectious Disease Society 43(2):385–9. of America. Clinical infectious diseases, 16. Jacobs MR et al. Susceptibilities of 1998, 26(4):811–38. Streptococcus pneumoniae and Haemo- 9. Bartlett JG et al. Practice guidelines for philus influenzae to 10 oral antimicrobial the management of community-acquired agents based on pharmacodynamic pa- pneumonia in adults. Infectious Disease rameters: 1997 U.S. Surveillance study. Society of America. Clinical infectious Antimicrobial agents and chemotherapy, diseases, 2000, 31(2):347–82 1999, 43(8):1901–8. 10. Hooper DC. New uses for new and old 17. Thornsberry C et al. Resistance surveil- quinolones and the challenge of resis- lance of Streptococcus pneumoniae, tance. Clinical infectious diseases, Haemophilus influenzae and Moraxella 2000, 30(2):243–54. catarrhalis isolated in the United States, 1997–1998. Journal of antimicrobial 11. Niederman MS et al. The cost of treating chemotherapy, 1999, 44(6):749–59. community acquired pneumonia. Clini- cal therapeutics, 1998, 20(4):820–37.

05 In vitro activity.pmd 42 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 43

18. Felmingham D, Gruenberg RN. The zae strains to oral antimicrobial agents Alexander Project 1996–1997: latest used in Saudi Arabia. Chemotherapy, susceptibility data from this international 1994, 40(6):399–403. study of bacterial pathogens from com- 27. Qadri SMH, Lee GC, Ellis ME. B- munity-acquired lower respiratory tract lactamase production in recent clinical infections. Journal of antimicrobial che- isolates of Haemophilus influenzae and motherapy, 2000, 45:191–203. their susceptibility to cefaclor and other 19. Memish ZA, Shibl AM, Ahmed QA. antimicrobial agents. Saudi medical Guidelines for the management of com- journal, 1993, 14:59–61. munity-acquired pneumonia in Saudi 28. Kambal AM, Abdul Khaliq MR, Arabia: a model for the Middle East re- Chowdhury NH. Susceptibility of Hae- gion. International journal of antimicro- mophilus influenzae to selected antimi- bial agents, 2002, 20(suppl. 1):1–12. crobial agents. Annals of Saudi 20. Vila J et al. Increase in quinolone resis- medicine, 1996, 16(5):582–6. tance in a Haemophilus influenzae 29. Memish ZA et al. Streptococcus strain isolated from a patient with recur- pneumoniae in Saudi Arabia: Antibiotic rent respiratory infections treated with resistance and serotypes among recent ofloxacin. Antimicrobial agents and che- clinical isolates. International journal of motherapy, 1999, 43(1):161–2. antimicrobial agents, 2004, 23(1):32–8. 21. Jordens JZ, Slack MPE. Haemophilus 30. El-Mouzan MI et al. Pneumococcal infec- influenzae: then and now. European tions in eastern Saudi Arabia: serotypes journal of clinical microbiology and in- and antibiotic sensitivity patterns. Tropi- fectious diseases, 1995, 14(11):935–48. cal and geographical medicine, 1988, 22. Chen DK et al. Decreased susceptibility 40(3):213–317. of Streptococcus pneumoniae to flouro- 31. Qadri SM, Kroschinsky R. Prevalence of quinolones in Canada. Canadian Bacte- pneumococci with increased resistance rial Surveillance Network. New England to penicillin in Saudi Arabia. Annals of journal of medicine, 1999, 341(4):233– tropical medicine and parasitology, 9. 1991, 85(2):259–62. 23. Low DE et al. Antimicrobial resistance 32. Rotimi VO, Feteih J, Barbor PRH. Preva- among clinical isolates of Streptococcus lence of penicillin-resistant Streptococ- pneumoniae in Canada during 2000. cus pneumoniae in a Saudi Arabian Antimicrobial agents and chemotherapy, hospital. European journal of clinical mi- 2002, 46(5):1295–301. crobiology and infectious diseases, 24. Davidson R et al. Resistance to 1995, 14(2):149–51. levofloxacin and failure of treatment of 33. Shibl AM et al. Penicilin-resistant and in- pneumococcal pneumonia. New En- termediate Streptococcus pneumoniae gland journal of medicine, 2002, in Saudi Arabia. Journal of chemo- 346(10):747–50. therapy, 2000, 12(2):134–7. 25. Chowdhury MNH, Mahgoub ES. Ampicil- 34. El-Tahawy AT. Antimicrobial resistance lin-resistant Haemophilus influenzae in of Streptococcus pneumoniae at a Uni- Riyadh, Saudi Arabia. Saudi medical versity Hospital in Saudi Arabia. Journal journal, 1982, 3:100–5. of chemotherapy, 2001, 13(2):148–53. 26. Shibl AM, Gaillot O. Susceptibility of clinically significant Haemophilus influen-

05 In vitro activity.pmd 43 10/25/2005, 1:21 AM 44 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

35. Al-Aqeeli AA, Guy ML, Al-Jumaah SA. 38. Memish ZA. Meningococcal disease and Streptoccous pneumoniae resistance to travel. Clinical infectious diseases, penicillin and ceftriaxone in a tertiary 2002, 34:84–90. care centre in Saudi Arabia. Saudi medi- 39. Babay HA. Isolation of Moraxella cal journal, 2002, 23(4):400–4. catarrhalis in patients at King Khalid Uni- 36. Appelbaum PC. Resistance among versity Hospital, Riyadh. Saudi medical Streptococcus pneumoniae: Implica- journal, 2000, 21(9):860–3. tions for drug selection. Clinical infec- 40. Zhanel GG et al. Animicrobial resistance tious diseases, 2002, 34:1613–20. in Haemophilus influenzae and Mora- 37. Gill MJ, Brenwald NP, Wise R. Identifica- xella catarrhalis respiratory tract iso- tion of an efflux pump gene, pmrA, asso- lates: Results of the Candadian ciated with fluoroquinolone resistance in Respiratory Organisms Susceptibility Streptococcus pneumoniae. Antimicro- Study, 1997 to 2002. Antimicrobial bial agents and chemotherapy, 1999, agents and chemotherapy, 2003. 47(6): 43:187–9. 1875–81.

WHO Expert Committee on Specifications for Pharmaceutical Preparations, thirty-eighth report This report presents the recommendations of an international group of experts convened by the World Health Organization to consider matters concerning the quality assurance of pharmaceuticals and specifications for drug substances and dosage forms. Of particular relevance to drug regulatory authorities and pharmaceutical manufacturers, this report discusses the latest volume of the Interna- tional pharmacopoeia and quality specifications for pharmaceutical substances and dosage forms, as well as quality control of reference materials, good manufacturing practices, inspection, distribution and trade and other aspects of quality assurance of pharmaceuticals, and regulatory issues. The report is comple- mented by a number of annexes, including recommendations on good trade and distribution practices for pharmaceutical starting materials, guidelines on the WHO scheme for the certification of pharmaceutical materials moving in international commerce, draft procedures for assessing quality control laboratories and procurement agencies for use by the United Nations agencies, and guidelines for preparing a laboratory information file and a procurement agency information file. Further information on this publication can be obtained from WHO Press: http://www.who.int/bookorders/ anglais/home1.jsp?sesslan=1

05 In vitro activity.pmd 44 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 45 Seroprevalence and risk factors for Toxoplasma infection in pregnant women in Jordan N.F. Jumaian1

ABSTRACT To determine the seroprevalence and risk factors for toxoplasmosis among pregnant women in Jordan, sera from 280 pregnant women were tested during the period January 2000–May 2001. Blood samples were taken after the first antenatal visit. Serum was separated and tested for Toxoplasma IgG antibodies using an indirect fluorescent antibody. Seroprevalence gradually increased with age, from 31.7% at 15–24 years to 90.0% at 35–45 years. Regression analysis showed that seroprevalence of toxoplasmosis is positively correlated with age and residence. Consumption of undercooked meat and contact with soil were significant risk factors.

Séroprévalence et facteurs de risque de l’infection à toxoplasme chez les femmes enceintes en Jordanie RÉSUMÉ Afin de déterminer la séroprévalence et les facteurs de risque de toxoplasmose chez les femmes enceintes en Jordanie, on a testé le sérum de 280 femmes enceintes pendant la période de janvier 2000 à mai 2001. Des échantillons sanguins ont été prélevés après la première consultation prénatale pour une recherche d’anticorps IgG anti-toxoplasmes par immunofluorescence indirecte, après séparation du sérum. La séroprévalence augmentait progressivement avec l’âge, passant de 31,7 % dans la tranche d’âge de 15- 24 ans à 90,0 % dans la tranche d’âge de 35-45 ans. L’analyse de régression a montré que la séropréva- lence de la toxoplasmose était corrélée positivement avec l’âge et le lieu de résidence. La consommation de viande insuffisamment cuite et le contact avec le sol étaient des facteurs de risque significatifs.

1Princess Iman Research and Laboratory Sciences Centre, King Hussein Medical Centre, Amman, Jordan (Correspondence to N.F. Jumaian: [email protected]). Received: 12/06/02; accepted: 20/10/03

06 Seroprevalence and risk.pmd 45 10/25/2005, 1:21 AM 46 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction Methods Toxoplasmosis is a disease caused by the On the basis of previously reported preva- protozoan parasite Toxoplasma gondii. lence of toxoplasmosis in the north of Jor- This disease is clinically insignificant in im- dan, we studied 280 randomly selected munocompetent adults. The immunologic pregnant women of a total of around response to primary infection is followed 25000 women who attended the obstetric by encystment of the parasite (latent toxo- outpatient clinic at Prince Rashid Military plasmosis), providing life-long immunity. Hospital between January 2000 and May Possible reactivation of latent infection in 2001. Patients at this hospital are mainly an increasingly immunosuppressed popula- dependents of military personnel. Sample tion, however, makes toxoplasmosis an im- size was calculated on the basis of previ- portant opportunistic infection [1]. In ously reported prevalence of toxoplasmosis addition, toxoplasmosis has long been in this area [11]. Expected prevalence was known as a major cause of perinatal mor- 26%, range 21%–31%. Sample size calcu- bidity [2]. Newly acquired T. gondii infec- lated at 95% confidence was 210. All the tion in pregnant women can be transmitted women in this study were living in Irbid to the fetus and cause mental retardation, and surrounding villages. Blood samples blindness, epilepsy and death. Toxoplasmo- were taken after the first antenatal clinic sis is, therefore, a serious clinical and pub- visit. Serum was separated by blood cen- lic health problem. The prevalence of trifugation at 3000 rpm for 5 minutes. antibodies to this disease shows consider- Samples were stored at –20 °C until analy- able variation from country to country and sis in the Department of Biostatistics for between different population groups within IgG anti-Toxoplasma antibodies. a country [3,4]. Data were collected from all pregnant High T. gondii seroprevalence among women by interview using a specially de- pregnant women has been reported as signed, pretested questionnaire. Interviews reaching 71% in France, 37% in Slovenia, were conducted in the clinic by staff nurs- 42% in Italy and 43.4% in Saudi Arabia [5– es or a medical technologist. Women were 8]. The risk factors for acquisition of T. first asked about age, place of residence gondii infection have not been previously (urban or rural) and occupation, then they reported for pregnant women in Jordan. were asked about contact with cats, con- Recent epidemiological studies have identi- sumption of raw or undercooked meat, fied the following risk factors: owning contact with soil and the probable mode of cats, eating raw or unwashed fruits and transmission of Toxoplasma to assess their vegetables [8], eating raw or undercooked knowledge of the source of infection. lamb, beef and minced meat products [9], The screening for Toxoplasma-specific animal farming [10] and having contact IgG antibodies was performed by indirect with soil [9]. fluorescent antibody test using a commer- This study was conducted to identify cial kit (Gull Laboratories Inc., Salt Lake the risk factors for Toxoplasma infection City, Utah, USA). during pregnancy with a view to improving After collection, data were coded and primary prevention among non-immune entered on the computer using SPSS, ver- pregnant women since the results may be sion 9.0. Then frequency tables, cross- useful to others working in this field. tabulation, chi-squared and the association

06 Seroprevalence and risk.pmd 46 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 47

between different variables were calculated Sociodemographic characteristics by logistic regression. Analysis of Toxoplasma seropositivity by The dependent variable was the stable age (Table 1) showed that the risk of Toxo- infection: infected (seropositive)/non- plasma infection among pregnant women infected (seronegative); all other variables was significantly greater among the oldest were considered independent. age group (P < 0.001). Correlation coefficient (r) and slope found by logistic regression were positive, indicating that increase Results in age corresponds with increase in serop- Prevalence of Toxoplasma ositivity. Seropositivity was 31.7% for seropositivity those aged 15–24 years, 48.3% for those Basic sociodemographic and epidemiologi- aged 25–34 years, and 90.0% for those cal data of the women who participated in aged 35–45 years. The study showed a di- this study are presented in Tables 1 and 2. rect correlation between age and positivity Overall prevalence of T. gondii infection for IgG antibodies (r = 0.286, slope = was 47.1% (133/280) on the basis of de- 1.147). The risk of Toxoplasma infection in tection of IgG antibodies, which are indica- the oldest age group was about 3-fold tive of previous infection. Prevalence of greater than in the youngest age group seropositivity among pregnant women was [odds ratio (OR) = 3.150]. The risk of tox- calculated on the basis of the first trimester oplasmosis was significantly greater in findings: number of seropositive pregnan- pregnant women from rural areas com- cies related to total number of pregnancies. pared with those from urban areas (P =

Table 1 Demographic characteristics for Toxoplasma gondii seropositive pregnant women

Characteristic Toxoplasma Toxoplasma P-value seropositive seronegative (n = 132) (n = 148) No. % No. %

Age group (years) < 0.0001 15–24 (n = 120) 38 31.7 82 68.3 25–34 (n = 120) 58 48.3 62 51.7 35–45 (n = 40) 36 90.0 4 10.0 Residence 0.002 Urban (n = 177) 71 40.1 106 59.9 Rural (n = 103) 61 59.2 42 40.8 Working outside the home 0.158 Yes (n = 87) 47 54.0 40 46.5 No (n = 193) 85 44.0 108 55.7

06 Seroprevalence and risk.pmd 47 10/25/2005, 1:21 AM 48 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 2 Transmission risk factors for Toxoplasma gondii seropositive pregnant women

Variable Toxoplasma Toxoplasma P-value seropositive seronegative (n = 132) (n = 148) No. % No. %

Consumption of undercooked meat < 0.0001 Yes (n = 64) 48 75.0 16 25.0 No (n = 216) 84 38.9 132 61.1 Contact with soil 0.022 Yes (n = 26) 18 69.2 8 30.8 No (n = 254) 114 44.9 140 55.1 Keep cats 0.742 Yes (n = 25) 11 44.0 14 56.0 No (n = 255) 121 47.5 134 52.5 Knowledge of transmission 0.012 Yes 2 12.5 14 87.5 No 130 49.2 134 50.8

0.002). The risk was about 2-fold greater ropositive, and those who did not (r = (r = 0.318, slope = 0.774, OR = 2.168). –0.092, slope = 1.016, OR = 2.762, P = There was no significant difference in 0.022). prevalence of Toxoplasma infection between housewives and women who were Knowledge about transmission of working outside the home (P = 0.158). toxoplasmosis Pregnant women who had knowledge Sanitary risk factor about the mode of disease transmission had The prevalence of Toxoplasma infection significantly lower prevalence of Toxoplas- was significantly higher among pregnant ma infection (12.5%) than with those hav- women who ate undercooked lamb and ing no knowledge (49.2%) (P = 0.012, r = goat meat, 75.0%, compared with those 0.015, slope =1.914, OR = 6.782). who did not, 38.9% (r = 0.235, slope = 1.551, OR = 4.714, P < 0.001) (Table 2). There was no significant difference in Discussion prevalence of Toxoplasma infection between pregnant women who kept cats in The present study demonstrates that preva- their home, 44.0% (11/25), and those who lence of anti-Toxoplasma IgG antibodies in did not, 47.5% (121/255) (P = 0.742). pregnant women in the north of Jordan is There was, however, a significant differ- 47.1%. This is in agreement with results ence between women who worked on the reported in studies done in some other land and had contact with soil, 69.2% se- countries [12–16].

06 Seroprevalence and risk.pmd 48 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 49

The high prevalence of anti-Toxoplas- density of cysts than skeletal muscle ma IgG antibodies in pregnant women ap- [24,25]. pears to be a reflection of chronic infection The significance of contact with soil for that might have originated through various acquisition of infection has been demon- risk factors, which will be discussed. strated by studies showing a higher preva- The risk of toxoplasmosis among preg- lence of Toxoplasma infection in pregnant nant women was significantly greater women who worked on the land or had among women in the older age group. This contact with soil (P = 0.022). Contact with finding is in agreement with that reported soil was identified as a risk factor for Toxo- by Bobic et al., which indicated that Toxo- plasma infection in pregnancy in this study plasma infection increased with age from and in 2 of 3 previous studies that adjusted 57% at 18 years to 93% at 50 years, with for confounders [8,20,26]. an overall mean of 77% (P = 0.022) [14]. Although in most previous studies cat Similar findings were reported by Wong et ownership has been associated with either al., who found that the seropositivity of increased risk for toxoplasmosis seroposi- Toxoplasma IgG increased with the age of tivity or no change in risk [27,28], in this pregnant women [15]. The relatively grad- study, there was no difference in preva- ual increase in seroprevalence associated lence of infection between those women with age suggest that soil exposure, which who kept cats in their home and those who is greatest during childhood years, may not did not (P = 0.742). In another study, it be the principal mechanism by which peo- was found that possession of cats de- ple in Jordan are exposed to T. gondii. creased the risk for seropositivity [29]. The prevalence of Toxoplasma infection The study showed there was a signifi- was significantly greater among pregnant cant difference between prevalence of in- women who ate undercooked meat (P fection in pregnant women who had < 0.0001). The association between eating knowledge about the mode of transmission raw or undercooked meat and Toxoplasma of toxoplasmosis and those who did not (P infection found in this study has also been a = 0.012). Comparing our results with simi- consistent finding in previous studies lar results obtained by Cook et al. [9], low- [18,19]. The type of meat, however, has er rates of exposure were observed among varied. In a Norwegian study, undercooked women who identified raw meat as a risk lamb but not beef was identified as a risk factor for toxoplasmosis. Therefore, health factor [20], whereas in northern France promotion strategies should be based on an beef and lamb were both risk factors [8]. understanding of the factors affecting Evidence from studies that used bioassay women’s behaviour [30]. Information giv- suggested that lamb and goat meat were en by clinicians to groups or via the media more commonly infected than beef may be more effective than written material [21,22]. The risk of meat being infected [31–33]. depends on the age of the animal, the pro- In summary, this study indicated that portion of time the animal has spent in- the seroprevalence of toxoplasmosis is doors, farm hygiene [21,23] and the positively correlated with age. Consump- specific tissues used: non-skeletal muscle tion of undercooked meat and contact with (heart, diaphragm and tongue) had a higher soil are significant risk factors.

06 Seroprevalence and risk.pmd 49 10/25/2005, 1:21 AM 50 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

References

1. Luft BJ, Remington JS. Toxoplasmic en- multi-centre case–control study. British cephalitis in AIDS. Clinical infectious dis- medical journal, 2000, 321(7254):142– eases, 1992, 15(2):211–22. 7. 2. Remington JS, Desmonts G. Toxoplas- 10. Weigel RM, Dubey JP, Dyer D. Risk fac- mosis. In: Remington JS, Klein JO, eds. tors for infection with Toxoplasma gondii Infectious diseases of the fetus and the for residents and workers on swine newborn infant, 3rd ed. Philadelphia, WB farms in Illinois. American journal of Saunders Company, 1990:89–195. tropical medicine and hygiene, 1999, 60(5):305–9. 3. Courvreur J, Desmonts G. Acquired and congenital toxoplasmosis. In: Harris AA, 11. Abdel-Hafez SK et al. Serodiagnosis of ed. Handbook of clinical neurology, vol. Toxoplasma gondii in habitually abort- 8(52). Microbial disease. Amsterdam, ing women and other adults from North Elsevier, 1988:351–63. Jordan. Folia parasitologica, 1986, 33(1):7–13. 4. Jackson MH, Hutchison WM, Siim JC. A seroepidemiological survey of toxoplas- 12. Franklin DM, Dror Z, Nishri Z. The preva- mosis in Scotland and England. Annals lence and incidence of Toxoplasma anti- of tropical medicine and parasitology, bodies in pregnant women. Israel journal 1987, 81(4):359–65. of medical sciences, 1993, 29(5):282–6. 5. Logar J, Novak-Antolic Z, Zore A. Sero- 13. El Hady HM. Toxoplasmosis among logical screening for toxoplasmosis in pregnant women in Abha, Saudi Arabia. pregnancy in Slovenia. Scandinavian Journal of the Egyptian Society of Para- journal of infectious diseases, 1995, sitology, 1991, 21(3):811–5. 27(2):163–4. 14. Ahmed MM. Seroepidemiology of Toxo- 6. Valcalvi PP et al. Prevalence of anti- plasma infection in Riyadh, Saudi Toxoplasma gondii antibodies in the Arabia. Journal of the Egyptian Society population of the area of Parma (Italy). of Parasitology, 1992, 22(2):407–13. European journal of epidemiology, 15. Ashrafunnessa et al. Seroprevalence of 1995, 11(3):333–7. Toxoplasma antibodies among the ante- 7. Abbas SMA, Basalamah AH, Sereborn natal population in Bangladesh. Journal FE. The prevalence of Toxoplasma of obstetrics and gynaecology research, gondii antibodies in Saudi Arabian 1998, 24(2):115–9. women and outcome of congenital infec- 16. Bobic B et al. Epidemiological indication among newborns in Saudi Arabia. tions for toxoplasmosis serology in preg- Saudi medical journal, 1986, 7:346–54. nancy. Jugoslavenska ginekologija i 8. Baril L et al. Facteurs de risque perinatologija, 1998, 34:17–21 [in d’acquisition de la toxoplasmose chez Serbo-Croat]. les femmes enceintes en 1995 (France) 17. Wong SY, Remington JS. Toxoplasmosis [Risk factors for acquiring toxoplasmosis in pregnancy. Clinical infectious dis- in pregnant women in 1995 (France)]. eases, 1994, 18(6):853–62. Bulletin epidémiologique hebdoma- daire, 1996, 16:73–5. 18. Smith JL. Foodborne toxoplasmosis. Journal of food safety, 1991, 12:17–57. 9. Cook A et al. Sources of Toxoplasma infection in pregnant women: European

06 Seroprevalence and risk.pmd 50 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 51

19. Stead L, Lancaster T. Group behaviour 27. Frenkel JK, Ruiz A. Endemicity of toxo- therapy programmes for smoking cessa- plasmosis in Costa Rica: transmission tion. Cochrane database of systematic between cats, soil, intermediate hosts reviews, 2002, (3). and humans. American journal of epidemiology, 1981, 113(3):254–69. 20. Kapperud G et al. Risk factors for Toxo- plasma gondii infection in pregnancy. 28. Seuri M, Koskela P. Contacts with pigs Results of a prospective case–control and cats associated with high preva- study in Norway. American journal of epi- lence of Toxoplasma antibodies among demiology, 1996, 144(4):405–12. farmers. British journal of industrial medicine, 1992, 49(12):845–9. 21. Dubey JP. Strategies to reduce transmission of Toxoplasma gondii to animals 29. Frenkel JK et al. Transmission of Toxo- and humans. Veterinary parasitology, plasma gondii in Panama City, Panama: 1996, 64(1–2):65–70. a five-year prospective cohort study of children, cats, rodents, birds and soil. 22. Smith JL. Foodborne toxoplasmosis. American journal of tropical medicine Journal of food safety, 1991, 12:17–57. and hygiene, 1995, 53(5):1458–68. 23. Faull WB, Clarkson MJ, Winter AC. Toxo- 30. Naidoo J, Wills J. Health promotion: plasmosis in a flock of sheep: some in- foundations for practice. London, vestigations into its source and control. Bailliere Tindall, 1994. Veterinary record, 1986, 119(20):491–3. 31. Silagy C, Ketteridge S. Physician advice 24. Dubey JP, Thulliez P. Persistence of tis- for smoking cessation. In: Cochrane col- sue cysts in edible tissues of cattle fed laboration, ed. Cochrane library, issue 2. Toxoplasma gondii oocysts. American Oxford, Update Software, 2000. journal of veterinary research, 1993, 54(2):271–3. 32. Stead L, Lancaster T. Group behaviour therapy programmes for smoking cessa- 25. Dubey JP. Long-term persistence of tion. In: Cochrane collaboration, ed. Toxoplasma gondii in tissues of pigs in- Cochrane library, issue 2. Oxford, Up- oculated with T. gondii oocysts and effect date Software, 2000. of freezing on viability of tissue cysts in pork. American journal of veterinary re- 33. Grilli R et al. Impact of mass media on search, 1988, 49(6):910–3. health services utilization. In: Cochrane collaboration, ed. Cochrane library, is- 26. Buffolano W et al. Risk factors for recent sue 2. Oxford, Update Software, 2000. Toxoplasma infection in pregnant women in Naples. Epidemiology and infection, 1996, 116(3):347–51.

06 Seroprevalence and risk.pmd 51 10/25/2005, 1:21 AM 52 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Device-related nosocomial infection in intensive care units of Alexandria University Students Hospital S.A. Sallam,1 M.A. Arafa,1 A.A. Razek,2 M. Naga3 and M.A. Hamid 4

ABSTRACT We studied the magnitude and determinants of device-related nosocomial infections in intensive care units (ICU) and identified the predominant microorganisms involved. Sputum, urine, blood cultures and chest X-ray were taken from each patient on admission and on appearance of any sign/symptom of infection. Out of 400 patients admitted to 3 units during the study, 45 (11.3%) developed nosocomial infections. The most important determinants of infection were previous admission to an ICU, whether in the same or another hospital, and duration of stay. Overall, 38 patients developed 43 device-related nosocomial infections (1.13 episodes per patient). The main infectied association with invasive procedures were ventilator-associated pneumonia followed by catheter-related urinary tract infection and bloodstream infections. Infections nosocomiales liées aux dispositifs médicaux dans les services de réanimation de l’Hôpital universitaire d’Alexandrie RÉSUMÉ Nous avons étudié l’ampleur et les déterminants des infections nosocomiales liées aux dispositifs médicaux dans les services de réanimation et identifié les micro-organismes prédominants impliqués. Pour chaque patient, on a procédé à des prélèvements d’expectorations et d’urine ainsi qu’à une hémoculture et à une radiographie des poumons à l’admission et lors de l’apparition de tout signe/symptôme d’infection. Sur les 400 patients admis dans les trois services pendant l’étude, 45 (11,3 %) ont contracté une infection nosocomiale. Les déterminants les plus importants de l’infection étaient l’admission antérieure dans un service de réanimation, que ce soit dans le même hôpital ou dans un autre établissement, et la durée du séjour. En tout, 38 patients ont contracté 43 infections nosocomiales liées aux dispositifs médicaux (1,13 épisode par patient). La pneumonie associée à la ventilation artificielle suivie par les infections des voies urinaires et les infections sanguines liées à un cathéter constituaient les principales infections associées à des procédures invasives.

1Department of Epidemiology, High Institute of Public Health; 2Department of Anaesthesia, Faculty of Medicine, University of Alexandria, Alexandria, Egypt (Correspondence to M.A. Arafa: [email protected]). 3Intensive Care Unit; 4Laboratory Department, Alexandria University Students Hospital, Alexandria, Egypt. Received: 27/02/03; accepted: 14/01/04

07 Device-related nosocomial.pmd 52 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 53

Introduction Although not all catheter-associated urinary tract infections can be prevented, it is be- Nosocomial infection is an infection ac- lieved that many could be avoided by the quired by a patient in a hospital or other proper management of the indwelling cath- health care facility that was not present or eter [13,14]. Catheter-related bloodstream incubating at the time of admission or that infections are associated with increased was the residual of an infection acquired morbidity, a mortality rate of 10% to 20%, during a previous admission [1–4]. Noso- prolonged hospitalization (mean of 7 days) comial infections have been recognized for and increased medical costs. over a century as a critical problem affect- In 1994, the first German national study ing the quality of health care and a principal was conducted on the prevalence of noso- source of adverse health care outcomes. comial infections in medical, surgical, ob- Today, nosocomial infections account for stetric and gynaecological departments as 50% of all major complications of hospital- well as ICUs; the overall rate reported was ization; the remainder are due to medication 3.5% [15]. The reported rate for Spain in errors, patient falls and other non-infec- 1992 was 9.9% [16] and for Belgium in tious adverse events [5,6]. 1998 was 9.3% [16]. During 1997, in Nor- Nososcomial infection constitutes a way, the reported prevalence was 6.1%. In major problem globally, with major social, Saudi Arabia, an overall nosocomial infec- economical, moral and personal effects, tion prevalence rate was reported to be that increases the morbidity and mortality 5.7% in 1991, 2.7% in 1995 and 2.2% in of hospitalized patients in intensive care 1997 [17]. In the United Arab Emirates, a units (ICUs) [7–9]. The highest infection rate of 4.7% was reported over an 18- rates are in intensive care patients. The month period [18]. rates in ICUs are approximately 3 times In Egypt, a study conducted in 1983 higher than elsewhere in hospitals. The and 1984 in Alexandria University Hospital sites of infection and the pathogens in- reflected overall rates of nosocomial infec- volved are directly related to treatment in tions of 21.4% and 20.3% respectively ICUs [10,11]. [19]. In 1995, nosocomial infection was The important risk factors for acquisi- 25% in ICUs, and 10% in the burns unit at tion of infection are invasive procedures, Ain Shams University Hospital [20]. Alex- which include operative surgery, intravas- andria University Students Hospital pro- cular and urinary catheterization and me- vides services to a population of more than chanical ventilation of the respiratory tract 50 000, including university students (un- [1]. A device-associated infection is an in- dergraduate and postgraduate), staff of dif- fection in a patient given a device (central ferent faculties, different employees venous catheter, mechanical ventilator or working in the university and their families. indwelling urinary catheter) that was in use On average, 500–600 cases are admitted within 48 hours before the onset of infec- every year to the ICUs of the hospital. Data tion [12]. regarding device-related nosocomial infec- Nosocomial pneumonia has been asso- tion in the ICUs of Alexandria Hospital in ciated with high fatality rates. It is associat- general and in Alexandria University Stu- ed with a mortality rate of up to 50% dents Hospital in particular are scarce. among ICU patients. Most infections of the Availability of accurate data on nosocomial urinary tract (66% to 86%) follow instru- infection is mandatory for proper preven- mentation, mainly urinary catheterization. tion and control of nosocomial infections.

07 Device-related nosocomial.pmd 53 10/25/2005, 1:21 AM 54 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

The current study aimed to describe the • Investigations: sputum, urine and blood magnitude and determinants of device-re- cultures were taken on admission and lated nosocomial infections in the ICUs of on appearance of any sign and/or symp- Alexandria University Students Hospital tom of infection. Isolation and identifi- and to identify the predominant micro- cation of any cultured organisms was organisms involved. performed and the antibiotic sensitivity of organisms was tested. Chest X-rays were also performed. Methods • Follow up: to study the development of The study was carried out at the 3 ICUs of device-related nosocomial infection all the Alexandria University Students Hospital: patients were followed-up daily during the general, coronary and intermediate their stay and for 2 days after discharge ICUs. All patients admitted to ICUs were from ICUs to the wards of the hospital. followed from admission to discharge and for 2 days after discharge from the ICU to Data analysis the general ward, during a period of 1 year Quantitative variables were expressed as from January 2000 to February 2001. The arithmetic mean and standard deviation total sample amounted to 400 patients. (SD). Odds ratio (OR) was calculated to Patients admitted were followed up dai- find the determinants of the risk of devel- ly for the development of device-related oping ICU device-related nosocomial infec- nosocomial infections: ventilator-associat- tion. Stepwise multiple logistic regression ed pneumonia (VAP), catheter-related uri- analysis was performed to explore the ef- nary tract infection (CR-UTI) and fect of dependent variables. catheter-related bloodstream infection The variables entered in the logistic re- (CR-BSI). The Centers for Disease Control gression analysis for the determinants of and Prevention case definitions of device- nosocomial infection were: previous ICU related nosocomial infection were used admission in the same hospital, previous [21]. ICU admission in other hospital, duration of stay in hospital and number of invasive de- Data collection and scoring vices. The variables entered in the logistic Questionnaires, observations and record regression analysis for the determinants of reviews were used to collect the following pneumonia and urinary tract infection information about patients onto a pre- were: duration of stay in hospital, duration structured sheet: of stay in ICU, level of consciousness, en- • Sociodemographic data. dotracheal tube insertion, duration of en- • Clinical background: diagnosis at entry, dotracheal tube insertion, use of ventilator, other associated problems, past medical level of head, presence of other infection, history, previous hospitalization and his- use of nasogastric tube and use of sedation. tory of present condition. • ICU data: length of stay, invasive devic- Results es used (mechanical ventilator, central venous catheter or urinary tract cathe- Profile of the sample ter) and length of exposure to invasive The study sample comprised 400 patients, device. 57.0% males and 43.0% females, with an

07 Device-related nosocomial.pmd 54 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 55

age range from 5 to 95 years [mean (SD) developed infection (P = 0.02 and P = age 52.0 (19.9) years]. More than half of 0.007 respectively). the patients (51.8%) were admitted to the Table 2 shows the crude odds ratio and general ICU, 29.5% were admitted to the 95% confidence interval of previous hospi- coronary ICU and 18.8% were admitted to talization and ICU admission as determi- the intermediate ICU. nants of infections. Patients who were Out of 400 patients, 45 (11.3%) devel- previously hospitalized and those who were oped nosocomial infection in the ICU. previously admitted to the ICU were at in- Nearly two-thirds of them (31 patients) creased risk of nosocomial infection. were admitted to the ICU without any in- After adjusting for all other variables, fection; the remaining 14 patients were ad- multivariate logistic regression analysis mitted to the ICU with an existing infection (Table 3) showed that the most important (community-acquired or nosocomial from determinants of nosocomial infection were another hospital or ICU) and then devel- previous admission to the ICU, whether in oped another type of infection (a new spe- the same or another hospital, and duration cies and a new organism detected). Of the of stay in the hospital. 45 patients with infections, 25 had previ- Comparing patients who developed ously been hospitalized (18 in the same nosocomial infection and those who did not hospital, 7 in another) and 7 had previously showed no significant difference with re- been in an ICU (3 in the same hospital, 4 in gard to mean age (t = 0.42). However, the another). mean duration of stay in hospital and in the No infections occurred after discharge ICU for patients who developed nosocomi- from ICU to wards during the 48 hours of al infection was significantly longer than follow-up. those who did not develop infection (t values –2.28 and –2.81 respectively). Effect of age and duration of stay Table 1 shows the mean age and duration Effect of using invasive devices of hospitalization for patients who did and Patients who had an invasive device insert- did not develop infection. Duration of stay ed had a higher risk of developing nosoco- in the ICU and duration of stay in hospital mial infection in the ICU than those without were significantly higher for patients who a device (Table 4). For the 17 patients given

Table 1 Mean age and duration of hospitalization for patients according to development of nosocomial infection

Variable Developed infection t-test P-value Yes N o (n = 355) (n = 45)

Age (years) 53.6 (23.6) 52.0 (19.5) 0.42 0.677 Duration of stay in ICU (days) 20.3 (40.9) 6.3 (10.5) 2.28 0.02 Duration of stay in hospital (days) 33.0 (45.5) 12.5 (16.9) 2.81 0.007

Values shown are mean (standard deviation). n = number of patients. ICU = intensive care unit.

07 Device-related nosocomial.pmd 55 10/25/2005, 1:21 AM 56 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 2 Crude odds ratio (OR) and 95% Table 4 Crude odds ratio (OR) and 95% confidence interval (CI) of previous confidence interval (CI) of invasive devices hospitalization and intensive care unit (ICU) as determinants of nosocomial infection for admission as determinants of nosocomial patients given different types of invasive infection device

Variable Crude OR 95% CI Device Crude OR 95% CI

Previous hospitalization Endotracheal tube Same hospital 1.51 0.73–3.12 insertion 40.3 5.45–21.19 Other hospital 15.20 3.88–61.66 Central venous Previous ICU admission catheter insertion 26.2 5.95–35.20 Same hospital 9.21 1.42–59.76 Urinary tract catheter Other hospital 18.42 2.78–150.44 insertion 21.0 3.50–18.52 ICU = intensive care unit.

an endotracheal tube the odds ratio for ac- central venous catheter or urinary tract quiring a nosocomial infection was 40.3, catheter for a duration more than 3 days for 152 patients with a central venous cath- and for those who had these devices insert- eter the odds ratio was 26.2 and for 184 ed once or more than once. patients with a urinary tract catheter it was Adjusting all other variables for their ef- 21.0. The risk of infection increased with fect on the prevalence of nosocomial pneu- the increasing number of invasive devices monia and nosocomial urinary tract used; the odds ratios were 4.79, 5.21, infection showed that the following vari- 31.68 for 1, 2 and 3 devices respectively. ables were significantly associated with the Table 5 shows that the risk of develop- outcome: patients who were in coma; en- ing ICU nosocomial infection increased for dotracheal tube insertion; and use of a ven- patients who had an endotracheal tube, tilator for pneumonia. The risk of infection increased with every incremental increase in the duration of stay in the hospital and duration of device insertion (Table 6). Table 3 Multivariate logistic regression model: adjusted odds ratio (OR) and 95% Device-related nosocomial confidence interval (CI) of some infections determinants of nosocomial infection Of the 45 patients who developed a noso- Variable Adjusted 95% CI comial infection in the ICUs, 38 had a de- OR vice-related infection. Table 7 shows that out of 58 infections, 43 were device-related Previous ICU admission nosocomial infections: 11 pneumonia, 30 in same hospital 5.96 1.01–35.64 urinary tract infection and 2 bloodstream Previous ICU admission infection. Thus 38 patients developed 43 in other hospital 8.83 1.25–62.34 device-related nosocomial infections (1.13 Duration of stay in episodes per patient). hospital 1.03 1.01–1.04 Of the 17 patients on mechanical venti- ICU = intensive care unit. lators, 11 (64.7%) developed VAP. Of 184

07 Device-related nosocomial.pmd 56 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 57

Table 5 Crude odds ratio (OR) and 95% confidence interval (CI) as determinants of nosocomial infection for patients given different types of invasive device

Variable Endotracheal tube Central venous catheter Urinary tract catheter Crude OR 95% CI Crude OR 95% CI Crude OR 95% CI

Device inserted once 7.88 3.24–20.01 10.63 4.06–29.45 9.99 4.70–21.34 Device inserted twice 20.29 1.97–193.70 134.44 24.74–212.48 23.85 3.00–219.33 Device < 3 days 1.51 0.17–13.13 5.02 0.64–31.28 1.59 0.18–11.72 Device 3 days 1.02 0.01–8.80 2.86 0.38–16.92 6.38 1.53–25.00 Device > 3 days 14.48 5.59–39.70 20.08 7.67–55.60 19.94 8.38–48.00

ETT = endotracheal tube.

with a urinary tract catheter, 30 (16.3%) of sensitivity tests showed that Klebsiella them developed CR-UTI. Of 152 patients spp. were sensitive to imipenem, and resis- with a central venous catheter, 2 (1.3%) tant to meropenem, amikacin and sulbac- developed CR-BSI. tam/cefoperazone. Pseudomonas spp. For the patients with VAP, 2 were con- were sensitive to amikacin and ceftazidime scious, 2 semi-conscious, 3 in a coma, 1 and resistant to meropenem and sulbactam/ progressed and 3 deteriorated. cefoperazone. As regards CR-UTI cases, Escherichia Microorganisms isolated: coli 16 infections (53.3%), Candida albi- sensitivity and resistance cans 11 (36.7%), Klebsiella spp. 2 (6.7%) The most common microorganisms isolat- and Pseudomonas spp. plus E. coli 1 ed from patients with VAP were Klebsiella (3.3%) were the most common; they were spp. in 6 infections (54.5%) and Pseudo- sensitive to imipenem and ketoconazole. monas spp. in 5 (45.5%). The antibiotic

Table 6 Adjusted odds ratio (OR) and 95% confidence interval (CI) of determinants of nosocomial pneumonia and urinary tract infection in the intensive care units (ICU)

Variable Pneumonia Urinary tract infection Adjusted 95% CI Adjusted 95% CI OR OR

Duration of stay in hospital 2.97 1.85–4.29 1.18 1.03–45.00 Coma 42.37 1.84–120.12 – – Device inserted 12.27 1.07–115.23 – – Duration of device insertion 2.18 1.50–3.18 15.67 4.55–53.97 Ventilators 21.05 1.71–29.80 – – Duration of stay in ICU – – 1.21 1.02–45

07 Device-related nosocomial.pmd 57 10/25/2005, 1:21 AM 58 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 7 Distribution of nosocomial infections day was more than 4 times higher in trans- developed in the intensive care unit ferred patients—however, the risk was according to type of infection and invasive similar between patients transferred from device use another hospital and patients transferred Type of infection No. of infections within the hospital [22]. The multivariate Invasive device Total analysis showed that intra-hospital transfer, Yes N o length of hospital stay over 7 days and having had at least one invasive procedure Pneumonia 11 11 22 were independent risk factors for infection. Urinary tract infection 30 4 34 In the present study, 38 patients devel- Bloodstream infection 2 0 2 oped 43 device-related nosocomial infec- Total 43 15 58 tions (1.13 episodes per patient). A similar study in France revealed 41 patients developed 60 nosocomial infections (1.46 episodes per patient): 33 urinary tract infections, 15 pneumonias and 12 central The microorganism isolated from both venous catheter-related infections [23]. A cases with CR-BSI was E. coli and it was significant association between nosocomial sensitive to imipenem and ceftriaxone and infection and the type of device, the num- resistant to ampicillin/sulbactam and ce- ber of invasive devices at the same time, furoxime. the number of insertions for the same device and the duration of invasive device Outcome was observed in both studies, as each item Nearly half of the patients who developed increased the risk of nosocomial infection nosocomial infection in the ICU died in parallel. (44.4%, 20/45). Mortality was highest for The current study revealed that VAP oc- patients with VAP (63.6%, 7/11). curred in 64.7% of patients on mechanical ventilators, with an associated mortality of 63.6%. This higher figure than other re- Discussion ports (9%–21%) [24] may be attributed to The relation between nosocomial infection the small number of cases on mechanical and mortality in hospitals remains unclear; ventilators (17) of whom 11 developed in- however, the highest rates of nosocomial fection. It may also be related to reasons infections are observed in ICUs. In this such as lack of knowledge and poor prac- study period 11.3% of patients admitted to tices of the medical and nursing staff and the ICUs of the Alexandria University Stu- inadequate supplies leading to reuse of sin- dents Hospital developed nosocomial infec- gle-use supplies. tion and 44.4% of them died. Previous Several studies have examined the po- hospital admission, previous ICU admis- tential risk factors for nosocomially ac- sion and length of stay were the most im- quired bacterial pneumonia. These risk portant determinants for developing factors may be: extremes of age; severe nosocomial infection. A study done in underlying conditions (including patients France showed that transferred patients on immunosuppressant therapy and post- had a greater risk of nosocomial infec- operative patients); administration of antimicrobials, H -receptor blockers and tion—the risk of being infected on a given 2

07 Device-related nosocomial.pmd 58 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 59

sedatives; admission to an ICU; underlying the quality of catheter care and host sus- chronic lung disease; coma; endotracheal ceptibility [14]. UTIs have been identified intubation; insertion of a nasogastric tube; in approximately 30% of patients with uri- supine position; use of mechanical ventila- nary catheters within 2 weeks and virtually tor support with potential exposure to con- 100% at 6 weeks [24]. Another study re- taminated respiratory equipment or contact ported infection rates ranging from 1%– with contaminated or colonized hands of 5% after a single brief catheterization to health care workers [25,26]. In the present virtually 100% for patients with indwelling study, the risk factors for VAP were dura- urethral catheters draining into an open tion of hospital stay; coma; insertion of an system for longer than 4 days [27]. In our endotracheal tube; frequency and duration study, the number of catheter insertions, of endotracheal tube use; and use of a ven- duration of catheterization and previous ad- tilator. We found Klebsiella and Pseudomo- mission to an ICU or general ward played a nas spp. were the pathogens associated significant role in determining the occur- with VAP and this is similar to a study in the rence of CR-UTIs. USA by Fridkin who found that 69% of The most common microorganisms pathogens associated with VAP were gram- isolated from the patients with CR-UTIs in negative pathogens, most commonly the current study were E. coli, C. albicans, Pseudomonas aeruginosa and Enterobacter and Klebsiella spp. and Pseudomonas spp. spp. [27]. respectively. Similar organisms were found Each year, urinary catheters are inserted in other studies, highlighting the fact that in more than 5 million patients in acute-care CR-UTIs comprise perhaps the largest in- hospitals and extended-care facilities. CR- stitutional reservoir of nosocomial anti- UTIs are the most common nosocomial in- biotic-resistant pathogens and are caused fection in hospitals and nursing homes, by a variety of pathogens [14,24]. comprising more than 40% of all institu- CR-BSI occupies the third rank of tionally acquired infections. Nosocomial nosocomial infection in ICUs [4]. It occurs bacteriuria or candiduria develops in up to in less than 1% of hospitalized patients, but 25% of patients requiring a urinary catheter the greatest risk occurs in ICU patients for ≥ 7 days, with a daily risk of 5% [14]. [24]. It is often associated with intravascu- Platt et al. suggested that nosocomial CR- lar catheters. Often attention is directed UTIs were associated with substantially in- only towards the fact that the catheter creased institutional death rates, unrelated breaks the integrity of the skin, while other to the occurrence of urosepsis [28]. environmental factors include the dressing, Although some studies consider CR- frequency of dressing changes, the topical UTI as the most common nosocomial in- antimicrobial agent and the indication for fection in ICUs [14,28], it was observed in antimicrobials [30]. The present study re- the present study that CR-UTI was the sec- vealed a similar prevalence (1.3%) as CR- ond most common nosocomial infection BSI occurred in 2 patients from a total of after VAP, affecting 16.3% of the patients 152 with central-line catheters, and E. coli with a urinary catheter. This finding was in were isolated in both cases. accordance with that of Fridkin et al. [27] Studies showed that since the mid- and Gikas et al. [29]. 1980s, an increasing proportion of nosoco- The risk of acquiring a UTI depends on mial bloodstream infections reported to the the method and duration of catheterization, Nosocomial Infections Surveillance Sys-

07 Device-related nosocomial.pmd 59 10/25/2005, 1:21 AM 60 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

tem (at the Centers for Disease Control and Establishing a relationship between se- Prevention) have been due to gram-posi- verity of illness, therapeutic activity, occur- tive, rather than gram-negative organisms. rence of nosocomial infections and out- It was largely due to major increases in 4 come requires separate analyses of illness pathogens: coagulase-negative staphylo- severity and therapeutic activity as causes cocci, Candida spp., enterococci and Sta- of nosocomial infections, and of nosocomi- phylococcus aureus. The distribution of al infections as causes of excess illness se- these pathogens varied by hospital size and verity and extra-therapeutic activity. affiliation [30].

References

1. Damani NN. Manual of infection control 9. El-Shazly M, Osman H, Shehata S. Study procedures. Oxford, Oxford University of mortality in the main university hospi- Press, 1997. tal in Alexandria. Bulletin of the High In- stitute of Public Health, 1997, 27(2): 2. Benenson AS, ed. Control of communi- 104–7. cable diseases manual, 6th ed. Wash- ington DC, American Public Health 10. Monitoring hospital-acquired infections Association, 1995. to promote patient safety—United States, 1990–1999. Morbidity and mor- 3. World Health Organization. Guidelines tality weekly report, 2000, 49(8):149–53. on prevention and control of hospital associated infection. New Delhi, WHO Re- 11. Salama O. Hospital infection magnitude, gional Office for South-East Asia, 2002. prevention and control [Masters thesis]. Department of Community Medicine, 4. Infectious diseases. In: Braunwald E et Faculty of Medicine, University of Alex- al., eds. Harrison’s principles of internal andria, Egypt, 1999. medicine, 13th ed. New York, McGraw– Hill, 1994. 12. Quality Indicator Project. Acute care measures. Indicator 1a: device-associ- 5. Gaynes RP. Surveillance of nosocomial ated infection in the ICU. Baltimore, infections: a fundamental ingredient for Maryland, Association of Maryland Hos- quality. Infection control and hospital pitals and Health Systems, 1998. epidemiology, 1997, 18(7):475–8. 13. Tambyah PA, Knasinski V, Maki DG. The 6. Infections nosocomiales: souci de direct costs of nosocomial catheter- spécialiste ou préoccupation de tous? associated urinary tract infection in the [Nosocomial infection: concern of spe- era of managed care. Infection control cialist or preoccupation of all?]. Swiss- and hospital epidemiology, 2002, NOSO, 1994, 1(1):1. 23(1):27–31. 7. Gaynes R et al. Feeding back surveil- 14. Wong ES. Guidelines for prevention of lance data to prevent hospital-acquired catheter-associated urinary tract infec- infection. Emerging infectious diseases, tions. Atlanta, Georgia, Centers for Dis- 2001, 7(2):295–8. ease Control and Prevention, 1981:13. 8. Nanivadekar SM. Nosocomial infection. 15. Ruden H et al. Nosocomial and commu- An epidemiologic and microbiological nity-acquired infection in Germany. Sum- perspective: guidelines for antibiotic mary of the results of the First National policy. Bombay hospital journal, 1997, 39(1): 314–9.

07 Device-related nosocomial.pmd 60 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 61

Prevalence Study (NIDEP). Infection, matched case–control study of ICU pa- 1997, 25(4):199–202. tients. American journal of respiratory and critical care medicine, 1998, 157: 16. Gastmeier P et al. Importance of the sur- 1151–8. veillance method: national prevalence studies on nosocomial infections and 24. Nosocomial infection. In: Kirby RR, Tay- the limits of comparison. Infection control lor RW, Civetta JM, eds. Handbook of and hospital epidemiology, 1998, 19(9): critical care, 2nd ed. Philadelphia, 661–7. Lippincott, 1997:821–31. 17. Abussaud MJ. Prevalence of nosocomial 25. Emori TG, Gaynes RP. An overview of infection in a Saudi Arabian teaching nosocomial infection, including the role hospital. Journal of hospital infection, of the microbiology laboratory. Clinical 1991, 17:235–8. microbiology reviews, 1993, 6:428–42. 18. Salama OE. A study of nosocomial infec- 26. Koeman M et al. Ventilator-associated tion in a general hospital. Journal of the pneumonia: recent issues on pathogen- Egyptian Public Health Association, esis, prevention and diagnosis. Journal 1996, 71(1,2):31–46. of hospital infection, 2001, 49(3):155– 62. 19. Zaki SA. Nosocomial infection in Alexan- dria university hospitals: control and pre- 27. Fridkin SK, Welbel S, Weinstein RA. ventive measures. Postgraduate doctor Magnitude and prevention of nosoco- Middle East, 1988, 11(2):88–96. mial infections in the intensive care unit. Infectious disease clinics of North 20. Rasslan O, Abd El Sabour M. Infection in America, 1997, 11(2):479–96. intensive care in Egypt: problems and control. Postgraduate doctor Middle 28. Platt R et al. Reduction of mortality asso- East, 1996, 19(4):108–11. ciated with nosocomial urinary tract infection. Lancet, 1983, 1:893–6. 21. Garner JS et al. CDC definitions for nosocomial infections. In: Olmsted RN, 29. Gikas A et al. Prevalence study of hospi- ed. APIC infection control and applied tal-acquired infections in 14 Greek hos- epidemiology: principles and practice. St pitals: planning from the local to the Louis, Mosby, 1996:A1–20. national surveillance level. Journal of hospital infection, 2002, 50(4):269–75. 22. Eveillard M et al. Association between hospital-acquired infections and pa- 30. Mermel LA et al. Guidelines for the man- tients’ transfers. Infection control and agement of intravascular catheter related hospital epidemiology, 2001, 22(11): infections. Clinical infectious diseases, 693–6. 2001, 32:1249–72. 23. Girou E et al. Risk factors and outcome of nosocomial infections: results of a

07 Device-related nosocomial.pmd 61 10/25/2005, 1:21 AM 62 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Rate of hepatitis B seropositivity following mass vaccination in the Islamic Republic of Iran M-R. Zali,1 K. Mohammad,2 A-A. Noorbala,2 B. Noorimayer1 and S. Sahraz1

ABSTRACT Universal vaccination of all neonates against hepatitis B virus has been implemented in the Islamic Republic of Iran since 1993. To evaluate the efficacy of the programme, 2 large seroepidemiologic surveys were conducted before and after mass vaccination on a representative sample of 1/1000 of the population. The overall seropositivity rate showed no significant decline between 1991 and 1999 but in the age group 2–14 years the rates reduced significantly (1.3% versus 0.8%, P < 0.05). Interestingly, we observed a significantly higher decline in hepatitis B virus carrier rate in rural (1.5% versus 0.6%) than urban areas (1.1% versus 0.9%). Universal vaccination significantly decreased the carrier rate among young children in this country.

Taux de séropositivité à l’hépatite B après la vaccination de masse en République islamique d’Iran RÉSUMÉ La vaccination de tous les nouveau-nés contre le virus de l’hépatite B est appliquée en République islamique d’Iran depuis 1993. Afin d’évaluer l’efficacité du programme, deux grandes enquêtes séro- épidémiologiques ont été réalisées avant et après la vaccination de masse auprès d’un échantillon représentatif de 1/1000 de la population. Le taux de séropositivité global n’a pas baissé significativement entre 1991 et 1999 mais dans le groupe d’âge de 2-14 ans, ce taux a diminué significativement (1,3 % vs 0,8 %, p < 0,05). Fait intéressant, on a observé une diminution du taux de portage du virus de l’hépatite B significativement plus importante en milieu rural (1,5 % vs 0,6 %) qu’en milieu urbain (1,1 % vs 0,9 %). La vaccination universelle a permis de réduire significativement le taux de portage parmi les jeunes enfants dans ce pays.

1Research Centre for Gastroenterology and Liver Disease, Shaheed Beheshti University of Medical Sciences, Tehran, Islamic Republic of Iran (Correspondence to M.R. Zali: [email protected]). 2Ministry of Health, Tehran, Islamic Republic of Iran. Received: 08/12/02; accepted: 12/06/03

08 Rate of hepatitis B.pmd 62 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 63

Introduction ineffective strategy of vaccinating carriers and high-risk individuals [5,12]. Hepatitis B is one of the most common viral It is estimated that over 35% of Iranians infections in the world. Approximately 350 have been exposed to HBV and about 3% million individuals are chronically infected are chronic carriers, ranging from 1.7% in carriers of the virus and at high risk of Fars (a province in the south of the coun- death from active hepatitis, cirrhosis and try) to over 5% in Sistan va Baluchestan primary hepatocellular cancer [1]. Each (the most south-eastern province) [13]. It year, approximately 1 million people die has been concluded that over 50% of Irani- from the acute and chronic sequelae of an carriers have contracted the infection hepatitis B virus (HBV) infection, making it perinatally, making this the most likely one of the major causes of morbidity and route of transmission of HBV in our coun- mortality worldwide [2–4]. The HBV carri- try [13]. er rate varies widely from 0.01% to 20% in In the Islamic Republic of Iran, mass different geographical regions of the world vaccination against HBV infection was be- [5]. The development of safe and highly gun in 1993. Neonates born after this time effective hepatitis B vaccines now provides were required to be vaccinated using re- the means to confer long-term immunity combinant HBV surface antigen. In this against HBV infection [5–7]. In May 1992, study, we assessed the impact of the mass the World Health Assembly, the governing vaccination programme on the carrier rates body of the World Health Organization, en- of HBV in the Iranian population by com- dorsed recommendations stating that all paring the rates before and after the pro- countries should have hepatitis B vaccine gramme began. integrated into their national immunization programmes by 1997 [4]. Postnatal horizontal transmission of Methods HBV in early childhood seems to be the predominant method by which high HBV car- Study group rier rates in the Middle East are maintained. The Health and Disease Survey in the Is- Infection in the first 5 years of life makes lamic Republic of Iran was conducted in 2 the largest contribution to the caseload of phases, in 1991 and 1999. These 2 large- chronic liver disease and thus to mortality scale studies were conducted on a 1/1000 from HBV [8]. Mass vaccination pro- nationwide sample of the Iranian population grammes to reduce the rate of chronic hep- of all age groups. Families from different atitis and its carrier state in regions with a geographical areas were selected to take high prevalence of HBV infection have been part in the studies. A clustering method was implemented in different parts of the world. used to perform sampling, with each clus- The results have been promising, and ter containing 8 families. ‘Family’ was de- marked reductions in the carrier rate fined as 1 or more people living together in among children have been reported in a common residence. The families were se- follow-up studies [4,9–11]. Even in coun- lected from both urban and rural locations tries with a low frequency of the disease, in different geographical areas. the authorities are strongly considering im- During the first health survey 39 841 plementing universal immunization pro- individuals were enrolled. The sample size grammes as a substitute for the current for the second health survey was 46 631.

08 Rate of hepatitis B.pmd 63 10/25/2005, 1:21 AM 64 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Data collection areas (2.0%) than in urban areas (1.5%) (P The Ministry of Health and Medical Educa- < 0.05, 95% CI 0.003–0.008). The HBV tion assigned a core coordinating team who carrier rate was 1.5% in females versus discussed the goals of the study and 1.9% in males. The difference between planned its design and conduct. In each men and women was statistically signifi- province, 1 person was selected as provin- cant (95% CI 0.00l–0.006). The age distri- cial supervisor who was responsible for bution of HBsAg carriers showed that coordinating the data gathering in that highest prevalence rates were among those province. Provincial supervisors checked aged 30–69 years, while the lowest preva- all data before submission to the core coor- lence (1.3%) was seen among children dinating team. Sample collection teams aged 2–9 years. There was a statistically were sent to each province, comprising 2 significant difference of HBsAg carrier physicians (1 male and 1 female), 1 techni- state between the age groups 2–14 years cian and 1 person trained in interview tech- old and the rest of the population (P < niques and the relevant questionnaire. 0.001). Blood samples were obtained from all Among 46 631 specimens obtained for individuals enrolled in the study and were screening of HBsAg in the second survey in sent as soon as possible to the central 1999, 806 (1.7%) were positive (95% CI laboratory. Samples were tested for pres- 1.6–l.9). The rate was 1.8% for urban, and ence of the hepatitis B surface antigen 1.6% for rural areas; this difference was (HbsAg) by immunoassay (Hepanostika®, not statistically significant. Women showed Biomerieux, Netherlands). a 1.5% rate of positivity and for men this was 1.9%; the difference was significant Statistical analysis (95% CI for the difference 0.002–0.006). The percentage of the population positive The carrier rate was 0.8% among children for HBsAg were calculated. Differences aged 2–14 years old. between rates were estimated for compara- Between the first and second health sur- ble groups and 95% confidence intervals veys, a significant reduction in HBV carrier (CI) were calculated as recommended by rate occurred in children between 2 and 14 Altman et al. [14]. All calculations were years old (1.3% versus 0.8% respectively, performed with SPSS, version 10.05 P < 0.05, 95% CI for the difference 0.003– (SPSS, Chicago, Illinois) and CIs were cal- 0.007) (Figure 1). This decline was uni- culated by the Confidence Interval Assess- formly seen both in males (1.4% versus ment Program (CIA, copyright 2000, 0.8%) (P < 0.05, 95% CI for difference Trevor Bryant). 0.003–0.009) and females (1.2% versus 0.8%) (P < 0.05, 95% CI for difference 0.001–0.007), but was not observed in oth- Results er age groups (Figure 2). Decline was more pronounced among children living in rural Among 39 841 individuals tested for areas (1.5% versus 0.6%) (P < 0.05, 95% HBsAg during the first health survey in CI for difference 0.006–0.0 13) compared 1991, 1.7% (677) had HBsAg detected in with those living in urban areas (1.1% ver- their sera (95% CI 1.6–1.8). The HBsAg sus 0.9%) (P < 0.05, 95% CI for differ- carrier rate was significant higher in rural ence –0.00l to 0.005) (Figure 3).

08 Rate of hepatitis B.pmd 64 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 65

Discussion Our comparison of serum HBsAg positivity rates in the general population of the Islam- ic Republic of Iran between 1991 and 1999 shows no change in the overall rate of HB- sAg positivity. To be more precise, the rate did not change in the age group 15–69 years during this time, while among children aged 2 to 14 years old a significant reduction in HBsAg seropositivity was observed. This result was seen both in male Figure 1 Hepatitis B surface antigen carrier and female subjects and in urban and rural rate in the Iranian population by age, areas. comparing health surveys in 1991 and 1999 Interestingly, after the mass vaccination programme the higher rate of HBsAg carri- Males ers in rural areas changed to a lower rate compared with urban areas. We speculate that this may be due to the particular design of the health system in rural parts of this country, which is delivered through a special network of health houses. In these houses, health staff prepare a specific record for each family in their village and monitor the vaccination status of any newborn in the area. Vaccinations may be delivered at home if the family lives far from a health house. In urban areas, vaccination provision is more dependent on families at- Females tending special facilities that provide such services. The different vaccination delivery systems may explain the more pronounced decline in HBsAg seropositivity rate in 2– 14-year-old children living in rural areas compared with those living in urban areas. Although during the first and second health surveys, no data were provided concerning complete vaccination coverage rates, other studies from the Iranian Ministry of Health and Medical Education show that vaccination coverage is overall above 95%. We think our data provide indirect evidence Figure 2 Hepatitis B surface antigen carrier that implementation of mass vaccination rate in the Iranian population by age and against hepatitis B virus may be effective in sex, comparing health surveys in 1991 and reducing the hepatitis B carrier rate in the 1999 population.

08 Rate of hepatitis B.pmd 65 10/25/2005, 1:21 AM 66 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Rural vaccination of predefined cohorts) may be necessary to deliver a more rapid decline in the hepatitis B carrier rate. Vaccination has been reported to result in adequate protection in about 90% of cases. Integration of HBV into the WHO Ex- panded Programme on Immunization for mass vaccination of infants in areas where HBV infection is endemic and morbidity is high would be the most effective means of providing the coverage necessary for effective control and prevention [3]. Follow- ing the neonatal vaccination programmes that were begun in 1989 in Saudi Arabia, Urban the overall HBsAg carrier rate in children under 12 years old fell from 6.7% in 1989 to 0.3% in 1997 [16]. Chen et al. [17] showed that 10 years after implementation of mass vaccination in Taiwan, the HBsAg seropositivity rate declined from 9.8% to 1.3%. Mass vaccination has been shown to be a safe and highly effective means for the control and prevention of HBV [4]. The 2 main challenges to eradication of hepatitis B in the next few decades are, first, the pres- Figure 3 Hepatitis B surface antigen carrier ence of a large worldwide reservoir of car- rate in the Iranian population by age and riers that requires a continuous effort to area, comparing health surveys in 1991 and sustain immunization programmes and, 1999 secondly, how to make vaccines and drugs available to countries with limited resources [12]. The problem of hepatitis B carriers The universal vaccination programme is of paramount importance as the eradica- against hepatitis B was launched in 1993. tion of hepatitis B can be achieved only af- As the Islamic Republic of Iran is consid- ter the 300 million carriers of the disease in ered a country with an intermediate rate of the world today have died or been cured HBV infection, there will not be a rapid de- [18]. cline in the overall carrier rate, and the rates of chronic liver disease and deaths resulting from HBV are unlikely to change over a pe- Acknowledgement riod of less than one decade [13,15]. Any- The authors would like to thank Dr Roya how, the rate of carrier state among Sherafat for her contribution in editing the children is a good proxy for evaluating the article. effectiveness of this programme. Other prevention strategies (such as successive

08 Rate of hepatitis B.pmd 66 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 67

References

1. Maynard JE. Hepatitis B: global impor- 10. Viviani S, et al. Hepatitis B vaccination in tance and need for control. Vaccine, infancy in The Gambia: protection 1990, 8(Suppl.):S18–20. against carriage at 9 years of age. Vac- cine, 1999, 17(23–24):2946–50. 2. Alter MJ, Mast EE. The epidemiology of viral hepatitis in the United States. Gas- 11. Coursaget P et al. Seven-year study of troenterology clinics of North America, hepatitis B vaccine efficacy in infants 1994, 23:437–55. from an endemic area (Senegal). Lan- cet, 1986, 2(8516):1143–5. 3. Maynard JE, Kane MA, Hadler SC. Glo- bal control of hepatitis B through vacci- 12. Rizzetto M, Zanetti AR. Progress in the nation: role of hepatitis B vaccine in the prevention and control of viral hepatitis Expanded Program on Immunization. type B: closing remarks. Journal of medi- Reviews of infectious diseases, 1989, cal virology, 2002, 67(3):463–6. 11(Suppl.3):S574–8. 13. Merat S et al. Hepatitis B in Iran. Archives 4. Kane M. Global program for control of of Iranian medicine, 2000, 3(4):192– hepatitis B infection. Vaccine, 1995, 201. 13(Suppl.1):S47–9. 14. Gardner MJ, Altman DG, eds. Statistics 5. Margolis HS, Alter MJ, Hadler SC. Hepa- with confidence. London: British Medical titis B: evolving epidemiology and impli- Journal, 1989. cations for control. Seminars in liver 15. Kew MC. Progress towards the compre- disease, 1991, 11:84–92. hensive control of hepatitis B in Africa: a 6. Hadler SC et al. Long–term immunoge- view from South Africa. Gut, 1996, nicity and efficacy of hepatitis B vaccine 38(Suppl. 2):S31–6. in homosexual men. New England jour- 16. Al-Faleh FZ et al. Seroepidemiology of nal of medicine, 1986, 315:209–14. hepatitis B virus infection in Saudi chil- 7. Margolis HS et al. Prevention of hepatitis dren 8 years after a mass hepatitis B B virus transmission by immunization: an vaccination programme. Journal of infec- economic analysis of current recom- tion, 1999, 38:167–70. mendations. Journal of the American 17. Chen HL et al. Seroepidemiology of Medical Association, 1995, 274:1201–8. hepatitis B virus infection in children: ten 8. Toukan AU. Hepatitis B in the Middle years of mass vaccination in Taiwan. East: aspects of epidemiology and liver Journal of the American Medical Asso- disease after infection. Gut, 1996, ciation, 1996, 276(22):1802–3. 38(Suppl. 2):S2–4. 18. Ghendon Y. WHO strategy for the global 9. Poovorawan Y et al. Impact of hepatitis B elimination of new cases of hepatitis B. immunisation as part of the EPI. Vaccine, Vaccine, 1990, 8(suppl.):S129–33, dis- 2000, 19(7–8):943–9. cussion S134–8.

08 Rate of hepatitis B.pmd 67 10/25/2005, 1:21 AM 68 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Behçet disease: clinical spectrum and association with hepatitis B and C viruses S. Farajzadeh,1 M.R. Shakibi,2 S. Darvish Moghaddam3 and Z. Rahnama1

ABSTRACT We studied 48 patients with Behçet disease to determine the clinical spectrum of the disease. We also compared the seropositivity of patients for hepatitis B (HBV) and C (HCV) infection with a healthy control group to determine whether there is an association. The major physical findings were oral aphthosis 93.8%, genital aphthosis 77.1% and ocular manifestations 64.6%. No patients were HCV antibody seropositive, but 3 of the control group (3.1%) tested seropositive. One patient (2.1%) and 2 in the control group (2.1%) tested positive for both HBV surface antigen and HBV core antibody. The differences were not statistically significant. There is, therefore, no case for recommending viral screening for HBV and HCV in Behçet disease patients at present. Maladie de Behçet : spectre clinique et association avec les virus des hépatites B et C RÉSUMÉ Nous avons étudié 48 patients atteints de la maladie de Behçet pour déterminer le spectre clinique de la maladie. Nous avons également comparé la séropositivité des patients pour l’infection par le virus de l’hépatite B (VHB) et C (VHC) par rapport à un groupe témoin de sujets sains afin de déterminer s’il y avait une association. L’aphtose buccale (93,8 %), l’aphtose génitale (77,1 %) et les manifestations ocu- laires (64,6 %) constituaient les principales constatations physiques. Aucun patient n’était séropositif aux anticorps anti-VHC, mais trois personnes dans le groupe témoin (3,1 %) présentaient une séropositivité. Un patient (2,1 %) et deux personnes dans le groupe témoin (2,1 %) ont été trouvés positifs pour l’antigène de surface du VHB et les anticorps anti-HBc. Les différences n’étaient pas statistiquement significatives. Il n’y a donc pas lieu à l’heure actuelle de recommander le dépistage du VHB et du VHC chez les patients atteints de la maladie de Behçet.

1Department of Dermatology; 2Department of Rheumatology; 3Department of Gastroenterology, Kerman Medical University, Kerman, Islamic Republic of Iran (Correspondence to S. Farajzadeh: [email protected]). Received: 10/12/03; accepted: 22/03/04

09 BehÁet disease.pmd 68 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 69

Introduction and major agents of viral hepatitis, we carried out a case–control study in Kerman, a Behçet disease, a multisystem vasculitis, is southern province in the Islamic Republic seen most frequently in Far Eastern and of Iran. Middle Eastern countries. Prevalence is estimated to be 1.67/10 000 population in the Islamic Republic of Iran [1]. Incidence is Methods relatively higher from eastern Asia to the The participants in this case–control study Mediterranean, roughly 1–10 in 10 000 included all patients with Behçet disease people whereas it is only 1–2 in 1 000 000 who were referred to the rheumatology and people in the United Kingdom and North dermatology clinics of Kerman Medical America [2]. The etiopathogenesis of this University from May 2001 to May 2002. chronic disease is unknown; it is suspect- Diagnosis of Behçet disease was based on ed, however, that some type of infection the 1990 international classification criteria such as human herpes virus 1 or Strepto- [14]. The control group was selected from coccus sp. initiates an autoimmune reaction healthy blood donors attending the Kerman in genetically predisposed individuals [3]. branch of the Iranian blood donation orga- Recently, the association between Behçet nization after history taking and clinical ex- disease and hepatitis C virus (HCV) [4–6] amination. Informed consent was given by and hepatitis B virus (HBV) [6,7] infection all participants in both the case and the con- has been studied. trol groups. There were no refusals to take Infection with HCV is widespread and part in the study. it is estimated that chronic infection affects A total of 48 patients with the disease 170 million people worldwide. [8]. Numer- (24 women and 24 men) and 96 healthy ous extra-hepatic disorders have been rec- volunteer blood donors (43 women and 53 ognized in association with HCV infection, men) participated in the study. The 2 among which dermatological diseases are groups were matched for age and sex. Peo- central [9–11]. While there is at least 1 re- ple with a history of transfusion of blood port which favours an association between and blood products, haemophilia, thalas- Behçet disease and HCV [12], other studies saemia and HBV vaccination as well as hae- dispute it [4–6]. modialysis patients and intravenous drug Additionally, HBV remains an important users were excluded from both groups. causative agent of chronic liver disease There were no exclusions from the test worldwide, and it is estimated that over group but 3 people were excluded from the 35% of Iranians have been exposed to the control group, 2 with a past history of virus [13]. Few studies have, however, blood transfusion and 1 who had previous been done concerning an association be- HBV vaccination. tween HBV and Behçet disease [6–7]. The History taking and physical examination high prevalence of HBV infection in the Is- (cutaneous examination by a dermatologist lamic Republic of Iran might be a possible and systemic examination by a rheumatolo- causative factor in the disease in this part gist) were done to detect the cutaneous and of the world [13]. other organ involvement. Liver disease was In order to determine the clinical spec- evaluated by the determination of serum turum of Behçet disease and whether there levels of alanine aminotransferase, aspar- is an association between Behçet disease

09 BehÁet disease.pmd 69 10/25/2005, 1:21 AM 70 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

tate aminotransferase, alkaline phosphatase Table 1 Seropositivity for hepatitis C virus and total and fractionated bilirubin. (HCV) among patients with Behçet disease Serum was tested for HCV antibodies and a control group using an enzyme-linked immunosorbent as- Outcome Patients Controls Total say (bioMerieux Kits, Marcey L’Etoile, (n = 48) (n = 96) France) and positive cases were confirmed No. % No. % by second-generation recombinant immu- noblot assay in both groups. Markers for HCV Ab + 0 – 3 3.1 3 HBV were evaluated via HBV surface anti- HCV Ab – 48 100 93 96.9 141 gen (HBs Ag) and total anti-HBV core anti- The difference between the 2 groups was not body (HBc Ab), also detected by the significant. enzyme-linked immunosorbent assay Fisher exact test, P = 0.55. method. The data were recorded and analysed using SPSS, version 10. The Fisher exact patient group who had elevated alanine ami- test was used for statistical comparison of notransferase and aspartate aminotrans- the 2 groups. ferase, the function tests were normal for both groups. No anti-HCV antibody was detected in Results the patient group, but 3 people in the control group were positive for antibodies (Ta- Mean age for the 48 people in the patient ble 1). One patient and 2 people in the group was 30.3 years (standard deviation control group tested positive for both HB- 11.0) and for the 96 in the control group sAg and anti-HBcAb. The results were not was 31.4 years (standard deviation 8.7). statistically significant (Table 2). The major physical findings in the patients were oral aphthosis (93.8%), genital aphthosis (77.1%), ocular manifestations Discussion (64.6%) and skin changes (62.5%). The most frequent ophthalmic involvement was There have been many reports on the clini- anterior uveitis (39.6%). Other ocular man- cal picture of Behçet disease from different ifestations included posterior uveitis (35.4%) and retinal vasculitis (22.9%). Skin manifestations were varied, the most Table 2 Seropositivity for hepatitis B virus common being pseudofolliculitis (62.5%). among patients with Behçet disease and a Less-common types included erythema no- control group dosum (39.6%), papules (22.9%), superfi- cial phlebitis (2.1%) and skin aphthosis Outcome Patients Controls Total (12.5%). Of the minor manifestations, the (n = 48) (n = 96) No. % No. % most frequent finding was joint involvement in the form of peripheral arthritis HBs Ag + 1 2.1 2 2.1 3 (66.7%). Other minor manifestations in- HBs Ag – 47 97.9 94 97.9 141 cluded central nervous system (20.8%) and renal (4.2%) involvement. Pulmonary, The difference between the 2 groups was not significant. gastrointestinal and cardiac involvements HBs Ag = hepatitis B surface antigen. were not found. Except for 1 person in the Fisher exact test, P = 1.0.

09 BehÁet disease.pmd 70 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 71

parts of the world [1,15–17]. In a large group and 1 in the patient group. The dif- survey by Davatchi et al. on 3443 Iranian ferences between the 2 groups were not patients, the findings were similar to those statistically significant. Indeed, in our in our study except for lower incidence of study the rate of seropositivity in people genital aphthosis (–12.9%) and a higher in- with Behçet disease and those in the control cidence of skin involvement (+11.1%) [1]. group was comparable with the rate for The differences may be related to sample chronic carriers in the Islamic Republic of size, since theirs was a nationwide study, Iran, 1.7% to 5.0% [13]. and diagnosis in their study being based on The results of 3 studies done in Turkey the clinical picture. Comparison between on association with HCV are similar [4–6]. other studies and our survey reveals some Hamuryudan, Sonsuz and Yurdakul studied differences in the frequency of the physical 13 patients with severe Behçet disease; all findings. In Saudi Arabia a higher incidence were negative for HCV antibodies [4]. The for genital aphthosis (+9.9%), and a lower study carried out by Ilter et al. on 35 pa- incidence of joint manifestations (–29.7%) tients with Behçet disease and 35 patients was found [15], and in England a higher with various cutaneous disorders also incidence of genital aphthosis (+13.9%) showed no positive correlation [5]. In an- and a lower incidence of ocular (–39.6%) other study, Asku et al. compared the fre- and articular (–19.8%) manifestations was quency of hepatitis A, B, C and E virus found [16]. In a study done in Turkey, the serology in 124 patients and a control incidence of ophthalmic (–17.2%) and joint group of healthy and diseased individuals. (–19.8%) manifestations was lower, while No difference was found in the frequency the incidence of skin involvement was of hepatitis A, C and E seropositivity. For higher (+15.3%) [17]. In another study unexplained reasons, however, the fre- which was done in Iraq, the incidence of quency of HBV seropositivity was signifi- ophthalmic (–16.6%) and joint (–18.7%) cantly lower among those with Behçet manifestations was lower, while the inci- disease [6]. One case was reported by dence of cutaneous manifestations was Munke, Stockman and Ramadori who pre- higher (+12.5%) [18]. These differences sented a Behçet disease patient with associ- may be explained by racial and geographi- ated HCV infection in Germany [12]. In a cal divergence, method of patient selection Turkish study, the incidence of HBs Ag, and the diagnostic criteria used by different anti-HBc Ab and anti-hepatitis A virus im- authors. munoglobulin (IgG, IgM) was no higher in Because of the unpredictable course of patients with Behçet disease than in the Behçet disease and the difficulty in manage- control group [7]. ment, some researchers have tried to dis- In conclusion, the results of our study cover the etiopathogenesis and conse- were in agreement with those of other re- quently a possible therapy for this pro- ports, and no significant difference was longed systemic disease. In this regard, found between people with Behçet disease some studies have been done on the possi- and healthy controls for HCV and HBV in- ble association with HBV and HCV [4–7]. fection. Therefore, we do not recommend In our study, we found 3 of the control serologic screening for HCV and HBV in group positive for HCV antibodies but none Behçet disease patients until more data are of the patient group. We also found 2 peo- collected. On the other hand, because there ple positive for HBV markers in the control may be some unknown epidemiologic and

09 BehÁet disease.pmd 71 10/25/2005, 1:21 AM 72 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

immunologic factors which could affect in different parts of the world, i.e. other the association of the disease with HCV and than Turkey and the Islamic Republic of HBV, we recommend doing more research Iran.

References

1. Davatchi F et al. Behçet disease. Analy- tology and Venereology, 1998, 10(1): sis of 3443 cases. APLAR journal of 12–21. rheumatology, 1997, 1(1):2–5. 10. Krengel S et al. Hepatitis C virus- 2. Suzuki Kurokawa M, Suzuki N. Behcet’s associated dermatosis: a review. disease. Clinical and experimental Hautarzt, 1999, 50(9):629–36. medicine, 2004, 4(1):10–20. 11. Gordon SC. Extrahepatic manifestations 3. Alpsoy E et al. Interferon alfa-2a in the of hepatitis C. Digestive diseases, 1996. treatment of Behçet’s disease: a random- 14(3):157–68. ized placebo-controlled and double- 12. Munke H, Stockmann F, Ramadori G. blind study. Archives of dermatology, Possible association between Behçet’s 2002, 138(4):467–71. syndrome and chronic hepatitis C virus 4. Hamuryudan V, Sonsuz A, Yurdakul S. infection. New England journal of medi- More on hepatitis C virus and Behçet’s cine, 1995, 332(6):400–1. syndrome. New England journal of medi- 13. Merat S et al. Hepatitis B in Iran. Archives cine, 1995, 333(5):322–3. of Iranian medicine, 2000, 3(4):192– 5. Ilter N et al. Behçet’s disease and HCV 201. infection. International journal of derma- 14. International Study Group for Behçet’s tology, 2000, 39(5):396–7. Disease. Criteria for diagnosis of 6. Aksu K et al. Prevalences of hepatitis A, Behçet’s disease. Lancet, 1990, 335: B, C and E viruses in Behçet’s disease. 1078–80. Rheumatology, 1999, 38(12):1279–81. 15. Al-Dalaan AN et al. Behçet’s disease in 7. Gurler A et al. Behçet’s hastali-ginda Saudi Arabia. Journal of rheumatology, hepatit markerlarinin rolu. In: Tuzun Y, 1994, 21(4):658–61. Kotogyan A, Serdaroglu S, eds. Ulusal 16. Chamberlain MA. Behçet’s syndrome in dermatoloji kongeresi serbest bildiriler 32 patients in Yorkshire. Annals of the kitabi. Istanbul, Teknografik Matbaacilik, rheumatic diseases, 1997, 36(6):491–9. AS, 1988:37–42 [in Turkish]. 17. Yurdakul S et al. The prevalence of 8. Dhumeaux D, Marcellin P, Lerebours E. Behçet’s syndrome in a rural area in Treatment of hepatitis C. The 2002 northern Turkey. Journal of rheumatol- French consensus. Gut, 2003, 52(12): ogy, 1988, 15(5):820–2. 1784–7. 18. Al-Rawi ZS et al. Behçet’s disease in 9. Hadziyannis SJ. Skin diseases associ- Iraqi patients. Annals of the rheumatic ated with hepatitis C virus infection. Jour- diseases, 1986, 987–90. nal of the European Academy of Derma-

09 BehÁet disease.pmd 72 10/25/2005, 1:21 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 73 Rapid reinfection by Giardia lamblia after treatment in a hyperendemic area: the case against treatment M.J. Saffar,1 J. Qaffari,1 A.R. Khalilian1 and M. Kosarian2

ABSTRACT We selected 405 children aged 1–10 years with Giardia lamblia infection but without abdominal or gastrointestinal complaints for the previous month. For 5 days, 204 received metronidazole 15 mg/kg/day and 201 received B-complex syrup. Stool samples were examined 2–3 weeks and 3 months after treatment and results were tested with chi-squared. Weight and height 6 months after treatment were compared with primary weight and height by Z-score and Student t-test. Metronidazole efficacy at 2–3 weeks was 85.3%. Three months after treatment, 60 were reinfected (34.5%) and 71 had spontaneously cleared (35.3%). Because of high reinfection, spontaneous clearing and treatment failure rates, and the lack of effect on nutritional status or growth, we do not recommend treatment for children with asymptomatic giardia infection.

Réinfection rapide par Giardia lamblia après traitement dans une zone hyperendémique : faut-il traiter ? RÉSUMÉ Nous avons sélectionné 405 enfants âgés de 1 à 10 ans ayant une infection à Giardia lamblia mais n’ayant pas eu de troubles abdominaux ou gastro-intestinaux pendant le mois qui précédait l’étude. Pendant cinq jours, 204 enfants ont reçu du métronidazole à la dose de 15 mg/kg/jour et 201 enfants ont pris un complexe de vitamines B en sirop. Des échantillons de selles ont été examinés 2-3 semaines et 3 mois après le traitement et les résultats ont fait l’objet d’une analyse en utilisant le test du khi-carré. Le poids et la taille six mois après le traitement ont été comparés avec le poids et la taille initiaux à l’aide du test t de Student sur les scores Z. L’efficacité du métronidazole à 2-3 semaines était de 85,3 %. Trois mois après le traitement, 60 enfants étaient réinfectés (34,5 %) et 71 avaient éliminé le parasite spontanément (35,3 %). En raison des taux élevés de réinfection, d’élimination spontanée du parasite et d’échec thérapeutique ainsi que de l’absence d’effets sur l’état nutritionnel ou la croissance, le traitement pour les enfants ayant une infection à Giardia asymptomatique n’est pas recommandé.

1Department of Paediatrics; 2Department of Statistics, Mazandaran Medical Sciences University, Boalisina Hospital, Sari, Islamic Republic of Iran (Correspondence to M.J. Saffar: [email protected]). Received: 25/06/03; accepted: 04/12/03

10 Rapid reinfection.pmd 73 10/25/2005, 1:21 AM 74 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction however, no adverse effects on growth or nutritional status were detected [3,10,11]. Throughout the world, Giardia lamblia is Because of differences in study designs, often one of the first parasites to infect in- interpreting these findings is difficult. fants and children. Most human infections There is general agreement that persons result from the ingestion of contaminated with symptomatic giardiasis should be water or by direct fecal–oral transmission treated, although the indications for treat- [1]. The prevalence of giardia in stool spec- ing children with asymptomatic giardia in- imens submitted for ova and parasite exam- fection remain controversial [12,13]. ination has been reported to be 2%–5% in Because diarrhoea is a common symptom industrialized countries and 20%–30% in of giardiasis, giardia infection in the ab- developing countries [2]. In the majority of sence of diarrhoea is often considered to be infected individuals, or approximately 60% asymptomatic. However, other symptoms depending on the population, the infection including flatulence, foul-smelling stool and remains asymptomatic. Asymptomatic in- abdominal pain may occur more frequently fection may be more common in children than diarrhoea and are often of more con- and in people with prior infection [2–4]. cern to parents and patients [14–16]. Symptomatic patients have diarrhoea with If many children who live in hyperen- loose, foul-smelling stool, flatulence, ab- demic areas are rapidly reinfected after dominal cramping, bloating, nausea, anor- treatment and if their infections clear spon- exia, malaise and weight loss. There is taneously without treatment, a policy of increased amount of fat and mucus in fae- treatment for all infected children with gia- cal samples, but blood is not present. Pa- rdia, with or without symptoms, may not tients usually have no fever. Although be prudent where public health resources giardia infection may resolve spontaneous- are scarce [13,17–19]. Our study aimed to ly, the illness lasts for several weeks and determine the prevalence of asymptomatic sometimes months if left untreated [5]. Pa- giardia infection and its spontaneous clear- tients with chronic giardiasis have pro- ing rate by evaluating the effect of metron- found malaise, diffuse epigastric pain and idazole on carrier, growth and nutritional abdominal discomfort. Malabsorption of status of infected children and also to de- fats, carbohydrates, sugars and vitamins in termine the rate of reinfection after effec- cases of giardiasis has been well docu- tive therapy for children with asympto- mented and may be responsible for sub- matic giardia infection, i.e. without diar- stantial weight loss, even when the infec- rhoea and without abdominal complaints, in tion is apparently asymptomatic [5–7]. Sari, Islamic Republic of Iran. According to clinical and epidemiological studies the effects of nondiarrhoeal or asymptomatic giardia infection on nutri- Methods tional status are inconsistent. In New Delhi, asymptomatic infected children had lower Our experimental study included children weight and height than uninfected children aged 1–10 years in day care centres and of the same age [8]. In Guatemala, treat- schools in Sari in the Mazandaran province ment of asymptomatic infection caused in the northern part of our country. The greater increases in weight and height than children were without any abdominal or in a control group [9]. In other studies, gastrointestinal complaints for the 1 month

10 Rapid reinfection.pmd 74 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 75

prior to our study. Stool specimens were tect the rate of giardia infection. The results collected by the centres or the parents. Gi- were compared with Student t-test. After 6 ardia infection was determined by 2 sepa- months, weight and height were charted rate examinations of the specimen directly again. Z-scores of weight and height of the after emulsification in saline by 2 techni- 2 groups after treatment were compared by cians in the Boalisina hospital laboratory. A Student t-test. portion of the stool specimen was pre- served by emulsification in methiolate– iodine–formalin solution and processed by Results both formalin-ether sedimentation and To identify the 405 asymptomatic giardia- zinc-sulfate flotation techniques. infected children in our study, we examined We chose 405 asymptomatic children 2500 stool samples from 2500 asymptom- with giardia infections for our study. We atic children, i.e. one sample from each then gathered from their parents informed child. Of these, 16.2% were positive. The consent and completed questionnaire about rate of infection increased with increasing signs and symptoms related to giardia in- age. Table 1 shows that the infection rate fection, educational level of the family, was 10.0% for children attending day care family size, water supply and sanitary con- centres, but was 17.7% for slightly older ditions of the home. We measured weight children in school. and height and then charted and matched Of the 405 children with asymptomatic children in the case and control groups. In- giardia infection who were enrolled in our fected children in the case group received study, 204 children in the case group re- metronidazole at 15 mg/kg per day for 5 ceived 15 mg/kg per day of metronidazole days and infected children in the control for 5 days and 201 children in the control group received B-complex syrup also daily group received B-complex syrup only. To for 5 days. Children were followed for 6 detect metronidazole efficacy, 2–3 weeks months duration. Stool specimens from the after treatment 3 stool samples from each children in the case group were examined child in the case group were examined. Ta- 2–3 weeks after treatment to examine the ble 2 shows that 30 of the 204 children in efficacy of metronidazole. At 3 months af- the case group (14.7%) were still infected, ter treatment, 3 stool samples from each yielding an efficacy rate of 85.3%. To de- child in each group were examined to de-

Table 1 Asymptomatic Giardia lamblia infection among children aged 1–10 years in Sari

No. of Attending day Attending Total children care centres school

Total 490 2010 2500 Girls 215 921 1136 Boys 275 1089 1364 Infected (%) 49 (10.0) 356 (17.7) 405 (16.2) Girls (%) 22 (10.2) 118 (12.8) 140 (12.3) Boys (%) 27 (9.8) 238 (21.8) 265 (19.4)

10 Rapid reinfection.pmd 75 10/25/2005, 1:22 AM 76 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

tect reinfection and spontaneous clearing, 3 0.52 (0.42), and of the controls, 0.46 months after treatment 3 stool samples (0.495), t = 1.3. Neither Z-score was sta- from each child in the case and control tistically significant. groups were examined. We found that 90 children in the case group were infected, i.e. 30 remaining infections from ineffec- Discussion tive treatment and 60 reinfections (reinfec- In our study, the asymptomatic giardia in- tion rate = 60 of 174, or 34.5%), and that fection rate was 16.2% among apparently 71 of the 201 children in the control group healthy children aged 1–10 years in Sari, were clear of infection (35.3% spontane- Islamic Republic of Iran. The prevalence ous clearing). The relative risk for reinfec- rate was not significantly different between tion was 1.89 (95% confidence interval = high and low socioeconomic classes and 1.59–2.26); for spontaneous clearing, rela- between sexes, but increased with increas- tive risk was 8.81 (df = 1, P < 0.01). After ing age (10% at ages 1–5 years to 17.7% at 6 months weight and height were charted ages 6–10 years). This indicated that giar- again. Mean (standard deviation) Z-score dia infection was as common as in other for weight of the cases was 0.115 (0.78), parts of the country, e.g. 25.8% for chil- and of the controls, 0.163 (0.576), t = 0.7. dren aged 1–10 years in Tehran and Mean Z-score for height of the cases was

Table 2 Outcomes for 405 children with asymptomatic Giardia lamblia infection treated with metronidazole (cases) or with B-complex syrup (controls)

Parameter Time Cases Controls (n = 204) (n = 201)

Girls (No.) 62 78 Boys (No.) 142 123 S/E positivitya (%) 3 weeks after treatment 30 (14.7%)b 127 (37%)c 3 months after treatment 60 (34.5%)d 130 (35.3%)c Mean weight in kg (No.) Before treatment 27.7 (204) 26.3 (201) 6 months after treatment 30.7 (183) 29.2 (188) Mean weight Z-score (SD) 6 months after treatment 0.115 (0.78) 0.163 (0.576), t = 0.7 (NS) Mean height in cm (No.) Before treatment 131.8 (204) 127.75 (201) 6 months after treatment 138.6 (183) 132.96 (188) Mean height Z-score (SD) 6 months after treatment 0.52 (0.42) 0.46 (0.495), t = 1.3 (NS)

a3 stool samples were examined for each child. bMetronidazole efficacy rate = 85.3%. cSpontaneous clearing rate. dReinfection rate = 34.5% of the 174 cases that were effectively treated. The number of children examined before and 6 months after differ as some were lost to follow-up. NS = not significant. SD = standard deviation.

10 Rapid reinfection.pmd 76 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 77

26.79% for children aged 6–12 years in Guatemalan children were not seen in our Shiraz (both examined 3 samples) [11,20]. study (the differences between our case It was also as common as in other coun- and control groups were not statistically tries, such as 21% for children in day care significant) [8,9]. It was, however, compa- centres and 26% for children aged under 2 rable with other studies that found no years in the United States of America and effect on the nutritional status of asymp- 95% for Peruvian children [17,18]. tomatic giardia-infected children [3,10,11]. The children in our study did not have In our study, the efficacy rate of met- gastrointestinal complaints attributable to ronidazole was 85.3%. Few drugs are giardia infection for 1 month before and 3– available for the treatment of giardiasis and 6 months after treatment. This indicated they include metronidazole, tinidazole, that most children tolerated giardia infec- furazolidone, quinacrine hydrochloride and tion well, especially in cases of reinfection, paromomycin and their efficacy rates vary as in prospective studies of day care centre from 60% to 100%. All occasionally have children, of rural Egyptian children and severe side-effects [12,22,23]. others [3,4,17,19]. Most new infections were not associated with diarrhoea and were well tolerated. Conclusion The reinfection rate 3 months after successful therapy was 34.5%. This was sim- Considering the 15% failure rate and proba- ilar to Shiraz, a hyperendemic area similar ble side-effects of therapy, the high rate of to Sari, where 3 months after therapy the reinfection, the high rate of spontaneous reinfection rate was 24.5% with a 1-year clearing, the lack of effects on nutritional cumulative reinfection rate of 97.6% [11]. status and growth rate, the strong tolerance It was also similar to Lima, Peru, where 6 of reinfection by children, and the econom- months after treatment the reinfection rate ic impact of treating all cases in developing was 98% [18]. These studies indicate that countries with limited resources, we do not the chance of reinfection for a child who recommend the routine treatment of all has been successfully treated is high. asymptomatic giardia-infected children Spontaneous clearing of infection with- [9,23]. Treatment is required if an infected out treatment after 3 months was 35.3% in child exhibits persistent or intermittent, un- our study and was comparable to or higher specific gastrointestinal symptoms, signs than other studies [21,11]. These studies of malabsorption or impaired growth, poor have suggested that many people with giar- weight gain, poor nutritional status and dia infection will be cleared of infection malnutrition even in the absence of other spontaneously. gastrointestinal symptoms. Treatment may The positive effects of anti-giardia ther- also be indicated for prevention of trans- apy on nutritional status and growth rates mission to close contacts who are at risk of as in studies of Indian children and of developing severe giardiasis [13,16].

References

1. Keating JP. Giardiasis. In: Feigin RD, 2. Ortega YR, Adam RD. Giardia: Overview Cherry JD, eds. Textbook of pediatric in- and update. Clinical infectious diseases, fectious diseases, 4th ed. Philadelphia, 1997, 25:245–50. WB Saunders, 1998:2400–3.

10 Rapid reinfection.pmd 77 10/25/2005, 1:22 AM 78 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

3. Pickering LK et al. Occurrence of Giardia 14. Spencer MJ et al. Parasitic infections in a lamblia in children in day care centers. pediatric population. Pediatric infectious Journal of pediatrics, 1984, 104(4):522– disease journal, 1983, 2(2):110–3. 6. 15. Hopkins RS, Juranek DJ. Acute giardia- 4. Rauch AM et al. Longitudinal study of sis: an improved clinical case definition Giardia lamblia infection in a day care for epidemiologic studies. American center population. Pediatric infectious journal of epidemiology, 1991, 133:402– disease journal, 1990, 9(3):186–9. 7. 5. Hill DR. Giardia lamblia. In: Mandell GL, 16. Katz M. Asymptomatic giardia infection: Bennett JE, Dolin R, eds. Mandell, Dou- letters to the editor. Pediatric infectious glas and Bennett’s: Principles and prac- disease journal, 1992, 11(3):248–9. tice of infectious diseases, 5th ed. 17. Bartlett AV et al. Controlled trial of Giar- Philadelphia, Churchill Livingstone, dia lamblia: control strategies in day care 2000:2888–94. centers. American journal of public 6. Wolfe MS. Giardiasis. Clinical microbiol- health, 1991, 81(6):1001–6. ogy reviews, 1992, 5:93–100. 18. Gilman RH et al. Rapid reinfection by 7. Hill DR. Giardiasis: Issues in diagnosis Giardia lamblia after treatment in a hy- and management. Infectious disease perendemic third world community. Lan- clinics of North America, 1993, 7:503– cet, 1988, 13:343–5. 25. 19. Sullivan PS et al. Illness and reservoirs 8. Gupta MC et al. Effect of giardia infection associated with Giardia lamblia infection on nutritional status of preschool chil- in rural Egypt: the case against treatment dren. Indian journal of medical research, in developing world environments of 1990, 92:341–3. high endemicity. American journal of epidemiology, 1988, 127(6):1272–81. 9. Gupta MC, Urrutia JJ. Effect of periodic antiascaris and antigiardia treatment on 20. Shirbazoo Sh, Aghamiri H. Giardia nutritional status of preschool children. lamblia infection prevalence in Tehran- American journal of clinical nutrition, ian children. Koosar medical journal, 1982, 36:79–86. 2000, 5(2):105–9 [in Persian]. 10. Ish-Horowicz M et al. Asymptomatic giar- 21. Levi GC, De Avila CA, Neto VA. Efficacy diasis in children. Pediatric infectious of various drugs for treatment of giardia- disease journal, 1989, 8(8):773–9. sis: A comparative study. American journal of tropical medicine and hygiene, 11. Alborzi A, Zerafati Z. Asymptomatic giar- 1977, 26:564–5. dia infection in children in a hyperendemic area. Iranian journal of medical 22. Davidson RA. Issues in clinical parasitol- sciences, 1994, 19(1,2):1–6. ogy: the treatment of giardiasis. Ameri- can journal of gastroenterology, 1984, 12. Nash TE. Treatment of Giardia lamblia in- 79(4):256–61. fections. Pediatric infectious disease journal, 2001, 20(2):193–5. 23. Murphy TV, Nelson JD. Five vs. ten days therapy with furazolidone for giardiasis. 13. Addiss DG, Juranek DD, Spencer HC. American journal of diseases of chil- Treatment of children with asymptomatic dren, 1983, 137:267–70. and nondiarrheal giardia infection. Pedi- atric infectious disease journal, 1991, 10(11):843–6.

10 Rapid reinfection.pmd 78 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 79 Asthma prevalence and severity among primary-school children in Baghdad D. Al-Thamiri,1 W. Al-Kubaisy 2 and S.H. Ali 1

ABSTRACT To measure asthma prevalence and severity among children in Baghdad, we randomly sampled 3360 primary-school children and had their parents complete standardized questionnaires. The response rate was 86%, male to female ratio was 0.75:1 and age range was 6–12 years. Prevalence of wheezing ever was 25.0%. Wheezing during the last 12 months was 19.9% and 2.9% of the children had more than 12 attacks. Parents reported nocturnal wheezing for 16.3% of the children and severe attacks that limited speech for 10.5%. Prevalence of asthma ever was 22.3%. Asthma was detected in 81.9% of those with wheezing in the last 12 months. Males were predominant among children with wheezing ever, whereas females were predominant among children with asthma ever. Prevalence rates of asthma and of severe asthma symptoms decreased with increasing age.

Prévalence et sévérité de l’asthme chez des écoliers du primaire à Bagdad RÉSUMÉ Afin de mesurer la prévalence et la sévérité de l’asthme chez les enfants à Bagdad, nous avons sélectionné de manière aléatoire un échantillon de 3360 écoliers du primaire et avons fait remplir des questionnaires standardisés à leurs parents. Le taux de réponse était de 86 %, le rapport des sexes masculin/féminin était de 0,75:1 et l’âge était compris entre 6 et 12 ans. La prévalence des sifflements respiratoires au cours de la vie était de 25,0 % ; au cours des douze derniers mois elle s’élevait à 19,9 % et 2,9 % des enfants avaient eu plus de 12 crises. Les parents ont rapporté des sifflements nocturnes pour 16,3 % des enfants et des crises sévères limitant la parole pour 10,5 %. La prévalence de l’asthme au cours de la vie était de 22,3 %. L’asthme a été dépisté chez 81,9 % de ceux ayant eu des sifflements respiratoires au cours des 12 derniers mois. Une prédominance masculine a été notée chez les enfants ayant eu des sifflements respiratoires au cours de la vie, tandis qu’on retrouve une prédominance féminine chez les enfants ayant souffert d’asthme au cours de la vie. Les taux de prévalence de l’asthme et de symptômes d’asthme sévère diminuaient lorsque l’âge augmentait.

1Department of Paediatrics; 2Department of Community Medicine, Iraqi College of Medicine, Baghdad, Iraq (Correspondence to W. Al-Kubaisy: [email protected]). Received: 03/03/03; accepted: 09/12/03

11 Asthma prevalence.pmd 79 10/25/2005, 1:22 AM 80 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction prevalence of asthma among schoolchildren in Iraq. Our study, therefore, estimat- Childhood asthma is a major clinical con- ed the prevalence of asthma and wheezing cern worldwide and represents a huge bur- among primary-school children and evalu- den on families and societies. It accounts ated the severity of asthma symptoms in for a large number of lost school days and our study group. may deprive children of both academic achievement and social interaction. Child- hood asthma also places a strain on health- ›Methods care resources as a result of doctor and hospital visits and cost of treatment [1]. We conducted a cross-sectional study of The prevalence of childhood asthma has primary-school children in Baghdad be- been reported to vary (1%–30%) in differ- tween October 2000 and June 2002. We ent populations [2]. These variations are obtained from the Ministry of Education a possibly due to differing exposures to res- list of the 1494 primary schools throughout piratory infections, indoor or outdoor pol- Baghdad, in both urban and rural areas. The lution and diet [1,2]. schools were of 3 types: for girls only, for Genetics and lifestyle and environmen- boys only, or coeducational. The number tal factors may also play a role in these vari- of students in these schools at the time of ations [3]. The prevalence of asthma seems our study was 1 002 004. Official approval to be higher in affluent than in non-affluent was obtained from the appropriate authori- populations [2,4]. The prevalence of child- ties prior to the study. We randomly chose hood asthma is increasing worldwide and 20 schools in both urban and rural regions consequently morbidity, mortality and cost of any of the 3 types of primary schools. A of care continue to increase [2,4,5]. The sample of 3360 primary-school children combination of the genetic basis of asthma was then selected by a cluster random and indoor allergens like dust mites and at- method. mospheric dusts to the increasing preva- Letters were sent to the parents of the lence of asthma has been reported [6]. In chosen children. The letters explained the the United States of America, 15 million purpose of the study and directions for the people have asthma, and of them, nearly attached questionnaire. We also visited the 4.8 million are children [5]. Multiple studies schools on Parent’s Day and further ex- support the “hygienic hypothesis” in asth- plained the questions and the directions. ma development; this hypothesis states that The standardized questionnaire that was childhood asthma develops as a result of distributed to the 3360 children asked their decreased exposure to infectious agents parents about risk factors for asthma, trig- during infancy and early childhood, result- gering factors for wheezing attacks and ing in the persistence of neonatal T helper seasonal variations. The questionnaire also lymphocyte 2 immunophenotype, thereby covered characteristics including birth pro- predisposing the child to atopic diseases cess, birth weight, feeding pattern and du- [7]. ration of breastfeeding, family history of Although many have studied asthma in asthma, level of education of parents and Iraq, the epidemiology of this disease is still cigarette-smoke exposure. Another section not fully understood. To the best of our included questions about drugs used by the knowledge, no study has yet described the child in the last 12 months. To distinguish

11 Asthma prevalence.pmd 80 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 81

between asthma and wheezing, we depend- returned, i.e. the response rate was 86.0%. ed upon responses to questions about the The study population ranged in age from 6 types of medications used for the child’s to 15 years, was in primary grades 1–6 and asthma during the last 12 months. We included 1242 boys (43%) and 1647 girls asked parents to respond with “yes” or (57%). The male to female ratio was “no” to the key question in the question- 0.75:1. naire to confirm the child as asthmatic: Table 1 shows the prevalence of asthma “Had their child ever been diagnosed by a and asthma symptoms. Wheezing ever was physician as having asthma?” In addition to reported for 723 children (25.0%), 574 this, we contacted parents who answered children (19.9%) had wheezing in the last “yes” and confirmed the diagnosis of asth- 12 months and only 84 (2.9%) reported ma with the type of medication used by the more than 12 wheezing attacks in the past child. These questions were added to the year. Nocturnal wheezing attacks disturb- first part of the International Study of Asth- ing sleep were reported for 470 (16.3%) ma and Allergies in Childhood (ISAAC) children and only 88 (3.1%) had more than core questionnaire, which was translated 4 attacks per month disturbing their sleep. into Arabic [8]. Severe wheezing attacks that limited We defined severe asthma as cases that speech were reported for 304 children met one or more of the following criteria: (10.5%). more than 12 wheezing attacks in the last The prevalence rate of wheezing ever 12 months; sleep disturbed by wheezing at- was significantly higher among boys tacks more than 4 times per month; or (28.3%) than girls (22.6%, ÷2 = 21.020, P wheezing attacks that limited speech [5,9]. = 0.007). Girls, however, had significantly Non-severe asthma was defined as meeting higher prevalence of asthma ever (22.5%) one or both of these criteria: less than 12 than boys (22.1%, ÷2 = 18.110, P = 0.034). wheezing attacks in the last 12 months or Girls also had significantly higher preva- sleep disturbed by wheezing attacks less lence rates of severe asthma symptoms, in- than 4 times per month [5,9]. cluding more than 12 wheezing attacks in Our limitations were that the authors the last 12 months (3.1% versus 2.6%) and themselves did not clinically examine the nocturnal attacks disturbing sleep more children because of the political situation in than 4 times per month (3.4% versus Iraq during the study and that there was 2.5%). The sex difference was not signifi- recall bias of the parents. cant for the prevalence of wheezing in the The data were analysed with SPSS, ver- last 12 months (Table 1). sion 10.0 and were presented in simple Physicians had previously diagnosed measures of frequency (%). The signifi- asthma in 644 children (22.3%). Of 723 cance of difference between proportions children with a history of wheezing ever, was calculated with chi-squared test, with physicians had diagnosed 171 (23.7%) P-value < 0.05 as the level of significance. asthma cases, whereas 473 (21.8%) of 2166 children without a history of wheezing were diagnosed with asthma. In con- Results trast, 81.9% of those with wheezing in the last 12 months (470 of 574 children) had Of the 3360 questionnaires that we sent been diagnosed by a physician as asthmatic out, 2889 completed questionnaires were (Table 2).

11 Asthma prevalence.pmd 81 10/25/2005, 1:22 AM 82 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 1 Prevalence of asthma and asthma symptoms by sex

Symptom Boys Girls Total P-value χ 2 (n = 1242) (n = 1647) (n = 2889) No. (%) No. (%) No. (%)

Wheezing ever 351 (28.3) 372 (22.6) 723 (25.0) 0.007* 21.020 Wheezing in last 12 months 270 (21.7) 304 (18.4) 574 (19.9) 0.055 16.630 < 4 episodes 149 (12.0) 155 (9.4) 304 (10.5) 0.366 4–12 episodes 88 (7.1) 98 (5.9) 186 (6.4) 0.066 > 12 episodes 33 (2.6) 51 (3.1) 84 (2.9) 0.000* Sleep disturbances 219 (17.6) 251 (15.2) 470 (16.3) 0.336 10.179 < 2 disturbances/month 126 (10.1) 135 (8.2) 261 (9.0) 0.023 3–4 disturbances/month 62 (5.0) 59 (3.6) 121 (4.2) 0.116 > 4 disturbances/month 31 (2.5) 57 (3.4) 88 (3.1) 0.004* Wheezing attacks limiting speech 147 (11.8) 157 (9.5) 304 (10.5) 0.171 11.575 Asthma ever 274 (22.1) 370 (22.5) 644 (22.3) 0.034* 18.110

*Significant at P < 0.05.

Table 3 gives the prevalence of asthma 6–7 years to 14.3% and 14.3% respectively and asthma symptoms by age. The preva- for those aged 14–15 years. lence of wheezing ever increased with in- Table 4 shows the distribution of asth- creasing age. The numbers of children who ma severity according to age. The propor- had experienced wheezing attacks during tion of children who experienced more than the last year and children with physician- 12 wheezing attacks during the last year diagnosed asthma, however, steadily de- varied little with age, while both prevalence creased with increasing age from 35.0% rates of nocturnal wheezing attacks that and 40.6% respectively for children aged disturbed sleep and attacks that limited speech decreased with increasing age from 25.9% and 21.2% respectively for children aged 6–7 years to 8.2% and 6.1% for those Table 2 Percentage distribution of children aged 14–15 years. with asthma ever

Category Number of Asthma Discussion children evera No. (%) Our study is the first nationally to measure asthma prevalence in Iraq. The higher Total 2889 644 (22.3) number of females in the study population Wheezing ever 723 171 (23.7) can probably be attributed to the effects of Without wheezing 2166 473 (21.8) the embargo at the time of our study as Wheezing in the last many boys had left school to work. It is 12 months 574 470 (81.9) evident that both wheezing and asthma are major public health problems as their prev- aAsthma ever as confirmed by the parent’s report of physician diagnosis.

11 Asthma prevalence.pmd 82 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 83

Table 3 Prevalence of asthma and asthma symptoms by age of primary-school children

Age group Population Wheezing Wheezing in Asthma (years) ever last 12 months ever No. (%) No. (%) No. (%)

6–7 320 18 (5.6) 112 (35.0) 130 (40.6) 8–9 582 39 (6.7) 142 (24.4) 158 (27.1) 10–11 1055 275 (26.1) 174 (16.5) 184 (17.4) 12–13 883 380 (43.0) 139 (15.7) 165 (18.7) 14–15 49 11 (22.4) 7 (14.3) 7 (14.3) Total 2889 723 (25.0) 547 (19.9) 644 (22.3)

alence rates are high (25.0% and 22.3%, thermore, it was higher than in Russia respectively). (5.1%), Malaysia (10.3%), India (15.7%) Although asthma prevalence in our and the Republic of Korea (8.7%) [15,21– study was similar to prevalence in Saudi 23]. Asthma prevalence in Iraq, however, Arabia (23%), it was higher than in other was lower than rates in the United Kingdom nearby countries like Turkey (14.1%), (29.7%) and the USA (23.6%) [24,25]. United Arab Emirates (UAE) (13%), Leba- Prevalence of wheezing ever was much non (12%) and Jordan (4.1%) [3,10–13]. lower than in some countries: France Moreover our prevalence was much higher (25.4%), the United Kingdom (35.8%) and than in many European countries: France the USA (34.5%), while it was higher than (14.9%), Norway (8.6%), Italy (4.6%), in Turkey (22.4%), the UAE (15.6%), Leb- Sweden (7%) and Spain (10%) [14–18]. It anon (23%), Jordan (8.3%), Italy (23.2%), was also higher than in Latin America Malaysia (12.5%) or India (20.8%) [3,11– (16.5%) and Brazil (20.4%) [19,20]. Fur- 14,16,21,22,24,25]. These differences in

Table 4 Prevalence of severe asthma symptoms by age of primary- school children

Age Number > 12 wheezing > 4 sleep Attacks (years) attacks disturbances limiting speech No. (%) No. (%) No. (%)

6–7 320 11 (3.4) 83 (25.9) 68 (21.2) 8–9 582 16 (2.7) 115 (19.7) 82 (14.1) 10–11 1055 26 (2.5) 150 (14.2) 78 (7.4) 12–13 883 29 (3.3) 118 (13.4) 73 (8.3) 14–15 49 2 (4.1) 4 (8.2) 3 (6.1) Total 2889 84 (2.9) 470 (16.3) 304 (10.5)

11 Asthma prevalence.pmd 83 10/25/2005, 1:22 AM 84 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

prevalence of asthma and wheezing in var- Kingdom (2.2%) and Lebanon (1%) ious regions of the world can be explained [8,12]. Likewise, of the children in our by the variability of environmental and ge- study who experienced attacks that dis- netic factors that predispose children to turbed sleep, only 3.1% were awakened wheezing [12]. more than 4 times per month. Studies in the Wheezing attacks in the last 12 months United Kingdom and France reported were reported for 19.9% of the children in wheezing attacks disturbing sleep more our study. This percentage was lower than than 1–2 times per week as 4.3% and 2.2% the prevalence rate of wheezing ever respectively [8,14]. Rates of wheezing at- (25.0%) and was similar to studies from tacks that disturbed sleep over 12-month different regions [3,12,14,16,21,22,24, periods were 10% in Turkey, 9% in Leba- 25]. This trend reflected the tendency of non and 23.2% in a city in Brazil [3,12,20]. wheezing to resolve as children grow older Several regions have reported frequen- [12]. Moreover, the majority of our cases cies of wheezing attacks that limited (470 of 574 children) who experienced speech ranging from 2.3% to 9.6%, i.e. wheezing during the last 12 months were frequencies lower than in our study diagnosed as asthmatics (81.9%) (Table 2). (10.5%) [3,8,12,14,20]. This variation This emphasizes the importance of wheez- could explained as either true severe epi- ing as a major clinical presentation of asth- sodes limiting speech or as overestimations ma. A similar study in Lebanon reported of the attacks by parents. One striking find- asthma in 51% of children with wheezing ing in our study was the significantly higher ever and in 85% of children with wheezing prevalence of asthma ever and severe asth- in the last 12 months [12]. Although only ma symptoms for girls than for boys. about 1% of children with asthma have se- Many regional studies have shown male vere asthma, its prognosis is very worri- predominance in prevalence of wheezing some as 90% of children with severe and asthma ever, whereas studies from asthma continue with asthma into adult- Turkey, Lebanon and India found similar hood [18]. In 1997, National Asthma Edu- rates of asthma and wheezing ever for both cation and Preventive Program (NAEPP) sexes [3,8,12–14,17,22]. Our clinical ob- guidelines recommended routine monitor- servation is that girls tend to overestimate ing of asthma patients to identify signs and asthma symptoms more frequently than symptoms for the classification of asthma boys which may explain the female pre- severity. Frequent exacerbations and at- dominance in our study. Similarly, in the tacks, limited speech and night-time symp- United Kingdom, one study identified a de- toms were designated as indicators of crease in male to female ratio with increas- severe asthma [5,9]. ing age [8]. Although 19.9% of the children in our In our study, the prevalence of asthma study experienced wheezing in the last 12 and wheezing in last 12 months decreased months, only 2.9% were severe cases who with increasing age, as it did in other stud- experienced more than 12 wheezing at- ies [8,23]. It has been estimated that tacks during last 12 months. Similarly stud- wheezing and asthma improve with age and ies in Turkey and France reported frequent that remission of asthma is about 50% [9]. attacks (> 12 attacks) with rates of 3.3% We, however, found that older children had and 4.5% respectively [3,14]. Lower fig- higher rates of wheezing ever, similar to a ures have been reported from the United study from India [22]. This may be ex-

11 Asthma prevalence.pmd 84 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 85

plained by incomplete recall of past wheez- tries. Significant sex differences were re- ing episodes. Prevalence of wheezing at- flected by male predominance in wheezing tacks limiting speech and disturbing sleep ever and female predominance in asthma. in our study markedly decreased with in- Both asthma prevalence and severe asthma creasing age (Table 4). This emphasizes symptoms decreased steadily with increas- that these attacks are better indicators of ing age. severity than are frequency of attacks. Fre- Further studies in other governorates quency of attacks showed no constant sig- are needed to determine the prevalence of nificant differences with age among our asthma and wheezing in the various geo- cases. Decreasing asthma severity with in- graphical regions of our country. creasing age was confirmed by one large study in the United Kingdom [8]. Acknowledgements Conclusions This work was funded in part by the Minis- try of Education. Our deepest gratitude and Asthma is a major public health problem appreciation are given to His Excellency, Dr among Iraqi children in Baghdad. Preva- Fahad Al-Shagra, Minster of Education, for lence rates of asthma and wheezing ever in his support and efforts. We would also like Baghdad are comparable to industrialized to thank Dr Khadim Al-Khazraji and his and developed countries. Occurrences of staff for their assistance and participation severe asthma symptoms were similar or in this study. higher but not lower than in other coun-

References 1. Von-Mutius E. The burden of childhood 6. Mochizuki H, Morikawa A. Childhood asthma. Archives of disease in child- asthma. Nippon–Rinsho, 2001, 59(10): hood, 2000, 82(suppl. 1):2112–5 [ab- 1919–24 [abstract]. stract]. 7. Tantisira KG, Weiss ST. Childhood infec- 2. Partridge MR, Alwan A. Prevention of tions and asthma at the crossroads of the asthma and approaches for enhanced hygiene and Barker hypothesis. Respi- care in the Eastern Mediterranean Re- ratory research, 2001, 2(6):324–7 [ab- gion. Eastern Mediterranean health jour- stract]. nal, 1997, 3(1):133–43. 8. Strachan DP et al. A national survey of 3. Ece A et al. Prevalence of asthma and asthma prevalence, severity, and treat- other allergic disorders among school- ment in Great Britain. Archives of dis- children in Diyarbakir, Turkey. Turkish ease in childhood, 1994, 70:174–8. journal of pediatrics, 2001, 43:286–92. 9. Sly M. Asthma. In: Behrman RE, 4. Al-Dawood KM. Epidemiology of bron- Kliegman RM, Jenson HB, eds. Nelson chial asthma among school boys in Al- textbook of pediatrics, 16th ed. Philadel- Khobar City, Saudi Arabia. Saudi phia, WB Saunders Company, 2000: medical journal, 2001, 22(1):61–6. 664–80. 5. Rorstad A, Bhatt-Mehta V. Outpatient 10. Al-Frayh AR et al. Increased prevalence management of asthma in children. of asthma in Saudi Arabia. Annals of al- Pharmacy times, 2000 May:55–66.

11 Asthma prevalence.pmd 85 10/25/2005, 1:22 AM 86 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

lergy, asthma and immunology, 2001, Allergologia et immunopathologia, 86(3):292–6 [abstract]. 1999, 27(2):46–53 [abstract]. 11. Al-Maskari F et al. Asthma and respira- 19. Mallol J et al. Prevalencia del asma en tory symptoms among schoolchildren in escolares chilenos. Estudio descriptivo United Arab Emirates. Allergie et de 24.470 ninos. ISAAC–Chile. [Preva- immunologie, 2000, 32(4):159–63 [ab- lence of asthma in Chilean students. De- stract]. scriptive study of 24 470 children. ISAAC–Chile.] Revista medica de Chile, 12. Ramadan FM et al. Prevalence of 2000, 128(3):279–85 [abstract]. asthma and asthma symptoms in children in urban Lebanon. Saudi medical 20. De Britto MC et al. Asthma prevalence in journal, 1999, 20(6):453–7. schoolchildren in a city in north-east Bra- zil. Annals tropical paediatrics, 2000, 13. Abuekteish F et al. Prevalence of asthma 20(2):95–100 [abstract]. and wheeze in primary-school children in Northern Jordan. Annals of tropical 21. Norzila MZ et al. Prevalence of child- pediatrics, 1996, 16:227–31. hood asthma and allergy in an inner city Malaysian community: Intra-observer re- 14. Fontaine V et al. Epidemiology of child- liability of two translated international hood asthma in the department of questionnaires. Medical journal of Ma- Calvados. Revue de pneumologie laysia, 2000, 55(2):33–9 [abstract]. clinique, 1999, 55(1):5–11 [abstract]. 22. Chhabra SK et al. Prevalence of bron- 15. Selnes A et al. Asthma and allergy in chial asthma in schoolchildren in Delhi. Russian and Norwegian schoolchildren. Journal of asthma, 1998, 35(3):291–6 Results from two questionnaire-based [abstract]. studies in the Kola Peninsula, Russia, and northern Norway. Allergy, 2001, 23. Lee SI et al. Prevalence symptoms of 56(4):344–8 [abstract]. asthma and other allergic diseases in Korean children: a nationwide question- 16. Peroni DG et al. Prevalence of asthma naire survey. Journal of Korean medical and respiratory symptoms in childhood sciences, 2001, 16(2):155–64 [abstract]. in an urban area of north-east Italy. Monaldi archives for chest disease, 24. Ng Man Kwong G et al. Increasing preva- 1998, 53(2):134–7 [abstract]. lence of asthma diagnosis and symptoms in children is confined to mild 17. Ronmark E et al. Asthma, type I allergy symptoms. Thorax, 2001, 56(4):312–4 and related conditions in 7 and 8 year [abstract]. old children in northern Sweden. Respi- ratory medicine, 1998, 92(2):316–24 25. Hu HB et al. Prevalence of asthma and [abstract]. wheezing in public schoolchildren: association with maternal smoking during 18. Garde-Garde JM, Medina Pomares J. pregnancy. Annals of allergy, asthma Mesa Redonda: Asma grave en and immunology, 1997, 79(1):80–4 [ab- Pediatría: Diagnostico y pronóstico. stract]. [Round Table: Severe asthma in paediatrics: diagnosis and prognosis.]

11 Asthma prevalence.pmd 86 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 87 Comparison of plasma and leukocyte vitamin C status between asthmatic and healthy subjects F. Shidfar,1 N. Baghai,2 A. Keshavarz,2 A. Ameri1 and S. Shidfar 3

ABSTRACT In a case–control study the vitamin C status of 50 adults with chronic controlled asthma was compared with that of 50 randomly selected healthy controls. Vitamin C intake was assessed by 3-day dietary recall, and plasma and leukocyte vitamin C concentrations were measured colorimetrically. A positive significant correlation was found between plasma vitamin C and dietary vitamin C intake. Plasma and leukocyte vitamin C levels were significantly lower in the asthma group. Plasma vitamin C was deficient (< 0.4 mg/dL) in significantly more patients than controls (38.0% versus 0%) and leukocyte vitamin C (< 20 µg/108 leukocytes) was deficient in 92.0% of asthmatics versus 8.0% of controls. A significant association was observed between duration of asthma and plasma vitamin C level.

Comparaison du statut en vitamine C plasmatique et leucocytaire entre des patients asthmatiques et des sujets sains RÉSUMÉ Dans une étude cas-témoins, le statut en vitamine C de 50 adultes asthmatiques chroniques ayant un asthme contrôlé a été comparé à celui de 50 témoins sains choisis au hasard. L’apport en vitamine C a été évalué par rappel des 24 heures sur 3 jours et les concentrations plasmatiques et leucocytaires en vitamine C ont été mesurées par colorimétrie. Une corrélation significative positive a été trouvée entre la vitamine C plasmatique et l’apport alimentaire en vitamine C. Le taux de vitamine C plasmatique et leucocytaire était significativement moins élevé dans le groupe des patients asthmatiques. Il y avait une déficience en vitamine C plasmatique (< 0,4 mg/dL) chez un nombre significativement plus important de patients que de témoins (38,0 % vs 0 %) et en vitamine C leucocytaire (< 20 µg/108 leucocytes) chez 92,0 % d’asthmatiques contre 8,0 % de témoins. Une association significative a été observée entre la durée de l’asthme et le taux de vitamine C plasmatique.

1Department of Nutrition, School of Health, Iran University of Medical Sciences, Tehran, Islamic Republic of Iran (Correspondence to F. Shidfar: [email protected]). 2Department of Nutrition, School of Health, University of Tehran, Tehran, Islamic Republic of Iran. 3Memorial Hospital, University of Massachusetts, Worcester, Massachusetts, United States of America. Received: 24/09/03; accepted: 26/02/04

12 Comparison of plasma.pmd 87 10/25/2005, 1:22 AM 88 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction lamic Republic of Iran. To further investigate the subject, we made a case–control Asthma is a chronic relapsing inflammatory study in Tehran with the hypothesis that disorder of the airways and a major health plasma and leukocyte vitamin C level are problem worldwide [1]. The prevalence lowered in asthmatic patients and that the and morbidity rate of asthma have in- duration of asthma affects plasma and leu- creased in the past decade, despite kocyte vitamin C levels. improved knowledge about its pathophysi- ology and treatment. As inflammatory cells generate and release reactive oxygen spe- Methods cies, asthmatic airways are liable to oxidative stress [2]. Moreover, inflammatory A case–control study was conducted to cells from asthmatic patients generate more compare the vitamin C status of asthmatics reactive oxygen species than those from and healthy people at Massih Daneshvari controls. The extent of oxidative stress will Hospital in Tehran, the Islamic Republic of depend, in part, on the antioxidant defences Iran, between October 2002 and January available within the respiratory tract lining 2003. fluid [3]. It has been suggested that antioxidants Subjects might have an etiologic role in asthma and, Among the asthmatic patients attending the if so, this could lead to the development of hospital, 50 non-smoking asthmatic pa- new therapeutic strategies. The data are tients (24 men and 26 women) were ran- strongest for vitamin C, which is one of the domly selected using a random number key antioxidant vitamins [4]. It is abundant table. Their mean [standard error (SE)] du- in the extracelluar fluid lining the lung, and ration of asthma was 8.0 (1.3) years. The adequate vitamin C intake has been associ- diagnosis of asthma was established from a ated with the protective effects of airways documented clinical history of recurrent responsiveness and lung function [1,5–7]. cough and/or wheezing and previous dem- Vitamin C reduces the number and severity onstration of improvement of symptoms in of attacks in patients with asthma and re- response to asthma medication. Subjects duces the severity of the bronchial re- were selected after each patient had been sponses to exercise [8]. In the search for a assessed clinically. The asthma was con- possible relationship between vitamin C and sidered to be under control if no changes asthmatic symptoms, both plasma and leu- were deemed necessary to their current kocyte levels of vitamin C have been stud- medication after consultation. Patients who ied [9–13]. There are hints that asthmatic had taken oral steroids in the 4 weeks be- subjects have low plasma and leukocyte fore the study were excluded. All patients concentrations of vitamin C but the rela- were intermittently taking inhaled short- tionship between vitamin C levels and dura- acting beta-adrenergic agents as the sole tion of asthma has not been demonstrated treatment. These patients were under ob- [10–12]. On the other hand, the epidemio- servation and followed up through periodic logical evidence about the role of dietary visits in the hospital for 2 months. vitamin C in asthma is controversial Among all of the hospital employees, 50 [10,11]. healthy non-smoking persons (26 men and No studies of the role of vitamin C in 24 women) were selected for a control asthma have yet been conducted in the Is- group. None of the controls was taking vi-

12 Comparison of plasma.pmd 88 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 89

tamin supplements and none had clinical forearm (2 cm3 into small tubes containing evidence of an acute respiratory infection. EDTA and 8 cm3 into large tubes containing All of the subjects (patients and con- a separator solution). The plasma was sep- trols) were matched for age, weight, height arated by centrifugation of the small tubes and other variables and were informed at 1000 g for 10 min at 4 °C and plasma about the aims and possible benefits that was pipetted into separate, capped vials for could be derived from the study. Informed vitamin C analyses. To isolate the leuko- written consent was obtained from each cytes, after precipitation of the red blood subject. The study protocol was approved cells, the supernatant was centrifuged at by the local ethical committee. 800–1000g for 5 min in large tubes and the precipitate was washed with distilled water Questionnaires and saline solution. Plasma and leukocyte A single 3-day dietary recall questionnaire samples were deproteinized with a trichlo- which has been validated previously against roacetic acid solution for the vitamin C as- weighed food intakes and in epidemiologi- says. cal studies of heart disease was used to es- Vitamin C in the protein-free superna- timate dietary intake [14,15]. This tant was determined by a colorimetric questionnaire was administrated by trained method using 2,4-dinitrophenylhydrazine personnel possessing at least a bachelor’s [17]. The cut-off values for deficient plas- degree in food and nutrition. The 3-day di- ma and leukocyte vitamin C levels were < etary recall was conducted using a stan- 0.4 mg/dL and < 20 µg/108 leukocytes re- dardized protocol and 51 3-dimensional spectively, according to NHANES II [16]. models to estimate portion size. Interview- ers coded the dietary recall questionnaires Statistical analyses using nutrient information obtained from Dietary records were analysed using Food the US Department of Agriculture (USDA). processor II software [18,19] and statisti- The recommended dietary allowance cal analysis was carried out using SPSS for (RDA) of vitamin C is 60 mg daily and < 45 Windows, version 9. The t-test was used to mg (i.e. less than 75% of this level) is rec- compare age, weight, height, daily dietary ognized as deficient intake and ≥ 45mg is vitamin C and plasma and leukocyte vita- recognized as normal intake according to min C between groups. The chi-squared the National Health and Nutrition Examina- and Fisher tests were used to compare cat- tion Survey II (NHANES II) [16]. egorical data (sex, literacy, employment). A sociodemographic and clinical ques- Pearson correlation was used for determin- tionnaire that has been extensively validated ing the correlation between quantitative in our department and elsewhere [14] was variables. Simple regression analyses were completed for each subject; this collected used to determine whether age, sex or du- information about sex, literacy, employ- ration of asthma were possible predictor ment, income, suspected risk factors for variables with respect to plasma vitamin C asthma, other respiratory and allergic dis- levels. ease and drugs used. Results Laboratory analyses Non-fasting blood samples (10 cm3) were The baseline characteristics of the 2 groups drawn from the antecubital space of the confirmed that they were well matched

12 Comparison of plasma.pmd 89 10/25/2005, 1:22 AM 90 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

(Table 1). The mean (SE) age of patients < 0.001) (Table 2). Leukocyte vitamin C with asthma was 44.6 (2.2) years and of was deficient (< 20 µg/108 leukocytes) in controls was 44.3 (2.1) years. There were 92.0% of patients, but only 8.0% of con- no significant differences in weight, height trols (P < 0.001). or age. The frequency distribution of the The mean daily dietary intake of vitamin participants in terms of sex, employment C was deficient (< 75% of RDA) in 50.0% and literacy were similar and showed no of the asthma group and 32.0% of con- significant differences between the 2 trols, although the difference between groups. The health status indicated that the groups was not significant (Table 2). Plas- subjects were different only with respect to ma vitamin C was deficient in 47.4% of asthma and had no other chronic diseases. women and 52.6% of men, and leukocyte Although there was no significant dif- vitamin C was deficient in 52.0% of wom- ference in daily dietary vitamin C intake been and 48.0% of men. There was no signif- tween the 2 groups, there was a significant icant association of sex, health status, difference in mean (SE) plasma vitamin C literacy, employment or severity of asthma concentration: 0.7 (0.006) mg/dL for asth- (mild or severe) with plasma or leukocyte matics versus 1.15 (0.006) mg/dL for con- vitamin C status. trols (P < 0.001) (Table 1). There were also Table 3 compares the percentage of significant differences in the mean (SE) participants with deficient/normal vitamin leukocyte vitamin C levels: 9.6 (1.1) µg/108 C levels and deficient/normal vitamin C in- leukocytes for the asthma group and 31.2 take. There was a significant association (7.5) for the controls (P < 0.001) (Table 1). between plasma vitamin C status and di- The distribution of participants by vita- etary vitamin C intake in the 2 groups com- min C status showed that plasma vitamin C bined (P < 0.03) (Table 3), but leukocyte was deficient in 38.0% of patients (< 0.4 vitamin C status had no significant associa- mg/dL) compared with 0% of controls (P tion with dietary vitamin C intake.

Table 1 Baseline characteristics of participants

Variable Asthma group Control group (n = 50) (n = 50) Mean (SE) Mean (SE)

Age (years) 44.6 (2.2) 44.3 (2.1) Weight (kg) 67.8 (1.4) 68.4 (1.3) Height (cm) 165.7 (1.0) 166.9 (1.0) Vitamin C status Plasma vitamin C level (mg/dL) 0.7 (0.006)*** 1.15 (0.006) Leukocyte vitamin C level (µg/108 leukocytes) 9.6 (1.1)*** 31.2 (7.5) Vitamin C daily intake (mg) 71.1 (9.8) 82.8 (7.5)

***P < 0.001, independent t-test. SE = standard error.

12 Comparison of plasma.pmd 90 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 91

Table 2 Frequency distribution of participants in terms of plasma and leukocyte vitamin C status and dietary vitamin C intake

Vitamin C status Asthma group Control group (n = 50) (n = 50) No. % No. %

Plasma vitamin C status*** Deficient 19 38.0 0 0 Normal 31 62.0 50 100.0 Leukocyte vitamin C status*** Deficient 46 92.0 4 8.0 Normal 4 8.0 46 92.0 Vitamin C intake Deficient 25 50.0 16 32.0 Normal 25 50.0 34 68.0

***P < 0.001, Fisher test.

A significant positive correlation was association with plasma and leukocyte vita- observed between plasma vitamin C level min C but duration of asthma had a signifi- and dietary vitamin C intake in the asthma cant association with plasma vitamin C group (r = 0.49, P < 0.001) and in the con- level (P < 0.03). Therefore each unit in- trol group (r = 0.55, P < 0.001) (Table 4). A crease in duration of asthma affliction (i.e. significant positive correlation was also each year of increase) was associated with found between plasma vitamin C level and 0.024 mg/dL decrease in mean plasma vita- leukocyte vitamin C level in the control min C in the asthma group but these chang- group (r = 0.47, P < 0.01). es was not significant for leukocyte vitamin Regression analysis showed that the age C levels. of asthmatics and control subjects had no

Table 3 Frequency distribution of plasma and leukocyte vitamin C status in relation to dietary vitamin C intake

Vitamin Plasma vitamin C status Leukocyte vitamin C status C intake Deficient* Normal* Deficient Normal No. % No. % No. % No. %

Deficient 12 63.2 29 35.8 23 46.0 18 36.0 Normal 7 36.8 52 64.2 27 54.0 32 64.0 Total 19 100.0 81 100.0 50 100.0 50 100.0

*P < 0.05, Fisher test.

12 Comparison of plasma.pmd 91 10/25/2005, 1:22 AM 92 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 4 Correlation between plasma and tissue stores of vitamin C [12,16,22], so leukocyte vitamin C levels and dietary leukocyte vitamin C would be a more sensi- vitamin C intake in the 2 groups of tive indicator of duration of asthma. This is participants consistent with our findings showing a greater difference in leukocyte than plasma Vitamin C level Vitamin C intake Asthma Control vitamin C between groups. Kelly et al. re- group group ported a decrease in vitamin C and tocopherol in mild asthmatic patients and Plasma vitamin C concluded that reliance on plasma mea- level r = 0.49*** r = 0.55*** surement alone is not a sufficient indicator Leukocyte vitamin C of vitamin C status and highlights the fact level r = 0.2 r = –0.1 that the nature of the relation between plas- ***P < 0.001, Pearson’s correlation, plasma vitamin ma and vitamin C pools is unknown [3]. C versus dietary intake. Vallance et al. reported that subjects who have undergone infarction, infection and surgery had leukocytosis complica- Discussion tions that result in a decrease in leukocyte vitamin C [26]. In the present study, none Vitamin C, as an important antioxidant, has of the patients had any chronic disease (ex- been hypothesized to have a role in prevent- cept asthma) or history of surgery, so there ing the increased production of oxidants in was no leukocytosis due to other diseases. asthma and the resulting oxidative stress in The results of this study were consistent the lungs [2,11]. Our study showed that with Olusi [21] and Aderele [22]. Olusi et mean plasma and leukocyte vitamin C con- al. found significantly higher concentra- centrations in a group of asthma patients tions of vitamin C in plasma and leukocytes were significantly lower than a control in controls than in either treated or untreat- group. This was consistent with the find- ed patients with asthma [21]. Similarly, in a ings of other studies [20–24]. In our asth- study comparing plasma vitamin C levels in ma group, 38% and 92% of participants children with asthma with levels in healthy had deficient levels of vitamin C in plasma controls, Aderele et al. found significantly and leukocytes respectively, compared lower levels in children with asthma. How- with 0% and 8% of controls. We also showed a significant positive ever, no relationship was demonstrated be- correlation between plasma vitamin C sta- tween vitamin C levels and severity or tus and dietary vitamin C intake in the 2 duration of asthma. Unfortunately, Ader- groups. Leukocyte vitamin C status ele’s study did not quantify vitamin C intake seemed to be more affected by asthmatic in the diet or in supplements taken by ap- status than plasma vitamin C whereas plas- proximately 50% of the children, making ma vitamin C was more influenced by di- the results difficult to interpret [22]. etary intake of the vitamin. Some studies Schwartz and Weiss assessed the rela- report that plasma vitamin C concentra- tionship between dietary vitamin C intake tions are more indicative of recent intake of and pulmonary function in 2526 randomly vitamin C than of body stores. Indeed, selected adults, of whom approximately plasma vitamin C has a linear relationship 3% were asthmatic [23]. Lower dietary vi- with intake of vitamin C. Leukocyte vita- tamin C intakes were shown to be directly min C concentration is more reflective of related to lower plasma vitamin C levels and

12 Comparison of plasma.pmd 92 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 93

lower forced expiratory volume at 1 sec- were not considered [28]. Trenga et al. in a

ond (FEV1). Although the accuracy of di- double-blind crossover study reported that etary recall may be low in their survey, the vitamin C plus vitamin E can decrease the population surveyed was large and the mar- ozone-induced bronchial hyper-respon- gin of error was small. siveness in adults with asthma [29]. Other studies showed that lower intake Some potential limitations of our study of vitamin C was associated with the pres- are the small sample size, measuring vita- ence of respiratory symptoms and pulmo- min C by the colorimetric method, which is nary diseases such as asthma [1,11,24–26]. less sensitive than high-pressure liquid In the present study, the 3-day dietary re- chromatography, and the timing of the call showed 50% of asthma patients report- study in autumn when the intake of citrus ing deficient vitamin C intake (below the fruits as a rich source of vitamin C is good. RDA) and 50% with vitamin C intake above In conclusion, this study showed that the RDA. But this study was done in au- both plasma and leukocyte vitamin C levels tumn when the availability of citrus fruits is are significantly lower in asthmatic patients high and repeating the study in other sea- compared with healthy subjects. Leuko- sons might show different results. cyte vitamin C reduction was greater in Powell compared antioxidant status in asthmatic patients. Plasma vitamin C levels 35 asthmatic patients with 35 healthy sub- were more influenced by diet. These find- jects and reported that there was no signif- ings, which may have important scientific icant difference in plasma vitamin C and public health implications, justify a between the 2 groups, which was inconsis- more detailed and larger prospective study tent with the results of the present study. of the effects of disease activity and treat- However, in Powell’s study, only plasma ment on vitamin C status. Ongoing and vitamin C and not leukocyte vitamin C was proposed interventional studies are needed measured and all the subjects were infants to define the role of vitamin C supplementa- [27]. In our study, the asthmatic patients tion in asthma and to define indications to were taking inhaled short-acting beta-adr- support its ongoing use. energic agents intermittently as the sole treatment, but neither these drugs nor oral corticosteroids have been observed to af- Acknowledgements fect vitamin C status in asthmatic patients The authors are grateful to all participants [3,27]. of the study for their helpful collaboration Vural and Uzun reported that red blood and give special thanks to all physicians and cell vitamin C levels in asthmatic patients staff members of Massih Daneshvari Hos- were significantly lower than a control pital, Tehran for their invaluable work in group, but smoking status, employment, the collection of data. other diseases and leukocyte vitamin C

References 1. Seaton A, Devereux G. Diet, infection 2. Shanmugasundaram KR, Kumar SS, and wheezy illness: lessons from adults. Rajajee S. Excessive free radical gen- Pediatric allergy and immunology, 2000, eration in the blood of children suffering 11(suppl. 13):37–40.

12 Comparison of plasma.pmd 93 10/25/2005, 1:22 AM 94 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

from asthma. Clinica chimica acta, 2001, with bronchial asthma. Turkish journal of 305(1–2):107–14. pediatrics, 2000, 42:17–21. 3. Kelly FJ et al. Altered lung antioxidant 13. Romieu I et al. Antioxidant supplementa- status in patients with mild asthma. Lan- tion and lung functions among children cet, 1999, 354:482–3. with asthma exposed to high levels of air pollutants. American journal of respira- 4. Kaur B, Rowe BH, Ram FS. Vitamin C tory and critical care medicine, 2002, supplementation for asthma. Cochrane 166(5): 703–9. database of systematic reviews, 2001, (4):CD000993. 14. Mirmiran P et al. Estimation of energy requirements for adults: Tehran lipid and 5. Mainous AG, Hueston WJ, Connor MK. glucose study. International journal for Serum vitamin C level and use of health vitamin and nutrition research, 2003, care resources for wheezing episodes. 73(3):193–200. Archives of family medicine, 2000, 9(3): 241–5. 15. Howat PM. Validity and reliability of reported dietary intake data. Journal of the 6. Bielory L, Gandhi R. Asthma and vitamin American Dietetic Association, 1994, C. Annals of allergy, 1994, 73(2):89–96. 94:169–73. 7. Omenaas E et al. Dietary vitamin C in- 16. Wolinsky I, Hickson JF. Vitamin C. In: take is inversely related to cough and Sauberlich HE, ed. Laboratory tests for wheeze in young smokers. Respiratory the assessment of nutritional status, 2nd medicine, 2003, 97(2):134–42. ed. Florida, CRC Press, 1999:11–35. 8. Cohen HA, Neuman I, Nahum H. Block- 17. Omaye ST, Turnbull JD, Sauberlich HE. ing effect of vitamin C in exercise-in- Selected method for the determination of duced asthma. Archives of pediatrics & vitamin C in animal cells, tissues and flu- adolescent medicine, 1997, 151(4): ids. In: McCormick DB, Wright LD, eds. 367–70. Methods in enzymology. NewYork, Aca- 9. Romieu I, Trenga C. Diet and obstructive demic Press, 1979:62–3. lung diseases. Epidemiologic reviews, 18. Ilich JZ, Brownbill RA, Tamborini L. Bone 2001, 23(2):268–87. and nutrition in elderly women: protein, 10. Smit HA, Grievink L, Tabak C. Dietary in- energy and calcium as main determi- fluences on chronic obstructive lung dis- nants of bone mineral density. European ease and asthma: a review of the journal of clinical nutrition, 2003, 57(4): epidemiological evidence. Proceedings 554–65. of the Nutrition Society, 1999, 58(2): 19. Jonnalegadda SS, Diwan S. Nutrient in- 309–20. take of first generation Gujarati Asian In- 11. Forastiere F et al. Consumption of fresh dian immigrants in the US. Journal of the fruit rich in vitamin C and wheezing American College of Nutrition, 2002, symptoms in children. SIDRIA Collabo- 21(5):372–80. rative Group, Italy (Italian Studies on 20. Vallance S. Leucocyte ascorbic acid and Respiratory Disorders in Children and the leucocyte count. British journal of nu- the Environment). Thorax, 2000, 55: trition, 1979, 41:409–11. 283–8. 21. Olusi SO et al. Plasma and white blood 12. Kalayci BT et al. Serum levels of antioxi- cell ascorbic concentrations in patients dant vitamins (alpha tocopherol, beta carotene and ascorbic acid) in children

12 Comparison of plasma.pmd 94 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 95

with bronchial asthma. Clinica chimica Prospective Investigation into Cancer acta, 1979, 92:161–6. and Nutrition. European respiratory journal, 2000, 16:397–403. 22. Aderele WI et al. Plasma vitamin C (ascorbic acid) levels in asthmatic chil- 26. Huang SL, Pan WH. Dietary fats and dren. African journal of medical sci- asthma in teenagers: analyses of the first ences, 1985, 14:115–20. Nutrition and Health Survey in Taiwan (NAHSIT). Clinical and experimental al- 23. Schwartz J, Weiss ST. Relationship be- lergy, 2001, 31(12):1875–80. tween dietary vitamin C intake and pulmonary function in the First National 27. Powell CV et al. Antioxidant status in Health and Nutrition Examination Sur- asthma. Pediatric pulmonology, 1994, vey (NHANES I). American journal of 18:34–8. clinical nutrition, 1994, 59:110–4. 28. Vural H, Uzun K. Serum and red blood 24. Hu G et al. Dietary vitamin C intake and cell antioxidant status in patients with lung function in rural China. American bronchial asthma. Canadian respiratory journal of epidemiology, 1998, 148:594– journal, 2000, 7:476–80. 8. 29. Trenga CA, Koeing JQ, Willams PV. Di- 25. Sargeant LA, Jaeckel A, Wareham NJ. etary antioxidants and ozone-induced Interaction of vitamin C with the relation bronchial hyperresponsiveness in between smoking and obstructive air- adults with asthma. Archives of environ- ways disease in EPIC Norfolk. European mental health, 2001, 56(3):242–9.

International statistical classification of diseases and health related problems, 10th revision, (ICD-10), 2nd edition This new edition of WHO’s International statistical classification of diseases and related health problems, 10th revision (ICD-10) has been fully updated. Originally published in the early 1990s, ICD-10 now incorporates all updates and other changes to this core health classification since 1996. The new edition has hundreds of updates, and includes new diseases such as SARS, to ensure that the classification is entirely suited to today’s needs. Improvements have been made to the rules and guidelines for mortality coding, making them easier to implement and which will result in the improved comparability of international mortality statistics. In addition, there are many additions and changes to the alphabetical index resulting in some 60 additional pages, making it far more comprehensive than that of the first edition. The three-volume second edition follows the same structure as the first edition, with the addition of a chapter XXII (Codes for special purposes) made necessary by the emergence of SARS and the need for an interim solution for the coding of this major health problem. The ICD-10 2nd edition is available electronically on CD and as a downloadable version. Further information on this publication can be obtained from WHO Press: http://www.who.int/bookorders/anglais/home1.jsp?sesslan=1

12 Comparison of plasma.pmd 95 10/25/2005, 1:22 AM 96 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Umbilical cord blood pH and risk factors for acidaemia in neonates in Kerman V. Modarressnejad 1

ABSTRACT A prospective cross-sectional study was carried out to determine the relationship and predictive value of umbilical cord blood pH for adverse neonatal outcomes. A total of 400 singleton term infants delivered by vaginal delivery or caesarean section were studied at a hospital in Kerman, Islamic Republic of Iran, in 2001. Mean (SD) umbilical cord blood pH was 7.25 ± 0.14 and 81 cases had acidaemia (pH < 7.1). Apgar score at 1 minute and fetal distress were significantly related to acidaemia. There was also a significant relation between meconium-stained amniotic fluid and acidaemia. Logistic regression analysis showed that Apgar score < 7 at 1 minute, meconium-stained amniotic fluid and fetal distress were significant risk factors for acidaemia in newborn infants. Umbilical cord blood acid–base alterations are related to subsequent adverse outcome events for neonates.

Le pH du sang de cordon ombilical et les facteurs de risque d’acidémie chez le nouveau-né à Kerman RÉSUMÉ Une étude transversale prospective a été réalisée en vue de déterminer la relation et la valeur prédictive du pH du sang de cordon ombilical pour l’issue néonatale défavorable. Au total, 400 accouche- ments uniques à terme par voie basse ou par césarienne ont été étudiés en 2001dans un hôpital à Kerman (République islamique d’Iran). Le pH moyen (E.T.) du sang de cordon ombilical était de 7,25 ± 0,14 et 81 cas présentaient une acidémie (pH < 7,1). Le score d’Apgar à une minute de vie et la souffrance foetale étaient significativement associés à l’acidémie. Il y avait également une relation significative entre le liquide amniotique teinté de méconium et l’acidémie. L’analyse de régression logistique a montré qu’un score d’Apgar inférieur à 7 à une minute de vie, qu’un liquide amniotique teinté de méconium et que la souffrance fœtale étaient des facteurs de risque d’acidémie significatifs chez le nouveau-né. Les altérations acido-basiques du sang de cordon ombilical sont liées à des évènements ultérieurs ayant une issue défavorable pour le nouveau-né.

1Department of Obstetrics and Gynaecology, Kerman University of Medical Sciences and Health Services, Kerman, Islamic Republic of Iran (Correspondence to V. Modarressnejed: [email protected]). Received: 21/05/03; accepted: 25/02/04

13 Umbilical cord blood.pmd 96 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 97 Introduction Method Umbilical cord blood pH is the best avail- The present study was a prospective able criterion for determining fetal hy- cross-sectional study from 1 May 2001 to poxaemia during labour and making 30 April 2002. A sample of 400 neonates decisions about necessary care after birth born in the delivery ward of Nik-Nafs Hos- [1,2]. Considerable effort has been expend- pital, Kerman, Islamic Republic of Iran, ed to determine risk factors and the role of was taken by selecting every other child intrapartum asphyxia for adverse neonatal born during the period of the study. The outcome [3]. Hypoxaemia is an arterial par- sample size was based on a pilot study of tial oxygen pressure less than normal. Hy- 30 newborns with 17% prevalence of poxaemia associated with ischaemia causes acidaemia and the maximum error of 5% hypoxic–ischaemic encephalopathy in the was determined as 395 subjects. fetus, leading to permanent lesions of the After admission for delivery, informed central nervous system. This may result in consent was taken from the pregnant fetal mortality or later problems of cerebral women and they were enrolled in the study palsy and mental retardation [4,5]. Arterial with the following inclusion criteria: single- blood gases and blood cord pH are useful ton pregnancy, no underlying disease and for measuring the degree of asphyxia and gestational age of 38–42 weeks. The moth- are also good criteria for assessing the like- er’s age, gestational age, gravidity and any lihood of neonatal asphyxia and cerebral history of high-risk pregnancy with its rea- palsy [6]. son and symptoms were recorded. Based on different studies, mean values Immediately after the delivery of the fe- of umbilical cord blood pH range from tus, the umbilical cord was clamped at a 7.25–7.28 [7]. In a study of 19 000 neo- distance of 15–20 cm and 1 mL of umbili- nates, an umbilical cord blood pH 7.04–7.1 cal cord arterial blood was aspirated with a was considered the extreme of the normal heparinized syringe and delivered on ice to range and pH < 7.1 as acidaemia [7]. Se- the hospital laboratory. Umbilical cord arte- vere acidaemia is one of the symptoms of rial blood pH was measured by a blood gas hypoxaemia. Therefore, hypoxic cases can analyser (AVL Compact) and pH < 7.1 was be identified by measuring the umbilical considered as acidaemia [2,9–11]. cord blood pH and if it is raised, high pres- After delivery, the cardiac and respira- sure oxygen should be deployed in order to tory status of the fetus were evaluated and prevent cerebral lesions [4]. fetal distress was noted (defined as late de- Studies so far have shown increased celeration or repeated early and variable de- risk of several adverse outcome events in celeration of fetal heart beat). Apgar scores the preterm infant with evidence of at 1 and 5 minutes and the neonate’s birth acidaemia at birth, but some authors have weight were recorded. Meconium in the demonstrated that umbilical cord gases amniotic fluid and consistency of the have little value in the determination of in- meconium (thick or thin) was visually as- fants at risk of adverse outcome [8]. sessed and noted. All these measurements The aim of this study was to determine were done by the same researcher using the relationship, and predictive value, of the same measuring instruments. umbilical cord artery blood pH in relation to Chi-squared analysis was used to deter- adverse neonatal outcomes. mine the significance of differences be-

13 Umbilical cord blood.pmd 97 10/25/2005, 1:22 AM 98 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

tween groups. SPSS, version 11.5 was onates without fetal distress (Tables 1 and used for the analysis. Odds ratios were 2). used for studying the risk of acidaemia There was no significant relation be- from the variables measured. The Mantel– tween rate of normal vaginal delivery or Haenszel model was used for evaluating the caesarean section and acidaemia (Tables 1 significance of odds ratios. After calculat- and 2), and in the cases of caesarean sec- ing the meaningful odds ratios in order to tion, there was no significant relation be- remove confounding factors, analyses tween the type of anaesthesia (general or were repeated using a logistic regression epidural) and acidaemia. model. The presence of meconium in the amniotic fluid had a highly significant relation with neonatal acidaemia; the risk of Results acidaemia was 5.8 times higher in neonates The mean birth weight of neonates was with meconium-stained amniotic fluid than 3.28 (SD 0.32) kg and mean umbilical cord those with clear fluid (Tables 1 and 2). blood pH was 7.25 (SD 0.1). The mean Moreover, the concentration of meconium gestational age was 38.8 (SD 1.4) weeks. had a direct relation with acidaemia. In the The range of Apgar scores at 1 minute was 69 cases with meconium-stained amniotic 5–9 and at 5 minutes 7–10. Of the 400 neo- fluid, the risk of acidaemia in those with nates, 81 had acidaemia (umbilical cord ar- thick meconium was 12.5 times higher terial blood pH < 7.1). (Tables 1 and 2). Neonate weight and gestational age had Among the variables studied, the Apgar no significant relationship with acidaemia score at 1 minute, fetal distress and meco- (P = 0.5). However, there was a highly sig- nium-stained amniotic fluid had a signifi- nificant relationship between 1 minute Ap- cant relationship with the risk of acidaemia. gar score and acidaemia as 68% of Logistic regression analysis showed that neonates with acidaemia had Apgar score < each variable in the presence of 2 other 7 at 1 minute whereas only 1% of those variables was a risk factor for the occur- with normal umbilical cord blood pH had rence of acidaemia. Based on the Mantel– Apgar score < 7 at 1 minute (Table 1). This Haenszel test, Apgar score < 7 at 1 minute, difference was also observed in Apgar fetal distress and meconium-stained amni- score at 5 minutes; that is, 51.8% of neo- otic fluid were significant risk factors for nates with acidaemia had Apgar score < 7 neonatal acidaemia. at 5 minutes, compared with only 0.6% in neonates with normal pH (P < 0.0001). The Discussion odds ratio showed that neonates with Ap- gar score < 7 at 1 minute had a risk of Arterial blood gases and umbilical cord acidaemia 222.8 times higher than neonates blood pH are valuable measures for deter- with Apgar score ≥ 7 (P < 0.0001) (Table mining hypoxaemia and for evaluating the 2). extent of asphyxia in the newborn infant. Fetal distress had a highly significant The pH of normal umbilical cord blood var- relationship with acidaemia (P < 0.0001) ies across different studies. In the present and the risk of acidaemia in neonates with study the mean value of 7.25 ± 0.1 was the fetal distress was 7.7 times higher than ne- same as Arikan et al. (7.25) [12] but dif-

13 Umbilical cord blood.pmd 98 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 99

Table 1 Association between umbilical cord blood pH and adverse outcomes for 400 singleton term neonates

Outcome Umbilical cord blood pH Yates corrected χ2 Acidic Normal Total No. No. No.

Apgar score at 1 min < 7 55 3 58 ≥ 7 26 316 342 Total 81 319 400 χ2 = 228.26, df = 1, P < 0.0001 Fetal distress Ye s 32 25 57 No 49 294 343 Total 81 319 400 χ2 = 50.46, df = 1, P < 0.0001 Type of delivery Caesarean 16 71 87 Normal vaginal 65 248 203 Total 81 319 400 χ2 = 0.11, df = 1, P < 0.736 Meconium-stained amniotic fluid Ye s 34 35 69 No 47 284 331 Total 81 319 400 χ2 = 41.35, df = 1, P < 0.0001 Concentration of meconium Thick 23 5 28 Thin 11 30 41 Total 34 35 69 χ2 = 18.21, df = 1, P < 0.0001

df = degrees of freedom. Acidic = pH < 7.1.

Table 2 Odds ratios for variables associated with acidaemia in 400 singleton term neonates

Variable OR 95% CI Mantel–Haenszel χ2

Apgar score at 1 min < 7 222.8 61.1–962.4 χ2 = 233.04, P < 0.0001 Fetal distress 7.68 4.03–14.7 χ2 =52.89, P < 0.0001 Caesarean section 0.86 0.45–1.63 χ2 = 0.24, P < 0.626 Meconium-stained amniotic fluid 5.87 3.22–10.73 χ2 = 43.39, P < 0.0001 Thick meconium 12.55 3.37–50 χ2 = 20.07, P < 0.0001

OR = odds ratio. CI = confidence interval.

13 Umbilical cord blood.pmd 99 10/25/2005, 1:22 AM 100 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

fered from Ramin et al. and Yeomans et al. study performed in the United States, the (7.28) [9,13,14]. This difference may be mean umbilical cord blood pH was 7.22 in because Yeomans et al. excluded all cases neonates born by elective caesarean, 7.2 in with fetal distress, meconium staining, neonates born by caesarean following pre- twin pregnancy and diabetic mothers [9], vious normal vaginal delivery and 7.4 in ne- and Ramin et al. studied cases with meco- onates born by normal vaginal delivery nium-stained amniotic fluid and preterm following previous caesarean. This differ- neonates separately [13,14]. Although neo- ence was statistically significant (P < 0.04) natal umbilical cord blood pH varies ac- [15]. Another study concluded that the um- cording to different conditions, it seems bilical cord blood pH of neonates born by that there is no difference in blood pH be- vaginal delivery from the breech position tween term and preterm neonates was lower than that for those born from [10,12,14]. the cephalic position (P < 0.01). This may As for the relation of umbilical cord be due to longer labour and the higher de- blood pH and acidaemia with meconium in gree of pressure on the umbilical cord for a the amniotic fluid there is disagreement fetus in the breech position [16]. among different studies. Yeomans et al. re- In our study, an Apgar score less than 7 ported a higher rate of acidaemia in cases at 1 minute was a high risk factor for the of meconium-stained amniotic fluid, and occurrence of acidaemia (OR = 222.8). suggested a weak correlation between Valentin et al. found a significant correla- meconium-stained amniotic fluid and um- tion between Apgar score and umbilical bilical cord blood pH [11]. Ramin et al. record blood pH [17]. But Hankins et al. in a ported no significant difference between study conducted on neonates with corio- neonates with meconium-stained amniotic amnionitis concluded that this infection fluid and normal amniotic fluid in both was not a risk factor for the occurrence of mean umbilical cord blood pH and acidaemia and in spite of lower Apgar score acidaemia [14]. In the present study in this group in comparison to the control acidaemia was 5.87 times more likely in group, the Apgar score itself was not an neonates with meconium-stained amniotic accurate criterion for birth asphyxia [13]. fluid with than in those with normal amni- According to the results of the present otic fluid. study, fetal distress, Apgar score of less There was no significant relation be- than 7 at 1 minute and meconium-stained tween birth weight and acidaemia and this amniotic fluid are risk factors of acidaemia. is in line with the results of other studies Each factor independently increased the [10,13]. There was also no significant rela- risk of acidaemia. Based on a logistic re- tion between gestational age and acidaemia, gression model, each of these factors in the a finding that agrees with other studies presence of 2 others significantly increased [10,12,13]. the risk of acidaemia. We conclude that In our study, there was no difference these factors can be used clinically to pre- between neonates born by caesarean sec- dict the risk of acidaemia, leading to better tion and those born by normal vaginal deliv- management of this condition in neonates. ery in the occurrence of acidaemia. In a

13 Umbilical cord blood.pmd 100 10/25/2005, 1:22 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 101

References

1. Riley RJ, Johnson JW. Collecting and 10. Hankins GD, Snyder RR, Yeomans ER. analyzing cord blood gases. Clinical ob- Umbilical arterial and venous acid–base stetrics and gynecology, 1993, 36(1): and blood gas values and the effect of 13–23. chorioamnionitis on those values in a cohort of preterm infants. American jour- 2. Vandenbussche FP, Oepkes D, Keirse nal of obstetrics and gynecology, 1991, MJ. The merit of routine cord blood pH 164(5Pt1):1261–4. measurement at birth. Journal of perinatal medicine, 1999, 27(3):158–65. 11. Yeomans ER et al. Meconium in the amniotic fluid and fetal acid–base status. 3. Victory R et al. Umbilical cord pH and Obstetrics and gynecology, 1989, 73(2): base excess values in relation to neona- 175–8. tal morbidity for infants delivered preterm. American journal of obstetrics 12. Arikan GM et al. Cord blood oxygen satu- and gynecology, 2003, 189(3):803–7. ration in vigorous infants at birth: what is normal? British journal of obstetrics and 4. Stoll BJ, Kliegman RM. The fetus and the gynaecology, 2000, 107(8):987–94. neonatal infant. Section 1: Non-infectious disorders. In: Behrman RE, 13. Ramin SM et al. Umbilical artery acid– Kliegman RM, Jenson HB, eds. Nelson base status in the preterm infant. Obstet- textbook of pediatrics, 16th ed. Philadel- rics and gynecology, 1989, 74(2):256–8. phia, WB Saunders, 2000:451–538. 14. Ramin SM et al. Acid–base significance 5. Du Plessis AJ, Volpe JJ. Perinatal brain of meconium discovered prior to labor. injury in the preterm and term newborn. American journal of perinatology, 1993, Current opinion in neurology, 2002, 15: 10(2):143–5. 151–7. 15. Swaim LS, Holste CS, Waller DK. Umbili- 6. Goldaber KG, Gilstrap LC. Correlations cal cord blood pH after prior cesarean between obstetric clinical events and delivery. Obstetrics and gynecology, umbilical cord blood acid–base and 1998, 92(3):390–3. blood gas value. Clinical obstetrics and 16. Daniel Y et al. Umbilical cord blood acid– gynecology, 1993, 36(1):47–59. base values in uncomplicated term vagi- 7. Cunningham FG et al., eds. Williams ob- nal breech deliveries. Acta obstetricia et stetrics, 21st ed. New York, McGraw–Hill, gynecologica scandinavica, 1998:77(2): 2001:385–403. 182–5. 8. Gleissner M, Jorch G, Avenarius S. Risk 17. Valentin L et al. Clinical evaluation of the factors for intraventricular hemorrhage fetus and neonate. Relation between in a birth cohort of 3721 premature in- intra-partum cardiotocography, Apgar fants. Journal of perinatal medicine, score, cord blood acid–base status and 2002, 28:104–10. neonatal morbidity. Archives of gynecology and obstetrics, 1993, 253(2):103– 9. Yeomans ER et al. Umbilical cord pH, 15. PCO2 and bicarbonate following uncomplicated term vaginal deliveries. Ameri- can journal of obstetrics and gynecology, 1985, 151(6):798–800.

13 Umbilical cord blood.pmd 101 10/25/2005, 1:22 AM 102 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Factor VII polymorphisms associated with plasma factor VII coagulant activity levels in healthy Tunisians M. Nour,1 F. Belhaj Slama,2 R.M. Maaroufi,2 M. Hammami1 and T. Mahjoub 2

ABSTRACT Factor VII gene polymorphisms may contribute to elevations in factor VII coagulant (FVIIc) levels that have been associated with cardiovascular risk. We therefore studied the association of two polymorphisms – R353Q polymorphism at codon 353 involving the catalytic region and the 10 base pair (bp) insertion polymorphism involving the promoter region – with FVIIc levels in 176 healthy Tunisians. The variant Q allele had a frequency of 0.213 (SD 0.021) whereas the frequency of the 10 bp insert allele was 0.250 (SD 0.023). Subjects with R/R genotype had significantly higher FVIIc levels than Q353 heterozygote and homozygote subjects (96.36 versus 59.52). FVIIc levels with the 10 bp insertion polymorphism were not significantly different. The Q353 allele of the factor VII gene polymorphism is associated with decreased factor VII and could be protective against cardiovascular disease.

Les polymorphismes du gène du facteur VII associés à la concentration plasmatique en activité coagulante du facteur VII chez des Tunisiens en bonne santé RÉSUMÉ Les polymorphismes du gène du facteur VII peuvent contribuer à l’élévation de la concentration en activité coagulante du facteur VII (FVIIc) qui a été associée au risque cardio-vasculaire. Nous avons donc étudié l’association de deux polymorphismes – R353Q au niveau du codon 353 intervenant sur la région catalytique et le polymorphisme d’insertion de 10 paires de bases (pb) impliquant la région du promoteur - avec les taux de FVIIc chez 176 Tunisiens en bonne santé. L’allèle Q variant avait une fréquence de 0,213 (E.T. 0,021) tandis que la fréquence de l’allèle d’insertion de 10 pb était de 0,250 (E.T. 0,023). Les sujets ayant un génotype R/R avaient une concentration en FVIIc significativement plus élevée que les sujets hétérozy- gotes et homozygotes porteurs de l’allèle Q353 (96,36 vs 59,52). La concentration en FVIIc avec le polymorphisme d’insertion de 10 pb n’était pas significativement différente. L’allèle Q353 du polymorphisme du gène du facteur VII est associé à la diminution du facteur VII et pourrait constituer une protection contre les maladies cardio-vasculaires.

1Research Unit, Faculty of Medicine, Monastir, Tunisia (Correspondence to M.Nour: [email protected]). 2Research Unit, Faculty of Pharmacy, Monastir, Tunisia. Received: 01/06/03; accepted: 20/11/03

14 Factor VII polymorphisms.pmd 102 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 103

Introduction with severe to very severe bleeding tenden- cies [9]. The relationships between the Arterial and venous thrombosis, which clinical presentation and FVII levels and the clinically manifest as stroke, myocardial in- associated molecular genetic defects lack farction or pulmonary embolism are a ma- apparent consistency [7,10]. Plasma FVII jor cause of death [1]. Genetic factors levels vary significantly in the general pop- contribute significantly to the development ulation and are influenced by environmental of these diseases. Since the advent of mo- factors [11,12]. In addition, several poly- lecular genetics, it has been a focus of in- morphisms within the FVII gene can affect terest to elucidate the role of mutations in either the function or its expression level, candidate genes and their impact on hae- i.e. a replacement of arginine by glutamine mostatic disorders such as arterial and in codon 353 (R353Q) arising from a single venous thrombosis. The genes encoding nucleotide substitution (G to A) at position proteins affecting haemostasis are intrigu- 10976 in exon 8 in the catalytic region of ing factors that may increase cardiovascu- the FVII gene and an insertion of a decanu- lar disease. Recently several reports have cleotide (designated as 0/10 bp) in the pro- focused on the association between the moter region of the gene at position –323 factor VII of the cascade coagulation and [13,14]. cardiovascular risk [1]. To our knowledge, the prevalence of Human factor VII (FVII) is a vitamin K- factor VII gene polymorphisms has not dependent glycoprotein, synthesized in the been studied previously in any country of liver and secreted in the blood as an inac- North Africa. It has been reported that dif- tive zymogen. Upon contact with tissue ferences in FVII activity levels and in geno- factor exposed by vascular injury, FVII is type frequencies depend on the ethnic cleaved to its two-chain active form, which groups [15]. We studied FVII gene poly- then activates factors IX and X leading to morphisms in healthy Tunisians with the the generation of thrombin [2]. The impor- objectives of determining the frequencies tance of FVII in normal haemostasis is il- of these gene polymorphisms in the popula- lustrated by the severe bleeding diathesis tion and clarifying the genotype association associated with low plasma levels of the of the R353Q and 0/10 bp insertion poly- protein [3]. It also appears to play an im- morphisms with plasma FVIIc levels. portant role in atherothrombosis [4]. More- over, several studies have demonstrated positive correlations between plasma FVII Methods level and ischaemic cardiovascular events [5]. Controversially, Iacoviella et al. sug- We recruited 176 unrelated healthy blood gested that certain polymorphisms of FVII donors who were living in Tunisia for our may protect against familial myocardial inf- study. All 112 men and 64 women were arction, possibly by modifying FVII coagu- aged 17–63 years and were ethnic Arabs. lant (FVIIc) levels [6]. No women were taking oral contraceptives The FVII gene contains nine exons and or other drugs that might affect FVII lev- is located on the chromosome 13q34 [7,8]. els. No one had cardiovascular disease, dia- The clinical presentation of patients with betes or thromboembolic diseases. FVII deficiency is variable and consists of Blood was collected by standard atrau- asymptomatic patients as well as patients matic venepuncture technique using 0.1

14 Factor VII polymorphisms.pmd 103 10/25/2005, 1:23 AM 104 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

mol/L citrate buffer as anticoagulant. FVIIc Allelic frequencies were calculated by was measured by a one-stage semi-auto- gene-counting method for statistical analy- mated bioassay in an ST4 coagulometer sis. Chi-squared goodness of fit was used (Diagnostica Stago, Asnieres, France). Re- to determine if genotype distributions were sults were expressed as a percentage of as expected by Hardy–Weinberg propor- activity of the standard plasma supplied by tions. Variable means of FVIIc were com- the manufacturers. pared by Student t-test. The level of To prepare DNA specimens and poly- significance was set at 0.05. merase chain reaction (PCR) assays for polymorphism detection, DNA was ex- tracted from ethylenediamine tetra-acetic Results acid (EDTA)-anticoagulant white blood We determined FVII levels in a sample of cells by standard methods [16]. PCR am- 176 Tunisian blood donors and studied the plification of the promoter region (10-bp relationships between 2 polymorphisms of insertion) and exon 8 (R353Q) of the factor the FVII gene. The promoter decanucle- VII gene was performed using the oligonu- otide insertion (–323 0/10-bp) and the cleotide primers and amplification condi- R353Q polymorphisms of FVII gene were tions as described by Marchetti et al. and evaluated. For the exon 8 polymorphism, Green et al., respectively [13,14]. The con- 66 of 176 individuals carried the Q353 allele ditions of restriction enzyme digestion for (57 heterozygotes and 9 homozygotes), the R353Q were as given by the manufac- whereas 110 carried homozygotes for the turers (Promega Corporation, Madison, R353 allele. There were 81 individuals with Wisconsin, United States of America). The 10-bp insertion (74 heterozygotes and 7 amplified products of the 0/10-bp promoter homozygotes) and 95 lacked the insertion. polymorphism did not require enzyme di- The allele frequency derived from the gestion. This polymorphism contains two analysis of 352 chromosomes was 0.213 alleles, the 0-bp allele and the 10-bp allele; (standard deviation = 0.021) for the Q353 the latter corresponds to the presence of a allele and 0.787 (SD = 0.021) for the R353 decanucleotide insertion. The PCR frag- allele. The corresponding frequency of the ment size of the 0-bp allele is 214-bp 10-bp insertion allele was 0.25 (SD = whereas that of the 10-bp allele is 224-bp. 0.023) and that of the 0 allele (absence of The PCR fragment size of the R353Q poly- insertion) was 0.75 (SD = 0.023). morphism is 312-bp and required MspI di- The genotype distribution of the R353Q gestion. Enzyme digestion generated in all or 0/10 bp promoter polymorphism in the PCR products 40-bp which is added by ei- Tunisians did not deviate significantly from ther: one fragment of 272-bp in homozy- a population at Hardy–Weinberg equilibri- gote R/R; two fragments, one of 205-bp um (P > 0.05). Table 1 shows FVIIc levels and the other 67-bp in homozygote Q/Q; or as a function of genotypes for each poly- three fragments of 272-bp, 205-bp and 67- morphism without the concomitant pres- bp in heterozygote R/Q. The amplified ence of the other. Table 2 shows FVII products of the 0/10-bp promoter poly- levels as a function of a combination of the morphism and enzyme digestion fragment two polymorphisms. of R353Q polymorphism were separated by 8% polyacrylamide gel.

14 Factor VII polymorphisms.pmd 104 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 105

Table 1 Adjusted plasma factor VII coagulant R353 allele (P < 0.0001). The genotypes (FVIIc) activity levels as a percent of with the 10-bp insertion were associated standard by genotype with lower FVII levels (P > 0.05).

Genotype Number FVIIc Mean (SD) Discussion R353Q polymorphism R/R 110 96.36 (32.14) During the past 20 years, substantial R/Q 57 54.08 (33.27) progress has been made in understanding Q/Q 9 75.83 (30.02) the multifactorial and multigenic nature of Presence of 1 Q allele haemostatic diseases. It is well established or 2 Q alleles 66 59.52 (33.41) that interactions between environmental Promoter polymorphism and genetic factors determine the risk for a 0/0 bp 95 88.34 (3303) thrombotic episode [17]. During the last 0/10 bp 74 71.97 (41.44) decade, evidence has accumulated that in- 10/10 bp 7 73.80 (21.50) dicates that raised plasma factor VII levels Presence of 1 or 2 increase the risk of ischaemic heart dis- decanucleotide allele(s) 81 72.08 (40.34) ease, even though in some reports this as- Total 176 sociation was not seen [18]. Conflicting SD = standard deviation. results have been reported about the role of common polymorphisms in influencing For the R353Q polymorphism, the FVII plasma levels [19,20]. We, therefore, presence of allele Q353 in heterozygous or explored the frequencies of two FVII gene homozygous state was associated with polymorphisms among healthy Tunisians. FVII levels lower by 20%–40% than the Our investigation was the first of its kind in Tunisia to attempt to determine the associations of these polymorphisms with FVII Table 2 Plasma factor VII coagulant (FVIIc) coagulant levels. activity levels in combined genotypes of The observed frequency of the Q allele R353Q and 0/10 bp promoter polymorphisms (0.213) among normal Tunisians was high- of the factor VII gene among Tunisians er than among Europeans (0.099) or Chi- Genotype Number FVIIc nese (0.045) [20,21,15]. The 10 bp Mean (SD) insertion allele frequency in healthy Tuni- sians (0.25) was also higher than those of R/R/0/0 66 95.65 (27.90) Europeans (0.139) and Chinese (0.036) R/R/0/10 34 97.35 (42.82) [20,21]. R/R/10/10 2 88.50 (2.12) We also evaluated the FVII genotype- R/Q/0/0 20 68.87 (44.88) phenotype relationships by the determina- R/Q/0/10 38 46.68 (24.12) tion of two polymorphisms (four alleles) in R/Q/10/10 0 – different functional regions of the gene. There was a trend towards lower FVIIc in Q/Q/0/0 12 77.00 (34.35) individuals carrying the Q allele. These Q/Q/0/10 2 84.50 (3.54) findings were consistent with other previ- Q/Q/10/10 2 59.00 (2.83) ous reports in which heterozygous or ho- Total 176 mozygous individuals for the Q allele have SD = standard deviation. 10%–38% lower FVIIc levels than their

14 Factor VII polymorphisms.pmd 105 10/25/2005, 1:23 AM 106 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

homozygous (R/R) counterparts [20,22]. disease. Genetic changes could have par- However, we found that the 10 bp insert ticular strength in very sensitive systems allele was not associated with significantly like the haemostatic system. Because of the lower FVIIc levels. dichotomy of this system, polymorphic From our data, it was apparent that the changes affecting haemostatic factors effect of genotypes was more pronounced could have mild but opposite effects in the at the codon 353 than at the promoter lo- pathogenesis of thrombotic and haemor- cus. For the latter, no significant genotype rhagic disorders. It has been reported that associations with FVIIc levels were ob- the frequencies of the Q353 allele associa- served. However, others have found asso- tion with lower FVII levels in patients who ciations between them [21,23]. It is had a myocardial infarction were less fre- possible that the association of factor VII quent than in control subjects. These find- polymorphisms with plasma levels varies ings suggested that genetic protection from with race and ethnicity. In our study, the myocardial infarction is provided by this al- effect of the R353Q polymorphism on lele [6,25]. plasma FVIIc levels appeared to be domi- We observed significant genotype asso- nant over the promoter polymorphism. The ciations with FVIIc levels among Tunisians Q allele may interfere with biological activ- with more pronounced and consistent ef- ity of the factor VII protein leading to the fect in the R353Q polymorphism. If the Q decreased plasma FVIIc levels. allele is indeed the functional mutation that The best way to establish which gene causes reduction of FVII activity in Tuni- polymorphism affects plasma factor VII sians, it might confer some protection would be to study it in vitro, in which each against thrombosis and coronary artery polymorphism is expressed in mammalian disease in this ethnic group. Future work cells alone or in combination with others, should seek to determine the action mecha- and the amount of factor VII produced by nism of polymorphism, alone or in associa- the different constructs are compared. At tion with others, on factor VII levels and the moment, studies on single polymor- evaluate its frequency in other ethnic phisms without the concomitant presence groups and in patients with atherothrom- of others in the constructs are not avail- botic disease. able, making it difficult to assign a functional effect on FVIIc levels to any [24]. Acknowledgements Genetic polymorphisms underlie the diversity of any species. Most of such inher- We are grateful to Dr S. Kallel and Dr S. ited changes in DNA structure are neutral, Hizem for providing patient samples. We but others could affect the function of pro- thank Mrs Naoufel Maatouk for her help. teins and with more or less severity, the ef- This work was financially supported by the ficiency of whole physiological system, Unités de Recherche UR/08/39 and UR08/ thus modifying susceptibility to a particular 50.

References 1. Endler G, Mannhalter C. Polymorphisms 2. Bauer KA. Activation of the factor VII-tis- in coagulation factor genes and their im- sue factor pathway. Thrombosis and pact on arterial and venous thrombosis. haemostasis, 1997, 78:108–11. Clinica chimica acta, 2003, 330:31–55.

14 Factor VII polymorphisms.pmd 106 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 107

3. Peyvandi F et al. Clinical manifestations terminants of activated factor VII. Arterio- in 28 Italian and Iranian patients with se- sclerosis thrombosis and vascular biol- vere factor VII deficiency. Haemophilia, ogy, 1996, 16:1170–6. 1997, 3:342–6. 13. Green F et al. A common genetic poly- 4. Peyvandi F et al. A novel polymorphism morphism associated with lower coagu- in intron 1a of the human factor VII gene lation factor VII levels in healthy (G73A): study of a healthy Italian popula- individuals. Arteriosclerosis and throm- tion and of 190 young survivors of myo- bosis, 1991, 11:540–6. cardial infarction. British journal of 14. Marchetti G et al. A polymorphism in the haematology, 2000, 108:247–53. 5´ region of coagulation of factor VII gene 5. Hultin MB. Fibrinogen and factor VII as (F7) caused by an inserted decanu- risk factors in vascular disease. Progress cleotide. Human genetics, 1993, 90: in haemostasis and thrombosis, 1991, 575–6. 10:215–41. 15. Liu Y et al. Genotype associations of fac- 6. Iacoviello L et al. Polymorphisms in the tor VII gene with plasma factor VII coagu- coagulation factor VII gene and risk of lant activity and antigen levels in healthy myocardial infarction. New England jour- Chinese. Blood coagulation and fibrin- nal of medicine, 1998, 338:78–85. olysis, 2002, 13:217–24. 7. O’Hara PJ et al. Nucleotide sequence of 16. Miller SA, Dykes DD, Polesky HF. A the gene coding for human factor VII, a simple salting out procedure of extract- vitamin K-dependent protein participat- ing DNA from human nucleated cells. ing in blood coagulation. Proceedings of Nucleic acids research, 1988, 16:1215– the National Academy of Sciences of the 7. USA, 1987, 84:5158–62. 17. Bertina RM. Molecular risk factors throm- 8. De Grouchy J et al. Regional mapping of bosis. Thrombosis, 1999, 82:601–9. clotting factors VII and X to 13q34. Ex- 18. Meade TW et al. Fibrinolytic activity, clot- pression of factor VII through chromo- ting factors, and long-term incidence of some 8. Human genetics, 1984, 66: ischaemic heart disease in the North- 230–3. wick Park Heart Study. Lancet, 1993, 9. Mariani G et al. Molecular and clinical 342:1076–9. aspects of factor VII deficiency. Blood co- 19. Doggen CJ et al. A genetic propensity to agulation and fibrinolysis, 1998, 9:S83– higher factor VII is not associated with 8. the risk of myocardial infarction in men. 10. Cooper DN et al. Inherited factor VII defi- Thrombosis and haemostasis, 1998, ciency: Molecular genetics and patho- 80:281–5. physiology. Thrombosis and haemo- 20. Girelli D et al. Polymorphisms in the fac- stasis, 1997, 78:151–60. tor VII gene and the risk of myocardial 11. Howard RR et al. Factor VII antigen lev- infarction in patients with coronary artery els in a healthy blood donor population. disease. New England journal of medi- Thrombosis and haemostasis, 1994, cine, 2000, 343:774–80. 72:21–7. 21. De Maat MPM et al. Factor VII polymor- 12. Scarabin PY et al. Population correlates phisms in populations with different risks of coagulation factor VII. Importance of of cardiovascular disease. Arteriosclero- age, sex, and menopausal status as de-

14 Factor VII polymorphisms.pmd 107 10/25/2005, 1:23 AM 108 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

sis, thrombosis and vascular biology, glutamine 353 variant. Thrombosis and 1997, 17:1918–23. haemostasis, 1996, 75:56–67. 22. Bernardi F et al. Factor VII gene polymor- 24. Hunault M et al. The Arg353Gln polymorphisms contribute about one third of phism reduces the level of coagulation factor VII level variation in plasma. Arte- factor VII. In vitro and in vivo studies. Arte- riosclerosis, thrombosis and vascular bi- riosclerosis, thrombosis and vascular biology, 1996, 166:72–6. ology, 1997, 17:2825–9. 23. Humphries S et al. Low plasma levels of 25. Shimokata K et al. Effects of coagulation factor VIIc and antigen are more strongly factor VII polymorphism and coronary associated with the 10 base pair disease. Thrombosis research, 2002, promoter (–323) insertion than the 105:493–8.

Genetics, genomics and the patenting of DNA. Review of potential implications for health in developing countries This new report, Genetics, genomics and the patenting of DNA, addresses the important ethical, legal, social and health issues raised by the patenting of DNA sequences – not only for the industrialized world, but also for developing countries. Science is now at a critical moment as it attempts to translate voluminous genetic data into practical health tools. It is therefore timely to consider the various incentives and barriers that exist for the creation of practical health interventions of value in poorer settings. The report emphasizes that genomics has the potential to offer great benefit to public health on a global scale, notes the present ambiguity in international agreements on intellectual property rights on the legal status of genetic “inventions”, and highlights the ongoing controversy surrounding the patenting of genetic sequences. The report pro- poses areas of further exploration that could provide a foundation for the establishment of informed policies. This document is available free on line at: http://www.who.int/genomics/FullReport.pdf

14 Factor VII polymorphisms.pmd 108 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 109 Factors affecting the quality of life of hypertensive patients R.M. Youssef,1 I.I. Moubarak1 and M.I. Kamel1

ABSTRACT Using the Hypertension Health Status Inventory and multivariate analysis, predictors of quality of life were determined for a random selection of 316 hypertensive patients. Controlling for the effects of demographic and socioeconomic factors and existing co-morbidity, a better quality of life was independently predicted by achieving a controlled blood pressure and absence of target organ complications. Neither the number of antihypertensive drugs received nor the dose frequency affected patients’ quality of life. Presence of drug side-effects independently predicted a lower quality of life in the physical and emotional domains but not on aspects of daily living. The independent predictors explained 25%–30% of the variation in the quality of life of hypertensive patients. The study highlights the role of achieving blood pressure control to ensure a better quality of life for hypertensive patients.

Facteurs affectant la qualité de vie des patients hypertendus RÉSUMÉ On a déterminé les facteurs prédictifs de la qualité de vie pour une sélection aléatoire de 316 patients hypertendus à l’aide du questionnaire Hypertension Health Status Inventory et de l’analyse multivariée. Après élimination des effets des facteurs démographiques et socioéconomiques et de la comor- bidité existante, les facteurs prédictifs indépendants d’une meilleure qualité de vie étaient la normalisation de la pression artérielle et l’absence de complications des organes cibles. Ni le nombre des antihypertenseurs administrés ni la fréquence de la dose n’affectaient la qualité de vie des patients. La présence d’effets secondaires médicamenteux permettait de prédire indépendamment une moindre qualité de vie dans les domaines physique et psychologique mais non pour les aspects de la vie quotidienne. Les facteurs prédictifs indépendants expliquaient 25 % - 30 % de la variation de la qualité de vie des patients hypertendus. L’étude souligne le rôle du contrôle de la pression artérielle pour assurer une meilleure qualité de vie aux patients hypertendus.

1Department of Community Medicine, Faculty of Medicine, University of Alexandria, Alexandria, Egypt (Correspondence to R.M. Youssef: [email protected]). Received: 21/04/02; accepted: 23/01/04

15 Factors affecting the quality.pmd 109 10/25/2005, 1:23 AM 110 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction function and social activity with a deterioration of the sense of well-being as the du- As life expectancies increase and patients ration of the illness increases. Most are less likely to die from common infec- research in this area is in the form of clini- tious diseases, the focus of health care in- cal trials of individual drugs, focusing on creasingly shifts towards the prevention how blood pressure can be controlled in a and control of chronic disorders. These are way that minimizes adverse effects on pa- unlikely ever to be cured, and so new indi- tients’ quality of life [9,10]. Little knowl- cators of therapeutic success are needed: edge is available about other factors that clinicians place the emphasis on biomedical could contribute to the improvement of hy- outcome or improvement in physiological pertensive patients’ quality of life. status, whereas behavioural scientists look The purpose of this study was to identi- at psychological outcome [1]. Currently, it fy the predictors of a better quality of life seems that the gap between the 2 catego- of hypertensive patients in Alexandria, ries is narrowing as both professionals be- Egypt, using a disease-specific inventory. come interested in the same outcome for chronic disorders: quality of life [1]. The development of generic and dis- Methods ease-specific research tools has made it Patients easier to quantify the concept of quality of The target population was patients with hy- life [2]. Disease-specific instruments have pertension attending the health insurance the advantage of providing information rel- organization clinics in Alexandria for their evant to the underlying pathological condi- monthly prescription refill. Administrative- tion [2, Irvine SH, unpublished report for ly, the health insurance organization divides Glaxo–Wellcome Italy]. Both types of in- Alexandria into 6 medical zones. In each struments not only measure the patient’s zone, the service is provided through a general well-being but also take into con- number of polyclinics, each of which has sideration a relatively detailed assessment different specialized clinics. A multistage of function in the social, domestic and random sampling technique was adopted to work settings [3]. select 1 health insurance polyclinic from Regardless of the type of condition, pa- each zone and 1 internal medicine clinic tients’ quality of life is adversely affected from each polyclinic. All medical zones by chronic illness [4] and this is an impor- were considered in this study to allow the tant consideration in the management of representation of patients from different even asymptomatic conditions such as hy- socioeconomic backgrounds. Six clinics pertension [5]. However, quality of life of were identified, from which 10% of pa- hypertensive patients is often of little inter- tients attending for a prescription refill dur- est to clinicians as the disease is neither ing October 2001 were enrolled. painful nor debilitating [6]. Nevertheless, research has shown that the quality of life Questionnaire of hypertensive patients is not the same as All patients were interviewed using a ques- normotensive subjects [7,8]. Robbins et al. tionnaire consisting of the following sec- [8] revealed that blood pressure in unmedi- tions: cated patients adversely affects cognitive

15 Factors affecting the quality.pmd 110 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 111

Section 1: Sociodemographic informa- tively correlated and Cronbach alpha reli- tion including age, sex, marital status, edu- ability for the total scale was 0.951. The cational attainment and occupation. total scale score ranged from 26 to 116 Section 2: Detailed health history rele- where higher scores indicated a better qual- vant to the onset of the hypertension, as- ity of life. sociated complications and co-morbidity, Three subscales are derived from the pharmacological management as well as Hypertension Health Status Inventory: any drug side-effects experienced. Blood • Physical condition (6 items) assesses pressure was recorded. For all patients, the constraints that are likely to reduce hy- recorded blood pressure reflects the effect pertensive patients’ viability and ability of the last dose of medication taken the pre- (scores range from 6 to 24; all items vious day as usually the first morning dose were positively correlated; Cronbach alis postponed until the check-up is complet- pha reliability 0.758). ed. Accordingly, controlled blood pressure • Daily living (10 items) reveals patients’ was defined as systolic ≤ 140 mmHg and perception of the extent to which their diastolic ≤ 90 mmHg. daily living and social interaction are re- Section 3: Assessment of the quality of strained (scores range from 10 to 32; all life using a modified version of the Hyper- items positively correlated; Cronbach tension Health Status Inventory, which alpha reliability 0.805). consists of 29 items designed to investigate • Emotional status (10 items) portrays general quality of life [11]. After data col- the emotional contribution made by pa- lection, the scale was validated by testing tients’ perceptions of their current for inter-item correlation and reliability health status (scores range from 10 to analysis. Three of the original 29 items 60; all items were positively correlated; were omitted (sexual life has been affected; Cronbach alpha reliability 0.946). fear about holding a job; and afraid that work makes things worse) as they were Data handling and analysis negatively correlated with all items of the The overall scores on the Hypertension scale. This may be because they were inap- Health Status Inventory as well as the plicable to patients in the cultural context of scores on the 3 subscales were trans- Egypt: low levels of sexual activity among formed using the natural log to achieve a single people, failure of married people to normal distribution. To examine the relation distinguish between getting old and the ad- between the subscales, the scores on the 3 verse effect of hypertension or its treat- subscales were positively and significantly ment on sexual function, high job security correlated. The scores on the physical sub- among the employed and low employment scale were significantly and positively cor- among women. related with those of the daily living The final scale therefore consisted of 26 subscale (r = 0.833, P < 0.0001) and the items, 16 items with a 6-point response emotional subscale (r = 0.837, P = scale (from “Always” to “Never” or from 0.0001). Similarly, the scores on the emo- “Severely” to “Not at all”), while 10 items tional subscale were significantly and pos- had “Yes/No” responses. According to the itively correlated with those of the daily phrasing, 12 statements were scored in a living subscale (r = 0.848, P = 0.0001). positive direction, while 14 were scored in Hence, the 3 scales were considered col- the reverse direction. All items were posi- lectively in the analysis.

15 Factors affecting the quality.pmd 111 10/25/2005, 1:23 AM 112 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Data were analysed using SPSS, version An earlier paper from this study has re- 10. Means, standard deviations (SD) and ported the medical history and pharmaceu- Pearson correlation coefficients were com- tical management of the patients [12]. The puted. Scores on the overall quality of life patients had suffered from hypertension scale were transformed using the natural for a mean (SD) duration of 10.2 (7.7) log to achieve normality. Multivariate analy- years; minimum 0.25, maximum 40.0 sis of variance (MANOVA) was used to years. Just over half the patients (53.2%) test for the main effect of the demographic had achieved controlled blood pressure and variables and co-morbidity on the 3 sub- 34.5% suffered hypertension-related com- scales of the quality of life. Full factorial plications, the most common being cardiac analysis of covariance (ANCOVA) and events (99/109). Co-morbidity was noted multivariate analysis of covariance (MAN- in 52.8% patients, mainly diabetes mellitus COVA) were used to identify independent (104/167), arthritis (31/167) and bronchial predictors of the quality of life, controlling asthma (26/167). for demographic variables and existing co- In respect to pharmacotherapy, three- morbid conditions. Significance of the re- quarters of patients (75.0%) were managed sults was judged at the 5% level. by a single drug while the others were receiving a combination of 2 (23.4%) or 3 (1.6%) drugs of different classes. A single Results morning dose was prescribed for 55.1% of the patients while 44.3% were on a twice A total of 316 patients were enrolled in this daily dose. Only 11 patients (1.9%) report- study: 59.8% men and 40.2% women. The ed side-effects of the drugs. mean (SD) age was 59.4 (9.2) years; mini- On the Hypertension Health Status In- mum 35, maximum 83 years. Nearly two- ventory, significantly higher scores in the thirds of patients (67.8%) were married, physical, daily living and emotional do- 29.7% were widows or widowers, 1.9% mains were observed among married pa- were divorced and 0.6% remained single. tients (P = 0.002), those below the age of The highest proportion (40.2%) of pa- 60 years (P < 0.0001), the educated (P < tients had not received any formal educa- 0.0001) and those employed at the time of tion while the others had accomplished the study (P < 0.0001). The scores of men basic (12.3%) or high school education were significantly higher than of women (P (26.6%) and 20.9% were university gradu- = 0.002). However, significant differences ates. Just less than half of the men (47.1%) were observed only in the domain of emo- had professional or semi-professional oc- tions (P = 0.019). The presence of co- cupations, 20.7% were skilled or semi- morbid conditions was associated with skilled workers, 24.3% were unskilled or significantly lower scores (P < 0.0001) in manual labourers and 7.9% were drivers or the physical, daily living and emotional traders. At the time of the study, 64.6% domains (Table 1). were pensioners. Regarding women, Tables 2 and 3 portray the total scores 54.3% were housewives while 45.7% were on the Hypertension Health Status Invento- employed outside the home. Of the latter, ry, as well as the scores in the physical, 62.1% were still working at the time of the daily living and emotional domains, in re- study. lation to medical conditions and drug use

15 Factors affecting the quality.pmd 112 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 113

Table 1 Scores on the quality of life subscales in relation to patients’ demographic characteristics and co-morbid conditions

Characteristic Mean quality of life subscale scores (SD) F3,312 P-value Physical Daily living Emotional

Age (years) < 60 (n = 151) 18.73 (2.54) 27.17 (4.52) 44.73 (9.40) 10.137 ≥ 60 (n = 165) 16.67 (3.61) 24.24 (5.59) 39.53 (11.49) P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 Sex Men (n = 189) 17.44 (3.56) 25.72 (5.44) 42.64 (11.28) 5.011 Women (n = 127) 17.20 (3.26) 24.42 (5.37) 39.12 (10.55) P = 0.002 P = 0.693 P = 0.053 P = 0.019 Marital status Married (n = 214) 17.79 (3.41) 25.96 (5.33) 42.98 (10.85) 4.953 Singlea (n = 102) 16.42 (3.32) 23.59 (5.35) 37.54 (10.81) P = 0.002 P = 0.002 P = 0.001 P < 0.0001 Education Educatedb (n = 189) 18.11 (3.31) 26.09 (5.06) 43.46 (10.32) 8.335 Uneducated (n = 127) 16.21 (3.32) 23.86 (5.71) 37.91 (11.46) P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 Employment Employed (n = 103) 18.73 (2.54) 27.17 (4.52) 44.73 (9.40) 9.709 Unemployed (n = 213) 16.67 (3.61) 24.24 (5.59) 39.53 (11.49) P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 Co-morbidity Absent (n = 149) 18.32 (3.28) 26.86 (4.96) 44.46 (10.49) 9.015 Present (n = 167) 16.49 (3.35) 23.72 (5.44) 38.34 (10.89) P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001

aSingle includes never married, divorced, widows and widowers. bEducated includes primary, preparatory, secondary, university and higher. n = number of respondents. SD = standard deviation.

after controlling for patients’ demographic cantly lower scores on the Hypertension characteristics and co-morbidity. Signifi- Health Status Inventory (P = 0.005). These cantly higher scores on the Hypertension patients had significantly lower scores in Health Status Inventory and in the 3 do- the physical (P = 0.008) and emotional (P mains of the quality of life were observed = 0.001) domains but not in the domain of among patients who endured hypertension daily living (P = 0.063). Significantly for less than 10 years, those who were free higher scores were associated with single from hypertension-related complications as dose frequency (P = 0.046) while no sig- well as those who achieved blood pressure nificant difference was observed when control. In respect to drug factors, patients comparing patients’ scores in the 3 do- who reported drug side-effects had signifi- mains. On the other hand, no significant

15 Factors affecting the quality.pmd 113 10/25/2005, 1:23 AM 114 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 2 Multivariate analysis of covariance for the medical condition and drug factors in relation to the scores on the overall quality of life

Medical condition/ Mean quality of FP-value drug factor life scores (SD)

Duration of hypertension (years) < 10 (n = 159) 88.43 (17.95) 12.631 P < 0.0001 ≥ 10 (n = 157) 79.06 (19.01) Complications Absent (n = 207) 89.84 (16.86) 45.539 P < 0.0001 Present (n = 109) 72.27 (17.64) Blood pressure Controlled (n = 168) 88.07 (18.30) 18.748 Uncontrolled (n = 148) 78.91 (18.76) P < 0.0001 Number of drugs 1 (n = 237) 84.78 (19.04) 2.830 > 1 (n = 79) 80.74 (18.86) P = 0.094 Dose frequency per day 1 (n = 174) 86.21 (18.31) 4.037 > 1 (n = 142) 80.79 (19.56) P = 0.046 Drug side-effects Absent (n = 305) 84.43 (18.61) 7.930 P = 0.005 Present (n = 11) 65.54 (6.89)

n = number of respondents. SD = standard deviation.

difference was observed in relation to the same independent predictors. However, ab- number of antihypertensive drugs received. sence of drug side-effects independently Considering patients’ medical condition predicted higher scores in the physical (P = as well as drug factors, after controlling 0.009) and emotional (P = 0.001) domains for demographic characteristics and co- but not in the domain of daily living (P = morbid conditions, the full factorial analy- 0.088). This model explained 31.6% of the sis of covariance (Table 4) identified variation of the scores in the physical do- achieving blood pressure control, absence main, 29.9% in the daily living domain and of complications and absence of drug side- 25.2% in the emotional domain. effects as independent predictors of higher scores on the Hypertension Health Status Inventory. These factors explained 30.2% Discussion of the variability in patients’ score on the Improving the quality of life of the popula- quality of life scale. Regarding the scale’s tion will become an increasingly important domains (Table 5), the full factorial multi- goal for health professionals, supplanting variate analysis of variance revealed the

15 Factors affecting the quality.pmd 114 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 115

Table 3 Multivariate analysis of covariance for the medical condition and drug factors in relation to the 3 quality of life subscales

Medical condition/drug Mean quality of life subscale scores (SD) F3,270 P-value factor Physical Daily living Emotional

Duration of hypertension (years) < 10 (n = 159) 18.29 (3.08) 26.54 (5.06) 43.59 (10.66) 5.642 P = 0.001 ≥ 10 (n = 157) 16.38 (3.52) 23.83 (5.49) 38.83 (11.09) P < 0.0001 P = 0.001 P = 0.003 Complications Absent (n = 207) 18.47 (3.01) 27.06 (4.72) 44.30 (10.06) 19.998 P < 0.0001 Present (n = 109) 15.21 (3.19) 21.67 (4.97) 35.38 (10.70) P < 0.0001 P < 0.0001 P < 0.0001 Blood pressure Controlled (n = 168) 17.97 (3.29) 26.36 (5.13) 43.73 (10.74) 6.817 Uncontrolled (n = 148) 16.63 (3.46) 23.89 (5.50) 38.38 (10.88) P < 0.0001 P = 0.001 P < 0.0001 P < 0.0001 Number of drugs 1 (n = 237) 17.52 (3.42) 25.56 (5.39) 41.69 (11.04) 1.949 P = 0.122 > 1 (n = 79) 16.81 (3.44) 24.10 (5.46) 39.83 (11.29) Dose frequency per day 1 (n = 174) 17.71 (3.28) 25.83 (5.20) 42.67 (10.53) 1.504 > 1 (n = 142) 16.91 (3.58) 24.42 (5.64) 39.46 (11.53) P = 0.214 Drug side-effects Absent (n = 305) 17.48 (3.39) 25.40 (5.36) 41.74 (10.82) 4.297 Present (n = 11) 14.81 (3.33) 21.37 (5.69) 31.50 (12.34) P = 0.006 P = 0.008 P = 0.063 P = 0.001

n = number of respondents. SD = standard deviation.

Table 4 Independent predictors of overall the current medical and social objectives quality of life [1]. On the community and individual levels, quality of life can be affected by health Independent predictor FP-value status [13] and socioeconomic indicators Absence of complications 38.086 P < 0.0001 [1,13]. It has been emphasized that low quality of life is related to poverty, non- Controlled blood pressure 15.513 P < 0.0001 completion of high school and unemploy- Absence of drug side- ment [1, Irvine SH, unpublished report]. effects 7.782 P = 0.006 This agrees with our study where better 2 Model R 0.409 quality of life in all domains was observed Adjusted R2 0.302 among hypertensive patients who were

15 Factors affecting the quality.pmd 115 10/25/2005, 1:23 AM 116 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 5 Independent predictors of the 3 quality of life domains

Independent predictor Quality of life domains F3,265 P-value Physical Daily living Emotional

Absence of complications P < 0.0001 P < 0.0001 P < 0.0001 15.700 P < 0.0001 Controlled blood pressure P = 0.005 P < 0.0001 P < 0.0001 5.708 P = 0.001 Absence of drug side-effects P = 0.009 P = 0.088 P = 0.001 4.206 P = 0.006 Model R2 0.420 0.405 0.366 Adjusted R2 0.316 0.299 0.252

better educated and those who were em- health problem. A similar problem con- ployed at the time of the study. The favour- fronted Stewart et al. who highlighted the able effect of social stability was importance of considering the primary di- demonstrated by the significantly higher agnosis along with any co-morbid condi- quality of life for patients who were married. tion in controlling for differences in This study and a previous one [Irvine SH, patient case-mix in studies of quality of unpublished report] have demonstrated an life [4]. age and sex variation in the quality of life of Identifying aspects of hypertension hypertensive patients. Better quality of life that contribute to a better quality of life overall was observed among men and those requires that the effects of demographic in the younger age group, and men scored factors and existing co-morbidity are re- significantly higher in the emotional contri- moved from the analysis. In this study, bution of the perception of illness suggesting after controlling for these factors, longer that men tolerate chronic illness without be- duration of the original illness was associ- ing emotionally affected to the same extent ated with a significant reduction in quality as women. of life. This was also reported by Robbins Stewart et al. in their study on the quality et al. who observed that the duration of of life of patients with chronic disorders in- hypertension is inversely proportional to dicated that patients with multiple ailments the dimension of general well-being showed greater decrements in functioning among women [8]. However, the effect and well-being than those with only one con- of duration of hypertension was eliminat- dition [4]. Hypertensive patients are no ex- ed from the model when other parameters ception to this finding as co-morbidity was were considered. Our study as well as associated with a deterioration of the quality previous ones [2,8, Irvine SH, unpub- of life in all domains. Each additional ailment lished report] demonstrated the impact of exerts its toll on a patient’s physical condi- achieving controlled blood pressure on the tion, limiting his/her day-to-day activities quality of life of hypertensive patients. In- with considerable repercussions on emotion- deed, blood pressure control independent- al status. Co-morbid conditions were en- ly predicted a better quality of life in all countered among more than half of our domains. A mutual interaction between patients, and this precluded any attempt to blood pressure control and quality of life exclude those with more than one chronic has been postulated. Uncontrolled blood

15 Factors affecting the quality.pmd 116 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 117

pressure alters hypertensive patients’ quali- the unfavourable consequences of hyper- ty of life through anxiety and depressive tensive drug side-effects on quality of life, reactions, and poor quality of life hampers particularly in the physical and emotional blood pressure control even with a thera- domains. It is therefore essential for clini- peutic regimen [2]. To prevent patients cians to select the drug best tolerated by from being trapped in this vicious circle, it individual patients to balance bio-physio- is important to consider dimensions of the logical needs with quality of life. quality of life along with pharmacological Predictors identified by this study ex- compliance in achieving blood pressure plained only a small proportion (25%–30%) control, as among this group of patients of the variation in the quality of life of hy- controlled blood pressure was achieved pertensive patients. This could be attributed with at least 90% pharmacological compli- to the fact that the precise elements con- ance [12]. Besides its direct effect on qual- tributing to the quality of life vary substan- ity of life, achieving and maintaining blood tially in different individuals. Other factors pressure control minimizes complications could have influenced the quality of life of resulting from target organ damage with these patients, such as: the quality and subsequent beneficial effects [14]. In our quantity of social support [1,19], stressful study, the presence of complications of hy- events recently encountered [19] and pa- pertension independently predicted lower tients’ quality of life before the start of ther- quality of life in all domains. apy [20]. However, our study points to the Clinical trials often aim to identify the importance of achieving blood pressure superiority of one antihypertensive drug control and minimizing the occurrence of over another in controlling blood pressure complications to ensure better quality of without impairing quality of life [9,10]. The life for hypertensive patients. In this re- nature of our study design did not allow the spect it is important to select the antihyper- investigation of the effect of different ther- tensive drug that is the best tolerated by an apeutic agents on quality of life. However, individual patient in order to achieve the neither the number of prescribed antihyper- most successful patient outcome. tensive drugs nor the dose frequency had any influence on patients’ quality of life. Acknowledgement McInnes [15] suggested that aggressive treatment with a combination of antihyper- The authors acknowledge Dr Fabio tensives does not necessarily impair quality Arpinelli, Health Outcomes Research Lead- of life but a large body of literature has doc- er, Medical Department, GSK Italy, for umented the role of drug side-effects providing the Hypertension Health Status [3,9,10,16–18]. The present study revealed Inventory.

References 1. Clark NM. Understanding individual and qualité de vie de l’hypertension collective capacity to enhance quality of arterielle. [Socioeconomic and quality of life. Health education & behavior, 2000, life repercussions of arterial hyperten- 27(6):699–707. sion]. Drugs, 1998, 56(suppl. 2):45–53. 2. Selke B, Marquis P, Lebrun T. Les reper- 3. Weber MA, Radensky P. Measurement of cussions socio-économiques et de short-term, intermediate, and long-term

15 Factors affecting the quality.pmd 117 10/25/2005, 1:23 AM 118 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

outcomes of treating hypertension. Car- among hypertensive patients. Eastern diology clinics, 1996, 14(1):131–42. Mediterranean health journal, 2002, 8(4/ 5):579–92 4. Stewart AL et al. Functional status and well-being of patients with chronic con- 13. Centers for Disease Prevention and ditions. Results from the Medical Out- Control. Community indicators of health comes Study. Journal of the American related quality of life—United States, Medical Association, 1989, 262(7):907– 1993–1997. Morbidity and mortality 13. weekly report, 2000, 49:281–5. 5. Plaisted CS et al. The effects of dietary 14. Girvin B, Johnston D. The implication of patterns on quality of life: a substudy of non-compliance with antihypertensive the Dietary Approaches to Stop Hyper- medication. Drugs, 1996, 52(2):186–95. tension trial. Journal of the American Di- 15. McInnes GT. Integrated approach to etetic Association, 1999, 99(suppl. 8): management of hypertension: promoting S84–9. treatment acceptance. American heart 6. Wiklund I. Quality of life and cost-effec- journal, 1999, 138(3Pt2):S252–5. tiveness in the treatment of hyperten- 16. Dahlöf B et al. Main results of the sion. Journal of clinical pharmacy and losartan versus amlodipine (LOA) study therapeutics, 1994, 19(2):81–7. on drug tolerability and psychological 7. Lawrence WF et al. Health status and general well-being. Journal of hyperten- hypertension: a population based study. sion, 1997, 15:1327–35. Journal of clinical epidemiology, 1996, 17. Weir MR et al. Efficacy, tolerability, and 49:1239–45. quality of life of losartan, alone or with 8. Robbins MA et al. Unmedicated blood hydrochlorothiazide, versus nifedipine pressure levels and quality of life in eld- GITS in patients with essential hyperten- erly hypertensive women. Psychoso- sion. Clinical therapeutics, 1996, 18: matic medicine, 1994, 56:251–9. 411–28. 9. Anderson RB, Hollenberg NK, Williams 18. Omvik P et al. Evaluation and quality-of- GH. Physical Symptoms Distress Index: life assessment of amlodipine and a sensitive tool to evaluate the impact of enalapril in patients with hypertension. pharmacological agents on quality of Journal of human hypertension, 1995, life. Archives of internal medicine, 1999, 9(suppl. 1):S17–24. 159(7):693–700. 19. Constant P. Quatre facteurs contribuent à 10. Testa MA. Methods and applications of la qualité de vie de l’hypertendu. quality-of-life measurement during anti- Quotidien du médicine, 1996, 5838:21. hypertensive therapy. Current hyperten- 20. Testa MA et al. Quality of life and antihy- sion reports, 2000, 2(6):530–7. pertensive therapy in men. A compari- 11. Hypertension Health Status Inventory, son of captopril with enalapril. The Quality of Life Instruments Database, Quality-of-Life Hypertension Study Mapi Research Trust, Lyon, France (http:/ Group. New England journal of medi- /www.qolid.org/public/HYPER.html, ac- cine, 1993, 328:907–13. cessed 25 January 2005). 12. Youssef RM, Moubarak II. Patterns and determinants of treatment compliance

15 Factors affecting the quality.pmd 118 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 119 Predicting serum gastrin levels among men during Ramadan fasting M.Z.A. Nomani1, A.H. Khan2, M.M. Shahda2, A.K. Nomani3 and S.A. Sattar 4

ABSTRACT Increased gastric acidity is common when fasting during Ramadan. Our study aimed to develop a regression equation to predict fasting serum gastrin levels using parameters commonly analysed in clinical laboratories. Fasting blood samples from six men were taken on days 1, 10, 19, 26 and 28 of Ramadan. Serum gastrin, total cholesterol, urea and uric acid were analysed. All 5 samples from each man were included in multiple regression analysis and the prediction equation obtained was: serum gastrin, pg/mL = 198.27 – 0.199 total cholesterol (mg/dL) + 2.525 urea (mg/dL) – 103.238 uric acid (mg/dL) + 10.923 uric acid (mg/dL)2 + 3.683 body mass index, r 2 = 0.75, P < 0.001. This equation might be used to estimate gastrin levels and plan dietary and medicinal measures to avoid high gastric acidity during Ramadan.

Prédiction du taux de gastrine sérique chez des hommes pendant le jeûne du ramadan RÉSUMÉ Une augmentation de l’acidité gastrique est courante pendant le jeûne du ramadan. Notre étude visait à établir une équation de régression pour prédire le taux de gastrine sérique à jeun en utilisant les paramètres couramment analysés dans les laboratoires cliniques. Des prélèvements sanguins à jeun ont été effectués chez six hommes au 1er, 10e, 19e, 26e et 28e jour de ramadan. La gastrine sérique, le cholestérol total, l’urée et l’acide urique ont été analysés. Les cinq échantillons de chaque homme ont tous été inclus dans l’analyse de régression multiple et l’équation de prédiction obtenue était la suivante : gastrine sérique, pg/mL = 198,27 – 0,199 cholestérol total (mg/dL) + 2,525 urée (mg/dL) – 103,238 acide urique (mg/dL) + 10,923 acide urique (mg/dL)2 + 3,683 indice de masse corporelle, r 2 = 0,75, p < 0,001. Cette équation pourrait être utilisée pour estimer le taux de gastrine et prévoir des mesures médicamenteuses et diététiques permettant d’éviter une forte acidité gastrique pendant le ramadan.

1Human Nutrition and Foods Program, 2Chestnut Ridge Psychiatric Hospital, 3School of Nursing; 4Department of Agricultural Biochemistry, West Virginia University, Morgantown, West Virginia, United States of America (Correspondence to M.Z.A. Nomani: [email protected]). Received: 02/04/03; accepted: 18/12/03

16 Predicting serum gastrin.pmd 119 10/25/2005, 1:23 AM 120 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction which mainly consisted of 240 mL of vegetable or pasta soup, 2 cups (410 g) of rice Several reports have indicated the problem and one and a half servings (160 g) of a of increased gastric acidity and episodic lamb and chicken combination. The meal peptic ulcers during the fasting month of also included either 3/4 cup of tabouli Ramadan [1,2]. During Ramadan changes (chopped parsley, lettuce, green onion in total cholesterol, uric acid and urea have leaves and bulgar with olive oil or corn oil been observed as well [3–9]. and lemon juice or vinegar dressing) or sal- A negative association between fasting ad (lettuce, tomato and cucumber with ol- serum gastrin level and gastric acidity has ive oil or corn oil and lemon juice or vinegar been observed [10]. Others have also noted dressing). To be sure that the subjects con- low plasma gastrin concentrations and in- sumed sufficient dietary fibre, they ate an creased intragastric acidity [11,12]. The apple [171.6 (SD 7.6) g] and a tangerine possibility of using fasting blood gastric [123.9 (SD 6.1) g] immediately after each level to predict gastric acidity has been sug- dinner. At sahur, the predawn meal, a bowl gested [10]. The objective of our study of vegetable salad (33 g each of lettuce, was to develop a prediction equation for shredded carrot, tomato and oil dressing) fasting serum gastrin level during Ramadan and a sandwich (two slices of bread, 58g, using blood parameters routinely analysed and lamb/beef meat or chicken, 50 g) or in clinical laboratories such as cholesterol, bran cereal (56g), and one and half cups of urea and uric acid. Screening for gastrin milk (366g) were consumed. Sweet des- for those who fast during Ramadan or for serts were consumed sparingly, such as 1– other reasons may help to define dietary 2 pieces of baklava on Fridays at dinner. and medicinal measures to prevent epi- Diet analysis software (Food processor for sodes of peptic and duodenal ulcer. Windows, version 7, ESHA Research, Sa- lem, Oregon, USA) was used to estimate Methods energy [2072 (SD 307) kcal/day], protein [9.6 (SD 17.5) g], fat [70.6 (SD 14.8) g], The study was approved by the Institution- carbohydrate [305.9 (SD 41.1 g)] and di- al Review Board of West Virginia Universi- etary fibre intake [24.2 (SD 7.6) g]. Blood ty, United States of America, for the samples were drawn at 16:00–16:15, i.e. protection of human subjects. Six healthy approximately one hour before iftar male college students and faculty members (breaking of the fast at sunset) on days 1, participated in the study from November 10, 19, 26 and 28 of Ramadan. Total fast- through December 2000. The mean and ing hours per day were approximately 11.5 standard deviation (SD) for age, weight hours (predawn to sunset) and blood sam- and body mass index [BMI, body weight in ples were taken after 10.5 hours of fasting. kilograms/(height in metres)2] were 33.5 Serum gastrin, total serum cholesterol, (SD 17.9) years, 81.1 (SD 10.5) kg and urea and uric acid were analysed by enzy- 26.5 (SD 1.4) kg/m2 respectively. matic methods [13–15]. Serum gastrin was The subjects broke their fast at sunset analysed by radioimmunoassay (ION Phar- by consuming 3 dates and 240 mL of or- maceuticals, Orangeburg, New York, ange juice or cow’s milk (3% fat). Immedi- USA). The data were statistically analysed ately after maghrib prayer, they ate dinner, with Statistica software.

16 Predicting serum gastrin.pmd 120 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 121

Results Gastrin = 13.065 – 0.2875 TC + 3.893 BMI (r2 = 0.4, P < 0.00l) There was no significant difference in Gastrin = 8.53 – 13.561 U + 0.5814 U2 + 4.5 mean serum gastrin levels with day of fast- BMI (r2 = 0.58, P < 0.001) ing: day 1 [57.3 (SD 22.l) pg/mL], day 10 Gastrin = 155.067 – 103.189 UA +11.514 UA2 [62.7 (SD 19.0) pg/mL], day 19 [54.5 (SD + 4.753 BMI (r2 = 0.48, P < 0.001) 15.2) pg/mL], day 26 [54.7 (SD 13.5) pg/ mL] and day 28 [59.3 (SD 20.3) pg/mL]. where: TC = total cholesterol, BMI = body Therefore, all 30 observations were con- mass index, U = urea and UA = uric acid. sidered as a single set of data for regression The equation with total cholesterol, urea analysis. The mean (SD) for serum gastrin, and BMI as predictors was: total cholesterol, urea, uric acid and BMI for the whole experimental period was Gastrin = – 24.6048 – 0.5205 TC – 0.0018 TC2 [57.7 (SD 17.3) pg/mL], [203 (SD 28) mg/ – 7.988 U + 0.383 U2 + 3.2171 dL], [14.1 (SD 3.2) mg/dL], [5.01 (SD BMI (r2 = 0.67, P < 0.001). 0.65) mg/dL] and [26.5 (SD 1.4) kg/m2] respectively. where: TC = total cholesterol, U = urea and A single predictor serum variable was BMI = body mass index. first used to predict the serum gastrin vari- The data were further analysed with a able. Gastrin (pg/mL) was negatively asso- stepwise forward multiple regression ana- ciated with serum total cholesterol level lytical model (Table 1). In the model, the (mg/dL): Gastrin 127.25 – 0.3419 total variance in serum gastrin was a good fit cholesterol (correlation coefficient, r = (R2 = 75.3%, i.e. the percent coeffi- 2 gastrin –0.55, coefficient of determination, r = cient of determination). Total cholesterol 0.3, P < 0.0l). contributed the most to the variance With regard to serum urea and uric acid (30.3%), followed by urea (25.5%), uric (mg/dL), there was a curvilinear increase in acid and its square (11.8%) and BMI increasing order in serum gastrin levels (7.5%). Partial regression coefficients for with these variables. the predictor variables were significant at P < 0.05 or P < 0.01 levels as well. The mul- Gastrin = 136.68 – 15.0896 U + 0.6419 tiple regression equation derived was: U2 (r2 = 0.44, P < 0.01) Gastrin = 210.34 – 77.1039 UA + 9.16 S. gastrin = 198.27 – 0.199 TC + 2.525 U – 2 2 UA (r = 0.34, P < 0.01) 103.238 UA + 10.923 UA2 + 3.883 BMI (r2 = 0.753, P < 0.001) where: U = urea and UA = uric acid. An association between gastrin and where: S. gastrin = serum gastrin, TC = to- 2 BMI was noted as well (r = 0.45, P < tal cholesterol, U = urea, UA = uric acid and 0.05). Serum gastrin level increased with BMI = body mass index. increasing BMI. In the next step of the analysis, multiple regression equations were developed that Discussion included BMI in the model. Inclusion of BMI improved r2 values in all 3 equations by Often because of changes in meal and sleep more than 10% as compared with earlier patterns people who fast experience irregu- equations with one predictor variable only. larity in bowel movements during the

16 Predicting serum gastrin.pmd 121 10/25/2005, 1:23 AM 122 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 1 Summary of stepwise forward multiple regression analysis, serum gastrin as dependent variable

Predictor variable Step Multiple Multiple R2 F to P- Variables in R R2 change enter value included

Total cholesterol 1 0.551 0.304 0.304 12.2 0.002 1 Urea 2 0.747 0.559 0.255 15.6 0.001 2 Body mass index 3 0.796 0.633 0.075 5.3 0.03 3 Uric acid2 4 0.82 0.672 0.039 3.0 0.09 4 Uric acid 5 0.868 0.753 0.081 7.8 0.01 5

month of Ramadan. In our study, subjects The gallbladder content is mainly bile salts had regular bowel movements and had no and cholesterol that are needed for emulsi- flatus or constipation. Furthermore, the fying dietary fat. The duodenum also se- men did not experience any feelings of cretes gastric inhibitory peptides. Gastric heaviness or discomfort in the chest or of juice is mainly hydrochloric acid and pro- acidity in the mouth, i.e. acid reflux. This tein digesting enzymes. With high gastric indicates that the intake of good sources of acidity (< pH 3), gastrin secretion declines, dietary fibre such as fruits after dinner and contributing to reduced or halted gastric salad in sahur helped to maintain the normal secretion. Generally protein-rich foods also transit time of food in the gastrointestinal contain high amounts of purines and pyri- tract (GIT). Total mean dietary fibre intake midines. Urea and uric acid are catabolic was 24.2 (SD 7.6 g/day) [20.5 (SD 6.4) g products of amino acid metabolism and pu- insoluble fibre and 3.7 (SD 1.3) g soluble rine metabolism respectively. Fasting also fibre]. Fibre, especially insoluble fibre, creates physiological stress [9]. Serum through GIT distension, contributes to mo- urea can be the result of catabolism of body tility of gastrointestinal tract churning, mix- proteins used as a source of energy or for ing of content and emptying of chyme gluconeogenesis [8]. [16]. The positive association between BMI Serum cholesterol, urea and uric acid and serum gastrin level in our study agrees association with gastrin can be explained with the findings of others [24,25]. Serum through the regulatory role of duodenal gastrin has been positively associated with hormone, cholecystokinin, and the role of BMI [24]. Significantly high serum gastrin dietary protein, fat and fibre in inducing levels have been noted among obese wom- hormonal and gastric secretions and gastric en compared with lean subjects [25]. emptying [16–23]. A negative relationship between fasting Dietary protein first buffers, then trig- serum gastrin (pg/mL) and percent of time gers gastrin secretion that stimulates gas- or duration of high acidity in 24 hours has tric juice secretion [19]. Cholecystokinin is been previously reported (pH < 4, r = secreted in response to the emptying of fat –0.55, P < 0.01) [10]. In a rabeprazole and other digesta (chyme) from stomach dose-related response study, decreased in- into the duodenum; the hormone triggers tra- gastric acidity (median 24-hour inte- emptying of gallbladder content [17,18]. grated acidity) and increased plasma

16 Predicting serum gastrin.pmd 122 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 123

gastrin (median 24-hour integrated gastrin) Based upon the findings of our study were noted [11]. Baak et al. found a signif- and a review of the literature, the diagram- icant elevation of serum gastrin with the in- matic representation of fasting serum gas- creased gastric pH [12]. Contrary to the trin to various serum variables may be some findings, Iraki et al. noted decreased described as follows: gastric pH and increased plasma gastrin during Ramadan [1,10–12]. In the Iraki re- ↓Cholesterol, ↑urea, ↑uric acid or↑BMl = port, gastrin level went up on day 10 of ↑Gastrin = ↑Gastric pH or ↓Acidity Ramadan and down on day 24 in comparison with levels before or after the Ramadan period. There was a significant difference Conclusion between the mean gastrin levels of these 2 The prediction equation generated by our days of Ramadan [day 10 gastrin = 76.0 study may be useful for the estimation of (SD 2.4) pg/mL, pH = 2.35 (SD 0.06); day fasting serum gastrin levels and for the 24 gastrin = 56.5 (SD 2.3) pg/mL, pH = planning of an appropriate diet. The diet 2.25 (SD 0.05)]. There seemed to be a pos- from all food groups which includes suffi- itive association between the two variables. cient dietary fibre through intake of fresh We, however, did not find a significant dif- fruit/s at the end of each meal or before ference in serum gastrin levels with various sleep is suggested. An appropriate diet and days of fasting, which may be due to pres- medicine regime may help avoid high gas- ence of sufficient dietary fibre in the diets tric acidity during Ramadan fasting and of our study subjects. thus reduce gastrointestinal discomfort and Gastric emptying is an important aspect the risk of gastric ulcer. of avoiding high gastric acidity. Dietary fi- Studies need to be conducted at institu- bre, especially insoluble fibre, plays an im- tions with 24-hour gastric sample collec- portant role in gastric emptying and chyme tion facilities to develop an equation movement in the digestive tract [16]. There between fasting serum gastrin levels and is a need for further Ramadan studies on blood parameters among people with opti- gastrin and gastric acid regulation on sub- mal nutrient and fibre intake and to develop jects with optimal dietary fibre and nutrient a gastrin–gastric acidity equation. intake.

References 1. Iraki L et al. Ramadan diet restrictions 3. Hallak MU, Nomani MZA. Body weight modify the circadian time structure in hu- loss and changes in blood lipid levels in mans. A study on plasma gastrin, insulin, normal men on hypocaloric diets during glucose and calcium and on gastric pH. Ramadan fasting. American journal of Journal of clinical endocrinology and clinical nutrition, 1988, 48:1197–210. metabolism, 1997, 82:1261–73. 4. Adlouni A et al. Fasting during Ramadan 2. Donderici O et al. Effect of Ramadan on induces a marked increase in high- peptic ulcer complications. Scandana- density lipoprotein cholesterol and de- vian journal of gastroenterology, 1994, crease in low-density lipoprotein choles- 29:603–6.

16 Predicting serum gastrin.pmd 123 10/25/2005, 1:23 AM 124 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

terol. Annals of nutrition and metabolism, 13. Allain CC et al. Enzymatic determination 1997, 41:242–9. of total serum cholesterol. Clinical chemistry, 1974, 20:470–5 5. Akanji AO, Mojiminiyi CA, Abdella N. Beneficial changes in serum apo A-1 14. Marsh WH, Fingernut B, Miller H. Auto- and its ratio to apo B and HDL in stable mated and manual methods of the de- hyperlipidaemic subjects after Rama- termination of blood urea. Clinical chem- dan fasting in Kuwait. European journal istry, 1965, 11:624–7. of clinical nutrition, 2000, 54:508–13. 15. Town MH et al. A sensitive colorometric 6. Gumaa KA et al. The effects of fasting in method for the enzymatic determination Ramadan. I. Serum uric acid and lipid of uric acid. Journal of clinical biochem- concentrations. British journal of nutri- istry, 1985, 23:591. tion, 1978, 40:573–81. 16. Roberfroid M. Dietary fiber, insulin, and 7. Nomani MZA. Dietary fat, blood choles- oligofructose: a review comparing their terol and uric acid levels during physiological effects. Critical reviews in Ramadan fasting. International journal food science and nutrition, 1993, 33:553. of Ramadan fasting research, 1997, 17. Konturek JW. Cholecystokinin in the con- 1:1–6. trol of gastric acid and plasma gastrin 8. Nomani MZA et al. Changes in blood and somatostatin secretion in healthy urea and glucose and their association subjects and duodenal ulcer patents with energy containing nutrients in men before and after eradication of Heli- on hypocaloric diets during Ramadan cobacter pylori. Journal of physiology fasting. American journal of clinical nutri- and pharmocology, 1994, 45: 3(suppl. tion, 1989, 49:1141–5. 1):3–66. 9. Afrasiabi A et al. Effects of low fat and low 18. Riber C et al. Gallbladder emptying and calorie diet on plasma lipid levels in the cholecystokinin response to fish oil and fasting month of Ramadan. Saudi medi- trioleate ingestion. Digestion, 1996, 57: cal journal, 2003, 24(2):184–8. 161–4. 10. Bonapace ES, Fisher RS, Parkman HP. 19. Inoue S, Nakagawa T. Gastrin response Does fasting serum gastrin predict gas- to protein test meal in gastric disease. tric acid suppression in patients on Tohoku journal of experimental medi- proton-pump inhibitors? Digestive dis- cine, 1980, 132:61–7. eases and sciences, 2000, 45:34–9. 20. Brooks FP. Effect of diet on gastric secre- 11. Williams MP et al. A placebo-controlled tion. American journal of clinical nutri- study to assess the effects of 7-day dos- tion, 1985, 42(5 suppl.):1006–19. ing with 10, 20 and 40 mg rabeprazole 21. Konturek SJ et al. Cholecystokinin in the on 24-h intragastric acidity and plasma inhibition of gastric secretion and gastric gastrin in healthy male subjects. Alimen- emptying in humans. Digestion, 1990, tary pharmacology and therapeutics, 45:1–8. 2000, 14:691–9. 22. Armand M et al. Dietary fat modulates 12. Baak LC et al. Repeated intravenous gastric lipase activity in healthy humans. bolus injections of omeprazole: effects American journal of clinical nutrition, on 24-hour intragastric pH, serum gas- 1996, 62:74–80. trin, and serum pepsinogen A and C. Scandanavian journal of gastroentre- 23. Borovicka J et al. Role of lipase in the roloy, 1991, 26:737–46. regulation of postprandial gastric acid

16 Predicting serum gastrin.pmd 124 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 125

secretion and emptying of fat in humans: 25. Baranowska B et al. Disturbed release of a study with orlistat, a highly specific li- gastrointestinal peptides in anorexia pase inhibitor. Gut, 2000, 46:774–81. nervosa and in obesity. Diabetes obesity and metabolism, 2000, 2:99–103. 24. Linstedt G et al. Hypergastrinemia—a risk factor for myocardial infarction? Clinical chemistry, 1985, 31:585–90.

Note from the Editor We would like to draw our readers’ attention to the Guidelines for Authors which are published at the back of each issue of the journal in English, Arabic and French. We wish to remind our potential authors of the importance of applying the editorial requirements of the EMHJ when preparing their manuscripts for submission for publication. We regret that we are unable to accept papers that do not conform to the editorial requirements.

16 Predicting serum gastrin.pmd 125 10/25/2005, 1:23 AM 126 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Spirometric changes following the use of pesticides P. Salameh,1 M. Waked,2 I. Baldi 3 and P. Brochard 3

ABSTRACT We compared the respiratory function of 19 pesticide factory workers and a control group of 43 other factory workers in Lebanon. The groups had no difference in smoking status. Baseline measurements

of respiratory function showed significantly lower forced expiratory volume and flow rates (FEV1, FEF25–75%,

and FEV1/FVC ratio) among subjects working with pesticides, i.e. obstruction may be linked to chronic exposure to pesticides. After 4 hours of work, all respiratory variables were still significantly lower in pesticide-exposed subjects, but no acute changes in respiratory function were seen. Pesticide-exposed workers

had 5.6 times higher risk of abnormal FEV1/FVC ratio and 16.5 higher risk for abnormal FEF25–75%. Duration of occupation in the pesticide factory was significantly correlated with abnormal respiratory measures.

Modifications spirométriques suite à l’utilisation de pesticides RÉSUMÉ Une étude réalisée au Liban a comparé la fonction respiratoire de 19 travailleurs des usines de pesticides et d’un groupe de 43 travailleurs d’autres usines. Il n’y avait pas de différences entre les groupes concernant le statut tabagique. Les mesures initiales de la fonction respiratoire ont montré des taux significativement inférieurs pour le volume expiratoire maximum seconde et le débit expiratoire maximal (VEMS, DEM25-75 %, rapport VEMS/CVF) chez les sujets travaillant avec des pesticides, c’est-à-dire que l’obstruction peut être liée à l’exposition chronique aux pesticides. Après quatre heures de travail, toutes les variables respiratoires étaient significativement inférieures chez les sujets exposés aux pesticides, mais aucune modification de la fonction respiratoire n’a été observée. L’odds ratio a montré que les ouvriers exposés aux pesticides avaient un risque 5,6 fois plus élevé d’avoir un rapport VEMS/CVF anormal et

16,5 fois plus élevé d’avoir un DEM25-75 % anormal. Il y avait une corrélation significative entre la durée de l’emploi dans l’usine de pesticides et les mesures respiratoires anormales.

1Faculty of Pharmacy, Lebanese University, Beirut, Lebanon (Correspondence to P. Salameh: [email protected]). 2Pulmonology Service, Saint George Hospital, Beirut, Lebanon. 3Faculty of Public Health, Institute of Public Health, Epidemiology and Development (ISPED), Victor Segalen Bordeaux 2 University, Bordeaux, France. Received: 05/06/04; accepted: 25/02/04

17 Spirometric changes.pmd 126 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 127

Introduction Acute toxic effects are well known from basic toxicological and clinical stud- Occupational exposure to pesticide pro- ies. Immunological and neurological ef- duces the greatest hazard for individual ex- fects on the respiratory system linked to posure to these toxic substances [1], by the local and systemic toxic actions of pes- inhalation, dermal or oral exposure [2,3]. ticides have been reported. Beside allergic Exposure levels depend on the nature and phenomena, the development of respiratory formulation of the pesticide, utilization problems is mainly due to the overwhelm- techniques, hygiene and environmental ing toxic effect on cell reparation [12]. Ox- conditions [4]. The physical, chemical and idative stress, present whenever some aerodynamic properties of the product, pesticides are inhaled, may initiate diseases anatomy of the respiratory tract and indi- such as chronic bronchitis, emphysema, fi- vidual physiology are the main determi- brosis and cancer [13]. Occupational asth- nants of particle deposition [3]. ma, reactive–airways dysfunction syn- In developed countries, legislation pro- drome, or even a transient hypersensitivity vides some protection against the dangers elicited by irritation may occur [14]. of occupational use of pesticides [2], This study in Lebanon aimed to evaluate whereas in developing countries, the safety the respiratory function variables of work- problems are neglected and workers may ers employed in a pesticides processing not be legally protected against the toxic factory in a developing country and com- effects of pesticides at work [5]. Inhalation pare the effect of acute exposure to pesti- of pesticide dust, vapours, mists and gases cides of these workers with a group of may therefore represent a significant occu- workers not occupationally exposed to pational hazard [6] with workers vulnerable pesticides. to acute toxic episodes and chronic intoxication [1]. Among occupationally exposed workers, the prevalence of symptoms and Methods the risk of acute intoxication increase with increasing number of working years [7]. The study compared the respiratory func- Few epidemiological studies have been tion of a group of factory workers occupa- conducted throughout the world to evalu- tionally exposed to pesticides and a control ate the short and long-term effects on the group not occupationally exposed. Work- respiratory system of occupational expo- ers were interviewed and respiratory func- sure to pesticides. Mortality and morbidity tion measurements were taken before and studies are commonly performed, whereas after a work shift. The study was conduct- respiratory effects have been less studied, ed in March 2000, when the external tem- due to the difficulty of evaluating respirato- perature in the region reached an average ry exposure [8–11] in addition to the con- daily maximum of 25 °C. founding effect of workers’ lifestyle (smoking, physical conditions, climate). Sample Moreover, pesticide residues can have The pesticide-exposed sample was taken physicochemical transformations into from a pesticides conditioning factory in products that are more or less active, and Hosrael at Byblos (Mount Lebanon). All the can be found in the dust to which workers factory workers, except the administrative are exposed [8].

17 Spirometric changes.pmd 127 10/25/2005, 1:23 AM 128 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

workers, were included in the study sam- suffered (chronic cough, expectoration, ple. wheezing and dyspnoea) were noted. The control sample was taken from workers at 4 nutritional products factories, Respiratory function chosen to ensure the absence of occupa- measurements tional exposure to respiratory pesticides: Respiratory measurements were per- Choueifat (fruit juice dilution and condi- formed before the morning shift (08.00 tioning), Harissa (cheese production from hours) and 4 hours later (12.00 hours). fresh milk), Bekaata (ice cream produc- Respiratory function measurements were tion) and Kfarkatra (sweets baking). All made using a portable spirometer, the non-administrative workers were included Spirovit SP-1 (Schiller, Switzerland) with in the control sample. disposable filters (SP-150), designed for The interviewees were informed that ambulatory pulmonary function measure- the study was to evaluate “their general ment. Calibration was carried out in the health status” and all of them gave their morning of each use. The device measures consent to participation. actual respiratory flow, at a precision of 2%, in addition to predicted values accord- Questionnaire ing to age, sex, height, weight, and race. A pretested questionnaire was used: the The regression equation used was that of standardized and validated American Tho- the European Coal and Steel Community racic Society Chronic Respiratory Diseases Standards for individuals over 18 years of Questionnaire [15], adapted to the local lan- age, generally used for Mediterranean guage. It was orally administrated to both countries [16], and Quanjer and Tammeling groups of workers by previously trained comparisons for those under 18 years [17]. interviewers who were aware of the expo- Dependent variables were based on res- sure status of the subjects and of the objec- piratory function measures before and after tives of the study. For each subject, the pesticides use: questionnaire was administered for 10 min- • forced vital capacity (FVC), i.e. volume utes at 07.30 hours in the morning. in litres from peak inspiration to the end The questionnaire included detailed in- of the forced exhalation; formation about workers’ social and demo- • forced expiratory volume at 1 second graphic characteristics, health status, (FEV1), i.e. the volume expired during 1 employment history and exposure to pesti- second while performing the FVC; cides. Several potential confounding vari- • FEV /FVC ratio; ables were assessed: previous occupation, 1 sex, age, body mass index (BMI), active • forced expiratory flow at 25–75% (FEF ) or maximal mid-expiratory and passive smoking, general health status 25–75% and family medical history. The types of flow rate (MMFR), which is the aver- pesticide products used at work, duration age expiration flow rate during the mid-

of exposure, tasks performed and protec- dle 50% of the FVC [19,20]. tive measures taken were recorded. Infor- Each spirometric test was repeated 3 mation was collected about previous times to allow the choice of the best values, occupational and non-occupational expo- according to the American Thoracic Soci-

sure to pesticides (agricultural, house, gar- ety criteria (2 values of FEV1 and FVC den). Respiratory symptoms regularly

17 Spirometric changes.pmd 128 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 129

should not differ by more than 5% or 100 were eliminated from the pesticide-exposed mL) [18]. group due to inadequate data; the remaining Spirometric measurements conducted number was 19. before work accounted for baseline respi- There were statistically significant dif- ratory values, affected by past chronic ex- ferences in age, sex and weight between posure to pesticides. Absolute and relative the 2 groups. All the pesticide workers values (ratio of actual to predicted values) (100%) were males compared with 77.7% were noted. Binary classification of respi- of non-exposed workers (P < 0.02). Pesti- ratory measures was made for baseline cide workers were older (mean age 39.8 variables, considering values < 80% of years versus 28.0 years, P < 0.001) and

those predicted as abnormal; for FEV1/FVC heavier in weight (84.6 kg versus 70.7 kg, ratio, any value below 80% was considered P < 0.001). They had also spent a longer abnormal [21]. Spirometric measurements time in the industry (mean 10.6 years ver- conducted after 4 hours of work evaluated sus 5.3 years, P < 0.001). a possible effect of acute occupational exposure to pesticides on respiratory volumes Smoking and medical history and flows. Relative changes from baseline No significant differences were seen in the were also calculated as: (2nd value – 1st percentage of past or current smokers, in value)/1st value. the mean number of cigarettes smoked or in the cumulative smoking exposure ex- Analysis pressed as mean number of pack-years Data were analysed using Stata software, (no. of years respondent has smoked on release 5 (Drive East University, USA). An average 1 pack per day) (Table 1). alpha risk of 5% and beta risk of 20% were There were no significant differences accepted. Bilateral statistical tests were between pesticide-exposed and non-ex- performed: chi-squared test for categorical posed workers in terms of self-reported variables, Mann–Whitney test for compar- respiratory symptoms and other diseases. ing means of continuous variables in cases None of the interviewed subjects reported of non-homogeneous variance or non-nor- having any chronic respiratory problems mal distribution, and Student t-test in cases (asthma or chronic bronchitis) diagnosed of homogeneous variance and normal dis- by a doctor. tribution. Spearman rank correlation was applied for correlation of continuous vari- Occupational and extra- ables. A Mantel–Haenszel test was used for occupational exposure to bivariate analysis, and a multiple regression pesticides for multivariate analysis. The current occupational use of liquid pesticides was assessed in the chemical plant where subjects were working: 12 subjects Results were employed in conditioning a pyrethroid pesticide (cypermethrin 25% in xylene so- Demographic characteristics lution) and 9 subjects were exposed to a The characteristics of both groups of carbamate pesticide (isopropoxyphenyl workers are presented in Table 1; 43 non- methylcarbamate 20% in a mixture of cy- exposed subjects and 21 pesticide-exposed clohexane 10% and xylene 60%). The subjects were assessed. Two subjects

17 Spirometric changes.pmd 129 10/25/2005, 1:23 AM 130 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 1 Demographic and social characteristics of workers occupationally exposed and not exposed to pesticides

Demographic/social Non-exposed Pesticide-exposed P-value characteristic workers workers (n = 43) (n = 19)

Mean (SD) Mean (SD) Age (years) 28.0 (9.7) 39.8 (10.1) < 0.001 Weight (kg) 70.7 (12.4) 84.6 (10.6) < 0.001 Height (m) 172.0 (8.4) 171.6 (4.0) 0.80 BMI (kg/m2) 23.8 (3.1) 28.6 (3.0) < 0.001 Time in industry (years) 5.3 (5.0) 10.6 (12.0) < 0.001 Smoking frequency Cigarettes/day (No.) 17.8 (10.9) 15.5 (8.3) 0.60 Pack-years (No.) 5.2 (7.4) 7.2 (11.6) 0.86 No. % No. % Smoking status Ex-smoker 1 2.3 2 10.5 0.38 Current smoker 20 46.5 8 42.1 Non-smoker 22 51.2 9 47.4

n = number of participants. SD = standard deviation. BMI = body mass index. aPack-years = no. of years respondent has smoked on average 1 pack per day. tasks performed consisted of filling small Respiratory function variables receptacles from large reservoirs of liquid Baseline measurements of respiratory pesticides, on a 6 metre long production function made in the morning before the line. The production rate was 5000 litres of working shift (Table 2) show significantly

finished product per 8 hours of work. The lower FEV1, FEF25–75%, and FEV1/FVC ra- tasks were performed manually and work- tios among subjects working with pesti- ers did not use any special protection (i.e. cides. After 4 hours of work, all respiratory no specialized masks, gloves or clothes), variables were still significantly lower in except for cotton masks worn by 8 of the pesticide-exposed subjects. Multiple re- pesticide workers. gression analysis indicated poorer spiro- No significant difference existed be- metric measures among pesticide-exposed tween exposed and non-exposed workers workers (P < 0.05) except for the FVC (P in self-reported previous occupational and > 0.05), for both the first measures (morn- non-occupational exposure to pesticides ing, pre-shift) and second measures (mid- (domestic use, residential and indirect ex- day, post-shift). No statistically significant posure), except that a higher percentage of differences were noted between workers exposed workers (66.8%) had suffered who wore cotton masks and those who did previous occupational exposure during ag- not (P > 0.05). ricultural work than non-exposed workers Table 3 shows the mean percentage dif- (0%) (P = 0.001). ference in respiratory function measures

17 Spirometric changes.pmd 130 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 131

Table 2 Mean baseline (morning, pre-shift) and follow-up (midday, post-shift) respiratory function measures of workers occupationally exposed and not exposed to pesticides

Respiratory measuresa Mean (SD) value P-value Non-exposed Pesticide-exposed workers workers (n = 43) (n = 19)

1st measure Relative FVC (%) 93.6 (11.3) 90.5 (13.7) 0.39

Relative FEV1 (%) 98.7 (10.8) 87.7 (14.0) < 0.001

FEV1/FVC (%) 89.7 (7.6) 80.5 (9.9) < 0.001

Relative FEF25–75% (%) 103.5 (21.2) 74.5 (26.0) < 0.001 2nd measure Relative FVC (%) 95.2 (9.3) 87.4 (12.7) 0.01

Relative FEV1 (%) 102.0 (11.1) 92.4 (16.4) 0.03

FEV1/FVC (%) 91.2 (6.4) 86.7 (6.1) 0.01

Relative FEF25–75% (%) 107.5 (26.1) 85.3 (21.9) < 0.001

n = number of participants. aRelative values = measured/predicted values. SD = standard deviation.

FVC = forced vital capacity; FEV1 = forced expiratory volume at 1 second;

FEF25–75% = forced expiratory flow at 25–75%.

(based on the difference between the sec- found between industrial pesticide use and ond and first measures for individual work- abnormal relative respiratory measures. ers). Both groups showed a trend towards Mantel–Haenszel adjustment for smoking, improved values for all respiratory function past exposure to pesticides outside the fac- measures on the second measures (midday, tory, age, sex, weight and BMI were made post-shift) compared with the first mea- for abnormal respiratory values. Adjusted sures (morning, pre-shift), except for the odds ratios (ORs) obtained were signifi- FVC which showed a decrease between cantly higher than 1, with no interaction the 2 measures in pesticide-exposed work- between the adjustment variables and oc- ers. cupation exposure to pesticides. A high per- Table 4 shows the proportion of work- centage of workers with abnormal ers in both groups with abnormal respirato- respiratory function measures were found ry function measures, i.e. baseline values in both groups, especially those exposed to < 80% of predicted values. A significantly pesticides at work. The ORs showed that higher proportion of pesticide workers had pesticide-exposed workers had 5.6 times

abnormal respiratory variables than non- the risk of abnormal FEV1/FVC ratio and exposed workers, 26.3% versus 4.7% for 16.5 times the risk for abnormal relative

relative FEV1, 47.4% versus 14.0% for FEF25–75% (Table 4). FEV1/FVC ratio and 57.9% versus 4.7% Correlations between the duration of for FEF25–75%. Significant correlations were work in the pesticides plant and baseline

17 Spirometric changes.pmd 131 10/25/2005, 1:23 AM 132 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 3 Percentage change between baseline (morning, pre- shift) and follow-up (midday, post-shift) respiratory function measures of workers occupationally exposed and not exposed to pesticides

Respiratory % change in value P-value measuresa Non-exposed Pesticide-exposed workers workers (n = 43) (n = 19)

Relative FVC +2.5 –2.5 0.05

Relative FEV1 +3.6 +5.3 0.18

FEV1/FVC +2.1 +8.9 < 0.01

Relative FEF25–75% +4.5 +20.9 < 0.01

n = number of participants. aRelative change from baseline = (2nd measure–1st measure)/1st measure.

FVC = forced vital capacity; FEV1 = forced expiratory volume at 1

second; FEF25–75% = forced expiratory flow at 25–75%.

respiratory variables are shown in Table 5. Discussion There was a significant contribution of duration of occupation in the pesticide factory The study participants were all healthy to respiratory measure variances (i.e. 44% workers who did not report any pre-exist- of the FEV /FVC ratio and 12% of the ing respiratory disease. Despite that, statis- 1 tically significant differences were found FEF25–75% could be explained by duration of occupational exposure) (P < 0.001). for forced respiratory volumes and flows

Table 4 Frequency of abnormal baseline respiratory function measures of workers (< 80% of predicted values) occupationally exposed and not exposed to pesticides

Respiratory Non-exposed Pesticide-exposed Adjusted ORb P-value measuresa workers (n = 43) workers (n = 19) (95% CI) No. % with No. % with abnormal abnormal value value

Relative FVC 5 11.6 4 21.5 2.0 (0.4–10.6) 0.44

Relative FEV1 2 4.7 5 26.3 7.3 (1.1–63.4) 0.02

FEV1/FVC 6 14.0 9 47.4 5.6 (1.4–23.8) < 0.001

Relative FEF25–75% 2 4.7 11 57.9 16.5 (2.5–140.3) < 0.001

n = number of participants. aRelative values = measured/predicted values. bAdjusted for smoking status, past exposure to pesticides, age, sex, weight and body mass index.

FVC = forced vital capacity; FEV1 = forced expiratory volume at 1 second; FEF25–75% = forced expiratory flow at 25–75%.

17 Spirometric changes.pmd 132 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 133

Table 5 Association of work duration in pesticides industry with abnormal baseline respiratory function measures for 19 occupationally exposed workers

Respiratory Spearman rank Variance P-value measuresa correlation (r2)

coefficient (rs) Relative FVC –0.47 0.22 0.02

Relative FEV1 –0.46 0.22 0.01

FEV1/FVC –0.66 0.44 < 0.001

Relative FEF25–75% –0.35 0.12 < 0.001

aRelative value = measured/predicted value.

FVC = forced vital capacity; FEV1 = forced expiratory volume at 1

second; FEF25–75% = forced expiratory flow at 25–75%.

before and after work shifts, with consis- indication of pesticides workers’ fatigue, tently lower values for pesticide-exposed an inadequate realization technique or the

workers for FEV1, FEF25–75% and FEV1/FVC occurrence of a mixed obstructive and re- ratio. This could be the effect of chronic strictive pattern of lung disease [22]. The professional exposure to pesticides, which explanation of these observations remains gave lower baseline and post-shift values. to be explored in further studies. Acute ef- The measured variables are particularly ro- fects of pesticide exposure have been re- bust tests and their low values provide clear ported in the literature; in a double-blind signs of pulmonary obstruction among randomized study on asthmatics, acute ex- workers in the pesticides industry. This is posure to insecticide aerosols induced res-

especially true for the FEF25–75%, which piratory symptoms and decreased gives an indication of peripheral airway ob- respiratory function measures [30]. struction. It is believed to be a more sensi- A temporal relationship existed between tive index of airway obstruction than the chronic exposure to pesticides and baseline

FEV1, especially in the small airways that respiratory measures among the study sub- are dependent on the elastic and resistant jects. Abnormal baseline respiratory func- properties of the distal airways than on the tion measures correlated with length of subject’s spirometer performance [14]. exposure to pesticides, with duration of It was not possible to demonstrate an work in the industry accounting for 12% to effect of acute exposure to pesticides on 44% of the contribution to variances in res- the respiratory system because respiratory piratory measures. function improved after 4 hours of work in A number of factors may have affected both groups of workers. This could be ex- the findings non-differentially. Workers did plained by a natural circadian increase of not report any respiratory disease con- respiratory volumes and flows [14], or by a firmed by a doctor and the respiratory learning effect of the spirometer breathing function measures only indicated infra- method. The increase was seen for all mea- clinical pulmonary impairment. Respiratory sures, except for FVC, which decreased. symptoms and previous occupational and The decrease in relative FVC could be an domestic exposure to pesticides were self-

17 Spirometric changes.pmd 133 10/25/2005, 1:23 AM 134 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

reported and the standardized questionnaire bamates, during acute intoxication, can may have not elicited all relevant informa- cause respiratory disturbances, such as tion. The sensitivity of the portable spirom- bronchial hypersecretion, cough, dyspnoea eter may have been low. and asphyxia [2,20]. This may be due to the Nevertheless, our results are compara- systemic effects of carbamates after inha- ble to those found by other authors. The lation, at doses very much lower than those proportion of pesticide workers with ab- causing similar responses after dermal or normal respiratory variables (ranging from oral exposure [12]. Pyrethroid derivatives

26.3% for relative FEV1 to 57.9% for are allergens, causing asthma-like attacks

FEF25–75%) was even higher than those seen and anaphylactic reactions [20,26]. In addi- by Kossmann et al., who evaluated respira- tion, organic solvents used in pesticide fab- tory functions of pesticides factory work- rication and dilution can themselves be ers and found obstructive pulmonary toxic agents leading to the observed respi- impairment among 20% of workers in ad- ratory abnormalities [20]. dition to chronic bronchitis [23]. A study In conclusion, chronic exposure of conducted by the same author demonstrat- workers to industrial pesticides seems to be ed abnormal respiratory flows among 25% associated with a decrease in respiratory of pesticide plant workers [24]. Moreover, function, while acute exposure showed no a case of occupationally related asthma clear effect. Causality between chronic ex- was reported in 1993 for a pesticide plant posure to pesticides and reduced pulmo-

worker; he had a FEV1/FVC ratio of 0.76 nary function variables cannot be and a FEV1 that significantly decreased af- demonstrated, but further studies could ter exposure to pesticides, in addition to provide firmer evidence for this relation- asthma symptoms. Cessation of exposure ship. Protective measures for exposed improved his condition [25]. workers should be implemented in pesti- In rural environments where pesticides cide factories, especially in developing are heavily used, occupational asthma is countries such as Lebanon. The introduc- frequent [26]. Several studies have been tion and enforcement of laws regarding conducted on the respiratory health of pro- pesticides manufacture and use is neces- fessional agricultural workers [22,27–29]. sary in order to control pesticide risks on Most of them demonstrated a chronic spe- the health of factory workers. cific effect of pesticides on respiratory function, in addition to chronic bronchitis and asthma problems. This demonstrates Acknowledgement the consistency of our results with those of We thank the Lebanese National Centre for other researchers. Scientific Research for their support. The results we obtained for chronic effects are also biologically plausible. Car-

References 1. Krieger R, Ross J, Thongsinthusak T. As- 2. Périquet A, de Saint Blanquat G. Les sessing human exposure to pesticides. pesticides dans l’alimentation humaine. Reviews of environmental contamina- Revue du praticien, 1991, 41(11):977– tion and toxicology, 1992, 128:1–15. 84.

17 Spirometric changes.pmd 134 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 135

3. Kennedy G, Valentine R. Inhalational 12. Dinsdale D. Pulmonary toxicity of anti- toxicology. In: Hayes AW, ed. Principles cholinesterases. In: Ballantyne B, Marrs and methods of toxicology, 3rd ed. New TC, eds. Clinical and experimental toxi- York, Raven, 1994:805–38. cology of organophosphates and carbamates. Oxford, Butterworth–Heine- 4. van Hemmen JJ. Agricultural pesticide mann, 1992:156–66. exposure data bases for risk assessment. Reviews of environmental con- 13. Witschi H, Last J. Toxic response of the tamination and toxicology, 1992, 126: lungs. In: Klaassen CD, ed. Casarett & 1–85. Doull’s toxicology: the basic science of poisons, 5th ed. New York, McGraw–Hill, 5. Nieuwenhuijsen M et al. Exposure to 1996:443–61. dust, noise and pesticides, their determinants and the use of protective equip- 14. Seaton A, Crompton G. Asthma: clinical ment among California farm operators. features. In: Seaton A, Seaton D, Leitch Applied occupational and environmen- AG, eds. Crofton and Douglas’s respira- tal hygiene, 1996, 11(10):1217–25. tory diseases, 5th ed. Oxford, Blackwell, 2000:77–151. 6. Legaspi J, Zenz C. Occupational health aspects of pesticides: occupational 15. Ferris BG. Epidemiology Standardiza- medicine. In: Zenz C, Dickerson OB, tion Project (American Thoracic Society). Horvath EP, eds. The chemical occupa- American review of respiratory disease, tional environment, 3rd ed. St. Louis, 1978, 118:1–120. Mosby, 1994:30–41. 16. Quanjer PH et al. Lung volumes and 7. Palacios-Nava ME et al. Sintomatología forced ventilatory flows. Report of the persistente en trabajadores industrial- Working Party Standardization of Lung mente expuestos a plaguicidas organo- Function Tests, European Community for fosforados. [Persistent symptomatology Steel and Coal. Official Statement of the in workers industrially exposed to orga- European Respiratory Society. Euro- nophosphate pesticides.] Salud publica pean respiratory journal, 1993, 16:5–40. de Mexico, 1999, 41(1):55–61. 17. Quanjer PH et al. Standardization of lung 8. do Pico GA. Hazardous exposure and function tests in paediatrics. European lung disease among farm workers. Clin- respiratory journal, 1989, 2(suppl. 4): ics in chest medicine, 1992,13:311–28. 121S–265S. 9. Bailey J. Safer pesticide packaging and 18. Standardization of spirometry—1987 formulations for agricultural and resi- update. Statement of the American Tho- dential applications. Reviews of environ- racic Society. American review of respi- mental contamination and toxicology, ratory disease, 1987, 136:1285–98. 1992, 129:17–27. 19. Tantum K. Pulmonary function testing in 10. Léonard A. Les mutagènes de l’envi- asthma. Aerosols, airways and asthma, ronnement et leurs effets biologiques. 1981,1:53–63. Paris, Masson, 1990:30–4. 20. Ecobichon D. Toxic effects of pesticides. 11. Blair A, Zahm HS. Patterns of pesticide In: Klaassen CD, ed. Casarett & Doull’s use among farmers: implications for epi- toxicology: the basic science of poisons, demiologic research. Epidemiology, 5th ed. New York, McGraw–Hill, 1996: 1993, 4(1):55–62. 643–89.

17 Spirometric changes.pmd 135 10/25/2005, 1:23 AM 136 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

21. Weinberger S, Drazen J. Disturbances of 26. Underner M, Cazenave F, Patte F. respiratory function. In: Braunwald E et L’asthme professionel en milieu rural al., eds. Harrison’s principles of internal [Occupational asthma in the rural envi- medicine, 15th ed. New York, McGraw– ronment.] Revue de pneumologie Hill, 2001:1446–53. clinique, 1987, 43:26–35. 22. Dosman J et al. Respiratory symptoms 27. Senthilselvan A, McDuffie HH, Dosman and pulmonary function in farmers. Jour- JA. Association of asthma with use of nal of occupational medicine, 1987, pesticides. Results of a cross-sectional 29:38–43. survey of farmers. American review of respiratory disease, 1992, 146:884–7. 23. Kossmann S, Konieczny B, Hoffmann A. The role of respiratory muscles in the 28. Nejjari C et al. Aspects épidémio- impairment of the respiratory system logiques du vieillissement respiratoire: function in the workers of a chemical apport de l’enquête PAQUID. [Epidemio- plant division producing pesticides. logic aspects of respiratory ageing: con- Przeglad lekarski, 1997, 54(10):702–6. tribution of the PAQUID survey.] Revue d’epidemiologie et de santé publique, 24. Kossmann S, Pierzchala W, Hefczyc J. 1997, 45:417–28. Nadreaktywnosc oskrzeli u pracowni- kow zakladow chemicznych produkuja- 29. Bener A. Respiratory symptoms, skin dis- cych pyliste srodki ochrony roslin. orders and serum IgE levels in farm [Bronchial hyperreactivity in chemical workers. Allergie et immunologie, 1999, plant workers employed in the produc- 31(2):52–6. tion of dust pesticides.] Wiado-mosci 30. Salome CM et al. The effect of insecticide lekarskie, 1999, 52(1–2):25–9. aerosols on lung function, airway re- 25. Royce S et al. Occupational asthma in a sponsiveness, and symptoms in asth- pesticide manufacturing worker. Chest, matic subjects. European respiratory 1993, 103:295–6. journal, 2000, 16:38–43.

Network of WHO Collaborating Centres in occupational health The Network of Collaborating Centres makes a substantial contribution to the goal of “occupational health for all”; they are the “on-the- ground” actors and play a key role in capacity building. WHO estimates that only about 10% to 15% of workers worldwide have some kind of access to occupational health services, and extend- ing coverage is a key challenge. The existing multilateral and bilateral collaboration is being further developed to transmit information and sharing of experience between countries and institutions. In 2001 a Workplan 2001–2005 was prepared and progress was reviewed at the Sixth Network Meeting held in Iguassu Falls, Brazil, in February 2003 which is set out in the Compendium of projects of the WHO Collaborating Centres in occupational health. Further information on WHO’s work in occupational health is available at: http:// www.who.int/occupational_health/en/

17 Spirometric changes.pmd 136 10/25/2005, 1:23 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 137 Health locus of control beliefs and smoking among male Kuwaiti government employees H.E. Badr1 and P.M. Moody1

ABSTRACT A cross-sectional multistage stratified cluster survey of 1798 Kuwaiti male adults investigated the relationship between health locus of control (HLC) beliefs, health beliefs about smoking and smoking status. Non-smokers had stronger external HLC and better health beliefs about smoking than smokers. Thus the Kuwaiti men who used tobacco believed themselves to be more in control of their own lives than did non-users. Stepwise logistic regression analysis revealed that low HLC, poor health beliefs about smoking, single marital status and low level of education were significant predictors of risk of smoking. Simple linear regression analysis showed a significant negative relationship between HLC and health beliefs about smoking among non-smokers and ex-smokers, but not among smokers.

Croyances en fonction du lieu de contrôle de la santé et tabagisme chez des fonctionnaires koweïtiens RÉSUMÉ Une enquête transversale utilisant un échantillonnage stratifié à plusieurs niveaux auprès de 1798 hommes koweïtiens a permis d’examiner la relation qui existe entre les croyances en fonction du lieu de contrôle de la santé ainsi que les croyances relatives aux effets sur la santé du tabagisme et le statut tabagique. Les non-fumeurs avaient un lieu de contrôle de la santé plus fréquemment externe et de meilleu- res croyances relatives aux effets sur la santé du tabagisme que les fumeurs. Les Koweïtiens qui consommaient du tabac croyaient ainsi qu’ils contrôlaient davantage leur vie que ceux qui n’en consommaient pas. L’analyse de régression logistique par étapes a révélé que le lieu de contrôle de la santé, les croyances relatives aux effets sur la santé du tabagisme, le célibat et le niveau d’instruction constituaient des facteurs prédictifs significatifs du risque de tabagisme. L’analyse de régression linéaire simple a montré une relation négative significative entre le lieu de contrôle de la santé et les croyances relatives aux effets sur la santé du tabagisme chez les non-fumeurs et les ex-fumeurs, mais pas chez les fumeurs.

1Department of Community Medicine and Behavioural Sciences, Faculty of Medicine, University of Kuwait, Safat, Kuwait (Correspondence to H.E. Badr: [email protected]). Received: 18/09/03; accepted: 25/02/04

18 Health locus of control.pmd 137 10/25/2005, 1:23 AM 138 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction viduals’ predispositions, however, show promise for understanding the decisions The notion of perceived locus of control is [9]. perhaps the most widely known of the psy- The decision to take up smoking is an chological constructs associated with be- individual one. Regardless of the myriad of liefs about control. Although this trait is no external or situational factors, no one forc- doubt distributed normally among people, es people to experiment with tobacco use. those who believe that they are influenced Even though contacts and interaction with by external forces are considered to have smoking peers and family members are an external locus of control. However, major predictors of eventual smoking up- those who have confidence that whatever take, the actual choice ultimately rests with happens to them, pleasant or unpleasant, is the individual [10,11]. substantially within their domain of influ- It is well known that different cultures ence are said to have a predominantly inter- have different health outcomes influenced nal locus of control [1]. In other words, by different beliefs. Previous studies have people who score external on locus of con- found an association and an influence of trol are more likely to attribute life’s suc- cultural origin over generalized locus of cesses and failures to factors such as fate, control. An interaction between culture, be- luck, and chance. Conversely, internal lief and health was apparent from these scoring people have a greater tendency to studies [12,13]. In addition, factors such attribute life’s outcomes to personal at- as occupational stressors, socioeconomic tributes relating to ability, effort, and per- status, gender, age, etc. are likely to influ- sonal power of control [2]. ence the outcome of health [14–16]. Also Suffice it to say that the development of for illustrating the interconnection between a health locus of control (HLC) scale has socioeconomic status and locus of control, led to research demonstrating that there is, Smith et al. mentioned “an internal locus of indeed, an association between internality control may facilitate selection into higher and the likelihood of making healthy choic- status groups and occupations by fostering es [3,4]. One particular health-related as- an increased sense of self-efficacy. Addi- pect in which locus of control is tionally, social roles linked with status may considered to be significant is cigarette influence locus of control” [17]. smoking [5]. Locus of control is an attri- Cigarette smoking is considered one of butional personality dimension that has the important public health problems in Ku- been linked to a host of behaviours includ- wait and the prevalence of smoking among ing cigarette smoking [6]. adult males has been estimated at 34.4% These aspects have been shown to dif- [18]. This is despite numerous community ferentially predict health-related behaviours interventions and government legislation in that people with a more internal HLC against smoking in 1995, suggesting that an generally adhere more closely to health reg- individual’s decision to use tobacco is a imens whereas more externally oriented in- critical point in the process of preventing dividuals are less likely to engage in smoking. A review of a body of literature health-protective behaviours [7,8]. It is re- indicates that HLC and its association with ported that simple knowledge of the dan- smoking have not been reported in the Ku- gers associated with tobacco use has little waiti culture. The aim of this study is to impact on adolescent tobacco uptake. Indi- assess the relationship between HLC be-

18 Health locus of control.pmd 138 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 139

liefs, health beliefs about smoking accord- or ex-smoker (smoked at least 100 ciga- ing to smoking status and different socio- rettes during their lifetime and reported demographic factors among Kuwaiti men quitting smoking) [19]. working in government ministries. It is an- Respondents completed the HLC scale, ticipated that such results may help the consisting of 11 items that measure exter- health planners to design messages or internal versus internal HLC. The scale was ventions that are more effective in prevent- scored in the external direction, with each ing tobacco use in Kuwait. item scored from 1 (strongly disagree) to 5 (strongly agree) for the externally worded items and reverse scored for the internally Methods worded items [20]. Respondents’ beliefs about cigarette Study design and sampling smoking and health were also evaluated us- procedure ing a set of 10 statements, e.g. “I believe The study was a cross-sectional multistage that cigarette smoking is not good for my stratified cluster survey of men working in health and I believe that quitting smoking government ministries. The first stratifica- prevents a person from getting lung can- tion was according to ministry; 6 ministries cer”. Each statement was scored using a 5- were systematically selected. The second point Likert scale, which ranged from level of stratification was according to the strongly disagree to strongly agree, with number of departments within each minis- high scores indicating better (beneficial) try; 3 departments were chosen from each health beliefs. ministry, with probability proportional to the number of employees in each depart- Data collection and analysis ment. In the last stage, all the male employ- Data were collected from April to Decem- ees in the chosen departments were ber 1996. A pilot study was performed on grouped in equal clusters, with 8 clusters 243 individuals using the translated Arabic chosen randomly for inclusion in the study. version of the scales and the necessary A total of 1918 male Kuwaiti adults aged 18 modifications were made accordingly. years and over formed the target population Data entry and analysis were done using of the study. Full details of the survey sam- SPSS, version 10. Mean and standard devi- ple have been reported earlier [18]. ation (SD) and odds ratios (OR) were calculated. Stepwise binary logistic regres- Study instrument sion, simple linear regression, and analysis An anonymous self-administered question- of variance were used to estimate the rela- naire was completed by the study sample. tionship between the studied variables at a Socioeconomic data were collected about level of significance of P < 0.05 and 95% age, marital status, monthly income and confidence interval (CI). level of education. Smoking behaviour was assessed as: current smoker (those who at the time of the survey had smoked at least Results 100 cigarettes during their lifetime and were currently smoking), non-smoker Description of the study sample (smoked less than 100 cigarettes during The total sample included 1918 individuals; their lifetime or had never smoked before) 1798 of them returned complete question-

18 Health locus of control.pmd 139 10/25/2005, 1:24 AM 140 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

naires, giving a response rate of 93.7%. difference between the 3 groups was sta- The mean age and standard deviation (SD) tistically significant (F = 124.17, P < was 33.2 (7.9); the median was 32 years. 0.0001). Almost one-quarter of the men was single (25.2%). Regarding the level of educa- Stepwise logistic regression tion, 24.1% of them had graduated from The relationship between smoking status university or held advanced degrees, and HLC, health beliefs about smoking and 40.4% held diplomas, 18.9% had second- socioeconomic factors (age, marital status, ary level education and 16.6% had interme- level of education and monthly income) diate level or below. was examined using stepwise multivariate binary logistic regression (Table 2). Adding HLC and health beliefs about HLC and health beliefs about smoking and smoking the 4 socioeconomic variables as indepen- The mean (SD) score for the HLC was dent variables to the dependent variable (0 29.83 (4.29) for the whole sample (Table = smokers and 1 = non-smokers) led to a 1). Non-smokers had higher scores than significant improvement in the prediction smokers [30.38 (SD 4.05) and 29.65 (SD of smoking status over the model (the clas- 4.34) respectively], indicating a stronger sification matrix = 68.3%). external HLC. Ex-smokers had the lowest Four independent variables out of the 6 score [28.68 (SD 4.57)], i.e. the weakest entered into the equation emerged as signif- external HLC. The difference between the icant predictors in the regression equation. 3 categories was statistically significant (F HLC was the first predictor, with non- = 19.45, P < 0.0001). smokers having higher external HLC Table 1 also shows that the mean (SD) scores than smokers (OR = 0.95, 95% CI: score for health beliefs about smoking was 0.92–0.98, P < 0.0001). Non-smokers also 41.97 (SD 6.54) among the studied group. had better scores on health beliefs about Ex-smokers had the highest mean health smoking and this emerged as the second belief score [44.44 (SD 4.99)] followed by predictor (OR = 0.9, 95% CI: 0.88–0.91, P non-smokers [43.27 (SD 5.71)]. Smokers < 0.0001). Marital status ranked as the had the lowest mean score for health be- third predictor for smoking, where single liefs about smoking [38.87 (SD 7.15)]. The men were at higher risk and had higher ex-

Table 1 Mean total scores for health locus of control (HLC) and health beliefs about smoking, according to smoking status for a sample of Kuwaiti men

Smoking status No. % Mean (SD) HLC Mean (SD) health score beliefs score

Smokers 618 34.4 29.65 (4.34) 38.87 (7.15) Non-smokers 861 47.9 30.38 (4.05) 43.27 (5.71) Ex-smokers 319 17.7 28.68 (4.57) 44.44 (4.99) Total 1798 100.0 29.83 (4.29) 41.97 (6.54) F =19.45, P < 0.0001 F =124.17, P < 0.0001

SD = standard deviation.

18 Health locus of control.pmd 140 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 141

Table 2 Stepwise logistic regression of different confounders among smokers and non-smokers

Variables Odds ratio β CI P-value

HLC score 0.948 –0.054 0.923 to 0.974 P < 0.0001 Health beliefs score 0.895 –0.111 0.879 to 0.911 P < 0.0001 Marital status 0.769 –0.262 0.599 to 0.988 P < 0.05 Level of education Intermediate and below 2.038 0.712 1.472 to 2.912 P < 0.0001 Secondary 1.738 0.553 1.243 to 2.432 P < 0.001 Diploma 1.411 0.344 1.056 to 1.885 P < 0.05

HLC = health locus of control. CI = confidence interval.

ternal HLC than ever-married men (OR = There was no significant relationship be- 0.77, 95% CI: 0.6–0.99, P < 0.05). The tween the 2 variables among smokers, but fourth predictor was the level of education. there was a negative significant relationship The model showed that the lower the level between them for both non-smokers (β = – of education, the higher the risk for smok- 0.14, CI: –0.24 to –0.05, P < 0.01) and ex- ing. The people with the lowest level of ed- smokers (β = –0.19, 95% CI: –0.31 to ucation (intermediate or below) were at –0.07, P < 0.01) (Table 3). higher risk compared with the reference group (university or above) (OR = 2.04, 95% CI: 1.43–2.91, P < 0.0001). Discussion The same method of analysis was used The HLC model suggests that those who for the smokers and ex-smokers, but HLC score high on the internal dimension, i.e. did not enter the equation and was not a who regard their health as largely within predictor for smoking in the 2 groups. their own control, are likely to engage in Health beliefs about smoking was the first predictor, with ex-smokers having better beliefs than smokers. Age and marital status were the second and third predictors in Table 3 Linear regression between HLC total score and health beliefs about smoking total the equation, with older and ever-married score among smokers, non-smokers and ex- men at lower risk for smoking than young- smokers er participants and single men. Smoking β CI P-value HLC and health beliefs about status smoking Smokers –0.02 –0.15 to 0.115 NS Separating smokers, non-smokers and ex- Non-smokers –0.14 –0.24 to –0.05 P < 0.01 smokers, the relationship between the 2 variables—HLC and health beliefs about Ex-smokers –0.19 –0.31 to –0.07 P < 0.01 smoking—for the 3 groups separately was NS = not significant. investigated using simple linear regression. CI = confidence interval.

18 Health locus of control.pmd 141 10/25/2005, 1:24 AM 142 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

health-maintaining behaviours. Conversely, became more aware about their health as a those who score high on the external di- result of preventive programmes they at- mension view their health as relatively inde- tended or started to view certain medical pendent of their behaviour and, problems as complications of smoking, and accordingly, are more likely to engage in thus developed a belief of their own self- health-damaging behaviours than those control about their health. with lower scores [21]. Sense of control Some studies have suggested that HLC has been recognized over the past 25 years “internals” are more likely to achieve and as a psychological construct with impor- maintain abstinence than “externals” there- tant implications for health [22]. by suggesting that ex-smokers are more The present study investigated the rela- likely to demonstrate a lower external HLC tionship between HLC, health beliefs about score than current smokers. This agrees smoking and smoking status in a represen- with the findings of this study, as ex-smok- tative sample of 1798 male Kuwaiti adults. ers had the weakest external HLC, lower The findings of the present study pro- than either smokers or non-smokers [28– vide limited support for the HLC model in 32]. relation to smoking behaviour. Non-smok- Regarding health beliefs about smoking, ers were found to have stronger external the study revealed that both non-smokers HLC than smokers. The relationship is and ex-smokers had stronger health beliefs counterintuitive but does coincide with the about the adverse effects of smoking on findings of other studies [23,24]. One ex- their health than smokers. This can add planation for the apparent contradiction more to the denial process of the smokers may involve a process of denial through that smoking did not harm their health. The which smokers believe they are in control evidence for this is the presence of a nega- of their health despite their smoking. This tive significant relationship between HLC belief in the lack of control of their health and health beliefs about smoking among serves to further rationalize their smoking. non-smokers and ex-smokers, but not The denial process has been suggested in among smokers. This is in accordance another study [25,26]. with the suggestion that a strong and con- The study revealed that HLC did not fol- sistent relationship would be found below the predicted pattern for tobacco users tween HLC and behaviour only among as non-smokers had a stronger belief in ex- people who score high for beliefs about ternal control than smokers. This may sug- their health [33]. gest that these non-smoking adults were It is well known that cigarette smokers more compliant with authority figures. acknowledge that smoking is damaging to This finding is compatible with the results health. One possible reason for such appar- of another study where non-smoking ado- ent discrepancy between health beliefs and lescents had a higher belief in powerful oth- smoking behaviour is that individuals differ ers than smokers, which may indicate a in how much they believe they are in con- reliance on the credibility, knowledge and trol of their own health and are able to con- curative abilities of health care providers trol their behaviour and lifestyle. A second [27]. reason is that health itself is only one of a On the other hand, smokers in our number of values that can motivate health- study had a stronger external HLC than ex- related behaviour [26]. smokers. Maybe this group of ex-smokers

18 Health locus of control.pmd 142 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 143

The non-smokers with stronger exter- is an important factor playing a principal nal HLC might be dependent on values ob- role in directing people’s health behaviours. tained from different cultures. Early These Kuwaiti men had a lower external childhood experiences and environments in HLC than non-smokers. Ex-smokers had which children are allowed no control, or the lowest external HLC. However, health ones in which they see individual control as beliefs about smoking were the highest for random, may contribute to the develop- ex-smokers and the lowest for smokers. ment of an external locus of control [34]. Socioeconomic differences (young age, Furthermore, it has been suggested that low level of education and single) were risk “individuals reared in a culture that values factors predicting smoking behaviour. independence, uniqueness, self-reliant indi- In the light of these findings, we recom- vidualism, and personal output of energy mend that HLC deserves sustained consid- are likely to be more internally oriented than eration in investigations of behaviour and individuals from a culture that tends to em- health as it reflects general attitudes to phasize a different set of values” [35]. In health. Smoking prevention efforts need to the United States of America, a person has tap into the psychological profile of smok- access to education and economic advan- ers who believe in their own control on tages not available worldwide. According health in order to help them to quit smok- to this country’s values, education and eco- ing. Health education programmes about nomic advantage increase a person’s healthy behaviours should start at an early chances of adopting an internal HLC [36]. age to help with constructing the internal This contradicts the findings of our study, belief of HLC. which reveals the effect of different cultural values on embedded HLC beliefs. Acknowledgements Conclusion and This research was funded by a grant from the Research Administration, University of recommendations Kuwait (project no. MC040). The study focused on the HLC beliefs and its relationship to smoking behaviour. HLC

References

1. Tones K. Health education, behaviour 3. Wallston KA, Wallston BS. Locus of con- change, and the public health. In: Detels trol and health: a review of the literature R et al., eds. Oxford textbook of public Health education monographs, 1978, health. Volume 2. Methods of public 6(2):107–17. health, 3rd ed. Oxford, Oxford University 4. Tones BK. Health promotion, empower- Press, 1997:791–814. ment and the concept of control. In: 2. Molloy GN et al. Locus of control of smok- Health education: politics and practice. ers, nonsmokers and nonpracticing Victoria, Australia, Deakin University smokers. Psychological reports, 1997, Press, 1992. 81:781–2.

18 Health locus of control.pmd 143 10/25/2005, 1:24 AM 144 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

5. Georgiou A, Bradley C. The develop- in England. Health education research, ment of a smoking-specific locus of con- 2000, 15:163–74. trol scale. Psychology and health, 1992, 16. Wardle J, Steptoe A. Socioeconomic dif- 6:227–46. ferences in attitudes and beliefs about 6. Rotter JB. Internal versus external locus healthy lifestyles. Journal of epidemiol- of control of reinforcement. A case his- ogy and community health, 2003, 57: tory of a variable. American psycholo- 440–3. gist, 1990, 45:489–93. 17. Smith PB, Dugan S, Trompenaars F. Lo- 7. Jessor R, Turbin M, Costa F. Protective cus of control and affectivity by gender factors in adolescent health behavior. and occupational status: a 14 nation Journal of personality and social psy- study. Sex roles, 1997, 36:51–7. chology, 1998, 75:788–800. 18. Moody PM et al. Factors associated with 8. Newsom J, Knapp J, Schulz R. Longitu- the initiation of smoking by Kuwaiti dinal analysis of specific domains of in- males. Journal of substance abuse, ternal control and depressive symptoms 1998, 10(4):375–84. in patients with recurrent cancer. Health 19. Cigarette smoking among adults— psychology, 1996, 15:323–31. United States, 1992, and changes in the 9. Nutbeam D et al. Evaluation of two definition of smoking. Journal of the school smoking evaluation programmes American Medical Association, 1994, under normal classroom conditions. Brit- 272:14–6. ish medical journal, 1993, 306:102–7. 20. Wallston BS et al. Development and vali- 10. Greenlund K et al. Cigarette smoking at- dation of the health locus of control titudes and first use among third-through (HLC) scale. Journal of consulting and sixth-grade students: the Bogalusa clinical psychology, 1976, 44:580–5. Heart Study. American journal of public 21. Wallston KA, Wallston BS, DeVellis R. health, 1997, 87:1345–8. Development of the Multidimensional 11. Hunter S et al. Psychosocial influences Health Locus of Control (MHLC) Scales. on cigarette smoking among youth in a Health education monographs, 1978, southern community: the Bogalusa 6:160–70. Heart Study. Morbidity and mortality 22. Steptoe A, Wardle J. Locus of control and weekly report, 1987, 36(suppl. 4):17S– health behaviour revisited: a multivariate 23S. analysis of young adults from 18 coun- 12. Marano HE. Depression: beyond seroto- tries. British journal of psychology, 2001, nin. Psychology today, 1999, 32:30–76. 92:659–72. 13. Duffy ME. Determinants of reported 23. Norman P. Health locus of control and health promotion behaviors in employed health behaviour: an investigation into Mexican American women. Health care the role of health value and behaviour for women international, 1997, 18:149– specific efficacy beliefs. Personality and 63. individual differences, 1995, 18:213–8. 14. Bates MS, Rankin-Hill L. Control, culture 24. Bennett P et al. Health locus of control and chronic pain. Social science and and value for health in smokers and non- medicine, 1994, 39:629–45. smokers. Health psychology, 1997, 16(2):179–82. 15. Parmenter K, Waller J, Wardle J. Demo- graphic variation in nutrition knowledge

18 Health locus of control.pmd 144 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 145

25. Booth-Butterfield M, Anderson RH, study of the sense of control in illness. Booth-Butterfield S. Adolescents’ use of Journal of health and social behavior, tobacco, health locus of control and self- 1983, 24:144–60. monitoring. Health communication, 32. Shipley RH. Maintenance of smoking 2000, 12(2):137–48. cessation: effect of follow-up letters, 26. Stenstrom U, Andersson P. Smoking, smoking motivation, muscle tension and blood glucose control, and locus of con- health locus of control. Journal of control beliefs in people with type 1 diabetes sulting and clinical psychology, 1981, mellitus. Diabetes research clinical prac- 49:982–4. tice, 2000, 50(2):103–7. 33. Wallston KA, Smith MS. Issues of control 27. Eiser J et al. Health locus of control and and health: the action is in the interac- health beliefs in relation to adolescent tion. In: Penny G, Bennett P, Herbet M, smoking. British journal of addiction, eds. Health psychology: a lifespan per- 1989, 84:1059–65. spective. London, Harwood, 1994. 28. Mlott SR, Mlott YD. Dogmatism and locus 34. Chorpita B, Barlow D. The development of control in individuals who smoke, of anxiety: the role of control in the early have stopped smoking and never environment. Psychological bulletin, smoked. Journal of community psychol- 1998, 124:3–21. ogy, 1975, 3:53–7. 35. McLaughlin SC, Saccuzzo DP. Ethnic 29. Ockene JK et al. Relationship of psycho- and gender differences in locus of con- logical factors to smoking behavior trol in children referred for gifted pro- change in an intervention program. Pre- grams: the effects of vulnerability factors. ventive medicine, 1982, 11:13–28. Journal for the education of the gifted, 1997, 20:268–83. 30. Rosenbaum M, Argon S. Locus of control and success in self-initiated attempts to 36. Hamid PN. Self-monitoring, locus of con- stop smoking. Journal of clinical psy- trol and social encounters of Chinese chology, 1979, 35:870–2. and New Zealand students. Journal of cross cultural psychology, 1994, 25: 31. Seeman M, Seeman TE. Health behavior 353–68. and personal autonomy: a longitudinal

18 Health locus of control.pmd 145 10/25/2005, 1:24 AM 146 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Sociodemographic factors and age at natural menopause in Shiraz, Islamic Republic of Iran S.M.T. Ayatollahi,1 H. Ghaem1 and S.A.R. Ayatollahi 1

ABSTRACT With increasing life expectancy, menopause is an increasingly important aspect of women’s health. We recorded the age at natural menopause among women in a population-based cross-sectional study in Shiraz, Islamic Republic of Iran, in summer 2000. Interviews with 948 randomly selected menopausal women showed the mean (standard deviation) age at menopause was 48.3 (5.3) years (95% CI: 48.0– 48.6), median 49 years. The sociodemographic and health behaviour factors that were significantly related to early mean age of menopause were: never married (44.7 years), low income level (47.4 years), low social class (45.8 years), tobacco use (47.9 years) and non-consanguineous husband (48.1 years).

Facteurs sociodémographiques et âge au moment de la ménopause naturelle à Chiraz (République islamique Iran) RÉSUMÉ Avec l’allongement de l’espérance de vie, la ménopause est un aspect de la santé de la femme de plus en plus important. Nous avons enregistré l’âge à la ménopause chez les femmes dans le cadre d’une étude transversale dans une population définie à Chiraz (République islamique d’Iran) durant l’été 2000. Les entretiens avec 948 femmes ménopausées sélectionnées au hasard ont montré que l’âge moyen (écart type) à la ménopause était de 48,3 (5,3) ans (IC 95 % : 48,0-48,6), avec une médiane de 49 ans. Les facteurs sociodémographiques et relatifs au comportement en santé qui étaient significativement liés à l’âge moyen précoce de la ménopause étaient les suivants : le fait de ne s’être jamais mariée (44,7 ans), le faible niveau de revenus (47,4 ans), la classe sociale inférieure (45,8 ans), la consommation de tabac (47,9 ans) et la non-consanguinité de l’époux (48,1 ans).

1Department of Biostatistics and Epidemiology, Shiraz University of Medical Sciences, School of Health, Shiraz, Islamic Republic of Iran (Correspondence to S.M.T. Ayatollahi: [email protected]). Received: 18/09/02; accepted: 07/10/03

19 Sociodemographic factors.pmd 146 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 147

Introduction city is both the most developed city in the southern region and the cultural capital of The menopause is an important stage in the the country. life of every woman. With a current life expectancy for women of nearly 70 years Sample in the Islamic Republic of Iran, women can Women who had lived in Shiraz for at least now expect to live one-third of their adult 1 year and had experienced natural meno- lives beyond menopause. Current knowl- pause were included in this study. edge about menopause and its potential ef- Administratively, Shiraz is divided into 7 fect on the subsequent quality of life for large mailing zones, each with distinct so- women remains incomplete and even con- cial, cultural, economic and health charac- troversial. Although menopause, as a major teristics. Each large postal zone is divided event affecting the reproductive system, is into many smaller zones depending on the usually included as an extraneous variable population density, such that the number of in research studies, little is known from re- houses is proportional to the size of each of liable studies about the epidemiology of the these zones. Since an adequate sampling event itself as an independent variable. Re- frame of the target population was unavail- ports of several studies show that the age at able, a 2-stage cluster sampling was ap- natural menopause is influenced by a num- plied. Within each larger postal zone, every ber of physical [1–6], sociodemographic other smaller postal zone was selected ran- [1,7–14] and behavioural characteristics domly. A random start point was then cho- [2,10–12,15,16]. sen from each of the selected zones and the In view of the importance of age of houses were visited by 2 research teams, menopause for women’s health, and con- each of which was responsible for visiting sidering the shortage of related information homes from one side of the street (left or in the Islamic Republic of Iran [13,17], we right). Applying this procedure, a total of performed a study to determine the average 948 women who had undergone natural age of natural menopause in a sample of menopause were selected from 79 clusters women in Shiraz city and to measure its each of size 12. Only 7 women refused to association with sociodemographic fac- participate in this study; they were replaced tors. Our data were compared with data by 7 other subjects randomly selected from from other population-based research. the related zones. This procedure did not have any impact on the results. Methods Questionnaire and interview Place of study A structured questionnaire was designed to Shiraz, one of the 5 principal cities of the study a range of factors: physical charac- Islamic Republic of Iran, is the centre of teristics of the woman (height, weight, arm Fars province. The city is located 1000 km circumference, handedness), sociodemo- south of the capital Tehran and 100 km graphic data (birthplace, marital status, north of the Gulf, at an altitude of 1500 m consanguinity, ancestorship, occupation, above sea level. It has a Mediterranean cli- education, income) and health behaviour mate and a population of 1.2 million, of (exercise, smoking). The woman’s age whom 84% are literate (Iran Literacy was taken from her birth certificate which Movement, unpublished report, 1998). The is an accurate record in the Islamic Repub-

19 Sociodemographic factors.pmd 147 10/25/2005, 1:24 AM 148 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

lic of Iran. The age of menopause was di- Means, standard deviations (SD), medians rectly recorded for those women who and 95% confidence intervals (CI) were could remember the date correctly; women calculated. For comparisons between 2 who could not remember were assisted by means, Student t-test was used, and to reference to dates of important events, the contrast the difference between means in date of last confinement, the age of last variables with more than 2 groups, analysis child and information provided by the hus- of variance with Duncan’s multiple range band or close relatives. Age at menopause was applied. (recorded to the nearest half year) was defined as the age of spontaneous cessation of regular menstrual bleeding for 1 year or Results more [18]. Around 60% of respondents The mean (SD) age at menopause in the could remember their menopausal age ac- 948 women was 48.3 (5.3) years (95% CI: curately and 40% had to estimate it by re- 48.0–48.6) with a median of 49 years. call. Table 1 presents mean and median age The weight of the subject was mea- at natural menopause according to some sured to the nearest 0.1 kg taken without physical (arm circumference, height, shoes and with minimum clothing. Height weight, BMI) and personal characteristics was measured to the nearest millimetre (birthplace, handedness and ancestorship). with the subject standing on a flat surface None of these factors was found to be re- erect against a wall using a tape measure lated to menopausal age. and headboard. Body mass index (BMI) Table 2 presents the mean and median was calculated. Mid upper arm circumfer- age at natural menopause according to ence was measured in mm on the left hand socioeconomic and health behaviour char- using a graded flexible plastic tape. acteristics. Educational level and marital The survey was conducted between status were not significantly related to age July and September 2000. On average, at menopause, although comparing women each interview lasted 30 minutes. One who were never and ever married did show member of each of the 2 research teams a significant difference. Social class (based were responsible for interviewing each on occupational status of the husband), subject face-to-face. Each team consisted family income and consanguinity were sig- of 3 members supervised by a trained mid- nificant factors. Exercise and cigarette wife with a bachelor degree. Other mem- smoking were not significantly related to bers of each team were senior experts of age at menopause, although tobacco use medical records who were trained especial- through smoking of water pipes (“hubble- ly for this research in how to fill in the bubble”, kalyan) was a significant factor. questionnaires. Thereafter they personally Thus the groups of women with an ear- visited all women under investigation and ly mean age of menopause were those who completed the questionnaires. were never married (44.7 years, P < Statistical analysis 0.006), of low income level (47.4 years, P The classification of the participants by < 0.002), of low social class (45.8 years, P socioeconomic status was done using prin- < 0.001), tobacco users (47.9 years, P < cipal component analysis. The information 0.014) and having a non-consanguineous was analysed using SPSS, version 10. husband (48.1 years, P < 0.027).

19 Sociodemographic factors.pmd 148 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 149

Table 1 Mean and median age at natural menopause according personal characteristics

Variable No. of % Mean 95% CI Median Test P–value women (SD) age age (years) (years)

Arm circumference (cm) ANOVA 0.073 < 29 454 47.9 47.9 (5.8) 47.5–48.3 49 29–32 274 28.9 48.7 (5.1) 48.4–49.0 50 > 32 220 23.2 48.6 (4.5) 48.3–48.9 49 Height (cm) ANOVA 0.879 < 150 379 40.1 48.1 (5.6) 48.2–48.9 49 150–155 331 35.0 48.3 (5.2) 47.9–48.6 49 > 155 236 24.9 48.4 (5.0) 48.1–48.7 49 Weight (kg) ANOVA 0.206 < 55 256 27.1 47.8 (6.0) 47.4–48.2 50 55–65 312 33.0 48.2 (5.5) 47.8–48.6 49 > 65 378 40.0 48.6 (4.6) 48.3–48.9 49 Body mass index (mg/kg2) ANOVA 0.152 ≤ 20 57 6.0 47.8 (6.5) 47.4–48.2 49 20–24.9 247 26.1 47.8 (5.8) 47.4–48.2 49 25–29.9 363 38.4 48.3 (5.2) 47.9–48.6 49 ≥ 30 279 29.5 48.8 (4.7) 48.5–49.0 50 Handedness Left 48 5.1 49.0 (5.0) 48.7–49.3 50 ANOVA 0.589 Right 869 91.7 48.2 (5.4) 47.8–48.5 49 Both 31 3.3 48.6 (4.7) 48.3–48.9 50 Birthplace t-test 0.142 Rural 251 26.5 47.7 (5.6) 47.3–48.1 48 Urban 697 73.5 48.5 (5.2) 48.2–48.8 49 Ancestorship t-test 0.161 Descendant of Propheta 79 8.2 47.5 (6.4) 47.1–47.9 49 Other 869 91.8 48.3 (5.2) 47.9–48.6 49

Total 948 100.0 48.3 (5.3) 48.0–48.6 49

aDescended from the Prophet Muhammad  SD = standard deviation. ANOVA = analysis of variance.

Using principal component (PC) analy- nents which explained 86% of the sis, 5 interrelated variables were analysed: variation: PC1 = +0.51FE +0.49ME father’s level of education (FE), mother’s +0.46SC +0.41WO +0.33FI (cultural sta- level of education (ME), family social class tus); PC2 = –0.26FE –0.14ME +0.27SC – (SC), woman’s occupation status (WO) 0.45FO +0.80FI (economic status); PC3 and family income (FI). These were re- = –0.09FE –0.48ME +0.46SC +0.60WO duced to 3 independent principal compo- +0.44FI (social class).

19 Sociodemographic factors.pmd 149 10/25/2005, 1:24 AM 150 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 2 Mean and median age at natural menopause according to sociodemographic and health behaviour characteristics

Variable No. of % Mean 95% CI Median Test P-value women (SD) age age (years) (years)

Marital status Single 17 1.8 44.7 (5.8) 44.3–45.1 45 ANOVA 0.074 Married 567 59.8 48.4 (5.3) 48.1–48.7 49 Widowed 344 36.3 48.2 (5.4) 47.8–48.5 49 Divorced 8 8.0 47.0 (4.7) 46.7–47.3 48 Separated 12 1.3 49.1 (2.8) 48.9–49.3 49 Never married 17 1.8 44.7 (5.8) 44.3–45.1 45 t-test 0.006 Ever married 931 98.2 48.3 (5.3) 47.9–48.6 49 Consanguinity No 661 71.0 48.1 (5.3) 47.8–48.4 49 t-test 0.027 Yes 270 29.0 48.9 (5.2) 48.6–48.9 50 Educational level (woman) Illiterate 463 48.5 48.1 (5.7) 47.6–48.6 49 ANOVA 0.774 Elementary 321 33.5 48.4 (4.7) 47.3–49.5 49 High school 131 13.8 48.8 (5.7) 47.8–49.8 49 University 40 4.2 48.7 (5.6) 47.0–50.4 50 Educational level (husband) Illiterate 556 58.6 48.1 (5.6) 47.7–48.4 49 ANOVA 0.765 Elementary 282 29.7 48.5 (4.9) 48.2–48.8 49 High school 72 7.6 48.8 (4.6) 48.5–49.1 49 University 38 4.0 48.7 (5.8) 48.3–49.1 50 Occupation (woman) Unemployeda 851 89.1 48.3 (5.2) 47.9–48.6 49 ANOVA 0.114 Unskilled manuala 16 1.7 45.3 (9.4) 40.7–54.7 48 Semi-skilled manualb 5 0.5 48.6 (3.6) 46.9–50.3 49 Clericalb,d 5 0.5 50.6 (6.7) 44.9–56.3 52 Managerialb,c 17 1.8 48.3 (4.4) 46.3–50.2 50 Professionalb,d 61 6.4 48.6 (5.1) 47.3–49.9 50 Occupation (husband) Unemployeda 83 8.8 46.6 (6.8) 46.2–47.0 47 ANOVA 0.001 Unskilled manuala 51 5.4 45.8 (6.6) 45.4–46.2 46 Semi-skilled manualb 165 17.4 47.7 (5.2) 47.4–48.0 49 Skilled manualb 208 21.9 48.6 (4.8) 48.3–48.9 49 Clericalb,d 164 17.3 48.3 (4.7) 48.0–48.6 50 Managerialb,c 176 18.6 49.5 (4.8) 49.2–49.8 50 Professionalb,d 101 10.7 49.1 (5.5) 48.7–49.4 50

19 Sociodemographic factors.pmd 150 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 151

Table 2 Mean and median age at natural menopause according to sociodemographic and health behaviour characteristics (concluded)

Variable No. of % Mean 95% CI Median Test P-value women (SD) age age (years) (years)

Income Poora 160 16.9 47.4 (5.0) 47.1–47.7 48 ANOVA 0.002 Lowa 314 33.1 47.9 (5.7) 47.5–48.3 49 Middlea 250 26.4 48.3 (5.2) 47.9–48.6 49 Highb 224 23.6 49.4 (4.9) 49.1–49.7 50 Exercise No 766 80.8 48.2 (5.4) 47.9–48.5 49 t-test 0.312 Yes 182 19.2 48.7 (5.1) 48.4–49.0 49.5 Cigarette smoking Yes 47 5.0 47.2 (5.2) 46.9–47.5 47 t-test 0.364 No 901 95.0 48.3 (5.3) 47.9–48.6 49 Water pipe smoking (no. per day) 0a 676 78.3 48.4 (5.3) 48.1–48.7 49 ANOVA 0.014 1–4a,b 209 22.0 48.5 (5.0) 48.2–48.8 49 5–14c 52 5.5 46.0 (6.1) 45.6–46.4 47 ≥ 15 11 1.2 46.8 (4.7) 46.5–47.1 46

a,b,cValues having different superscripts letters are significantly different (Duncan multiple comparison test). SD = standard deviation. ANOVA = analysis of variance.

Discussion The mean age at menopause in our study was 48.3 years. It is generally ac- Ascertaining the average age at natural cepted that the average age at menopause is menopause remains a topic of general inter- about 51 years in industrialized countries est, so too does the influence of biological [2,5,11,12,20], but data are inconsistent and social factors and intra- and inter-cul- for the developing world because of meth- tural differences. There is considerable un- odological problems. Definitions may differ certainty as to what factors affect the across studies so that pooling information timing of menopause. Genetic and racial from several published sources may be in- factors have recently been proposed to be appropriate. Table 3 presents a worldwide the most important determinants of age at comparison of age at natural menopause natural menopause [19]. In addition to ge- during the past quarter century. Mean age netic factors, several behavioural (smok- at natural menopause in our study was low- ing, nutrition, sociodemographic factors) er than most studies in Europe, Australia [1,8–13,16,19] and anthropometric factors and the United States of America [1,6,16] are also associated with age at [2,4,11,12,19], Japan [21] and China [8], menopause.

19 Sociodemographic factors.pmd 151 10/25/2005, 1:24 AM 152 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 3 Worldwide comparison of age at natural menopause during the past quarter century

Continent/country Year of No. of Mean Median Reference study women age age (years) (years)

Africa Egypt 1999 289 46.7 – [1] Nigeria 1990 563 48.4 48 [9] Kenya (west) 1997 1 078 – 48.3 [23] America USAa 1989 2 014 – 50.7 [2] Mexico 1996 472 46.5 47 [7] USAa 1997 185 – 51.5 [20] USA 2001 14 620 – 51.4 [11] Asia China 1991 124 49.8 50 [8] Islamic Republic of Iran (north) 1992 400 47.3 – [13] Islamic Republic of Iran (centre) 2000 404 49.3 – [17] Lebanon 2001 298 – 49.3 [10] UAE 1998 742 47.3 48 [16] Australia Australia 1998 5 961 – 51 [12] Europe Finland 1994 1 713 – 51 [19] Italya 2000 4 300 50.9 – [5]

aProspective studies (others retrospective). USA = United States of America. UAE = United Arab Emirates.

but higher than Egypt and the United Arab though our underweight subjects experi- Emirates [1,16]. enced menopause earlier than obese wom- Age at natural menopause was not asso- en, this relationship was not statistically ciated with weight in our study, which con- significant. curs with an American study [2]; however, Left-handed women have been reported a significant association was observed in an to experience menopause earlier than right- Egyptian study [1]. Height showed no sig- handed subjects [3,4]. Our study did not nificant association with menopausal age, confirm this association, which supports which contradicts with the results of a pro- the findings of Pavia et al. [22]. spective study carried out in the United The mean age at natural menopause of States [2]. Several studies have established never married women was significantly a significant association between age at lower than ever married subjects, which natural menopause and BMI [1,5,6]. Al-

19 Sociodemographic factors.pmd 152 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 153

concurs with the findings of others mon among this group of older women [10,11,14]. (28.7% of the sample) and showed a sig- Lower social class was highly associat- nificant association with menopausal age ed with early menopausal age in our study (P < 0.014). (P < 0.001), which is in accordance with Although the majority (> 80%) of our other findings [2,7,12]. Some other stud- study population were illiterate or had ele- ies, however, did not establish this associa- mentary education, the cultural and reli- tion [8–10,13]. Three independent factors gious importance of menopause means that explained 86% of the variation in meno- women in this study were able to remem- pausal age: cultural status, economic status ber their menopausal age accurately and its and social class. precision was quite reliable. While 2 other Cigarette smoking has been shown in limited clinic-based studies on menopausal many studies to be an associated factor that age have been carried out in the Islamic lowers the age of natural menopause Republic of Iran, this is the first popula- [2,5,10–12,15,16,20], whereas a few stud- tion-based study and we conclude that the ies have not identified smoking as a deter- results may be generalized to other urban minant of menopausal age [9,13]. Our populations in the country. study falls in the latter category. This may be because cigarette smoking is not cultur- Acknowledgement ally acceptable among women in the Islam- ic Republic of Iran and is therefore This work was supported by grant no. 79- uncommon. However, smoking of water 1143 from Shiraz University of Medical pipes (“hubble-bubble”, kalyan) was com- Sciences.

References 1. Hidayat NM et al. Correlates of age at cross-sectional study. Maturitas, 2000, natural menopause: a community-based 34(2):119–25. study in Alexandria. Eastern Mediterra- 6. Monninkhof EM, Van der Schouw YT, nean health journal, 1999, 5(2):307–19. Peeters PH. Early age menopause and 2. Brambilla DJ, McKinlay SM. A prospec- breast cancer: are leaner women more tive study of factors affecting age at protected? A prospective analysis of the menopause. Journal of clinical epidemi- Dutch DOM cohort. Breast cancer re- ology, 1989, 42(11):1031–9. search and treatment, 1999, 55(3):285– 91. 3. Dane S, Reis N, Pasinliogu T. Left- handed women have earlier age of 7. Garrido-Latorre F, Lazcano-Ponce EC, menopause. Journal of basic and clini- Lopez-Carrillol L. Age of natural meno- cal physiology and pharmacology, 1999, pause among women in Mexico City. In- 10(2):147–50. ternational journal of gynaecology and obstetrics, 1996, 53(2):159–66. 4. Leidy LE. Early age at menopause among left-handed women. Obstetrics 8. Chang C, Chow SN, Hu Y. Age of meno- and gynecology, 1990, 76(6):1111–4. pause of Chinese women in Taiwan. In- ternational journal of gynaecology and 5. Meschia M et al. Determinants of age at obstetrics, 1995, 49:191–2. menopause in Italy: results from a large

19 Sociodemographic factors.pmd 153 10/25/2005, 1:24 AM 154 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

9. Okonofua FE, Lawal A, Bamgbose JK. 16. Bener A et al. Consanguinity and the age Features of menopause and meno- of menopause in the United Arab Emir- pausal age in Nigerian women. Interna- ates. International journal of gynaeco- tional journal of gynaecology and logy and obstetrics, 1998, 60(2):155–60. obstetrics, 1990, 31(4):341–5. 17. Tabatabaie BA. Mean age at menopause 10. Reynolds RF, Obermeyer CM. Age at in Yazd and hormone use. Journal of natural menopause in Beirut, Lebanon: Shaheed Sadouqi University of Medical the role of reproductive and lifestyle fac- Sciences, 2000, 8(1):31–4 [in Persian]. tors. Annals of human biology, 2001, 18. Sowers MR, La Pietr MT. Menopause: its 28(1):21–9. epidemiology and potential association 11. Gold EB et al. Factors associated with with chronic diseases. Epidemiology re- age at natural menopause in a multi- views, 1995, 17:287–302. ethnic sample of midlife women. Ameri- 19. Luoto R, Kaprio J, Uutela A. Age at natu- can journal of epidemiology, 2001, ral menopause and socio-demographic 153(9):65–74. status in Finland. American journal of 12. Do KA et al. Predictive factors of age at epidemiology, 1994, 139(1):64–76. menopause in a large Australian twin 20. Bromberger JT et al. Prospective study of study. Human biology, 1998, 70(6): the determinants of age at menopause. 1073–91. American journal of epidemiology, 1997, 13. Akbarian A et al. Age at menopause and 145(2):124–33. its influencing factors in Rey, Teheran. 21. Kono S et al. Age of menopause in Japa- Journal of Feiz, 1999, 11:46–52 [in Per- nese women: trends and recent sian]. changes. Maturitas, 1990, 12(1):43–9. 14. Sievert LL, Waddle D, Canalik K. Marital 22. Pavia M et al. Handedness, age at me- status and age at natural menopause: narche, and age at menopause. Obstet- considering pheromonal influence. rics and gynecology, 1994, 83(4): American journal of human biology, 579–82. 2001, 13(4):479–85. 23. Noreh J et al. Median age at menopause 15. Parazzini F, Negri E, La Vecchia C. Re- in a rural population of western Kenya. productive and general lifestyle determi- East African medical journal, 1997, nants of age at menopause. Maturitas, 74(10):634–8. 1992, 15(2):141–9.

19 Sociodemographic factors.pmd 154 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 155 Epidemiology of dysmenorrhoea among adolescent students in Mansoura, Egypt A-H. El-Gilany,1 K. Badawi 2 and S. El-Fedawy 2

ABSTRACT To examine the prevalence, determinants, impact and treatment practices of dysmenorrhoea, we studied 664 female students in secondary schools in urban and rural areas. Data was collected through a self-administered questionnaire. About 75% of the students experienced dysmenorrhoea (mild 55.3%, moderate 30.0%, severe 14.8%). Most did not seek medical advice although 34.7% treated themselves. Fatigue, headache, backache and dizziness were the commonest associated symptoms. No limitation of activities was reported by 47.4% of student with dysmenorrhoea, but this was significantly more reported by students with severe dysmenorrhoea. Significant predictors of dysmenorrhoea were older age, irregular or long cycle and heavy bleeding.

Épidémiologie de la dysménorrhée chez des élèves adolescentes à Mansoura (Égypte) RÉSUMÉ Pour examiner la prévalence, les déterminants, l’impact et les pratiques de traitement de la dysménorrhée, nous avons mené une étude auprès de 664 élèves d’écoles secondaires en zone urbaine et rurale. Des données ont été recueillies à l’aide d’un auto-questionnaire. Environ 75 % des élèves ont eu une dysménorrhée (légère 55,3 %, modérée 30,0 %, sévère 14,8 %). La plupart d’entre elles n’ont pas recher- ché un avis médical bien que 34,7 % se soient soignées elles-mêmes. Fatigue, céphalées, dorsalgies et vertiges étaient les symptômes les plus courants qui y étaient associés. Aucune limitation des activités n’a été rapportée par 47,4 % des élèves souffrant de dysménorrhée, mais ce point était signalé significativement plus souvent par les élèves souffrant de dysménorrhée sévère. Les facteurs prédictifs de la dysmé- norrhée étaient l’âge plus avancé, un cycle menstruel irrégulier ou long et des menstrues abondantes.

1Department of Community Medicine, Faculty of Medicine; 2Students’ University Hospital, University of Mansoura, Mansoura, Egypt (Correspondence to A-H. El-Gilany: [email protected]). Received: 21/04/03; accepted: 23/03/04

20 Epidemiology of dysmenorrhoea.pmd155 10/25/2005, 1:24 AM 156 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 Introduction Methods Dysmenorrhoea, or painful menstruation, This study was carried out during the peri- is defined as a severe, painful cramping od November 2001–April 2002 in Mansou- sensation in the lower abdomen [1]. It may ra district, Egypt. A cross-sectional survey be accompanied by headache, dizziness, di- was carried out. The target population was arrhoea, a bloated feeling, nausea and vom- female secondary-school students enrolled iting, backache and leg pains. Primary in government schools. Approval of the lo- dysmenorrhoea occurs in the absence of cal directorate of education was obtained. recognizable pelvic pathology and com- The eastern and western educational zones monly begins when the ovulatory menstru- (municipal subdivisions of the city) as well al cycle starts. The pain starts a few hours as the rural sector were represented. Both before menstruation and lasts for up to 72 general and technical secondary schools hours. It is usually most severe on the first were represented. One general secondary day of menstruation and gradually dimin- school was randomly selected from each ishes. It is caused by increased endometrial of the eastern and western zones (the num- prostaglandin production and almost al- ber of schools is similar in each zone) as ways first occurs in women younger than well as 1 school from the rural sector. One 20 years [1–4]. technical commercial school and 1 nursing Secondary dysmenorrhoea, on the oth- school were selected from Mansoura city. er hand, is associated with pelvic condi- All social strata as well as the urban and tions or pathology that cause pelvic pain in rural sectors of the community were repre- conjunction with the menses. This usually sented in this distribution. From each se- appears later in a woman’s reproductive life lected school, 1 class (cluster) from each and can occur with anovulatory cycles. It grade was randomly selected. A total of 15 often lasts for 5 to 7 days each month, and classes were studied. A total of 694 stu- progressively increases in severity [1–4]. dents were registered in the chosen class- Dysmenorrhoea has a negative effect es. Of these, 664 (95.7%) participated in on a woman’s life. It may be so severe as to the study. Others were either absent confine the woman to bed. During adoles- (3.6%) or refused to complete the ques- cence, dysmenorrhoea leads to high rates tionnaire (0.7%). of absence from school and non-participa- In cooperation with the school authori- tion in activities. Mild to moderate cases ties, female investigators spent about 45– can usually be treated by reassurance and 60 minutes in each class. The students paracetamol [1,2,5]. were briefed about the study, encouraged Population studies on dysmenorrhoea to participate and motivated to express their are scarce for Egyptian women, and practi- experiences. It was emphasized that all data cally non-existent for adolescent girls. In collected were strictly confidential. Stu- this study, therefore, we aimed to estimate dents were requested to complete a self- the prevalence of dysmenorrhoea and to administered, anonymous questionnaire study its determinants and impact as well covering family background, age at me- as treatment practices among adolescent narche, duration and amount of bleeding, students in Mansoura, Egypt. cycle length, pain during menstruation

20 Epidemiology of dysmenorrhoea.pmd156 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 157

(dysmenorrhoea) during the previous 3 Table 1 Sociodemographic characteristics of months, severity of the pain (mild, moder- secondary school students in Mansoura ate or severe, subjectively assessed), duration of pain, any associated symptoms, Variable No. % impact on daily activities and treatment tak- Residence en, if any. Dysmenorrhoea was defined as Urban 353 53.2 lower abdominal pain associated with men- Rural 311 46.8 strual periods. The social score and family Family social class social class were calculated according to High 175 26.4 Fahmy and El-Sherbiny [6]. Middle 124 18.7 Data were analysed using SPSS, version Low & very low 365 55.0 9. The chi-squared test was used as a test Religion of significance. Factors significantly af- Muslim 651 98.0 fecting prevalence of dysmenorrhoea on Christian 13 2.0 univariate analysis were entered into multi- Age (years)a variate logistic regression analysis. P 14 109 16.4 ≤ 0.05 was considered to be statistically 15 217 32.7 significant. 16 240 36.1 17+ 98 14.8 Ever menstruating Results Yes 642 96.7 No 22 3.3

The sociodemographic characteristics of b,c the students in the study are shown in Table Age at menarche (years) < 12 47 7.1 1. The majority (96.7%) ever menstruated. 12 178 26.8 The mean and median ages at menarche 13 244 36.7 were 12.9 years and 13.0 years respective- 14+ 173 26.1 ly. The vast majority of the students Total 664 100 (98.0%) were Muslim. Just over half aRange 14–18 years; mean 15.5 years (standard (53.2%) were from an urban residence. deviation 0.99). The overall prevalence of dysmenor- bRange 10–16 years; median 13.0 years; mean rhoea (assumed to be primary dysmenor- 12.9 years (standard deviation 1.03). c rhoea, as secondary is rare at this age) was Among ever menstruating adolescents. 74.6%; it was significantly more frequent among students from a rural residence, in those from low and very low social class- Dysmenorrhoea was mild in the majori- es, those of older age, those who said they ty of cases; only 14.8% of students with had an irregular cycle, those who stated dysmenorrhoea reported having severe they had heavy bleeding, those with long forms. In the majority of cases, the dura- duration of bleeding (≥ 6 days) and those tion of the pain was less than 24 hours. The with long cycles (≥ 30 days) (Table 2). most frequent symptoms associated with On logistic regression analysis, the sig- dysmenorrhoea were fatigue, headache, nificant determinants of dysmenorrhoea backache, dizziness and anorexia/vomiting. were older age, cycle irregularity, heavi- In 23.8% of cases there were no associated ness of menstrual flow and longer cycle symptoms. About half the students with length (Table 3). dysmenorrhoea did not take any medica-

20 Epidemiology of dysmenorrhoea.pmd157 10/25/2005, 1:24 AM 158 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 2 Prevalence and determinants of dysmenorrhoea among ever menstruating adolescents in Mansoura

Variable Total Dysmenorrhoea Significance test No. %

Residence Urban 335 233 69.6 χ2 = 9.5; P = 0.002 Rural 307 246 80.1 Family social class High 163 107 65.6 χ2 = 12.5; P = 0.002 Middle 119 85 71.4 Low & very low 360 287 79.7 Age (years) 14 97 60 61.9 χ2 = 31.4; P < 0.001 15 210 142 67.6 16 237 187 78.9 17+ 98 90 91.8 Age at menarche (years) < 12 47 38 80.9 χ2 = 1.8; P = 0.6 12 178 133 74.7 13 244 184 75.4 14+ 173 124 71.7 Cycle regularitya Regular 429 298 69.5 χ2 = 18.1; P < 0.001 Irregular 213 181 85.0 Bleeding amounta Drops 28 11 39.3 χ2 = 25.6; P < 0.001 Average 503 373 74.2 Heavy 111 95 85.6 Bleeding duration (days) < 4 63 40 63.5 χ2 = 6.8; P = 0.033 4–5 400 296 74.0 ≥ 6 179 143 79.9 Cycle length (days) < 30 245 166 67.8 χ2 = 9.8; P = 0.002 ≥ 30 397 313 78.8 Overall 642 479 74.6

aSubjectively reported.

tion; the others reported using herbs/home The impact of dysmenorrhoea on daily remedies (36.7%) and analgesics/non- activities is shown in Table 5. No limitation steroidal anti-inflammatory drugs was reported by 47.4% overall, but by only (NSAIDs)/antispasmodics (34.7%) (Table 2.8% of those who stated they had severe 4). These drugs were mostly self-pre- dysmenorrhoea. The activities the students scribed; only 13 students (2.7%) consulted reported as most often being limited were a physician or pharmacist. daily home chores (42.8%), going out of

20 Epidemiology of dysmenorrhoea.pmd158 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 159

Table 3 Logistic regression analysis of factors affecting dysmenorrhoea among ever menstruating adolescent students (n = 642)

Variable β P OR 95% CI

Age (years) 14a –1 15 0.28 0.3 1.32 0.78–2.23 16 0.78 0.005 2.18 1.27–3.72 17+ 1.89 < 0.001 6.59 2.82–15.4 Cycle regularity Regular –0.61 0.009 0.54 0.34–0.86 Irregular a –1 Menstrual flow Drops a –1 Average 1.31 0.002 3.71 1.63–8.47 Heavy 1.88 0.0002 6.54 2.47–17.29 Cycle length (days) <30 a –1 ≥ 30 0.39 0.048 1.48 1.00–2.19 Constant –0.61 –2 log likelihood 657.7 Model χ2 69.7; P < 0.001

OR = odds ratio, CI = confidence interval. aReference group.

the home (41.5%), participation in social health status, quality of life and social inte- activities (39.0%) and participation in gration among women in developing coun- sports (34.4%). Limitations on activities tries are scant. The lack of data and the were significantly more frequently reported private nature of menstruation perpetuate among students who reported having se- the belief that menstrual complaints do not vere dysmenorrhoea. warrant the attention of the public health community [7,8]. Dysmenorrhoea is the commonest gy- Discussion naecologic disorder among female adolescents and is one of the commonest Over the past decade there has been a para- gynaecologic complaints in young women digm shift in the field of population studies, who present to doctors today [3,9,10]. moving from a relatively singular focus on Dysmenorrhoea among adolescents is usu- family planning to a broader focus on re- ally of the primary type [1–4,11,12]. productive health. Menstrual problems are In our study, 74.8% of adolescent stu- generally perceived as only minor health dents reported pain with menstruation dur- concerns and thus irrelevant to the public ing the previous 3 months. This is health agenda. Data on the frequency of comparable to previously reported preva- menstrual dysfunction and its impact on

20 Epidemiology of dysmenorrhoea.pmd159 10/25/2005, 1:24 AM 160 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 4 Clinical presentation and treatment of perception and variability in pain threshold. dysmenorrhoea among ever menstruating Correlation between ethnicity and pain per- adolescent students in Mansoura ception has previously been reported [10]. Duration of pain was less than 24 hours Characteristic No. % in 64.9% of cases and only 8.6% of adoles- Severity cents reported pain lasting for more than 48 Mild 265 55.3 hours. Banikarim, Chacko and Kelder re- Moderate 143 30.0 ported that 90% of adolescents had men- Severe 71 14.8 strual cramps lasting for 48 hours or less Duration of pain [10]. In 23.8% of cases, dysmenorrhoea < 24 hours 311 64.9 was not associated with other symptoms 24–48 hours 127 26.5 (fatigue, headache, backache, dizziness, > 48 hours 41 8.6 anorexia/vomiting, abdominal pain/disten- Associated symptomsa sion and diarrhoea). Most of the studies we None 114 23.8 reviewed reported the same duration of Fatigue 339 70.8 pain as in our study but the associated Headache 297 62.0 symptoms occurred with different fre- Backache 272 56.8 quencies [2,3,10,13,15]. Dizziness 144 30.1 It has been reported that the risk of dys- Anorexia/vomiting 66 13.8 Abdominal distension/ menorrhoea is higher in women with irreg- bloating 39 8.1 ular, prolonged or heavy menstrual flow as Diarrhoea 18 3.8 well as early age of menarche [2,14, Treatmenta 16,20]. In our study we found that preva- None 237 49.5 lence of dysmenorrhoea was significantly Rest/relaxation 204 42.6 higher among adolescents aged 16 and 17 Herbs/home remedies 166 36.7 years compared to those aged 14 years, Analgesics/NSAIDs/ those who had irregular cycles and those antispasmodicsb 176 34.7 reporting heavier menstrual flow or a pro- Total 479 74.6 longed cycle (≥ 30 days). Treatment of dysmenorrhoea should be aCategories are not mutually exclusive. bSelf-prescribed in 166 (97.3%). directed at providing relief from the cramp- NSAIDs = non-steroidal anti-inflammatory drugs. ing pelvic pain and associated symptoms. Non-steroidal anti-inflammatory drugs and oral contraceptives are reported as provid- lence in both industrialized and developing ing the most effective treatment [7]. The countries that ranged from 20% to 93% for use of oral contraceptives by unmarried the same age group [2,3,5,7,13–19]. The girls is, however, culturally unacceptable in severity of dysmenorrhoea varied greatly. our traditional and conservative communi- In our study 14.8% of adolescents with ty. dysmenorrhoea reported their pain as se- In our study, only 2.7% of adolescents vere. In other countries, severe dysmenor- consulted a physician or pharmacist. This rhoea was reported by 15%–53% of is consistent with other findings that most adolescents [7,10,14,17,18,20]. These dif- adolescents with dysmenorrhoea self- ferences in the degree of pain severity may medicate with the over-the-counter prepa- be related to cultural differences in pain rations; few consult health care providers

20 Epidemiology of dysmenorrhoea.pmd160 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 161

Table 5 Impact of dysmenorrhoea on daily activities

Activity limiteda Dysmenorrhoea Total χ 2 Mild Moderate Severe No. % No. % No. % No. %

No limitation 169 63.8 56 39.2 2 2.8 227 47.4 89.0 Daily home chores 73 27.5 72 50.3 60 84.5 205 42.8 79.0 Going out of the home 60 22.6 71 49.7 68 95.8 199 41.5 128.6 Participation in social events 51 19.2 66 46.2 70 98.6 187 39.0 152.5 Participation in sports 47 17.7 49 34.3 69 97.2 165 34.4 156.5 Concentration in class 28 10.6 32 22.4 57 80.3 117 24.4 147.9 Homework tasks 16 6.0 38 26.6 49 69.0 103 21.5 134.7 School attendance 13 4.9 41 28.7 43 60.6 97 20.3 116.4 Total 265 100 143 100 71 100 479 100

aCategories are not mutually exclusive. P < 0.001.

[5,10]. We found that rest/relaxation, herb- daily activities. Activities most commonly al/home remedies and/or drugs were used limited due to dysmenorrhoea were daily by 42.6%, 36.7% and 34.7% of partici- home chores, going out of the home, par- pants respectively. The drugs included anticipation in social events, participation in algesics, NSAIDs and antispasmodics, sports, concentration in class, homework mostly self-prescribed. tasks and attending school. All the limita- Banikarim, Chacko and Kelder reported tions were significantly more frequent that treatment for dysmenorrhoea in His- among students with severe dysmenor- panic adolescents included rest (58%), rhoea compared to those with mild or mod- medication (52%), heating pad (26%), tea erate pain. Banikarim, Chacko and Kelder (20%), exercise (15%) and/or herbs (7%) reported that activities limited by dysmen- [10]. It has been reported that the most orrhoea among adolescents included con- common medications used by women with centration in class (59%), sports (51%), dysmenorrhoea were analgesics (53%) and class participation (50%), socialization NSAIDs (42%) [17]. (46%), homework (35%), test-taking skills Although not life threatening, dysmen- (36%) and grades (29%) [10]. In a study in orrhoea can be particularly disruptive to a Morocco, menstrual pain was often cited woman’s daily life and productivity. In the as the main single cause of school absen- absence of appropriate pain relief, women teeism among adolescent girls [16]. with severe dysmenorrhoea may not be Reproductive health information and able to carry out their normal activities education programmes for adolescents are [7,8]. being introduced in many countries, and In our study, 47.4% of students with these could be an important means of pro- dysmenorrhoea reported no limitation of viding information about treatment options

20 Epidemiology of dysmenorrhoea.pmd161 10/25/2005, 1:24 AM 162 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

for menstrual disorders [8]. Young girls and menstrual problems in particular. It is may be more open than older women to essential to make treatment available for discussing menstruation. girls. Many girls may feel shameful and re- The introduction of a reproductive luctant to report dysmenorrhoea and con- health component into school health educa- sequently, do not seek medical advice. It is tion programme could help in providing in- one of the roles of school health care pro- formation, education and support to viders to ask about and screen for dysmen- students regarding reproduction in general orrhoea and offer treatment if necessary.

References 1. Pearce JM. Disturbances of the men- 8. Walraven G et al. Menstrual disorders in strual cycle. In: Varma TR, ed. Clinical rural Gambia. Studies in family planning, gynaecology. London, Arnold, 1991: 2002, 33(3):261–8. 100–17. 9. Jamieson DJ, Steege JF. The prevalence 2. Herbst AL et al. Comprehensive gyne- of dysmenorrhoea, dyspareunia, pelvic cology, 2nd ed. Chicago, Mosby Year pain and irritable bowel syndrome in pri- Book Medical Publishers, 1996:1063. mary care practices. Obstetrics and gynecology, 1996, 87(1):55–8. 3. Ryan KJ, Barbieri RL. The menstrual cycle. In: Ryan KJ, Berkowitz R, Barbieri 10. Banikarim C, Chacko MR, Kelder SH. RL, eds. Kistner’s gynecology: principles Prevalence and impact of dysmenorrhea and practice, 6th ed. Chicago, Mosby on Hispanic female adolescents. Ar- Year Book Medical Publishers, 1995:15. chives of pediatrics & adolescent medicine, 2000, 154(12):1226–9. 4. Rapkin AJ. Pelvic pain and dysmenorrhea. In: Berek JS, Adashi EY, Hillard PA, 11. Koltz MM. Dysmenorrhea, endometriosis eds. Novak’s gynecology, 12th ed. and pelvic pain. In: Lemke DP et al. eds. (Middle East ed). Giza, Egypt, Mass Pub- Primary care of women. Norwalk, Con- lishing Co., 1996:399. necticut, Appleton & Lange, 1992:420– 32. 5. Davis AR, Westhoff CL. Primary dysmenorrhea in adolescent girls and treatment 12. Harlow SD, Park M. a longitudinal study with oral contraceptives. Journal of pedi- of risk factors for the occurrence, dura- atric and adolescent gynecology, 2001, tion and severity of menstrual cramps in 14(1):3–8. a cohort of college women. British journal of obstetrics and gynaecology, 1996, 6. Fahmy SI, El-Sherbini AF. Determining 103(11):1134–42. simple parameters for social classifica- tions for health research. Bulletin of the 13. Wood RP, Larsen L, Williams B. Social High Institute of Public Health, 1983, and psychological factors in relation to 13(5):95–108. premenstrual tension and menstrual pain. Australian & New Zealand journal 7. Harlow SD, Campbell OMR. Menstrual of obstetrics & gynaecology, 1979, 19(2): dysfunction: a missed opportunity for im- 111–5. proving reproductive health in developing countries. Reproductive health 14. Andersch B, Milsom J. An epidemiologic matters, 2000, 8(15):142–7. study of young women with dysmenor-

20 Epidemiology of dysmenorrhoea.pmd162 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 163

rhea. American journal of obstetrics and of menstrual discomfort. Archives of pe- gynecology, 1982, 144(6):655–60. diatrics & adolescent medicine, 1997, 151(9):905–12. 15. Thomas KD, Okonofua FF, Chiboka O. A study of the menstrual patterns of ado- 19. Klein J, Litt I. Epidemiology of adolescent lescents in Ile-Ife, Nigeria. International dysmenorrhoea. Pediatrics, 1981, 68(5): journal of gynaecology and obstetrics, 661–4. 1990, 33(1):31–4. 20. Di Cintio E et al. Dietary habits, repro- 16. Montero P et al A. Characteristics of ductive and menstrual factors and risk of menstrual cycles in Moroccan girls: dysmenorrhoea. European journal of prevalence of dysfunctions and associ- epidemiology, 1997, 13(8):925–30. ated behaviour. Annals of human biol- 21. Harlow SD, Park M. A longitudinal study ogy, 1999, 26(3):243–9. of risk factors for the occurrence, dura- 17. Hillen TI et al. Primary dysmenorrhoea in tion and severity of menstrual cramps in young Western Australian women: a cohort of college women. British jour- prevalence, impact and knowledge of nal of obstetrics and gynaecology, 1996, treatment. Journal of adolescent health, 103(11):1134–42. 1999, 25(1):40–5. 18. Campbell M, McGrath P. Use of medication by adolescents for the management

Adolescent health and development One in every five people in the world is an adolescent – defined by WHO as a person between 10 and 19 years of age. Out of 1.2 billion adolescents worldwide, about 85% live in developing countries and the remainder live in the industrialized world. Adolescents are generally thought to be healthy: by the second decade of life, they have survived the diseases of early childhood, and the health problems associated with ageing are still many years away. Yet adolescents are still a vulnerable sector of the population and many do die pre- maturely. WHO, along with its partners, UNICEF and UNFPA, advocate an accelerated approach to promoting the health and development of young people in the second decade of life. Further information on the work of the Adolescent Health and Development Team within the WHO Department of Child and Adolescent Health and Develop- ment is available at: http://www.who.int/topics/adolescent_health/ en/

20 Epidemiology of dysmenorrhoea.pmd163 10/25/2005, 1:24 AM 164 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005 A study of domestic violence among women attending a medical centre in Sudan A.M. Ahmed1 and A.E. Elmardi 2

ABSTRACT To investigate domestic violence in the Sudanese family, we studied 394 literate, married women attending the Arda Medical Centre, Omdurman, from October 2001 to February 2002. Through self- administered questionnaires, the women provided data on sociodemographic characteristics and abuse by the husband. Abuse was reported by 164 women (41.6%), who suffered 525 violent episodes in the previous year, classified into controlling behaviour (194), threatening behaviour (169) and physical violence (162). Frequency of violent episodes varied from 1 (25%) to > 6 (20.7%). Violence during pregnancy was reported by 27 women (16.5%). Provoking events included suspicion of illicit relations, talking back and inadequate home care. Common reactions reported by the women included staying quiet, crying and resistance.

Étude de la violence familiale chez des femmes consultant dans un centre médical au Soudan RÉSUMÉ Afin d’étudier la violence dans la famille soudanaise, nous avons procédé à une enquête auprès de 394 femmes mariées instruites qui sont venues consulter au Centre médical Arda d’Omdurman d’octobre 2001 à février 2002. Au moyen d’auto-questionnaires, les femmes ont fourni des données sur les caractéris- tiques sociodémographiques et les mauvais traitements infligés par le mari. Des mauvais traitements ont été signalés par 164 femmes (41,6 %), qui ont vécu 525 épisodes de violence au cours de l’année précédente, classés comme comportements autoritaires (194), comportements menaçants (169) et violence physique (162). La fréquence des épisodes de violence variait de 1 (25 %) à plus de 6 (20,7 %). La violence pendant la grossesse a été rapportée par 27 femmes (16,5 %). Parmi les événements qui déclenchaient la violence figuraient les suivants : le soupçon de relations illicites, répondre insolemment et ne pas bien s’occuper du foyer. Les réactions courantes signalées par les femmes étaient le silence, les pleurs et la résistance.

1Faculty of Medicine, University of Bahr Elghazal, Khartoum, Sudan (Correspondence to A.M. Ahmed: [email protected]). 2Department of Community Medicine, Faculty of Public Health, University of Khartoum, Khartoum, Sudan. Received: 18/03/03; accepted: 15/03/04

21 A study of domestic.pmd 164 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 165

Introduction prevalence than has been assumed. A prevalence of more than 40% has been reported Violence is defined as a behaviour toward in American and British studies [5–7]. Over another person which is outside the norms one third of women attending general prac- of conduct, and entailing a substantial risk tices had experienced violence [7]. The of causing physical or emotional harm [1]. healthcare costs associated with manage- The main types of violence include child ment of family violence injuries in the Unit- abuse, peer assault (including domestic ed States of America has been estimated at violence), stranger assault, abuse of the US$ 857 million annually [8]. In 1992, 12% elderly, state violence and war crimes. of all homicides resulted from intrafamilial Throughout history, in most societies violence [9]. women have held a much lower status than Violence against women is a product of men [2]. This situation renders them pro- the interaction of factors at individual, fam- foundly disadvantaged in terms of power, ily, community and society levels [1]. At wealth and personal freedom. The low val- the individual level, these factors include ue placed on women’s lives in many coun- being abused as a child, witnessing vio- tries can be evidenced by sombre lence at home, having an absent or rejecting indicators such as high maternal mortality, father and frequent use of alcohol or drugs. high level of illiteracy and unequal access At the family level, marital conflicts and the to government services [2]. Owing to these dominant male control of wealth and family factors, and as a result of their gender- issues are considered strong predictors of rooted subordinate status, women are per- abuse [10]. At the community and society petual victims of family, community and levels, the factors which interplay to pro- government violence. Examples of violence duce violence include poverty; unemploy- against women include spousal abuse, ment; lack of support and isolation of rape, exploitation of labour, trafficking, family and women; linkage of the concept forced prostitution, genital mutilation, debt of masculinity to male honour or domi- bondage and infanticide [1]. nance; acceptance of violence as a way to The term “domestic violence” is usually resolve conflicts; and social tolerance of used to define violence exerted toward the physical punishment of women [10]. Vio- woman by a family member (most com- lence may evolve from the socially accept- monly the husband or the intimate male able gender norms. Men are the family partner) [1]. It can be seen as a pattern of masters (or even women’s owners!) be- psychological, economic and sexual coer- cause they provide financially. Women are cion of one partner in a relationship by the expected to tend the house, mind the chil- other that is punctuated by physical assault dren and show obedience to their hus- or credible threat of bodily harm [3]. bands. A breach in the woman’s role or Physical violence can be anything from challenging the men’s rights may provoke pushing or punching to firearms injuries. violence. Apart from physical assaults, domestic vio- Violence against women has profoundly lence can be seen as a set of learned, con- grave consequences. It compromises their trolling behaviours and attitudes of health and erodes their self-esteem. In addi- entitlement that are culturally supported tion to causing injuries, violence may result and produce relationships of entrapment in physical disability, depression, unintend- [4]. Domestic violence has a much greater ed pregnancy, sexually transmitted infec-

21 A study of domestic.pmd 165 10/25/2005, 1:24 AM 166 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

tion, adverse pregnancy outcome and sui- ical centre receives patients mainly for fol- cidal thoughts [11]. Domestic violence can low-up of chronic diseases (of which dia- lead to marital conflict and hinder women’s betes mellitus and hypertension are the participation in the social and economic commonest) first seen by junior doctors in well-being of the society [12]. This is smaller health units. The clinic receives pa- worsened by the fact that men may per- tients from both urban and suburban areas ceive the empowerment of their wives as a of Omdurman (the largest city in Sudan). threat to their own control [12]. We recruited 492 women as eligible for Worldwide, awareness of the extent of our study, 394 of them gave their consent domestic violence has been increasing to participate, a response rate of 86.8%. Of since the first report about battered women the respondents, 164 women (41.6%) gave appeared in the early 1970s [13]. Since a history of 1 or more forms of abusive then, an efficient multidisciplinary system behaviour by their husbands in the previous has evolved in the industrialized countries year. The remaining 230 women (88.4%) to help victims. In the developing coun- served as a control group for our study in tries, including Sudan, neither the extent of regard to the sociodemographic character- the problem nor any protective measures istics of the abused. have been established. In Sudan, women The women eligible for the study were constitute 50% of the population (15 mil- all the married and literate women seen lion out of 30 million) [2]. They have a life consecutively during the study period. expectancy of 52 years and an illiteracy Women who were too ill to complete the rate of 60%; both figures are lower than questionnaire were excluded. The partici- those for their male counterparts [2]. pants provided oral consent—we avoided This study is the first effort to study written consent so as to preclude any domestic violence in Sudan. Up to now the chance of identifying the origin of any par- majority of health workers have been un- ticular questionnaire administered. Three aware of or indifferent to their role in help- female research assistants, who had been ing victims of violence (they do not screen trained in interview techniques, assisted in for or consider the role of violence in their recruiting the women and answering que- patients’ health problems). Our study ries raised by them in regard to completion aimed at investigating the problem of do- of the questionnaire. For each eligible mestic violence among women attending a woman we offered a brief description of medical centre for chronic diseases in re- the nature of the study. We clearly ex- gard to extent, patterns and determinants. plained that they would be asked sensitive questions about their relationship with their husbands. In our society, close relation- Methods ships of women (including sexual life) are This study was done among the women at- only allowed with the husbands following a tending the medical clinic of the Arda Med- formal marriage. Otherwise they may face ical Centre in Omdurman, Sudan in the death or a prison penalty according to the period 31 October 2001–28 February sharia (the Islamic legal code). 2002. The centre has referral clinics in all Each woman, after giving her consent the clinical specialties, and is run by con- to participate in the study, was given an sultant physicians and surgeons. The med- anonymous, self-administered question-

21 A study of domestic.pmd 166 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 167

naire in a private setting with strict confi- lence, frequency of abuse, her reaction to dentiality assured. The questionnaire was the assault, whether she had sought medi- developed from those used in international cal or other help, whether she had been studies which we modified to suit hurt by other family members (e.g. father, Sudanese women. It was piloted on 30 brother), whether she had ever been as- women before being used in this study. saulted during pregnancy, whether she had For the purposes of our study, domestic even been asked by a doctor about abuse. violence was defined as an assault, threat The analysis of the interval variables of or intimidation perpetrated by a husband the characteristics of both abused and non- [14]. Abusive behaviours were categorized abused women was done on frequency into threats, controlling behaviours and data using the Student t-test. The signifi- physical assaults. The physical assaults cance levels were determined at P < 0.005. were classified into minor (e.g. throwing objects, shoving), moderately severe (e.g. beating up, producing contusions) and se- Results vere (e.g. causing head and internal inju- Table 1 shows the sociodemographic char- ries). In the questionnaire we required acteristics of the abused (164 women, information on sociodemographic charac- 49.6%) and control (230 women, 58.4%) teristics such as age, education level, em- groups and their husbands. Comparing the ployment status (categorized as full time or 2 groups, the abused women were young- intermittently employed or unemployed) er, were of lower education status, had and annual household income. At the time been married a shorter time, and the major- of our study, the average family in Sudan ity were unemployed (P < 0.001 for these needed US$ 2500 per annum to cover the variables). Being less educated, the abused very basic requirements of living. We also group had a lower chance of being em- asked about the age, education level and ployed. The husbands in both groups were employment status of the husbands and older than their wives, but the age gap was whether they were using alcohol or illegal greater in the abused women. Compared to drugs. the husbands of the control group they Each woman was asked if her husband were less educated, had less chance of em- pushed her, hit her, kicked her or physically ployment, and were more likely to be alco- hurt her in some other way during the pre- hol or drug abusers (P < 0.001). vious year. The women were also asked if Table 2 shows the pattern of abuse they had received threats of physical injury. among the victims of domestic violence in We enquired about the controlling behav- our study. Of the 164 abused women, 112 iours adopted by their husbands (shouting (68.3%) had experienced 1 or more forms or screaming at her, restricting her social of controlling behaviour by their husbands, life, checking her movements, etc.). Ques- 119 (72.6%) had been threatened with tions on violence related to sexual issues physical injury and 79 (48.2%) had experi- were perceived to be too embarrassing for enced physical violence (of whom 26 had the first fifteen women (who refused to sustained injuries, including bruises, that give clear answers). These were then delet- needed medical attention). These women ed from the survey. reported 162 episodes of violence, of Each abused woman was asked about which 107 (66.1%) were mild, 47 (29%) the immediate provoking factors for vio- were moderately severe and 8 (4.9%) were

21 A study of domestic.pmd 167 10/25/2005, 1:24 AM 168 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Table 1 Sociodemographic characteristics of abused and control groups and their husbands

Characteristic Abused women Control group (n = 164) (n = 230)

Mean (SD) Mean (SD) Age (years) 29 (11) 36 (10) Duration of marriage (years) 6 (4) 9 (5) Husband’s age (years)a 34 (6) 38 (9) No. % No. % Education level 6–8 years 74 45.1 39 17 9–12 years 79 48.2 160 69.6 Graduate 11 6.7 31 13.4 Annual household income < US$ 2500 133 81.1 122 53.1 < US$ 2500 31 18.9 108 46.9 Employment status Employed full time 30 18.3 134 58.3 Intermittently employed 12 7.3 35 15.2 Unemployed 122 74.4 61 26.5 Husband’s education level 6–8 years 39 23.8 17 7.4 9–12 years 98 59.7 171 74.3 Graduate 27 16.5 42 18.3 Husband’s employment status Employed full time 70 42.7 172 74.8 Intermittently employed 31 18.9 37 16.1 Unemployed 63 38.4 21 9.1 Alcohol or drug abuse of husband 78 47.5 39 16.9

P < 0.001. aNot significant.

severe (resulting in burns and head and in- injury nor had their treating doctors asked ternal injuries). Our study group experi- them about the possibility of domestic vio- enced a total of 525 episodes of abuse lence. during the year prior to the visit. Forty-one In accordance with our operational def- women (25%) experienced 1 episode of vi- inition of domestic violence, we were only olence, 53 (32.3%) experienced 4–5 epi- investigating violent behaviours by the hus- sodes, 34 (20.7%) experienced 4–5 band. Even so, 42 of the abused group episodes and 34 (20.7%) experienced ≥ 6 (25.6%) reported other perpetrators (fa- episodes. Of the 26 women who had inju- ther, brother or other relatives). Also, more ries, only half had sought medical help. than a third of the abused group gave a his- They had not mentioned the real causes of tory of violent abuse before marriage. The

21 A study of domestic.pmd 168 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 169

Table 2 Patterns of abusive episodes among 7.3%), deciding to ask for divorce (10 164 women (some experienced more than women, 6.1%) and contacting the police (3 one form of abuse) women, 1.8%). Domestic violence during pregnancy Abusive behaviour Episodes was reported by 27 women (16.5%) but no Controlling behaviour 194 one stated that it had caused miscarriage. Shouting at her 49 Criticizing her in public 42 Restriction her social life 39 Discussion Checking her movements 24 Keeping her short of money 21 Our data provides compelling evidence of Other 19 the profound magnitude of domestic vio- Threatening behaviour 169 lence among the women who participated Throwing things 4 in the study. Our study group was highly Threatening with a fist 39 selective and not representative of the fe- Threatening the children 31 male population of Sudan, so the results of Threatening with a weapon 19 this study (49.6% of the participants re- Physical violence 162 ported abuse) should not be taken as an in- Shoving 41 dication of the prevalence rate of domestic Punching (body) 34 violence in Sudan as a whole. Nevertheless, Punching (face) 31 it is comparable with the results of interna- Kicking her on the floor 19 tional studies, where prevalence of domes- Forcing her to do something 8 tic violence varied between 40% and 60%, Trying to choke her 6 or even higher [14–16]. Burning 2 Even if all non-respondents in our study Use of a weapon 2 were women who had not experienced Other 12 abuse, then 1 in every 3 women attending our clinic would have experienced domestic violence. Moreover, it could be claimed that this figure is an underestimate due to immediate provoking events for violence sensitivity of the issue, meaning that some included suspicion of illicit relations, talk- women may hesitate to speak openly. ing back, not obeying the husband, not The surprisingly high response rate in having food prepared on time, refusing sex, our study, 86.8%, indicates that many failure to care for the home or children ad- women are willing to disclose such issues equately, going out of the home without when asked in a supportive fashion with permission and questioning the husband strict confidentiality assured [17]. This about his money or illicit relations. In about may give an indication that populations at 10% of the violent episodes, there were no risk might cooperate in future screening evident reasons. and intervention programmes if done in a The abused women reported several re- similarly supportive and sensitive way. actions to the violent behaviour including staying quiet (89 women, 54.3%), crying Patterns of abuse (32 women, 19.5%), resistance (18 wom- In contradiction to some European and en, 11%), telling a relative (12 women, American studies, firearms injuries and

21 A study of domestic.pmd 169 10/25/2005, 1:24 AM 170 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

stabbings were rarely found among our exclusively granted to the husband, and study group [14–16]. The assaulters in our abusers tend to refuse to divorce their vic- society seemed to prefer methods that tims. cause no severe physical injuries. The high Nevertheless, not all abused women re- contribution of verbal assaults such as main passive to violence. Some of them try shouting or yelling is probably because they to seek help (despite the fear of stigma) are culturally acceptable in our society. The from relatives or doctors. In a few cases of typical violent episode in our study involved domestic violence the police are called, es- a combination of assault, threats and verbal pecially for life-threatening injuries (when abuse. The frequency of violent acts is the police are required to be notified as a similar to that reported internationally prerequisite to medical care). Apart from [17,18], more than 70% of the women re- receiving physical treatment for their inju- porting violence in all cases. Repetition of ries, the victims received no other help. acts of assault is facilitated by the fact that victims are always readily available to the Risk factors abusers, and that the assaults are carried Our study clearly indicates that risk factors out in private. Victims may show little or no for the occurrence of domestic violence in- resistance so as to minimize the renewed clude young age, poverty, unemployment, aggression or injuries. poor education, pregnancy and alcohol and drug abuse by the perpetrators. Violence Reactions to abuse occurs in a context of a relatively wide gap The response of an abused woman to vio- in age and education level between the hus- lence is limited by the options available to band and wife. These risk factors are not her [19]. As elsewhere, the strongest rea- exclusive: any man can be a perpetrator and son that compels women to remain in an any woman can be a victim. In the pres- abusive relationship is probably lack of fi- ence of the risk factors, very minor or ab- nancial support [19]. The majority of these surd things such as talking back or not women are poorly educated and profes- having food ready on time are enough to sionally unskilled and therefore have no ac- trigger a violent episode. cess to jobs except marginal or illegal In contrast to the European or American employment. Consequently, many victims studies, we should emphasize the study of capitulate to the abusers’ demands and may the cultural factors and socially acceptable succumb to psychological and physical norms that cause or aggravate violence. problems (after an initial period of denial The Islamic religion, prevalent in Sudan, and self-blame) [20]. favours male supremacy and confers on In our society, divorce (or even being men the right to correct their “erring” unmarried) is socially unacceptable, espe- wives. Our social norms (through the per- cially when requested by the wife. Divor- ception that men have “ownership of wom- cees suffer social isolation or even further en”) predispose women to abuse by linking violence from other members of the family masculinity to male honour, and the accep- (as they are not allowed to live indepen- tance of violence or female chastisement to dently in a separate home). Moreover, the resolve conflicts. These norms are not pe- divorce is not easily obtained. According to culiar to Sudan, and may prevail in many Islamic law (sharia) the right to divorce is cultures worldwide [21,22].

21 A study of domestic.pmd 170 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 171

In accord with international studies, undeniable. However, some researchers most of the abused women were young (in suggest that employment itself does not their late teens and twenties) [18,23], but protect couples from violence [29]. Stress- the mean age for battered wives in our so- ful work experiences have been associated ciety was lower than in western societies, with wife abuse [28]. It has also been sug- and this may be because of the younger age gested that the increase in female employ- at marriage [14]. Older wives (e.g. those ment may generate tensions that increase over 45 years) were reported as being less the likelihood of marital conflict [29]. In a likely to be abused. If abuse occurred at conservative society like ours, in a situation this age, it may have even graver conse- where the wife is working and the husband quences. By this age, the woman’s parents is unemployed, the latter’s cultural role as may have died and families may have bro- family supporter is breached, and this may ken up; they have many children and have then generate family conflict. almost no chance of earning money. Low level of education is linked to un- Although domestic violence can occur employment and poor income [27]. Poor in all socioeconomic groups, many studies, education may be an indicator of poor including ours, found a higher prevalence communication skills, which have been in the poorer groups [24,25]. We found linked to domestic violence (especially in that 81.1% of abused women lived in fam- the context of a large gap in education sta- ilies whose annual income was < US$ 2500 tus in the couple) [30]. We found that compared to 53.1% of the control group. 23.8% of the abused women were married Poverty, and the resultant feelings of hope- to men who had ≤ 8 years of education lessness and crowding, increases relation- compared to 7.4% of the women in the ship conflicts and reduces the woman’s control group. power [26]. It also reduces the ability of A strong association has been found be- men to live in a manner that they regard as tween alcohol and drug abuse and the oc- successful [26]. Poverty frustrates men currence of domestic violence [27]. These owing to the loss of their cultural role as factors adversely affect income and em- money providers. The situation is aggravat- ployment status and thus aggravate marital ed when husbands, as a controlling behav- difficulties [23]. In the women we studied, iour, tend to keep their wives short of prevalence of alcohol or drug abuse in hus- money [18]. Poverty is the most common bands of abused women was almost 3 barrier to leaving an abusive relationship. times that in husbands of the control group. Unemployment (recent or long-term) Pregnancy is reported to increase the and the stress of looking for work increase risk of domestic violence and to alter the the risk a man will physically abuse his wife pattern of assault to be more severe [7,31]. [27,28]. In our study, almost 40% of the Up to 41% of antenatal attendees in Ameri- husbands of abused women were unem- can studies report a history of violence at ployed compared to just under 10% of the some point in their past pregnancies [31]. husbands of women in the control group. Just over 15% of the abused women we Unemployment is also consistently related studied reported violence during pregnan- to alcohol and drug abuse, leading to fur- cy. Pregnancy, because of the hormonal ther risk of domestic abuse [23]. The asso- and psychological changes which occur, ciation of unemployment with poverty is may trigger violent assaults resulting from

21 A study of domestic.pmd 171 10/25/2005, 1:24 AM 172 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

minor events such as refusal to have sex or Conclusions inadequate home care. A word remains to be said about the Domestic violence is a significant problem perpetrators. Compared to husbands in in our society. There is a need to implement non-violent relationships, perpetrators of a national multidisciplinary programme to abuse tend to be younger, unemployed and deal with this problem, with the coopera- more troubled [24]. These factors are not tion of health workers, nongovernmental the sole predictors of abuse. Any man may organizations, government departments of be an abuser and many perpetrators, apart health and social care, universities and oth- from abusing their wives, have normal rela- er concerned bodies. This programme tionships in their home or work environ- should aim at augmenting professional, ments. Therefore, universal screening for public and government awareness of the partner abuse should be encouraged. problem and providing medical services, social support, protection and legal assis- Limitations of the study tance to the victims of abuse. We should be We are aware of the limitations of our aware that unless evidence of benefit ex- study. Our clinic represents the maximum ists, many women might hesitate to coop- of diversity in terms of age, social class erate with programmes addressing and education level. Nevertheless, we re- domestic violence. ceive very few patients from rural areas, Relevant modifications in medical where women are placed at a lower social school curricula and doctors’ training pro- status, and some abusive behaviours are grammes are needed. As an interim mea- normal. sure, comprehensive guidelines for identi- There was also a possibility of selection fication and medical management of bias owing to the exclusion of illiterate abused women should be issued. women. The sensitivity of the issue and the We suggest that future studies on do- need for confidentiality compelled us to use mestic violence use a large sample size self-administered questionnaires to collect (preferably community-based) and apply the data. new methods to diminish under-reporting More than two thirds of our patients in (e.g. inclusion of rural and illiterate wom- the clinic suffer chronic disease (mainly en). Also, future studies should cover past hypertension and diabetes mellitus). Such violent experiences and adequate periods of women are particularly prone to abuse follow-up, and should assess the impact of [2,17]. The husband may, for example, personality and other psychological char- hide his wife’s insulin or other medication, acteristics on violence. Perpetrators should not help during hypoglycaemic episodes or also be studied as well as the less common refuse to arrange transport for follow-up possibility of abuse of husbands by their appointments [17]. wives.

References

1. Crowell NA, Burgess AW, eds. Under- 2. Ahmed AM et al. Impact of diabetes mel- standing violence against women. litus on Sudanese women. Practical dia- Washington DC, National Academy betes international, 2001, 18(4):115–8. Press, 1996:96–120.

21 A study of domestic.pmd 172 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 173

3. Schecter S. Interviewing battered 13. Gelles RJ. The violent home: a study of women: guidelines for the mental health physical aggression between husbands practitioner in domestic violence cases. and wives. Beverly Hills, California, Washington DC, National Coalition Sage Publications Inc., 1974. against Violence, 1987:25–65. 14. Bradley F et al. Reported frequency of 4. Stark E et al. Wife abuse in the medical domestic violence: cross-sectional sur- setting: an introduction for health per- vey of women attending general prac- sonnel. Washington DC, National Clear- tice. British medical journal, 2002, ing House on Domestic Violence, 324(7332):271–4. 1981:3–40 (Monograph Series No. 7). 15. Bacon LB et al. A survey of domestic vio- 5. Marais A et al. Domestic violence in pa- lence in a university emergency depart- tients visiting general practitioners— ment. Journal of the Arkansas Medical prevalence, phenomenology and Society, 2001, 98(6):180–2. association with psychopathology. South 16. Stanko E et al. Counting the costs: esti- African medical journal, 1999, 89(6): mating the impact of domestic violence 635–40. in the London Borough of Hackney. Lon- 6. McCauley J et al. The “battering syn- don, Crime concern, 1997:15–41. drome.” Annals of internal medicine, 17. Eisenstat SA, Bancroft L. Domestic vio- 1995, 123(10):737–46. lence. New England journal of medicine, 7. Richardson JO et al. Identifying domestic 1999, 341(12):886–92. violence: a cross-sectional study in pri- 18. Abbot J et al. Domestic violence against mary care. British medical journal, 2002, women. Incidence and prevalence in an 324(7332):274–7. emergency department population. 8. Healthy people 2000: national promo- Journal of the American Medical Asso- tion and disease prevention objectives. ciation, 1995, 273(22):1763–7. Washington DC, US Department of 19. Duttan MA. Battered women’s strategic Health and Human Services, Public response to violence. In: Edelson JA, Health Service, 1991:5–35. Eisikovits ZC, eds. Future interventions 9. Uniform crime reports: crime in the with battered women and their families. United States. Washington DC, Federal London, Sage Publications, 1996:105– Bureau of Investigation, 1994:3–12. 24. 10. Koss MP et al. No safe haven: male vio- 20. Landenburger KM. The dynamics of lence against women at home, at work leaving and recovering from an abusive and in the community. Washington DC, relationship. Journal of obstetric, gyne- American Psychological Association, cologic, and neonatal nursing, 1998, 1994:135–7. 27(6):700–6. 11. Campbell JC. Health consequences of 21. Heise LL. Violence against women: an intimate partner violence. Lancet, 2002, integrated ecological framework. Vio- 359(9314):1331–9. lence against women, 1998, 4(3):262– 90. 12. Carillo R. Battered dreams: violence against women as an obstacle to devel- 22. Orpinas P. Who is violent? Factors asso- opment. New York, United Nations De- ciated with aggressive behaviors in Latin velopment Fund for Women, 1992: America and Spain. Revista pan- 23–71.

21 A study of domestic.pmd 173 10/25/2005, 1:24 AM 174 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

americana de salud publica, 1999, 5(4– 27. Kyriacou DN et al. Risk factors for injury 5):232–44. to women from domestic violence against women. New England journal of 23. Rodriguez E et al. The relation of family medicine, 1999, 341(25):1892–8. violence, employment status, welfare benefit and alcohol drinking in the 28. Barling J, Rosenbaum A. Work stressors United States. Western journal of medi- and wife abuse. Journal of applied psy- cine, 2001, 174(5):317–23. chology, 1986, 71(2):346–8. 24. Hoffman LK et al. Physical wife abuse in 29. Kornblit AL. Domestic violence—an a non-western society: an integrated emerging health issue. Social science theoretical approach. Journal of mar- and medicine, 1994, 39(9):1181–8. riage and the family, 1994, 56:131–46. 30. Dutton DG, Strachan CE. Motivational 25. Straus MA, Gelles RJ. Societal changes needs for power and spouse-specific and change in family violence from 1975 assertiveness in assaultive and to 1985 as revealed by two national sur- nonassaultive men. Violence and vic- veys. Journal of marriage and the family, tims, 1987, 2(3):145–56. 1986, 48:465–79. 31. Mezey GC, Bewley S. Domestic violence 26. Jewkes R. Preventing domestic vio- and pregnancy. British medical journal, lence. British medical journal, 2002, 1997, 314(7090):1295. 324(7332):253–4.

Preventing violence. A guide to implementing the recommendations of the World Report on Violence and Health Each of the six parts of this guide corresponds to one of the six country-level recommendations of the World report on violence and health. Each part has provided a set of suggestions for activities in the areas of policy formulation, system development and programme implementations that, if acted upon, will assist countries in following these recommendations and thereby help to prevent interpersonal violence and improve care services for violence victims. This document, Preventing violence, has been prepared by WHO’s Department of Injuries and Violence Prevention, in consultation with violence prevention experts from around the world, and provides conceptual, policy and practical advice on how to implement each of the six country-level activities. This document is available free on line at http://whqlibdoc.who.int/publications/2004/ 9241592079.pdf

21 A study of domestic.pmd 174 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 175

Medical school’s experience with Arabization 1993–2003: Gezira University SUMMARY This paper examines the importance of teaching in the mother tongue and looks at the political decisions taken in this respect. The preparations for Arabization in the Faculty of Medicine, Gezira University (in 1993) are reviewed and the experience of implementation from 1994 to 2002 is analysed by questioning the students and professors. The successes, failures, problems and obstacles are discussed in detail. The paper concludes with discussion and recommendations on how to boost success in Arabization based on previous experience.

L’expérience d’une école de médecine en matière d’arabisation 1993-2003 : Université de Gezira RÉSUMÉ Cet article examine l’importance de l’enseignement dans la langue maternelle et passe en revue les décisions politiques prises à cet égard. Les préparations en vue de l’arabisation à la Faculté de Médecine de l’Université de Gezira (en 1993) font l’objet d’un examen et l’expérience relative à la mise en place de 1994 à 2002 est analysée à partir de questions posées aux étudiants et aux professeurs. Les succès, échecs, problèmes et obstacles rencontrés sont examinés en détail. L’article se termine par une discussion et des recommandations quant au moyen de favoriser le succès de l’arabisation sur la base de l’expérience accumulée.

1A.A. Mahmadani, D.E. Al Geely, O.T.M. Othman. Faculty of Medicine, University of Gezira, Wad Medani, Sudan (Correspondence to A.A. Mahmadani: [email protected]). Received: 22/08/04; accepted: 12/10/04

22 Medical schools experience with Arabization.pmd175 10/25/2005, 1:24 AM 176 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

22 Medical schools experience with Arabization.pmd176 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 177

22 Medical schools experience with Arabization.pmd177 10/25/2005, 1:24 AM 178 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

22 Medical schools experience with Arabization.pmd178 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 179

22 Medical schools experience with Arabization.pmd179 10/25/2005, 1:24 AM 180 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

22 Medical schools experience with Arabization.pmd180 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 181

Analyse La transition sanitaire en Tunisie au cours des 50 dernières années A. Ben Hamida,1 R. Fakhfakh,1 W. Miladi,1 B. Zouari1 et T. Nacef 1

RÉSUMÉ Cette étude décrit les principaux aspects des trois types de transition – démographique, socio- économique et sanitaire – observés en Tunisie au cours des 50 dernières années. Sur le plan dé- mographique, le taux brut de mortalité et le taux de mortalité infantile sont passés respectivement de 19 pour 1000 et de 150 pour 1000 en 1956 à 5,7 et 26,2 actuellement, l’espérance de vie à la naissance passant de 50 à 72 ans pour la même période. Sur le plan socio-économique, la population urbanisée est passée de 25 à 62 %, le taux d’alphabétisation des adultes de 15 à 73 % et le revenu par habitant a été multiplié par 5 en termes réels. Au plan épidémiologique, la morbidité prévalente dans les années 60 de type infectieuse et périnatale a régressé tandis que les maladies chroniques et dégénératives ont progressé. Au cours des 50 dernières années, la part des dépenses de santé dans le PNB est passée de 3,8 % à 6,2 %. Les implications futures de ces changements pour le système de santé tunisien et la nécessaire adaptation tant au niveau de la prise en charge thérapeutique que de la prévention des risques y sont discutées.

Health transition in Tunisia over the past 50 years ABSTRACT We describe the dramatic demographic, socioeconomic and health changes witnessed in Tunisia over the past 50 years. Demographically, the gross mortality rate and the infant mortality rate have gone from 19 per 1000 and 150 per 1000 respectively in 1956 to 5.7 per 1000 and 26.2 per 1000 now, and life expectancy at birth going from 50 to 72 years for the same period. Socioeconomically, the urban population has risen from 25% to 62%, the literacy rate from 15% to 73%, and the per capita income has increased 5-fold in real terms. Epidemiologically, the infectious and perinatal diseases prevailing in the 1960s have decreased whereas chronic and degenerative diseases have risen. The proportion of the GNP related to health expenditure has risen from 3.8% to 6.2%. The implication of these changes on the Tunisian health system and the need to adapt in terms of curative care and prevention of risks are discussed.

1Faculté de Médecine de Tunis, Tunis (Tunisie). Reçu : 24/06/02 ; accepté : 07/10/03

23 La transition sanitaire.pmd 181 10/25/2005, 1:24 AM 182 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction velles (recherche par exemple d’une ren- tabilité financière accrue de chaque em- En ce début du XXIe siècle et du troisième ployé, en laissant de côté toute considéra- millénaire, jamais les progrès accomplis par tion humaine d’équilibre ou de bien-être l’homme, en particulier dans le domaine personnel). Ceci va avoir une répercussion sanitaire et scientifique, n’ont été aussi por- sur les modes de vie. Cette « triple transi- teurs de défis et de paradoxes. Ainsi, la ré- tion », qui touche tous les pays du monde, duction parfois spectaculaire de l’incidence s’inscrit en fait dans un long processus et de la charge des maladies transmissibles d’ensemble, parti de l’Europe occidentale ainsi que des maladies périnatales dans de depuis la révolution industrielle au XVIIIe nombreuses régions du monde a permis siècle et qui s’est lentement propagé par la l’allongement de l’espérance de vie à la suite aux autres régions du monde, à des naissance, tant dans les pays industrialisés moments et à des rythmes différents, selon que dans les pays en développement ; cet leur degré d’urbanisation, d’industrialisa- état de fait a entraîné la survie d’un nombre tion et aussi de socialisation (collectivisa- croissant de personnes de plus en plus ex- tion des moyens de production) [3]. posées au risque de maladies chroniques En Tunisie, pays en développement non transmissibles – sans oublier la persis- « intermédiaire », plusieurs changements tance du risque infectieux – associées au se sont produits depuis l’indépendance du développement de nouvelles pathologies, pays en 1956 [4,5]. Cette étude se propose qualifiées de maladies de la « civilisation ». de décrire les principaux aspects de ces Qu’il s’agisse des « maladies de trois types de transition (socio- dégénérescence » (cardiopathies is- économique, démographique et sanitaire) chémiques, hypertension artérielle, diabète, observés dans le pays au cours des cancers, etc.) ou des « maladies de so- 50 dernières années, ainsi que leurs impli- ciété » (accidents, troubles mentaux, infec- cations futures, pour le système de santé, tions sexuellement transmissibles et SIDA, en particulier en termes d’organisation et de affections associées aux risques profes- coût. sionnels, additifs alimentaires, radiations, etc.), ces affections sont en fait plus chro- Méthodes niques, plus invalidantes et surtout plus coûteuses [1,2]. Les années de recensement général de la Ces changements du profil de la mor- population ont été choisies comme années bidité sont associés à des changements de référence, et à défaut, les années les plus d’ordre démographique et socio-écono- proches sont prises en compte. mique et sont qualifiés de « transition Les indicateurs ont été choisis en fonc- épidémiologique » ( ou « sanitaire »), tion de leur aptitude à refléter chacune des « transition démographique » et de « tran- transitions dans les 3 domaines de l’étude. sition socio-économique », cette dernière Ainsi, les principaux indicateurs utilisés étant notablement accélérée par ce que l’on pour caractériser l’évolution de la situation appelle actuellement la « mondialisation », démographique furent le taux de natalité et concept qui repose sur une vision écono- les taux de mortalité globale et spécifique miste du monde, considérant ainsi la sphère avec les corrections usuelles utilisées par économique au-dessus de la sphère sociale, l’Institut National de la Statistique (INS). et recouvre toute une série de réalités nou-

23 La transition sanitaire.pmd 182 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 183

L’étude de la transition socio- enregistrée avec un taux de raccordement à économique s’est basée essentiellement sur l’eau courante passant de 14,2 % en 1966 à l’évolution de quelques indicateurs, té- 68,3 % en 1994 et 75,2 % en 1999 (Tableau moins de la modification du mode de vie du 1). Tunisien depuis l’indépendance, tels que la Sur le plan économique, la part de consommation alimentaire, les conditions l’emploi agricole est passée de 74,7 % en de logement, le revenu par habitant, la sco- 1956 à 22,0 % en 1999, celle des services larisation, l’activité professionnelle, les de 7,1 % en 1956 à 43,9 % en 1997. La habitudes de vie (comme le tabagisme). proportion des femmes actives est passée Quant à l’étude de la transition épidé- de 6 % à 25 % depuis 1966. Le PNB par miologique en Tunisie, elle a porté en parti- habitant a augmenté, en dinars courants, de culier sur : TND 74 en 1961 à TND 1850 en 1995 et le • l’évolution des principales causes de taux de pauvreté a diminué de 33 % en mortalité par groupes de maladies ; 1966 à 4,4 % en 1994. • l’évolution de l’incidence de certaines De même, un changement au niveau des maladies transmissibles ; habitudes et du mode de vie est observé avec une évolution vers un modèle alimen- • l’évolution des principaux facteurs de taire caractérisé par une ration plus riche en risque cardio-vasculaire, en particulier lipides et en protéines d’origine animale ; le de la prévalence du diabète et de rapport « protéines végétales/protéines ani- l’hypertension artérielle (HTA), ainsi males » passe de 85,7/14,3 à 73,8/26,1 en- que l’évolution de l’incidence de cer- tre 1965 et 1996-97. tains cancers, des maladies respiratoires et de la santé mentale. Transition démographique À côté des diverses publications scienti- (Tableau 2) fiques (sources bibliographiques), les don- La population totale du pays a été multipliée nées ont été recueillies auprès de par 2,5 depuis l’indépendance, passant de différentes sources [6]: les résultats des re- 3,78 à 9,44 millions. La croissance naturel- censements publiés par l’Institut National le s’est nettement réduite, passant de 3,2 % de la Statistique, les documents et rapports en 1966 à 1,45 % en 1999. Le taux brut de publiés par le Ministère de la Santé, les rap- natalité en 1999 a été de 16,9 pour mille ports des enquêtes menées par différentes contre 50 pour mille en 1956. Au niveau de institutions, les publications par les orga- la structure par âge de la population, on nismes des Nations Unies (OMS , PNUD, note le recul du poids relatif des moins de UNICEF) et la Banque mondiale. 15 ans, qui passe de 46,5 % en 1966 à 34,8 % en 1994, en même temps qu’une tendance progressive au vieillissement avec Résultats une proportion des personnes âgées de plus Transition socio-économique de 60 ans de 5,1 % en 1956 contre 9,0 % Le pourcentage de la population urbanisée en 1999. est passé de 25 % en 1956 à 62,4 % en Le taux brut de mortalité (TBM) et le 1999 ; celui de l’alphabétisation des adul- taux de mortalité infantile (TMI) se sont ré- tes, de près de 15 % à 73 % depuis duits, passant respectivement de 18-20 à l’indépendance. Une amélioration progres- 5,7 pour mille et de 150-200 à 26,2 pour sive des conditions de vie a été également mille entre 1956 et 1999. Le taux de morta-

23 La transition sanitaire.pmd 183 10/25/2005, 1:24 AM 184 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Tableau 1 Évolution de quelques indicateurs socio-économiques en Tunisie, 1956-1999 [6]

Indicateurs 1956 1966 1975 1984 1994 1999

PNB par habitant (en TND) 70 74 (1961) 1120 1500 1850 (1995) Population active par secteur (%) Agriculture 74,7 45,9 39,0 28,1 21,9 22,0 Industrie 10,5 19,1 29,9 36,7 34,9 34,1 Tertiaire (services) 7,1 21,9 31,1 35,2 43,1 43,9 (1997) Urbanisation (en % de la population totale) ~ 25 40 47,5 52,8 61,0 62,4 Alphabétisation des adultes (%) 15,3 32,1 45,1 53,8 68,3 73,0 Accès à l’eau potable (%) - 14,2 26,4 49,4 68,3 75,2 Apport en protides - 85,7/14,3 85,5/14,5 80,8/19,2 79,2/20,8 73,8/26,1 alimentaires d’origine (1965) (1985) (1995) (1996-97) végétale/animale (%)

lité maternelle est passé de 134,2 à 68,9 sance est passée de 50 ans en 1956 à pour 100 000 naissances vivantes depuis 72,1 ans en 1999. 1966. Enfin, l’espérance de vie à la nais-

Tableau 2 Évolution de quelques indicateurs démographiques en Tunisie, 1956-1999 [6]

Indicateurs 1956 1966 1975 1984 1994 1999

Population totale (en millions) 3,78 4,58 5,61 7,03 8,78 9,44 Structure par âge (%) 0-4 ans 18,3 18,6 16,0 14,6 11,0 9,0 5-14 ans 24,2 27,9 27,8 25,1 23,8 21,9 15-59 ans 52,4 48 50,4 53,6 56,9 60,1 60 ans et plus 5,1 5,5 5,8 6,7 8,3 9,0 Taux de croissance naturelle 2 (1946- 3,2 2,67 2,58 1,8 1,45 par an (%) 1956) Taux brut de natalité (‰) ~ 50 44 36,6 32,4 22,7 16,9 Taux brut de mortalité (‰) 18 - 20 15 9,9 6,3 5,7 5,7 Taux brut de mortalité infantile 150 à ~ 200 140 88 51,6 30,6 26,2 (‰) (1976) (1985) (1995) Taux de mortalité maternelle pour 100 000 naissances vivantes 134,2 88,0 - 68,9 -

23 La transition sanitaire.pmd 184 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 185

Transition épidémiologique des changements complexes des profils de La proportion des décès infantiles et juvé- morbidité et de mortalité, observés durant niles (parmi l’ensemble des décès) est une large période de temps, parmi les popu- passée respectivement de 31 à 11,7 % et de lations [1-3]. Cette transition épidé- 21,7 à 2,8 % entre 1966 et 1999, alors que miologique se manifeste en particulier par celle des âgés de plus de 50 ans s’élève en une modification de la structure d’âge de la même temps de près de 33 % à 70 %. La mortalité (à partir d’âges jeunes vers des part de la mortalité attribuable aux maladies âges avancés) et un changement dans les cardio-vasculaires passe de 11 % (3e rang) principales causes de décès (et aussi de la à 23 % (1er rang) en moins de 30 ans ; celle morbidité), marqué par une augmentation des tumeurs s’élève de 3 % à 12 % au progressive des maladies chroniques et cours de la même période [4]. dégénératives (ou « de civilisation ») au En même temps, un changement du détriment des maladies transmissibles et de profil de morbidité est observé (Tableau 3), celles de la période périnatale. Ce concept marqué d’une part par le recul net voire de transition épidémiologique reconnaît en l’éradication de certaines maladies trans- fait, dans sa théorie originelle, certaines missibles. Ainsi, l’incidence de la tubercu- propositions de base, en particulier lose est passée de 34,8 pour 100 000 l’existence de plusieurs modèles de transi- habitants en 1970 à 23,6 pour 100 000 en tion (au moins trois) se distinguant surtout 1997. De même, aucun cas de poliomyélite par une dissemblance du développement n’a été enregistré depuis 1992. Nous assis- socio-économique entre les nations [3]. Le tons d’autre part à l’émergence des mala- modèle classique ou occidental décrit la dies non transmissibles de type chronique transition dans les pays occidentaux au et dégénératif. Ainsi, la prévalence de cours des avant-derniers siècles, d’une l’hypertension artérielle chez l’adulte (selon mortalité élevée et d’un taux élevé de nais- des études menées depuis 1980) varie entre sances (30-35 %) à un taux bas (inférieur à 15 % et 21 %, celle du diabète serait passée 20 pour mille)[3]. La transition dans ces de 3,5 % à 10 % en l’espace de 20 ans. pays a été strictement en rapport avec la Enfin, on note une prépondérance chez révolution industrielle et sociale (exemple l’homme des cancers liés au tabagisme, France et Royaume-Uni). À coté de ce avec le cancer du poumon largement en modèle, Omran décrit le modèle accéléré tête, dont l’incidence (pour 100 000 habi- qui caractérise la transition au Japon, en tants) est évaluée en 1994 et 1999 à 27,9/ Europe de l’Est et en Union soviétique. 17,5 (Nord/Sud). Le cancer du sein chez la Dans ce modèle, la transition de mortalité femme est de loin le plus fréquent, s’est déroulée pendant une courte période, représentant en effet plus de 20 % de tous par opposition au modèle classique. Le les cancers féminins, son incidence (pour modèle retardé (ou «contemporain ») sem- 100 000) variant entre 24,3 au nord et 12,7 ble décrire la transition du XXe siècle, in- au sud du pays. complète, de la majorité des pays en développement. Pour le modèle retardé, bien que les progrès atteints en survie in- Discussion fantile aient été considérables, les taux de La transition épidémiologique décrite par mortalité sont encore relativement élevés. Omran au début des années 70 se réfère à La baisse de mortalité n’a commencé

23 La transition sanitaire.pmd 185 10/25/2005, 1:24 AM 186 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Tableau 3 Évolution de quelques indicateurs épidémiologiques en Tunisie,1956-1997 [6]

Indicateurs 1956 1968 1975 1984 1995 1997

Tuberculose Risque annuel d’infection (%) 3,02 - - 0,5 - - Incidencea - 34,8 48,6 35,9 27,06 23,60 (1970) Fièvre typhoïdea - 20,5 11,7 13,8 3,78 1,32 Paludisme - 5397 cas - 0 49 casd 23 casd (0,25) Poliomyélite antérieure 525 0,7 1,9 (1971) 23 (1987) 0 0 aiguë a (1959) Rougeolea - - 3,5 (1980) 35,6 6,8 4,0 Leishmaniose cutanéea 0,2e 0,0 (1980) 21,5 11,9 34,7 Hépatites viralesa - 13,4 25,2 (1970) 50,8 - 64,30 Infection à VIHa - - - 0,7 (1986) 0,9 0,9 Rhumatisme articulaire aigua - 18 (1967) - 8,7 (1985) - 1,95 (1999) Hypertension artérielleb 9,9 13,56 18,8 22,6 (en milieu urbain) (≥ 3 ans) (≥ 3 ans) (≥ 20 ans) (35-65 ans) (1980) Diabèteb (en milieu 3,5 1,7 10,0 12,8 urbain) (≥ 20 ans) (tous âges) (≥ 20 ans) (35-65 ans) (1986) Bronchite chroniqueb 10 - 3,8 (20-60 ans) (≥ 25 ans) Cancersc -- - 12 Poumon (H) 27,9 (N) 17,56 (S) (1994) Sein (F) 24,3 (N) 12,73 (S) (1994)

aTaux d’incidence déclarée (cas pour 100 000 habitants). bPrévalence globale (en %). cIncidence standardisée (pour 100 000) selon le Registre du cancer du Nord tunisien (1994) et le Registre du cancer du Sud tunisien (1997). dCas non autochtones. eCas déclarés de leishmaniose viscérale (1968).

qu’entre la troisième et la cinquième décen- dans de très courtes périodes, par rapport nie du XXe siècle. Ce déclin s’est manifesté au modèle occidental. Le modèle retardé est de manière spectaculaire après la fin de la caractérisé en outre par l’existence d’un Deuxième Guerre mondiale ; il s’est opéré chevauchement des phases de transition,

23 La transition sanitaire.pmd 186 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 187

marqué par l’émergence des maladies non entre 15 % et 21 % et est proche de cer- transmissibles et la persistance des mala- tains pays occidentaux tels que le Cana- dies transmissibles. da (14 %). Notons aussi une prépondé- À travers notre étude basée sur rance chez l’homme des cancers liés au ta- l’analyse d’indicateurs socio-économiques, bagisme, avec le cancer du poumon large- démographiques et sanitaires, approche ment en tête, son incidence restant comportant des limites liées au manque de toutefois largement inférieure à celle de données valides dans notre contexte de certains pays industrialisés, tels que la pays en développement, la Tunisie serait France ou le Canada (respectivement 55,1 dans la situation du modèle de transition et 82,5 pour 100 000 habitants) [9]. De épidémiologique « retardée » du fait de la même, l’incidence du cancer du sein chez coexistence de maladies infectieuses et la femme, représentant plus de 20 % de périnatales, quoiqu’en recul, avec tous les cancers féminins, reste bien loin de l’émergence des maladies chroniques et celle des pays occidentaux, comme la dégénératives. France ou encore le Canada (respective- En effet, dans ce pays, la proportion des ment 68,2 et 71,7 pour 100 000) [9]. Par décès infantiles et juvéniles (parmi ailleurs, les taux de prévalence observés en l’ensemble des décès) a reculé de façon matière de troubles mentaux (dépressifs considérable au profit de celle des âgés de notamment) sont comparables à ceux rap- plus de 50 ans. Parallèlement, la part de la portés dans la littérature internationale, mortalité attribuable aux maladies cardio- avec une prévalence globale de la dépres- vasculaires passe du 3e rang au 1er rang en sion majeure sur la vie entière de 8,2 % et moins de 30 ans et celle des tumeurs de la schizophrénie de 0,57 % (Hachemi). s’élève de 3 % à 12 % au cours de la même Cette transition épidémiologique serait période, en raison de la forte réduction de la due en partie aux deux autres types de tran- morbidité périnatale et infantile [4]. sition, à savoir la transition socio- En même temps, un changement du économique et la transition démographique. profil de morbidité est observé, marqué La transition socio-économique, dont le d’une part par le recul des maladies trans- fait le plus marquant a été le passage d’une missibles « traditionnelles », considérées de économie rurale de subsistance (agriculture véritables fléaux autrefois (paludisme, bil- et artisanat) à une économie urbaine « de harziose, trachome, tuberculose, diarrhées marché », une économie industrielle et la infectieuses, etc.), et de celles de la petite transition ainsi d’une vie rurale à une vie enfance (poliomyélite, tétanos néonatal, urbaine, fut notablement accélérée par le diphtérie, etc.) grâce surtout aux pro- phénomène actuel de la « mondialisation » grammes de lutte de masse, associés à (qui date en fait de plusieurs siècles), mar- l’évolution socio-économique du pays, et quant le début d’une nouvelle ère où tous d’autre part par l’émergence des maladies les pays sont ouverts les uns aux autres. Ce non transmissibles de type chronique et phénomène semble s’étendre en fait pour dégénératif, d’étiologie multifactorielle et désigner un concept recouvrant toute une au coût de prise en charge nettement plus série de réalités nouvelles appelées à trans- élevé [7,8]. former profondément les modes de vie à Ainsi, l’étude des facteurs de risque travers un renforcement des interdépen- cardio-vasculaire montre un profil compa- dances entre les nations. Elle implique, dans rable à celui des pays occidentaux. La prévalence de l’hypertension artérielle varie ce sens, la diffusion surtout du modèle de la

23 La transition sanitaire.pmd 187 10/25/2005, 1:24 AM 188 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

modernisation avec ses exigences d’indus- peut comporter pour le financement des trialisation, de marchandisation et d’urbani- dépenses sociales dont celles de la santé. sation [3]. La transition démographique désigne en À ce titre, le mode de vie des Tunisiens fait le passage, à l’occasion d’un processus a beaucoup changé depuis le début du siè- de « modernisation » globale et au bout cle et notamment durant les 40-50 der- d’un certain délai, d’un régime traditionnel nières années [10-18] : alors qu’en 1921, « d’équilibre » marqué par une mortalité et seul 1 Tunisien sur 4 vivait en zone urbaine, une fécondité fortes, à un régime cette proportion est passée à 1 sur 3 en « d’équilibre » à mortalité et à fécondité 1956 et serait de 2 sur 3 en 2015. En 1956, basses [3]. Ce concept reconnaît également l’alphabétisation ne concernait que 20 % dans sa version originale l’influence, en des adultes alors qu’elle est actuellement de particulier, de l’entrée dans la croissance 65 % et la quasi-totalité de la population économique moderne sur le déclenchement saura lire et écrire dans les 10 à de la baisse séculaire de la fécondité. Les 20 prochaines années. De même, un démographes tunisiens (Seklani et coll) ont changement au niveau des habitudes et du décrit le phénomène en Tunisie depuis les mode de vie est observé avec une tendance années 60-70. En effet, à l’image des autres en particulier à la sédentarité, l’obésité et pays en développement, la Tunisie a connu une évolution vers un modèle alimentaire dès le début des années 1950-60 un taux caractéristique des peuples dits « civi- d’accroissement démographique important lisés », avec une ration plus riche en lipides voisin de 3 % (2,7 %), plaçant le pays dans et en protéines d’origine animale. De plus, une période transitionnelle sur le plan dé- urbanisation et élévation du revenu favo- mographique et aboutissant à un temps de risent la propagation de l’habitude ta- doublement de la population de 25 ans envi- bagique, avec une prévalence globale du ron. En effet, tandis que la mortalité a com- tabagisme variant autour de 30 % et ex- mencé à baisser dès le milieu du siècle, cédant même 50 % chez l’homme [5]. passant de 27 pour 1000 en 1920 à 16 pour Sur le plan économique, on assiste au 1000 en 1966, 13 pour 1000 en 1975 et 6 recul progressif du secteur agricole et mi- pour 1000 en 1994, la natalité est restée nier en faveur des secteurs industriels et quant à elle élevée jusqu’en 1966 (47 pour surtout tertiaire [19], avec une participation 1000) pour ne commencer à baisser qu’en de plus en plus importante de la femme 1975 (35,8) et descendant sous le seuil de dans le marché de l’emploi [15]. En même 30 en 1994 avec un indice synthétique de temps, on note une amélioration du niveau fécondité (ISF) de 2,4 – soit un niveau de vie du Tunisien, avec un revenu moyen proche de ceux des pays industrialisés par habitant multiplié par 5 en termes réels [20]. Ces résultats en matière de fécondité [15,16] et le taux de pauvreté officiel n’est ont pu être obtenus grâce à la conjugaison plus que de 4,4 % à la fin des années 90 de plusieurs facteurs, dont la politique de contre 33 % en 1966 [18]. Enfin, et dans le planification familiale et l’amélioration du cadre de l’ouverture du pays au marché niveau économique et social de la popula- mondial (mondialisation), un plan tion en général. Il faut ajouter à cela le recul d’ajustement structurel (PAS) a été adopté de l’analphabétisme, l’accès de plus en plus en 1986, visant surtout à libéraliser grand des femmes au marché de l’emploi et l’économie du pays et à réduire les dépen- l’élévation de l’âge moyen du premier ma- ses publiques [19], avec le risque que cela riage [15].

23 La transition sanitaire.pmd 188 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 189

Ce décalage entre la baisse de la morta- du secteur public qu’au niveau du sec- lité et celle de la fécondité situe la Tunisie teur privé. En effet, les implications dans la deuxième phase de la période de la pour une politique future de santé, y transition démographique. Cette dernière compris le choix d’un modèle de ré- sera probablement achevée dans les forme éventuelle du système de soins, prochaines décennies. Un nouvel état notamment au niveau du mode de fi- d’équilibre sera alors établi entre la natalité nancement, devrait tenir compte des et la mortalité avec une croissance faible de tendances futures résultant de la triple la population. L’une des principales con- transition en question. En particulier, séquences de cette baisse de la fécondité l’élévation projetée en matière sera le vieillissement de la population, fai- d’urbanisation, d’éducation, etc. serait sant de sorte que le pourcentage de la po- à l’origine d’un accroissement de la de- pulation âgée de 60 ans, actuellement plus mande de la population en services de faible que celle des pays industrialisés (près santé et surtout en meilleure qualité de de 20 %), mais plus importante que celle ces services ; de même, d’autres ten- des pays en développement (près de 7,5 %) dances socio-démographiques, notam- [21] serait de 15 % vers l’an 2030 [14]. ment l’allongement de l’espérance de Dans le contexte de la triple transition vie à la naissance, associé à une meilleu- précitée, le système de soins a évolué en re prise en charge des personnes âgées, effet en trois étapes : la prévalence relativement élevée de cer- • la première étape, celle des deux tains facteurs de risque, tels que le ta- premières décennies de l’indépendance bagisme en particulier, vont très (années 60-70), a été celle de la forma- probablement marquer le profil futur tion du personnel de santé et du déve- des problèmes de santé. Ce dernier se- loppement de l’infrastructure dans le rait en effet dominé, à côté d’une mor- secteur public, tant au niveau hospitalier bidité infectieuse persistante, par les (universitaire notamment), qu’au niveau maladies de « civilisation » de type de la première ligne (campagnes de chronique et dégénératif (cardiopathies masse et activités préventives) [22]. ischémiques, diabète, hypertension • La seconde étape (années 80-90) fut artérielle, cancers, maladies rhumatis- celle de l’accroissement de la couver- males, respiratoires chroniques, al- ture sanitaire et des dépenses globales lergiques, troubles mentaux, séquelles de santé (passant de 3,8 % à 6,2 % du d’accidents, etc.) dont la complexité et PIB depuis 1980). Cette évolution s’est le coût de prise en charge préventive et accompagnée d’une restriction du bud- thérapeutique seront manifestement get de l’État alloué à la santé publique et plus élevés que ceux de la période ac- d’un développement important, voire tuelle [22,23]. anarchique, d’un secteur privé de soins et d’une industrie pharmaceutique dont les prix sont peu contrôlés [22]. Conclusion • La troisième étape, celle des deux Au total, la Tunisie passe par une période de prochaines décennies (2000-2020), transition sanitaire globale depuis devrait être celle de l’amélioration de la l’indépendance, marquée sur le plan épidé- qualité des services et de la maîtrise des miologique par le passage d’une morbidité dépenses de santé et ce, tant au niveau

23 La transition sanitaire.pmd 189 10/25/2005, 1:24 AM 190 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

infectieuse, dont la prise en charge a un bon ci se révèle en particulier par l’accroisse- rapport coût-efficacité, à une morbidité ment au niveau de l’offre des services de chronique et dégénérative, dont le coût de santé, de la demande en soins de santé et prise en charge est nettement plus élevé. surtout au niveau des dépenses globales de Cette transition épidémiologique impose santé, le tout s’inscrivant dans un contexte une nécessaire adaptation du système de de restriction du budget pour la santé pu- soins, lui-même en pleine transition ; celle- blique.

Références 1. Bobadilla JL et al. The epidemiologic Washington, Banque mondiale, 1993 : transition and health priorities. In: 500. Jamison DT et al., eds. Disease control: 9. Hsaïri M et al. Estimation à l’échelle priorities in developing countries. New nationale de l’incidence des cancers en York, Oxford Medical Publications, 1993: Tunisie. Tunisie Médicale, 2002, 80:54– 51–63. 64. 2. Meslé F, Vallin J. Transition Sanitaire: 10. Ennaceur M. La politique sociale de la tendances et perspectives. Revue Tunisie depuis l’indépendance et sa Médecine/Sciences, 2000, 16:1161–71. place dans le développement. Revue 3. Chesnais JC. La transition démogra- « Travail et Développement », 1987, 10: phique : étapes, formes, implications 153–93. économiques. Études de séries 11. Zerab D. La Tunisie. Les clés de la temporelles (1720-1984) relatives à 67 réussite. Jeune Afrique, 1995, 1791:27– pays. Population, 1986, 6:1223–8. 30. 4. Njah A et al. Changements socio- 12. Recensement général de la population démographiques de la population et des logements 1975. Tunis, Institut tunisienne et transition épidémiolo- National de la Statistique, 1976. gique. Maghreb Médical, 1992, 253:26– 9. 13. Recensement général de la population et de l’habitat 1994. Tunis, Institut Na- 5. Fakhfakh R et al. Profil de santé de la tional de la Statistique, 1995. Tunisie : État actuel et tendances. Tunisie Médicale, 2002, 1:12–17. 14. Projections au niveau national. Vol. I. Tunis, Institut National de la Statistique, 6. Miladi W. Transition épidémiologique et 1996 évolution du système de santé en Tunisie [Thèse Med]. Tunis, Faculté de 15. Résultats de l’enquête nationale sur Médecine de Tunis, 2002. l’emploi de 1997. Tunis, Institut National de la Statistique, 1999. 7. Mosley WH, Bobadilla JL, Jamison T. The health transition: Implications for health 16. Premiers résultats de l’enquête policy in developing countries. (Conclu- nationale sur les dépenses des sion). In: Jamison DT et al., eds. Disease ménages 2000. Cinq premiers mois de control : priorities in developing coun- l’enquête. Tunis, Institut National de la tries. New York, Oxford Medical Publica- Statistique, 2001 [en arabe]. tions, 1993:673–99. 17. Enquête nationale sur le budget, la 8. Rapport sur le développement dans le consommation et le niveau de vie des monde 1993 : investir dans la santé. ménages 1995. Résultats de l’enquête

23 La transition sanitaire.pmd 190 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 191

alimentaire et nutritionnelle. Tunis, 21. PNUD. Rapport mondial sur le Institut National de la Statistique, 1997, développement humain 1999. Genève, Vol B. Deboeck Université, 1999. 18. Annuaire statistique de la Tunisie 1999. 22. Études stratégiques : perspectives Tunis, Institut National de la Statistique, d’évolution du secteur de la santé 1999, Vol. 42. (1996-2010). Rapport final. Tunis, Ministère de la Santé Publique, 1996: 19. Nabli MK et al.. Trade, finance and com- 64. petitiveness in Tunisia. In: Fanelli JM, Rohinton M, eds. Finance and competi- 23. Rapport sur la santé dans le monde tiveness in developing countries. Lon- 2000. Pour un système plus performant. don, Routledge, Taylor and Francis Genève, Organisation mondiale de la Group, 2002:222–56. Santé, 2000. 20. Omran AR. The epidemiologic transition theory. A preliminary update. Journal of tropical pediatrics, 1983, 29:305–16.

Statistiques sanitaires mondiales 2005 L’Organisation mondiale de la Santé (OMS) collecte et fait la synthèse d’un grand nombre de données quantitatives dans divers domaines de la santé par le biais de ses bureaux dans les pays, de ses bureaux régionaux et de ses départements au Siège. Elle utilise ensuite ces données pour les estimations, les plaidoyers, l’élaboration des politiques et l’évaluation. Elle les diffuse aussi largement dans le cadre de ses publications officielles ou par des mécanismes plus informels, sous forme électronique ou par des documents imprimés. Dans la présente publication, nous allons nous intéresser à un ensemble de base d’indicateurs sanitaires, retenus en fonction de leur disponibilité actuelle et de la qualité des données. On y trouve également la majorité des indicateurs sanitaires retenus pour suivre les progrès vers la réalisation des objectifs du Millénaire pour le développement (OMD). Ce groupe n’ est pas conçu pour saisir tous les aspects significatifs dans le domaine de la santé ; il tend plutôt à donner un cliché instantané de la situation sanitaire actuelle dans les pays. Enfin, il est important de souligner que cet ensemble n’est pas définitivement fixé : certains indicateurs seront ajoutés au fil des ans ou prendront davantage d’importance tandis que d’autres perdront de l’intérêt. Les indicateurs rassemblés dans les Statistiques sanitaires mondiales reprennent avant tout les estimations les plus récentes pour chaque pays (depuis 1995). Pour des information sur cette publication, consulter le site Internet des Éditions OMS à l’adresse suivante : http://www.who.int/bookorders/anglais/detart1.jsp?sesslan=1&codlan=2&codcol=15&codcch=636

23 La transition sanitaire.pmd 191 10/25/2005, 1:24 AM 192 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Review Health management information system: a tool to gauge patient satisfaction and quality of care B.T. Shaikh1 and F. Rabbani 1

ABSTRACT The health management information system (HMIS) is an instrument which could be used to improve patient satisfaction with health services by tracking certain dimensions of service quality. Quality can be checked by comparing perceptions of services delivered with the expected standards. The objective of the HMIS would be to record information on health events and check the quality of services at different levels of health care. The importance of patient assessment is a part of the concept of giving importance to patient’s views in improving the quality of health services. Expected benefits include enhancing patient satisfaction through improved communication; greater provider sensitivity towards patients; enhanced community awareness about the quality of services; and overall better use of services in the health system.

Le système d’information pour la gestion sanitaire : un instrument potentiel pour mesurer la satisfaction des patients et la qualité des soins RÉSUMÉ Le système d’information pour la gestion sanitaire est un instrument qui pourrait être utilisé pour améliorer la satisfaction des patients vis-à-vis des services de santé en suivant certaines dimensions de la qualité des services. On peut vérifier la qualité en comparant la perception des services fournis avec les niveaux attendus. L’objectif du système d’information pour la gestion sanitaire serait d’enregistrer des don- nées sur certains événements de santé et de contrôler la qualité des services à différents niveaux de soins de santé. L’importance de l’évaluation par les patients s’inscrit dans un concept de prise en compte de l’opinion du patient pour l’amélioration de la qualité des services de santé. Les avantages escomptés comprennent une plus grande satisfaction des patients grâce à une meilleure communication, une sensibilité plus importante des prestataires à l’égard des patients, une sensibilisation communautaire accrue concernant la qualité des services et une meilleure utilisation des services dans le système de santé d’une manière générale.

1Health Systems Division, Department of Community Health Sciences, Aga Khan University, Karachi, Pakistan (Correspondence to B.T. Shaikh: babar. [email protected]). Received: 04/11/03; accepted: 08/03/04

24 Health management information.pmd192 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 193

Introduction judgements as valued in themselves rather than as surrogate measures of other dimen- Background sions of quality [8]. Although arguments Assuring and promoting quality in health have been advanced that patients lack the care services continues to be a priority for expertise to evaluate medical care, several any health care system. Besides evaluation theoretical, practical and empirical argu- of health status through morbidity and ments provide a strong rationale for assess- mortality estimates, there has been equal ing patient perceptions [9]. The assessment emphasis on quality of care indicators in of patient satisfaction has various advan- health systems research. The rising de- tages. It is more sensitive in gauging vari- mand for assessment of quality in the ous quality indicators across health care health care system can be attributed to ris- delivery systems, has been found to be less ing costs, constrained resources and evi- expensive [10] and is more reliable than dence of variations in clinical practice [1]. other methods such as physician peer re- The World Health Organization has also view [11–13]. The importance of patient emphasized the importance of quality in the assessment in the health services is a part delivery of health care, defined by the crite- of the concept of giving importance to pa- ria of effectiveness, cost and social accept- tient views in improving the quality of ability [2]. A quality health system is a health services, i.e. “democratizing” and client-centred, integrated, responsive and “consumer sovereignty” [14,15]. In- cost-effective system that includes the creased interest in patient satisfaction sur- continuum of care from health promotion veys is, therefore, evidence of the shift and prevention. Such a system is based on from traditional doctor–patient relation- a regular assessment of people’s needs and ships to the provider–client attitude [16– wants, and monitors itself based on a phi- 18]. Moreover, such surveys are a form of losophy of continuous quality improve- health care evaluation independent of exter- ment. Assessment of quality usually nal pressures and bias [19–21]. focuses on technical concerns: the “struc- tural quality,” which is the availability of a Rationale structure to provide health care [3] and the A health management information system “process quality” through which care is de- (HMIS) is an essential tool for strengthen- livered [4,5]. This assessment of quality is ing planning and management in the health based on the application of professional facilities. Any conventional HMIS enables standards integrating patients’ views and monitoring of service delivery in terms of their experiences [6]. The assessment of access, coverage, expenditure, human re- patient satisfaction may be represented as a sources, disease profiles and health out- comparative balance between their percep- comes. The use of an HMIS in support of tions of the service delivered and the ex- health systems performance assessment pected standard they had set for the service and to address deficiencies and gaps in the [7]. services has often been recommended by The definition of quality may also in- researchers [22]. The question is whether clude a broader approach to satisfaction use of a traditional HMIS to track quality of and a stronger role for patient judgements. care dimensions would bring about im- This broader approach affirms patient provement in patient satisfaction of health

24 Health management information.pmd193 10/25/2005, 1:24 AM 194 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

facilities specifically and the quality of care Gauging quality of care with a provided in general. health management information This article is based on a literature re- system view, not only from local journals but also One approach to quality in service provi- from some regional and international ones. sion used extensively in the marketing and A few publications of the World Health Or- retailing world has been found interesting, ganization and the World Bank on quality of and is compatible with determining quality care and HMIS were also consulted. To re- of care in health service provision too. view the best practices, articles from less- Adapted from the SERVQUAL tool [27], developed and developing countries were some important dimensions for service also included. Official policy documents of quality assessment are: reliability, respon- the Government of Pakistan were also tak- siveness, assurance and empathy (Table 1). en into consideration to quote certain de- The aim is to improve the level of client sat- tails. isfaction with the service by tracking these Our aim is to ascertain the utility of an dimensions. HMIS and its use as a tool to monitor qual- Each of the 15 items listed for the 5 ity of care along with its customary usage. quality of care dimensions carries an equal The objective of the HMIS would then be weight and could be categorized as satis- not only to record information on health fied, somewhat satisfied or dissatisfied. As- events, but also to check the quality of the signing scores of 2, 1 and 0 respectively, services at different levels of health care. based on patient responses, provides a method of measuring the level of quality. The maximum possible score would be 30 Application of a health for a patient rating all 15 items as satisfac- management information tory. A cumulative score of 24 or above system (80% or above) could be considered excel- lent satisfaction, 18–23 (60–79%) satisfac- Dimensions of quality of service in tion and below 18 (< 60%) poor the health care sector satisfaction with quality of services. It has been observed that people’s own ex- Through a periodic use of this instrument perience, since it influences their satisfac- incorporated in the HMIS, an analysis will tion with the care they receive, improves help to show the variation in the satisfac- quality of service and in turn its utilization tion level and allow the management to plan [23]. Various approaches have been out- accordingly. This analysis is simple and lined for measuring quality in health care easy, and can be done by facility staff to such as availability, affordability, accom- monitor the quality of their services on a modation acceptability and physical acces- regular basis. sibility [24]. Two studies, focused mainly on family planning services, show other el- Expected benefits of adapting the ements of quality of care such as right of health management information choice, right to information, technical system competence of the provider, interpersonal Benefits for the community communication, follow-up services and Enhanced community awareness will trig- appropriate constellation of services in the ger an increase in the quality of health care, same outlet [25,26].

24 Health management information.pmd194 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 195

Table 1 Dimensions of quality of service

Dimension Element

Reliability Seen according to expectations Solved the problem Given/received the required treatment Responsiveness Hospital staff gives prompt attention Hospital staff do not keep one waiting Hospital staff help according to the need Assurance Hospital is trustworthy Doctor is qualified Hospital staff are courteous Empathy Hospital staff are caring Give/get individual attention Doctor calls person by name Tangible Doctor’s office is clean Hospital staff use standard instruments Doctor’s prescription is easy to understand

resulting in increased patient satisfaction. at health service outlets based on the feed- This will in turn promote the appropriate back of their clients. use of the health services, which will be catering to the precise needs and aspira- Benefits for policy-makers/local tions of the community. government Such an HMIS could provide a model for Benefits for health providers decision-making, where the effective use One survey in Burkina Faso found that pa- of quality indicators is often limited [28]. tients felt that services were geared more The attention of the local government au- towards their needs after providers had re- thorities could be drawn to improving the ceived training on patient-friendly services working and living conditions of health per- [26]. Training vis-à-vis gender sensitiza- sonnel and to designing capacity-building tion, interpersonal communication skills strategies for them in order to create a and confidence-building will enhance the sense of motivation and spirit in serving the provider’s performance. Various tested and poor population. This will eventually com- verified training curricula, designed to plement the health system reforms already strengthen the skills of health workers, in place in Pakistan to refurbish the system could be used to help providers deal with in its entirety with the benchmarks of equi- their own beliefs and biases. The providers ty, accountability and good governance. could benefit from the information generated by an HMIS. It could help them re-orient Benefits for health system development their service by adopting a more empathetic Given the complex nature of the health care client-centred approach, adjusting their at- delivery system in Pakistan, it is essential titude and introducing a convivial ambiance for the various sectors to plan and work

24 Health management information.pmd195 10/25/2005, 1:24 AM 196 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

together to improve the health status of the perspective [32,33]. The dissatisfaction population. Data derived through such an with primary care services leads many peo- information system will help towards un- ple to go “health care shopping” or to jump derstanding the perceptions of the popula- to higher level hospitals for primary care, tion and the factors influencing health leading to considerable inefficiency and services utilization, and will eventually car- loss of control over efficacy and quality of ry strong prospects for the development of services. Therefore, there is sufficient data a health system catering to the needs of the to demonstrate the need to assess patient local population. Hence, improved cover- satisfaction with these public and private age and quality of the services should be a health facilities to bring about an overall desired outcome. improvement in quality of services delivered. The national health policy document The health management of 2001 emphasized good governance to information system in Pakistan achieve quality health care [34], however, Given the large health infrastructure in Pa- an authentic system of information has al- kistan, both public and private, catering to ways been flawed throughout the health a population of more than 140 million peo- care system, and there is no mechanism in ple, there had long been a need to develop operation to check the quality of services. and establish a national HMIS which would The majority of facility-based HMIS sys- be able to collect, process, analyse and pro- tems currently only track type and quantity vide feedback on all health-related data, in- of services and relate these to improvement cluding information on input, process and in health status indicators. output indicators. In Pakistan, this was developed in the early 1990s by the HMIS cell Discussion in the Ministry of Health. This system has now been implemented in a phased manner An exercise to track the quality of care de- and more than 90% of primary health care livered at the facility level ought to sensitize facilities are using it [29]. management and health providers to the as- pirations of their patients. Any improve- Current situation in Pakistan ment in patient satisfaction and quality of Despite a sound health infrastructure com- services could be then demonstrated prising basic health units, rural health cen- through the periodic use of an HMIS. Al- tres, maternal and child health centres, though it seems very much like any com- maternity homes, dispensaries, etc. [22,30] mercial marketing strategy, through regular in the public sector, the deficiency in quali- tracking of these aspects of quality of care, ty has resulted in a general lack of confi- health providers and hospitals would re- dence, with most people relying on private ceive timely feedback on areas of deficien- health care providers [31]. In private hospi- cy and this will stimulate mechanisms to tals and outlets, the quality of services has improve patient satisfaction. The mecha- often been compromised, and unfortunate- nisms developed using an HMIS can be ly, has seldom been assessed. The respon- easily adopted in resource-constrained siveness of the provider in such for-profit public and private health care settings, par- facilities has been questionable too. Only a ticularly in developing countries. few studies have been identified in the re- Providing good quality health care is of gion which have looked into evaluation of critical importance for the future economic health care services from the consumer’s

24 Health management information.pmd196 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 197

and social well-being of our country. Qual- [4]. This gap needs to be bridged. This ac- ity, however, has many dimensions. From count illustrates that health providers can an ethical standpoint, programme manag- increase quality of care by reducing barriers need to satisfy the client’s right to ob- ers to care through patient-friendly servic- tain competent, compassionate, high es. An HMIS providing information on quality care. Clients tend to value the ap- quality of services will assist the newly pearance of the service setting, privacy, re- formed district health systems in our coun- spectful treatment and convenience, while try to develop and evaluate patient friendly providers emphasize facility environment, policies and services with the ultimate aim programme infrastructure and workload of improving the quality of health care.

References 1. Campbell SM, Roland MO, Buetow SA. Massachusetts, Ballinger Publishing, Defining quality of care. Social science & 1976:19. medicine, 2000, 51(11):1611–25. 9. Vuori H. Patient satisfaction—an at- 2. World health report 2000. Health sys- tribute or indicator of quality of care? tems: improving performance. Geneva, Quality review bulletin, 1987, 13(3): World Health Organization, 2000. 106–8. 3. Garner P, Thomason J, Donaldson D. 10. Davies AR, Ware JE Jr. Involving con- Quality assessment of health facilities in sumers in quality of care assessment. rural Papua New Guinea. Health policy Health affairs, 1988, 7(1):33–48. and planning, 1990, 5:49–59. 11. Goldman RL. The reliability of peer as- 4. Bryce J et al. Assessing the quality of fa- sessment of quality of care. Journal of cility based child survival services. the American Medical Association, Health policy and planning, 1992, 7: 1992, 267(7):958–60. 155–63. 12. Fitzpatrick RM. Surveys of patient satis- 5. Nicholas DD, Heiby JR, Hatzell TA. The faction I: important general consider- quality assurance projects: introducing ations. British medical journal, 1991, quality improvement to primary health 302(6781):887–9. care in less developed countries. Quality 13. Conner JM, Nelson EC. Neonatal inten- assurance in health care, 1991, 3(3): sive care: satisfaction measured from a 147–65. parents’ perspective. Pediatrics, 1999, 6. Haddad S et al. Patient perception of 103(1 suppl.):336–49. quality following a visit to a doctor in a 14. Calnan M. Towards a conceptual frame- primary care unit. Family practice, 2000, work of lay evaluation of health care. So- 17(1):21–9. cial science & medicine, 1988, 27(9): 7. Aharony L, Strasser S. Patient satisfac- 927–33. tion: What we know about and what we 15. Grol R. Improving the quality of medical still need to explore. Medical care re- care: Building bridges among profes- view, 1993, 50(1):49–79. sional pride, payer profit and patient sat- 8. Palmer RH. Definitions and data. In: isfaction. Journal of the American Green R, ed. Assuring quality in medical Medical Association, 2001, 286 (20): care: the state of the art. Cambridge, 2578–85.

24 Health management information.pmd197 10/25/2005, 1:24 AM 198 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

16. Reeder LG. The patient-client as a con- ies in family planning, 1990, 21(2):61– sumer: some observations on the chang- 91. ing professional–client relationship. 26. Creel LC, Perry RJ. Improving the quality Journal of health and social behaviour, of reproductive health care for young 1972, 13(4):406–12. people. Washington DC, Population Ref- 17. Savage R, Armstrong D. Effect of a erence Bureau. Population Council, general practitioner’s consulting style on 2003 (New perspectives on quality of patients’ satisfaction: a controlled study. care No. 4). British medical journal, 1990, 301 27. Parasuraman A, Zeithaml VA, Berry LL. (6758):968–70. SERVQUAL: a multi-item scale for mea- 18. Bollam M, McCarthy M, Modell M. suring consumer perception of service Patient’s assessments of out of hour care quality. Journal of retailing 1988, 64 in general practice. British medical jour- (1):12–40. nal, 1988, 296:829–32. 28. Duran-Arenas L et al, The development 19. Kincey J, Bradshaw P, Ley P. Patients’ of quality information system: a case satisfaction and reported acceptance of study of Mexico. Health policy and plan- advice in general practice. Journal of the ning, 1998, 13(4):446–58. Royal College of General Practitioners, 29. National health management informa- 1975, 25(157):558–66. tion system (HMIS). Karachi, Govern- 20. Roghmann K, Hengst A, Zastowny T. Sat- ment of Pakistan, Ministry of Health, isfaction with medical care: its measure- Health Division, 2002 (http://www. ment and relation to utilization. Medical pakistan.gov.pk/health-division/ care, 1979, 27:461–77. informationandservices/hims.jsp, accessed 20 October 2003). 21. Weiss BD, Senf JH. Patient satisfaction survey instrument for use in health main- 30. Ministry of Health/WHO annual report of tenance organizations. Medical care, health services in Pakistan. Islamabad, 1990, 28(5):434–45. Government of Pakistan, 1996. 22. Strengthening of health information sys- 31. Health, Nutrition and Population Unit. tems in countries of the South-East Asia The World Bank’s work in Pakistan. Region. Report of an intercountry con- Washington DC, World Bank, 2001. sultation. New Delhi, World Health Orga- 32. Andaleeb SS. Service quality perception nization Regional Office for South-East and patient satisfaction: a study of hospi- Asia, 2001 (SEA-HS-219). tals in a developing country. Social sci- 23. Gilson L, Alilio M, Heggenhougen K. ence and medicine, 2001, 52:1359–70. Community satisfaction with primary 33. Mendoza Aldana JM, Piechulek H, Al- health care services: an evaluation un- Sabir A. Client satisfaction and quality of dertaken in the Morogoro region of Tan- health care in rural Bangladesh. Bulletin zania. Social science & medicine, 1994, of the World Health Organization, 2001, 39(6):767–80. 79(6):512–7. 24. Hartigan P. The importance of gender in 34. National health policy 2001. The way for- defining and improving quality care: ward: agenda for health sector reform. some conceptual issues. Health policy Islamabad, Government of Pakistan, and planning, 2001, 16(suppl. 1):7–12. Ministry of Health, 2001. 25. Bruce J. Fundamental elements in the quality of care: a simple framework. Stud-

24 Health management information.pmd198 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 199

Analyse Les problèmes de l’eau dans la Région de la Méditerranée orientale T. Zeribi1

RÉSUMÉ Dans le domaine de l’eau, la Région OMS de la Méditerranée orientale se trouve confrontée à des défis redoutables du fait de la hausse de la demande en eau, aussi bien pour la consommation que pour l’irrigation d’une agriculture de plus en plus productive et polluante, de la rareté de la ressource en eau, de la sécheresse, de l’érosion, de la pollution, d’une gestion inappropriée, de politiques inadaptées et de con- sidérations juridiques et institutionnelles. À ces difficultés s’ajoutent les risques de conflits régionaux liés au manque d’une gestion « commune » des eaux transfrontalières, objet de convoitise entre pays limitrophes. Cette problématique touchant la disponibilité de l’eau, la santé et le développement est analysée dans le rapport et l’auteur, sur la base de la distribution des ressources hydriques par secteurs économiques et l’énorme pourcentage destiné à l’agriculture, pose la question de savoir si les pays de la Région sont prêts à revoir leurs stratégies en matière de priorités hydriques dont les aspects santé, agriculture, autosuffisance alimentaire ne sont pas les moindres.

Water problems in the Eastern Mediterranean Region SUMMARY The Eastern Mediterranean Region of the World Health Organization is confronted with formida- ble water problems due to: increased water demand both for consumption and for irrigation in agriculture that is becoming more productive and more polluting; scarce water resources; drought, erosion and pollution; inappropriate management; inadequate policies; and institutional and legal considerations. Added to these problems are the risks of regional conflicts because of the lack of “shared” management of cross-border waters which are an object of contention between neighbouring countries. This report analyses the issues relating to water availability, health and development on the basis of the distribution of water resources, and their use by industry and the huge proportion for agricultural use. It raises the question whether countries in the Region are ready to review their strategies on water priorities, particularly in the areas of health, agriculture and food self-sufficiency.

Ancien Représentant de l’OMS au Maroc (Correspondence à adresser à: [email protected]). Reçu : 19/03/01 ; accepté : 10/02/02

25 Les problFmes de l eau.pmd 199 10/25/2005, 1:24 AM 200 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction tion de la contamination et de l’évaporation dans les réservoirs, par le recyclage, Des Pharaons de l’antique Égypte l’entretien des réseaux, la lutte contre le jusqu’aux dernières révolutions indus- gaspillage et la culture de variétés moins trielles, l’épopée universelle est avant tout exigeantes en eau ou plus tolérantes au sel. celle de l’eau. De sa quête et de sa conquête Un habitant des États-Unis utilise 900 li- sont nés des civilisations, des économies, tres d’eau par jour, mais dans certaines des modes de vie qui ont influencé de façon zones de la Région on doit se contenter de décisive le cours de l’histoire, au point de 30 litres. Un Israélien consomme quatre transformer un élément naturel en un pro- fois plus qu’un Palestinien. Le gouverne- duit industriel indispensable à tous. « Nous ment de l’Afrique du Sud s’est fixé comme buvons 90 % de nos maladies » a dit Pas- première priorité la fourniture d’au moins teur, « et l’homme, être de chair et de sang, 25 litres d’eau par jour (pour la boisson et est avant tout fait…d’eau ». l’hygiène) à chacun de ses citoyens, en Sans céder au catastrophisme, il faut abordant la question sous l’angle de la di- pourtant constater que les ressources se gnité et de l’équité [2]. Les pays d’Afrique raréfient en maintes régions, que le niveau du Nord et du Proche-Orient, avec moins des réserves baisse, que la pollution est de 2000 m3 par habitant et par an, vivent largement répandue et que la désertification dans des conditions de « stress hydrique » avance. Ces phénomènes ne suscitent pas qui handicapent leur vie économique et sole même sentiment d’urgence que les ciale [3]. changements climatiques, la déforestation Le dessalement de l’eau de mer n’est ou la couche d’ozone. Le manque d’eau a possible, pour le moment, que dans cer- pourtant un coût humain exorbitant : mal- tains pays (où l’énergie est très bon nutrition, maladies hydriques, exode rural, marché), et ce procédé n’empêche pas la charges accrues pour les femmes, etc. La consommation d’eau d’atteindre des prospective des crises de l’eau demeure la niveaux alarmants en Arabie saoudite et au comparaison entre les ressources et les be- Koweït, par exemple. soins en eau, mais à la condition que Même s’il est surtout question des l’évaluation des unes, l’estimation et la pro- problèmes qui affectent la Région, l’eau jection des autres soient pertinentes et bien représente également un défi majeur pour adaptées à cette comparaison. l’Europe. • Trente pour cent des eaux de surface Disponibilité de l’eau sont menacées par la contamination. Plus de 60 % des terres agricoles Si la population de la Région de la Méditer- présentent des concentrations jugées ranée orientale a été multipliée par trois au préoccupantes en pesticides et éléments cours du siècle dernier, la demande en eau, nutritifs, composés qui constituent une elle, a été multipliée par sept, et la surface menace très sérieuse pour les eaux sou- des terres irriguées par six [1]. De plus, au terraines, les fleuves et finalement les cours des cinquante dernières années, la eaux côtières. pollution des aquifères a réduit les réserves • Vingt pour cent des eaux sont perdues hydriques d’un tiers. Les ressources dis- en moyenne dans les réseaux de distri- ponibles pourraient cependant être utilisées bution (30 % au Royaume-Uni, 40 % en beaucoup plus efficacement par la réduc- Sardaigne).

25 Les problFmes de l eau.pmd 200 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 201

• La fréquence des catastrophes naturel- travaux de recherche ont touché à des élé- les (crues, sécheresse) a été multipliée ments aussi importants que : par 3 ou 4 lors des trois dernières • le manque de temps et d’énergie qui décennies, vraisemblablement à la suite peut affecter la sélection par la des changements globaux. maîtresse de maison d’une source • Les déficits en eau, dont certains sont d’eau adéquate pour les besoins essen- susceptibles de conduire à des conflits, tiels familiaux ; ne sont pas encore chroniques mais • l’impact des facteurs sociaux et de la pourraient le devenir au cours du siècle santé de la femme sur la qualité de l’eau, prochain, y compris au Royaume-Uni et la nutrition et l’hygiène familiale ; aux Pays-Bas [4]. • le rôle des structures familiales et com- Il faut souligner de plus que l’évaluation munautaires qui responsabilisent les des ressources reste handicapée par femmes dans le choix de la source l’inégalité des connaissances hydrologiques d’eau, le transport, le stockage et dans la Région. « La collecte et l’évaluation l’usage de l’eau [5]. des données sur les ressources en eau Dans un aide-mémoire publié par douce et sur leur utilisation sont, dans bien l’OMS, on note que dans le monde, un en- des régions, rudimentaires...Face à l’éven- fant meurt d’une maladie associée à l’eau tualité d’une crise de l’eau, améliorer nos toutes les huit secondes [6]. Chaque année, connaissances devient impératif ». (John plus de cinq millions de décès sont dus à Rodda, Président de l’Association interna- des maladies associées à une eau de bois- tionale des Sciences hydrologiques 1996). son ou à un milieu domestique conta- miné(e) et à une évacuation insatisfaisante L’eau et la santé des excreta. L’eau contaminée par l’homme ou par L’eau est essentielle à la santé et même à la des déchets chimiques ou industriels peut survie de l’homme et il n’est pas exagéré de provoquer toute une série de maladies dire qu’elle représente l’un des droits fon- transmissibles par absorption ou par con- damentaux de l’être humain. Sans eau saine tact (voir Tableau 1) : et sans assainissement, il ne peut y avoir de • maladies causées par l’absorption d’une développement réel. Une communauté ra- eau contaminée par des excreta hu- vagée par les maladies diarrhéiques, la dra- mains ou animaux ; cunculose ou la schistosomiase est • maladies causées par une hygiène per- incapable de regarder au-delà de ses sonnelle insuffisante ; problèmes immédiats en direction de son • maladies causées par les parasites que bien-être socio-économique. L’eau saine l’on trouve chez des organismes interest la clé de la santé, et celle-ci est la condi- médiaires vivant dans l’eau ; tion préalable au progrès, à l’équité sociale • maladies causées par des insectes vec- et à la dignité humaine. teurs qui se reproduisent dans l’eau. Malgré le volume important de publications sur le statut de la femme et son rôle L’OMS considère que l’approvisionne- majeur dans l’approvisionnement en eau, le ment en eau et l’assainissement peuvent transport et l’utilisation de l’eau pour les être considérés comme un processus besoins humains et domestiques, peu de comportant trois éléments interactifs.

25 Les problFmes de l eau.pmd 201 10/25/2005, 1:24 AM 202 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Tableau 1 Estimations de la morbidité et de la mortalité dues aux maladies associées à l’eau

Maladie Morbidité (nombre Mortalité Lien avec d’épisodes par an, (nombre de l’approvisionnement en ou comme indiqué) décès par an) eau et l’assainissement

Maladies 1 000 000 000 3 300 000 Étroitement liées à une évacuation diarrhéiques inadéquate des excreta, à une hygiène personnelle et domestique médiocre et à une eau de boisson contaminée.

Helminthiases 1 500 000 000a 100 000 Étroitement liées à une évacuation inadéquate des excreta et à une hygiène personnelle et domestique médiocre.

Schistosomiase 200 000 000a 200 000 Étroitement liée à une évacuation inadéquate des excreta et à l’absence d’une source proche d’eau saine. Dracunculose 100 000a,b - Étroitement liée à la mauvaise qualité de l’eau de boisson. Trachome 150 000 000c - Étroitement lié à une hygiène insuffisante du visage et souvent à l’absence d’une source proche d’eau saine. Paludisme 400 000 000 1 500 000 Lié à une exploitation de l’eau, à un stockage, à un entretien des points d’eau et à un écoulement qui laissent à désirer. Dengue 1 750 000 20 000 Liée à une exploitation de l’eau, à un stockage, à un entretien des points d’eau et à un écoulement qui laissent à désirer. Poliomyélite 114 000 - Liée à une évacuation inadéquate des excreta, à une hygiène personnelle et domestique médiocre et à la mauvaise qualité de l’eau de boisson. Trypanosomiase 275 000 130 000 Liée à l’absence d’une source proche d’eau saine. Filariose à W. 72 800 000d - Liée à une exploitation de l’eau, à un bancrofti stockage, à un entretien des points d’eau et à un écoulement qui laissent à désirer. Onchocercose 17 700 000a,d 40 000e Liée à des problèmes d’exploitation de l’eau dans les projets à grande échelle.

aNombre de sujets actuellement infectés. bSoudan non compris. cNombre de cas de maladies évolutives. On compte quelque 5,9 millions de cas annuels de cécité ou de complications graves du trachome. dDont 270 000 cas de cécité (estimation). eMortalité due à la cécité. Source : [6].

25 Les problFmes de l eau.pmd 202 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 203

L’élément le plus fondamental consiste à et des usines. Les programmes destinés disposer d’une eau de boisson saine et de aux municipalités ont englouti plusieurs moyens adéquats d’évacuation des excreta. milliards de dollars. Les zones rurales et les Il faut pour cela que chacun dispose de 20 à producteurs agricoles continuent à ne re- 40 litres d’eau par jour à une distance rai- cevoir que des miettes de ces sommes. sonnable de son domicile. Une eau de bois- La pollution agricole a la particularité son saine suppose une bonne protection d’être surtout diffuse, puisqu’elle résulte des sources d’eau ainsi que des moyens de l’effet combiné de plusieurs facteurs. La hygiéniques pour transporter l’eau et la dégradation des sols provoque l’entraîne- stocker à domicile. L’OMS a émis des di- ment de matières fertilisantes et de pesti- rectives qui visent à servir de point de dé- cides qui aboutissent dans les eaux de part à l’élaboration de normes nationales surface et souterraines. Plusieurs plans qui, correctement appliquées, garantiraient d’eau se retrouvent avec des niveaux de une eau saine par l’élimination ou la réduc- phosphore, de nitrites et de nitrates tels tion à une teneur minimale des substances qu’il y a prolifération d’algues et, par voie présentes dans l’eau et réputées dangereu- de conséquence, mort des rivières et lacs. ses pour la santé. Les concentrations de pesticides dans l’eau Les principales méthodes de lutte contre dépassent largement les normes accepta- les maladies liées à l’eau sont l’accès à un bles. approvisionnement en eau saine, l’amélio- Une intervention efficace nécessite plus ration de l’hygiène personnelle et domes- qu’un constat général. Il ne suffit pas de tique et une participation communautaire savoir que la surfertilisation est la cause accrue. Pour être réellement efficaces, ces principale de pollution diffuse, encore faut- méthodes doivent cependant s’accompa- il connaître l’ampleur du phénomène et sa gner d’autres mesures telles que la lutte cartographie. contre la pollution et le drainage approprié À titre de repère, la protection de des eaux de surface. Étant donné que la l’environnement dans les zones rurales protection des sources d’eau naturelles est commande une réduction de l’usage des la base d’un approvisionnement permanent pesticides. On avance de plus en plus des en eau saine, cette mesure devrait toujours solutions de lutte biologique et génie géné- faire partie intégrante des programmes sa- tique, mais la distance est loin du labora- nitaires et environnementaux. toire au champ.

Conflits L’eau et le développement L’eau peut aussi être source de conflits. Is- raël et ses voisins sont toujours aux prises à Pollution cause des eaux du Jourdain, du Litani et du Il est illusoire de penser améliorer de façon château d’eau du Golan. L’accord de Taba significative la qualité des cours d’eau – le (dit aussi Oslo II), signé à Washington le meilleur indice de pollution – sans une di- 28 septembre 1995 entre l’Autorité palesti- minution notable du niveau de la pollution nienne et l’État hébreu, officialise le par- diffuse d’origine agricole. Or dans la Ré- tage léonin des aquifères de Cisjordanie : gion de la Méditerranée orientale, depuis 82 % aux Israéliens et 18 % seulement aux plus de vingt ans, les budgets sont allés à Palestiniens. Rien n’est définitivement réglé l’assainissement des eaux usées des villes

25 Les problFmes de l eau.pmd 203 10/25/2005, 1:24 AM 204 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

pour ce qui est des eaux du Nil, du Tigre et – en somme les crises « futuribles » – sui- de l’Euphrate. vant des scénarios assez contrastés, sans Le plus grand réseau d’irrigation du se prononcer sur leur probabilité : celle-ci monde se trouve au Pakistan et dépend de dépendra de la possibilité de réalisation de l’Indus, dont le bassin se situe en partie en chaque scénario, donc en partie des choix Inde. Les gigantesques projets hy- de politique de l’eau présents, que la drauliques asiatiques des Trois Gorges ou prospective vise précisément à éclairer. du Xialangdi en Chine, et du Bakun à Sa- Les inadéquations offre/demande rawaken (Malaisie), par exemple, suscitent prospectées et les situations critiques résul- de vifs débats quant à leur coût, leur oppor- tantes sont à analyser à différents points de tunité, leurs conséquences humaines et vue : écologiques. Les États fédérés, aux États- • économiques : augmentation des coûts Unis et en Inde, s’opposent sur la réparti- de maîtrise et de mobilisation des res- tion des ressources en eau. En Europe, les sources, de production et de traitement, travaux réalisés sur le Danube obligent la ainsi que de conservation, à la charge de Hongrie et la Slovaquie à redessiner leur tous les acteurs économiques ; néces- frontière matérialisée par le fleuve ; quant à sité d’adaptation et de révision des de- leur différend relatif à la centrale de Gab- mandes (avec réduction de certaines) ; cikovo, il a été soumis au Tribunal interna- incidences sur les niveaux de vie et les tional de la Haye. productions : restrictions sectorielles et À qui profite le Nil ? À l’Égypte et au conjoncturelles possibles, handicaps du Soudan, sans doute, mais l’Éthiopie, qui développement ; compte plus de 55 millions d’habitants, ai- • sociaux : inégalités accrues à l’intérieur merait aussi en tirer parti. De leur côté, la du territoire entre populations urbaines Mauritanie, le Sénégal et le Mali s’opposent et rurales, zones de montagne et espa- sur la manière de se répartir l’eau de leur ces côtiers qui s’approprient l’accès à la fleuve commun frontalier, qui traverse au ressource et influencent les stratégies total sept pays. d’aménagement du territoire ; faisabilité La répartition inégale de cette ressource sociale d’une maîtrise des demandes fondamentale à la vie des populations, et à (par les prix et la formation) ; l’activité humaine en général, fait craindre • écologiques : pressions accrues sur que certains pays ne fassent partie des l’environnement et les milieux aqua- zones de conflits latents ayant l’eau pour tiques, et par contrecoup sur l’intégrité origine, parmi les 300 que l’Organisation des ressources en eau elles-mêmes ; in- des Nations Unies a recensées de par le cidences des changements climatiques monde. Les grands fleuves de la Région, sur les ressources. qui traversent différents pays, pourraient être la raison de discordes majeures [7]. Il va de soi qu’un exercice au niveau ré- Des crises de l’eau sont-elles en gional ou même mondial ne sera révélateur perspective et à craindre au XXIe siècle que de situations moyennes pour chaque dans la Région ? Quelles crises de l’eau de- pays. Il devrait être le prélude à des études main ? et analyses plus fines dans tous les pays L’objectif de l’exercice de prospective menacés où une telle prospective n’a pas est d’explorer la géographie, les échéances encore été engagée ou ne l’a été que som- et les ampleurs possibles des crises de l’eau mairement.

25 Les problFmes de l eau.pmd 204 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 205

Eau virtuelle réutilisées après traitement mais ne représentent qu’une petite fraction des ap- Une des données qui émergent de façon ports). évidente de l’analyse de la distribution des Au cours des cinquante dernières an- ressources hydriques par secteurs écono- nées, les surfaces irriguées, traditionnelle- miques est l’énorme pourcentage destiné à ment déjà considérables en Égypte, ont l’agriculture, qui oscille entre 70 et 80 % continué à croître fortement dans plusieurs [8]. De nombreux facteurs concourent à pays et ont été multipliées par plus de deux cette priorité, et en premier lieu les choix de en Algérie, en Israël, en Syrie, en Tunisie politique économique des pays qui ont été ou en Turquie. La part réservée à dictés par l’objectif prioritaire de la sécurité l’irrigation est généralement sans commune alimentaire. mesure avec celle des apports de la produc- Dans la Région de la Méditerranée tion de l’agriculture irriguée au PIB des orientale, l’agriculture absorbe les deux pays. Bien entendu, ce rythme d’expansion tiers de la consommation d’eau, mais elle ne saurait se poursuivre et la plupart des doit compter avec l’intense concurrence pays prévoient eux-mêmes de réduire peu à des villes, de l’industrie et du tourisme. Les peu la portion de leurs ressources con- spécialistes pronostiquent qu’elle ne sacrée à l’irrigation [9]. parviendra pas à conserver, au XXIe siècle, Reste à savoir si ces objectifs ambitieux la part de la ressource qu’elle s’adjuge ac- de réduction de la demande agricole pour- tuellement. ront être atteints face à l’accroissement On peut affirmer que la rareté crois- considérable des besoins alimentaires et au- sante de l’eau douce est actuellement un tres de ces pays. Cette situation d’ensemble obstacle majeur à la production alimentaire, donne lieu à un débat, en particulier sur le à la santé des écosystèmes, à la stabilité so- prix de l’eau d’irrigation, généralement très ciale et à la paix. L’Algérie, l’Égypte, la faible, d’autant que le maintien, et à fortiori, Libye, le Maroc et la Tunisie, qui manquent l’augmentation des allocations de ressour- d’eau, importent d’ores et déjà plus du tiers ces à l’agriculture pourraient constituer un de leurs céréales. handicap au développement d’autres sec- L’examen des demandes sectorielles en teurs économiques, y compris le tourisme eau des pays méditerranéens montre qui demande des quantités d’eau relative- l’importance considérable qu’y tient ment importantes en saison estivale, au l’agriculture par rapport aux usages do- moment même où les ressources sont les mestiques et industriels et aux impératifs plus basses, et dans les régions littorales, là écologiques. C’est évidemment l’irrigation où elles sont généralement les plus éloi- qui prend cette part majeure, et même gnées ou les plus fragiles. écrasante, des quantités d’eau totales uti- A l’heure actuelle, près de la moitié des lisées, dépassant 80 % des prélèvements quantités d’eau prélevées pour l’irrigation dans presque tous les pays et environ 65 % grâce à des aménagements généralement pour l’ensemble de la région [9]. Cette si- coûteux est perdue par évaporation dans les tuation est d’autant plus sérieuse que la ma- retenues (1m3 sur 8 à Assouan), par infil- jeure partie de l’eau d’irrigation est tration dans les canaux, par de mauvais consommée et ne retourne pas localement réglages des apports aux champs, par des dans le cycle hydrologique (mis à part les apports excessifs encouragés par la quasi- eaux polluées de drainage qui peuvent être gratuité de l’eau, par des choix de cultures

25 Les problFmes de l eau.pmd 205 10/25/2005, 1:24 AM 206 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

trop consommatrices, etc., alors que la ne peut atteindre vu les conditions agro- productivité des périmètres irrigués est af- climatologiques – mais plutôt comme capa- fectée par la salination des sols due à un cité à agir sur la balance alimentaire de drainage insuffisant. L’amélioration de façon à ce que les exportations permettent l’irrigation pourrait fournir les trois quarts l’achat sur les marchés mondiaux du quota du « gisement » d’eau qu’il est concevable de produits alimentaires que le pays n’est d’économiser par une meilleure gestion de pas capable de produire sur la base des res- la demande dans la région méditerranéenne. sources dont il dispose. Ceci suppose à la fois des investissements Il faut souligner, à titre d’expérience, importants pour améliorer le transport et les que l’ouverture au monde du marché du riz réseaux, l’introduction de méthodes fines au Japon, imposée par les accords du d’irrigation (aspersion, goutte à goutte, GATT en 1993, a mis à la retraite bon nom- contrôle électronique) ainsi que le change- bre de riziculteurs, leur produit étant trop ment, toujours socialement difficile, de cher comparé à celui de leurs homologues comportements ancestraux, y compris en thaïlandais ou chinois. Cette mise en ce qui concerne le juste prix de l’eau. jachère forcée des rizières a provoqué, au Les principaux projets ont été conçus grand dam des riverains, des inondations dans le but d’assurer des denrées alimen- destructrices. Pareillement, l’accord de taires de base à une population en forte libre-échange nord-américain (ALENA) croissance. Pendant les années 80, la rend compétitifs, sur les marchés des détérioration du prix des produits agricoles États-Unis et du Canada, les fruits et a entraîné une chute des entrées provenant légumes mexicains, gros consommateurs de l’agriculture. On a donc enregistré un d’eau dans un pays qui en manque, ce qui changement des schémas de cultures, provoque de graves tensions sociales [10]. lequel a privilégié la production maraîchère Pourtant, un rapport régional discutant et fruitière, le coton et les oléagineux. la notion de sécurité alimentaire dans le cas Cependant, les gouvernements n’ont des pays arides comme ceux du Proche- pas soutenu les changements en cours, car Orient préconise que « ces États peuvent ils étaient préoccupés par la croissante assurer leurs besoins en eau et leur sécurité dépendance alimentaire et par les retom- alimentaire en développant des économies – bées que celle-ci pouvait avoir sur la stabi- basées sur le commerce, le tourisme et lité politique du pays. Les interventions de l’industrie – susceptibles de leur faire ga- l’État ont pris la forme de contrôles admi- gner suffisamment d’argent pour importer nistratifs sur les schémas de culture de « l’eau virtuelle » bon marché disponible adoptés, en altérant le libre fonctionnement sur le marché mondial sous forme de du marché et en favorisant les cultures à céréales, de légumes, de viande, de lait…» consommation hydrique élevée et à bas [10]. taux de rentabilité. On a en effet favorisé La question de l’eau dans la Région de la les cultures de grain et de sucre pour la Méditerranée orientale est d’une extrême consommation interne et de coton pour les sensibilité et elle est au coeur du devenir de exportations. la Région. Elle est en passe d’être une arme Le concept de sécurité alimentaire pour de développement dont l’usage crée tant le les pays de l’aire ne doit pas être entendu désarroi que l’indignation. Elle est source comme obtention de l’autosuffisance ali- de vie ; elle est aussi source de tension et de mentaire – un objectif entre autres que l’on conflit. C’est dire que la distribution équita-

25 Les problFmes de l eau.pmd 206 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 207

ble des ressources en eau dans la Région driques dont l’aspect santé, agriculture, au- est au coeur du devenir des nations. Ques- tosuffisance alimentaire ne sont pas les tion explosive, parce qu’aucun pays n’est moindres ? autosuffisant en eau. En effet, si les Depuis plus d’un demi-siècle, les pro- mesures adéquates ne sont pas prises, il n’y positions sur le partage ou la distribution de aura pas suffisamment d’eau d’ici la fin de l’eau dans plusieurs zones de la Région de la deuxième décennie de ce siècle pour la Méditerranée orientale ont fait l’objet de répondre aux besoins fondamentaux. plusieurs analyses et rapports mais Ce problème a souvent retenu l’élément essentiel qu’est la coopération à l’attention de plusieurs instances interna- l’échelle régionale pour la gestion interpays tionales et à titre indicatif, le Fonds des ressources hydriques n’a pas connu un monétaire international a publié un rapport sort heureux ! A ce jour, cette coopération analysant les causes hydriques des conflits horizontale, action d’envergure et en pro- transfrontaliers et les incidences des dis- fondeur, n’a été ni décidée ni mise en œuvre putes entre les usagers de l’eau [11]. [12]. L’analyse de l’auteur met en avant un L’OMS pourrait être garante, en co- modèle mathématique emprunté du docu- opération avec la FAO, d’une telle action ment « Optimal Water Management and d’envergure. Sa crédibilité régionale, son Conflict Resolution: The Middle East action dans les projets de soins de santé de Project » par Franklin M. Fisher et al. base, de besoins fondamentaux du déve- Les pays de la Région, au-delà de leurs loppement, de protection de l’environne- relations tumultueuses, sont confrontés au ment, de lutte contre la pollution, d’alimen- besoin de promouvoir une coopération en tation en eau et d’assainissement, de la vue de gérer de manière rationnelle les fai- qualité et des normes de l’eau potable, de bles ressources hydriques disponibles. salubrité de l’environnement, etc., justifient Face à la menace que représente le manque amplement ce choix. d’eau, ces pays n’ont plus d’autre choix C’est au Bureau régional de sensibiliser que de coopérer. ses partenaires du système des Nations La question qui se pose au-delà de la Unies et les Pays Membres et de préparer distribution des quantités d’eau limitées est une nouvelle stratégie régionale dont les la suivante : les pays sont-ils prêts à revoir composantes eau-santé-développement leurs stratégies en matière de priorités hy- sont les éléments moteurs.

Références 1. Forum mondial de l’eau. Communication Disponible à l’adresse suivante : du Directeur général de la FAO. Forum http://www.oieau.fr/ciedd/contributions/ Mondial de l’Eau, 21-23 mars 1997, at2/contribution/batisse.htm (consulté le Marrakech (Maroc). 9 février 2005). 2. Forum mondial de l’eau, 21-23 mars 4. Martin JM. Commission européenne, in- 1997, Marrakech (Maroc). tervention au Forum mondial de l’eau. Forum mondial de l’eau, 21-23 mars 3. Batisse M. Eau et développement du- 1997, Marrakech (Maroc). rable dans le bassin méditerranéen. Conférence internationale « Eau et 5. An anthology on women, health and en- Développement Durable » Paris, 19- vironment: water. Available at : http:// 21 mars 1998 (Documents de travail). www.who.int/docstore/peh/archives/

25 Les problFmes de l eau.pmd 207 10/25/2005, 1:24 AM 208 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

women/womwater.htm (accessed : 9 11. Ferragina E. Le contrôle des ressources February 2005). hydriques au Moyen-Orient : conflits ou coopération. International Solidarity 6. Eau et assainissement. Genève, Movement, 9 février 2005 Disponible à Organisation mondiale de la Santé, l’adresse suivante : http://www.ism- Aide-mémoire No 112, mars 1996. france.org/news/article.php?id=1088&- 7. Missaoui M. L’eau enjeu stratégique du type=analyse&lesujet=Eau (consulté le siècle. La Presse, Tunis, 2 août 2000. 9 février 2005). 8. Plan Bleu pour la Méditerranée. Contri- 12. Optimal water management in the bution au Forum mondial de l’eau, 21- Middle East and other Regions. Finance 23 mars 1997, Marrakech (Maroc). and Development. A quarterly magazine of the IMF, 2001, 38(3). Available at : 9. Sommet mondial de l’alimentation, 13- http://www.imf.org/external/pubs/ft/ 17 novembre 1996, Rome (Italie). Docu- fandd/2001/09/fisher.htm (accessed : 9 ment d’information technique. Dis- February 2005). ponible à l’adresse suivante : http:// www.fao.org/wfs/index_fr.htm (consulté 13. Zeribi T. Rôle de l’OMS dans la consoli- le 9 février 2005). dation de l’action nationale dans le domaine de l’eau. 9e Congrès de l’Union 10. Bouguerra ML. Bataille planétaire pour des Distributeurs de l’Eau en Afrique, « l’or bleu ». Le Monde diplomatique, 1998, Casablanca (Maroc). novembre 1997.

Water for life – making it happen As we enter the International Decade for Action Water for Life 2005– 2015, this report makes clear that achieving the target of the Millen- nium Development Goals (MDGs) for access to safe drinking water and basic sanitation will bring a payback worth many times the in- vestment involved. It will also bring health, dignity and transformed lives to many millions of the world’s poorest people. At US$11.3 billion a year, the dollar costs of achieving the MDG drinking water and sanitation target are affordable; the human costs of failing to do so are not. The first part of this report charts the effect that lack of drinking water and sanitation has on people’s lives at different stages, highlighting the gender divide and threat posed by HIV/ AIDS. The second part looks at a range of interventions that are being advocated and analyses their potential impact on progress towards the MDG drinking water and sanitation target. The report concludes with statistical tables showing the increase needed to achieve the MDG drinking water and sanitation target and drinking water and sanitation coverage estimates at the regional and global level. Further information on this publication can be obtained from WHO Press: http://www.who.int/bookorders/anglais/detart1.jsp? sesslan= 1&codlan=1&codcol=15&codcch=629

25 Les problFmes de l eau.pmd 208 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 209

Review Views on oral health care strategies N. Beiruti1

ABSTRACT The Oral Health Programme (ORH) is a health promotion and disease prevention initiative. ORH should be integrated into primary health care programmes by building policies suited to each country and based on the common risk factor approach. Dental caries and periodontal diseases are highly prevalent in the Eastern Mediterranean Region, especially among children; therefore, school-based programmes are recommended. Although cost-effective, water and salt fluoridation are often unavailable and topical fluorides are recommended. Governments and industry must ensure availability of affordable fluoride toothpaste. Fluoride toothpaste should also be used to control periodontal diseases. The atraumatic restorative treatment approach should be used to treat dental caries. The Basic Package of Oral Care (BPOC) for deprived communities outlines this approach in detail. Continuous training and research are recommended for personnel to keep pace with changes in methods of prevention and treatment procedures.

Point de vue sur les stratégies de soins de santé bucco-dentaire RÉSUMÉ Le programme de santé bucco-dentaire est une initiative de prévention de la maladie et de promotion de la santé. La santé bucco-dentaire devrait être intégrée aux programmes de soins de santé primaires en établissant des politiques adaptées à chaque pays et basées sur l’approche des facteurs de risque communs. Les caries dentaires et les parodontopathies sont très répandues dans la Région de la Méditerranée orientale, notamment chez l’enfant ; des programmes scolaires sont donc recommandés. Bien que rentable, la fluoruration de l’eau et du sel est souvent inexistante et les fluorures topiques sont recom- mandés. Les pouvoirs publics et l’industrie doivent assurer la disponibilité de dentifrices au fluor abordables. Les dentifrices au fluor devraient être aussi utilisés dans la lutte contre les parodontopathies. La méthode de restauration non traumatique devrait être utilisée pour le traitement des caries dentaires. Cette méthode figure dans les éléments de base pour soins bucco-dentaires (BPOC) destinés aux communautés défavo- risées. La formation continue et la recherche sont recommandées pour permettre au personnel de suivre l’évolution des méthodes de prévention et des procédures de traitement.

1Former Director of the WHO Demonstration, Training and Research Centre for Oral Health, Damascus, Syrian Arab Republic (Correspondence to N. Beiruti: [email protected]). Received: 21/08/03; accepted: 30/11/03

26 Views on oral health.pmd 209 10/25/2005, 1:24 AM 210 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction crease and periodontal disease prevalence remain high, while in developed countries, This paper is for oral health, health and dental caries rates have been reduced and non-health personnel in the Eastern Medi- periodontal disease prevalence has de- terranean Region of the World Health Orga- creased to moderate or low levels [4,5]. nization (WHO) who are involved in planning, re-planning and evaluating oral health programmes. The following views Oral health and primary health and guidelines could be useful in establish- care ing oral health strategies for the next 2 de- In 1981, the WHO adopted a global strate- cades. gy of Health for All by the Year 2000. This In most countries, populations of very effectively established an agenda for a pri- young and school-age children receive mary health care approach based on practi- considerable benefit from preventive health cal, scientifically sound and socially measures such as vaccinations and exami- accepted methods and technology. It was nations of hearing, sight and physical de- to be made universally accessible to all indi- velopment in addition to other curative viduals and families in the community at a procedures. Oral health care is not deliv- cost that each community and each coun- ered to the same extent and is not given the try, in the spirit of self-reliance and self-de- same priority, perhaps because oral diseas- termination, could afford to maintain as it es are not as life-threatening as some other developed [6]. diseases are. Oral diseases are important Oral health is an integral part of primary public health problems, nonetheless, be- health care based on community participa- cause of their high prevalence, public de- tion and self-reliance with emphasis on the mand for treatment, their effect on promotion of health and the prevention of individuals and society in terms of pain, diseases [7]. discomfort, social and functional limitations and handicaps, and their effect on general health and quality of life. In addi- Global oral health goals by the tion, the financial effect on the individual year 2000 and the community is very high [1,2]. Oral health is related to well-being and In 1979, in the context of Health for All by quality of life as measured along functional, the Year 2000, the WHO Global Oral Health psychosocial and economic dimensions. Programme set the first global goal that 12- Diet, nutrition, sleep, psychological status, year-old children should not have more social interaction, school and work are af- than 3 decayed, missing or filled teeth fected by impaired oral and craniofacial (DMFT) by the year 2000. This goal was health. Reduced oral health-related quality modified at the WHO Eastern Mediterra- of life is associated with poor clinical status nean Regional Office (EMRO) Meeting on and reduced access to care [3]. Strengthening Oral Health in Primary Dental caries and periodontal diseases Health Care, 1993, in Nicosia, Cyprus, to are the most widespread conditions among be 1.5 DMFT in the Eastern Mediterranean populations, especially children, in develop- Region. Then, in 1981, the Fédération Den- ing countries. The WHO data confirm that taire Internationale (FDI) and the WHO dental caries prevalence continues to in-

26 Views on oral health.pmd 210 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 211

formulated these global goals for other age sizes building oral health policies that lead groups [8]: to the effective control of risks to oral • Ages 5–6 years: 50% caries-free. health based on the common risk factor • Age 18 years: 85% retain all teeth. approach [12]. The common risk factor approach ad- • Ages 35–44 years: 50% reduction in the dresses risk factors common to many number of persons with no teeth. chronic conditions such as heart disease, • Ages 65 years and over: 25% reduction cancer, obesity, diabetes, stroke, injury and in the number of persons with no teeth. oral disease within the context of the wider During the last 20 years, most EMR socio-environmental milieu. The major risk countries have established National Oral factors for those chronic diseases are Health Plans and the WHO EMRO has smoking, diets high in saturated fats and helped to develop the Oral Health Pro- sugars and low in fibre, fruit and vegeta- gramme (ORH) in the Region to achieve bles, stress, alcohol abuse, environmental these goals. Dental caries and periodontal hygiene, injury and sedentary lifestyle. The diseases are still the most prevalent condi- common risk factor approach is a health tions among populations. Oral health care promotion approach. It is more likely to be in most countries is difficult to obtain even effective because it tackles causes com- if available and tooth extraction is the pre- mon to a number of chronic diseases and dominant mode of treatment. Conventional incorporates oral health into general health oral care usually emphasizes technical and strategies [13]. curative solutions that are expensive and are an option only for the affluent sector of the population. Often traditional dentistry Oral health promotion has overlooked the importance of commu- Oral health promotion should have the nity-oriented prevention, as exemplified by highest priority and should follow the prin- the improvement in oral health in develop- ciples defined in the Ottawa Charter for ing countries [9,10]. Health Promotion [14]. Health promotion means building healthy public policies, cre- Oral Health by the year 2020 ating supportive environments, strengthening community action, developing coping All countries in the Region should review skills and re-orienting dental services. the oral health situation to assess their Health promotion policy must take into achievements and the problems that they consideration the uneven distribution of face. The planning or re-planning process health and disease, the uneven distribution should be done in the next two decades and of health hazards in the physical and social must take into account new concepts in environment, personal behavioural risk fac- delivering oral health care. tors, personal opportunities to adopt health- The WHO has adopted the renewal of ier lifestyles, and the uneven distribution the Health for All strategy for the 21st cen- and quality of health care [14]. In addition, tury. The ORH should be integrated with the Jakarta Declaration advocated health general health into the frame of primary promotion strategies that could develop and health care. Recently, the WHO reoriented change lifestyles and social, economic and the ORH initiative towards health promo- environmental conditions that determine tion and disease prevention [11]. It empha- health [15].

26 Views on oral health.pmd 211 10/25/2005, 1:24 AM 212 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Guiding principles for planning 12.Learning must continue throughout the career of the health professional. The WHO report on Oral Health for the 21st Century suggested the following guiding principles for oral health planning in the Oral health strategy next two decades [7]: 1. Oral health is an essential part of general Oral health strategy should be based on es- health, human function and quality of tablishing a national oral health plan as part life. of the general health plan with measurable goals that are suited to the country. Very 2. Oral health status should be improved young children and schoolchildren are the and maintained in the most economical most important target groups in a society manner consistent with quality and ac- for delivering oral health care because: cess. The community has a social responsi- 3. Prevention is preferable to treatment as • bility for its children. a general rule. • If the children are maintained in a state 4. Individuals should be motivated to do as of good oral health, it will be easy to much as possible for themselves to maintain their oral health in adult life. achieve and maintain oral health. Regular dental attendance patterns in 5. Caries and periodontal diseases can be • early life may be continued after school prevented and controlled. age. 6. Community methods of prevention Most children regularly attend schools, should be supportive of individual and • which are logical places for administer- personal care and in some situations are ing the ORH. more efficient. Health education at school age can bring 7. Oral health care should be provided in • about changes of behaviour and can the context of comprehensive care. easily be disseminated to the family. 8. Oral health care providers should be School-based oral health programmes prepared and motivated to consider must be implemented in all EMR countries general health and should participate in in the health promotion, or common risk the provision of general health care. factor, approach. The specific oral health 9. The type, number and distribution of care strategy includes health education and oral health care personnel should be health promotion, integration of health pro- maintained at levels consistent with motion activities, prevention of oral diseas- need, quality, cost and access neces- es and supportive curative procedures. sary to achieve the desired oral health status. Health education and promotion 10.Planning, health care practices and edu- Health education and health promotion cational programmes should be appro- should be the basis of any programme to priate for the population or situation in improve oral health and must be shared by question. individuals, families, health professionals, 11. Research, evaluation and education are institutions and local and national authori- essential for the continued advancement ties. Health education can increase knowl- of oral health. edge and reinforce desired behaviours, but

26 Views on oral health.pmd 212 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 213

to be successful it must be integrated with part of general hygiene and developing pop- other economic, social, legal and environ- ulation policies rather than high-risk strate- mental influences on health, all of which gies [18]. affect public access to and acceptance of preventive programmes. Health promotion Prevention of oral diseases provides preventive procedures while si- Dental caries and periodontal diseases are multaneously offering the opportunity to highly prevalent in all EMR countries, espe- educate individuals, communities and other cially among children. Health and education health professionals about the value of the authorities should cooperate to apply sim- measure introduced. It aims to put ple, economic, effective and efficient pre- preventive-oriented decisions into practice ventive programmes. The choice of [16]. preventive measures will depend upon oral The Action-Oriented Health Education disease prevalence, existing legislative reg- Programme in primary schools, which was ulations, attitudes of health authorities, developed by the WHO EMRO, the United cost–effectiveness and acceptance by the Nations Children’s Fund (UNICEF), the beneficiaries [16]. United Nations Educational, Scientific and For dental caries prevention, fluoride Cultural Organization (UNESCO) and the use is the most effective measure. Water Islamic Education, Scientific and Cultural fluoridation and salt fluoridation are very Organization (ISESCO), was an alternative cost-effective methods and provide sub- method to increase awareness of school stantial protection against caries for the children towards health and oral health. whole community. But because they are This method depended mainly on the par- not available in most developing countries, ticipation of children in preparing and pre- effective alternative means need to be intro- senting health knowledge and skills. This duced for populations, especially for approach was further developed into what schoolchildren. The most common meth- is known as the “community school” and ods are to apply topical fluorides in the the “health promoting school”. It calls for form of solutions, gels, varnishes and parents, teachers, children and health per- toothpastes [19]. sonnel to work together and to take action The WHO report on fluorides and oral to improve general and oral health as part of health states that fluoride toothpaste is one the society’s development [17]. of the most important delivery systems for fluoride [19]. The caries-reducing effect of Integration of health promotion fluoride is almost exclusively topical, activities which explains the anti-caries efficacy of Oral health messages and activities must be fluoride toothpaste. The toothpaste is often included in general health messages and ac- too expensive and many people cannot af- tions. The new concept of health promo- ford to use it regularly; therefore, govern- tion should replace the traditional dental ments and the toothpaste industry should care model that primarily emphasizes cura- ensure the availability of fluoride toothpaste tive services and individual responsibility. at an affordable cost to consumers, espe- The integration of oral health in general cially for schoolchildren [4]. health could be implemented by tackling The application of fissure sealants to the causes that are common to a number of occlusal surfaces of teeth significantly re- chronic diseases, including oral hygiene as duces the onset of carious lesions at these

26 Views on oral health.pmd 213 10/25/2005, 1:24 AM 214 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

sites. Sealants are well-documented aids proach as part of primary oral health care for the retention of teeth and for caries pre- [21,22]. Clinical research has shown that vention and there is ample evidence of their this approach is effective in treating caries long-term efficacy [16,20]. Because of in both primary and permanent dentitions their high cost and sensitive application, when compared with the conventional ap- however, they are not used as a public proach [23,24]. The adoption of this ap- health measure for children in most coun- proach in oral health strategies in EMR tries. countries will improve the oral health sta- Another goal is to prevent periodontal tus. diseases. The aim is to keep the mouth free of long-term plaque accumulations and to restrict periodontal diseases to no more Basic package of oral care than the earliest and reversible stages of In 2001, the WHO/ORH in cooperation gingivitis. Oral hygiene is fundamental to with the WHO Collaborating Centre for oral the control of periodontal diseases and flu- health in Nijmegen, The Netherlands, de- oride toothpaste should be used as a clean- veloped an integrated community-based ing agent, as well as for caries prevention. programme, the Basic Package of Oral Preventing other oral conditions should Care (BPOC) for deprived communities also be considered when planning oral [10]. It meets the principles of primary health strategies. These conditions include health care, based on proven and effective oral cancer, oral injuries, fluorosis, oral oral health measures, and it is applicable for mucosal diseases, acquired defects of den- developing and developed countries. The tal hard tissues and dentofacial anomalies. package has three essential components of oral health service, i.e. Oral Urgent Treat- Supportive curative procedures ment (OUT), Affordable Fluoride Tooth- For many reasons, most EMR countries paste (AFT) and Atraumatic Restorative face difficulties in treating decayed teeth Treatment (ART). Oral health promotion is among schoolchildren with the convention- considered an integral part of BPOC. EMR al approach. These reasons include the lim- countries could develop a BPOC based on ited number of dental units, the high cost of the perceived needs of local populations establishing new dental units, lack of main- and on existing supporting environmental tenance and spare parts, problems with conditions [4,22]. electricity and water supplies, schools scattered throughout urban and rural areas and controversy concerning amalgam fill- Training and research ings [9]. The atraumatic restorative treatment Continuous training activities for oral (ART) approach for dental caries has been health, health and non-health personnel developed in the last 10 years and is the should be considered in oral health plans to least invasive of the minimal intervention keep up with the rapid pace of change in approaches among preventive and curative scientific knowledge, methods of preven- procedures. It does not involve drills, water tion and treatment procedures, in addition or electricity. It depends on the use of hand to introducing evaluation and research instruments to remove soft decayed tissues methods. Oral health research should be and to fill the clean cavity with glass iono- encouraged to develop oral health care in mer cements. The WHO supports this ap- the community.

26 Views on oral health.pmd 214 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 215

References

1. Assessment of the role of Western den- 10. Frencken JE, Holmgren CJ, van tistry: The limits to conventional dentistry. Palenstein Helderman WH. Basic pack- In: Mautsch W, Sheiham A, eds. Promot- age of oral care. Nijmegen, The Nether- ing oral health in deprived communities. lands, WHO Collaborating Centre for Berlin, German Foundation for Interna- Oral Health Care Planning and Future tional Development, 1995. Scenarios, College of Dental Science, University of Nijmegen, 2001. 2. Dolan TA, Gooch BF, Bourque LB. Asso- ciations of self-reported dental health 11. The world health report 2002: reducing and general health measures in the risks, promoting healthy life: overview. Rand Health Insurance Experiment. Geneva, World Health Organization, Community dentistry and oral epidemi- 2002. ology, 1991, 19:1–8. 12. Petersen PE. The world oral health report 3. Oral health in America: A report of the 2003: continuous improvement of oral Surgeon General. Rockville, Maryland, health in the 21st century – the approach United States of America, Department of of the WHO Global Oral Health Health and Human Services, National Programme. Geneva, World Health Or- Institute of Dental and Craniofacial Re- ganization, 2003. search and the National Institutes of 13. Sheiham A, Watt RG. The common risk Health, 2000. factor approach: a rational basis for pro- 4. Pakhomov GN. Future trends in oral moting oral health. Community dentistry health and disease. International dental and oral epidemiology, 2000, 28:399– journal, 1999, 49:27–32. 406. 5. WHO global databank. Geneva, World 14. Ottawa Charter for Health Promotion. Health Organization, 2000. First International Conference on Health Promotion, Ottawa, 21 November 1986. 6. Primary health care report of the Interna- Geneva, World Health Organization, tional Conference on Primary Health 1986 (WHO/HPR/HEP/95.1). Care, Alma-Ata, USSR, 6–12 September 1978 jointly sponsored by the World 15. The Jakarta Declaration on Leading Health Organization and the United Na- Health Promotion into the 21st Century. tions Children’s Fund. Geneva, World Geneva, World Health Organization, Health Organization, 1978. 1997. 7. Oral health for the 21st century. Geneva, 16. Prevention of oral diseases. Geneva, World Health Organization, 1994 (WHO/ World Health Organization, 1987 (WHO ORH/Oral C21.94). offset publication, No. 103). 8. Towards a better oral health future. 17. Sulieman O. The action-oriented school Geneva, World Health Organization, health curriculum in the Eastern Mediter- 1993 (WHO/ORH/WHD/93). ranean Region. Eastern Mediterranean health journal, 1998, 4(suppl.):S130–7. 9. Oral health situation analysis in EMRO countries. Cairo, World Health Organiza- 18. The Oral Health Alliance. The Berlin tion, Regional Office for the Eastern Declaration 1992 on oral health and oral Mediterranean, 2000 (WHO/EMRO, health services in deprived communi- 2000 unpublished). ties. In: Mautsch W, Sheihm A, eds. Pro-

26 Views on oral health.pmd 215 10/25/2005, 1:24 AM 216 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

moting oral health in deprived communi- 23. Taifour D et al. Effectiveness of glass- ties. Berlin, German Foundation for Inter- ionomer (ART) and amalgam restora- national Development (DSE), 1995. tions in the deciduous dentition: results after 3 years. Caries research, 2002, 19. Fluorides and oral health. Geneva, 36:437–44. World Health Organization, 1994 (WHO Technical Report Series, No. 846). 24. Taifour D et al. Comparison between res- torations in the permanent dentition 20. Simonsen RJ. Pit and fissure sealant: produced by hand and rotary instrumen- Review of the literature. Pediatric den- tations—survival after 3 years. Commu- tistry, 2002, 24:393–414. nity dentistry and oral epidemiology, 21. Frencken JE et al. Atraumatic restorative 2002, 33:122–8. treatment (ART): rationale, technique and development. Journal of public health dentistry, 1996, 56:135–40. 22. Oral health programme. Milestone. Geneva, World Health Organization, 1998.

The objectives of the WHO Global Oral Health Programme The objectives of the WHO Global Oral Health Programme, one of the technical programmes within the Department of Chronic Dis- eases and Health Promotion, have been reoriented according to the new strategy of disease prevention and promotion of health. Greater emphasis is put on developing global policies in oral health promotion and oral disease prevention, coordinated more effectively with other priority programmes of the Department of Chronic Diseases and Health Promotion and other clusters and with external partners. Several principles form the basis for the work carried out. The WHO Oral Health Programme works towards building oral health policies on effective control of risks to oral health, based on the common risk factors approach. The focus is on modifiable risk behaviours related to diet, nutrition, use of tobacco and excessive consumption of alcohol, and hygiene. Further information on the WHO Global Oral Health Programme is available at: http:// www.who.int/oral_health/en/

26 Views on oral health.pmd 216 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 217

Report Group purchasing of pharmaceuticals and medical supplies by the Gulf Cooperation Council states T.A.M. Khoja1 and S.A. Bawazir 2

ABSTRACT An important issue in health care today is the cost of essential pharmaceuticals and medical supplies. To control the increase of health care expenses, in 1976 the Gulf Cooperation Council states began to study the idea of establishing a group purchasing programme for pharmaceuticals and medical supplies. This paper demonstrates the elements of the programme, how it works, what obstacles it faces and how other countries can profit from this experience. It also discusses the future of the group purchasing programme in the light of globalization and how the international changes under the World Trade Organization agreements will affect the programme in future.

Les achats groupés de produits pharmaceutiques et fournitures médicales par les États du Conseil de Coopération du Golfe RÉSUMÉ Le coût des produits pharmaceutiques et fournitures médicales essentiels représente aujourd’hui un problème important dans les soins de santé. Afin de maîtriser l’augmentation des dépenses des soins de santé, les États du Conseil de Coopération du Golfe ont commencé en 1976 à envisager la mise en place d’un programme d’achats groupés des produits pharmaceutiques et des fournitures médicales. Le présent article montre les éléments de ce programme, son fonctionnement, les obstacles auxquels il est confronté et la manière dont les autres pays peuvent tirer parti de cette expérience. Il examine également l’avenir du programme des achats groupés à la lumière de la mondialisation et l’impact futur sur ce programme des changements internationaux qui s’opèrent dans le cadre des accords de l’Organisation mondiale du Com- merce.

1Executive Office of the GCC States for Health Ministers, Riyadh, Saudi Arabia. 2Department of Clinical Pharmacy, College of Pharmacy, Riyadh, Saudi Arabia (Correspondence to S.A. Bawazir: [email protected]). Received: 03/09/03; accepted: 30/11/03

27 Group purchasing.pmd 217 10/25/2005, 1:24 AM 218 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction purchasing programme in the light of globalization and how the international trade An important issue facing those in charge agreements will negatively or positively im- of health care in different parts of the pact on the programme in the future. world is how to finance the necessary pharmaceuticals and medical supplies. The Council of Health Ministers for Anyone following the progress of financial the GCC funds assigned by a country to cover the Structure health care expenses of its citizens realizes The GCC countries—Bahrain, Kuwait, the increasingly large amounts of money Oman, Qatar, Saudi Arabia and the United spent. The rising costs are due to a number Arab Emirates—have a common basis of of causes, the most important of which is language, religion, geography, history, en- progress in standards of health care: mod- vironment, economy, society and heritage. ern techniques that require new and expen- A common interest in building strong rela- sive equipment and medical supplies, and tionships between these countries is there- the increasing range of new drugs and im- fore obvious. The Council of Health munizations available against disease. Ministers for the GCC was established to In 1976 AD (1396 AH), to control the coordinate the cooperation between its increase of health care expenses, the Gulf member states in all fields of health: pre- Cooperation Council (GCC) states started vention, intervention and rehabilitation. In exploring the idea of establishing a group addition, the Council aims to disseminate purchasing programme for pharmaceuti- health knowledge and improve health cals and medical supplies. The first group awareness among the general public of medical supplies competition in GCC states GCC countries. was in 1978, hospital equipment in 1982, The Executive Board carries out all sec- serum and vaccinations in 1985, chemicals retariat work for the Council and the Exec- in 1992, medical rehabilitation in 1996, and utive Director represents the Executive lastly laboratory equipment and blood Board in all communications and organiza- banks in 2001. tions. The establishment of the Executive The group purchasing programme has Board was a result of frequent meetings achieved several advantages for the mem- before 1975 among some health ministers ber states, the most important of which is in the region about the importance of coor- the decreased cost price and operation dination and cooperation among the health price, the acceleration of the purchasing ministries of Gulf countries. Consequently, process, the establishment of unified spec- the meetings became annual events under ifications, the promotion of information ex- the name Health Ministers’ Conference for change and the development of pharma- the Arab Countries in the Gulf. In 1981, the ceutical policies among the GCC states. health ministers decided to change “Con- This paper will demonstrate the ele- ference” to “Council”. The Health Minis- ments of the programme, how it works, ters’ Council of the Arab Countries in the the obstacles it faces and how other coun- Gulf now comprises the health ministers of tries can profit from this experience. We the 6 states in the GCC, and it is convened will also discuss the future of the group twice annually.

27 Group purchasing.pmd 218 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 219

Objectives History of group purchasing The objectives of the Council of Health The escalation of health care costs is one of Ministers for the GCC are as follows: the major challenges facing health care ad- • Achieving coordination and integration ministration. Several reasons have contrib- by strengthening cooperation in various uted to such escalations, including the rapid fields of health among the GCC states growth and expansion in the health care and by setting and developing policies sector, the increase in the cost of pharma- and health regulations that aim to devel- ceuticals and hospital supplies and changes op the health services in the GCC states. in the pattern of diseases. • Unification of common goals in interna- These changes are leading to the estab- tional gatherings, and strengthening re- lishment of group purchasing schemes in lations with all regional and international the health care industry. As early as 1909, organizations. the Hospital Superintendents Club of New York was addressed about the establish- • Setting up a unified drug policy for the ment of a purchasing agency for New York region that allows full control of import- hospitals. The first group was formed in ed drugs and helps establish national 1910. In 1918, the Cleveland Hospital As- drug industries that realize self-suffi- sociation began its group purchasing oper- ciency for the GCC states. ations. Between 1962 and 1974, more than • Encouraging joint studies and research, 40 groups were organized in the United and presenting recommendations and States. giving technical advice to the GCC In 1981, 5 countries—Morocco, Mau- states in the various fields of health in ritania, Algeria, Tunis and the Libyan Arab such a way that helps realization of its Jamahiriya—established a bulk purchasing objectives. group. The group announced its first ten- • Achieving coordination and integration der for 144 drugs in 1982. Unfortunately, in the field of group purchasing of the system did not develop well due to drugs, medical supplies and its expan- many logistical problems. sion to include other medical needs. • Deciding on plans, programmes and The Gulf experience joint work projects in the field of health. Group purchasing: origins and • Endorsing the general plan and the ex- development ecutive programmes related to carrying The idea of group purchasing in the Gulf out studies and research in the fields of started in February 1976 when the minis- health problems, control of endemic ters of health in the GCC states requested diseases, health education, and environ- the Executive Office to form a technical mental health sanitation in the GCC committee whose function was to: states. • Study the setting up a unified system • Organizing exchange of experiences in for registration and control of drugs, as the fields of education and medical well as preparing a directory of drugs in training in a way that realizes developing the GCC states. and unifying systems of education in • Study the possibility of unifying the faculties of medicine and health insti- process of purchasing some drugs for tutes.

27 Group purchasing.pmd 219 10/25/2005, 1:24 AM 220 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

the ministries and government institu- lence testing in the group purchasing tions. The General Secretariat was to tenders. compile a detailed programme with re- 2. The 44th Conference in 1998 agreed the gard to requesting and awarding tenders basics principles of the drug tender and importing, taking into consideration awarding and endorsed the project of the local regulations. central registration of drugs in the GCC • Study the possibility that the member states. states may make use of the drug control 3. In the 46th Conference in 1999, the ac- laboratories in both Saudi Arabia and ceptance of analysis certificates issued Kuwait. from the certified reference laboratories Between 1980 and 1989, several resolu- for registration was endorsed. It was tions were issued that aimed to support the agreed that drugs manufactured in any group purchase scheme. Gulf state be treated in terms of pricing • In 1985, the member states were called according to the same procedures for upon to secure at least 60% of their national drugs in each state. needs from all items submitted through 4. The 47th Conference in 1999 endorsed group purchasing. that the company (registered in 2 states) • In 1987, it was agreed that the member whose laboratories were assigned as states should obtain at least 60% of their reference laboratories should participate drug needs, and for locally produced in group purchasing of drugs. The cen- drugs at least 20% of their needs, tral registration system for drugs in the through group purchasing. This was to GCC states was also agreed upon. ensure that prices would not rise for the 5. In the 48th Conference in 2000, the ten- other member states. der for laboratory supplies and blood • In 1991, a consultative committee from bank supplies was introduced. The Ex- the Faculty of Pharmacy, King Saud ecutive Board was also requested to re- University was formed to study the view the basis and the mechanisms for means of developing group purchasing. awarding group purchasing tenders. It The consultative committee held a num- was reaffirmed that member states ber of meetings with representatives of should commit themselves to the previ- the member states taking part in group ously issued ministerial resolutions, i.e. purchasing to define the scientific basis to share in the group purchasing tenders for improving group purchase. no less than 60% of their annual needs. After 1991, a series of resolutions were 6. In the 49th Conference, the rules of taken at the conferences of the Council of pricing drugs in the GCC states were Health Ministers for the GCC: endorsed. 1. The 40th Conference in 1996 agreed to 7. In the 50th Conference, the principles collect 0.5% of the amount of the ten- of awarding drug tenders were en- ders awarded in the group purchasing dorsed and it was agreed that compa- tenders to establish a fund for support- nies taking part in tenders should have a ing joint research among the member local agent in each state. The regula- states. It also agreed to prohibit drug tions of qualifying companies for medi- preparations from companies that do cal supplies were endorsed. The not fulfil the conditions of bioequiva- directory of specifications of hospital

27 Group purchasing.pmd 220 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 221

supplies and the medical rehabilitation Advantages of group purchasing directory was prepared and the labora- These can be summarized as follows: tories and blood banks tender was submitted as a separate tender for the year Cost savings 2001. It was also agreed to widen par- Group purchasing is an important method ticipation in the group purchasing pro- of saving money for the ministries of health gramme to other governmental and by reducing the cost of pharmaceutical private sector non-governmental insti- products. A report on hospital purchasing tutions through the relevant ministry in and inventory management by the United each state. States General Accounting Office advocated participation in group purchasing as a Objectives of group purchasing mechanism for controlling pharmaceutical The group purchasing programme was in- costs. According to a study performed by troduced in the GCC states to realize the the GCC Executive Office in 1992, a total following objectives: of US$ 33 million was saved by the 5 GCC • Securing financial surplus through pur- states. Furthermore, more than US$ 11 mil- chasing large amounts of supplies for a lion was saved by 3 GCC states in 2001. smaller price. Transparency • Certifying companies that follow good Group purchasing leads to transparency in manufacturing practices and that are the examination of tender awards. registered according to the rules and regulations set by the Executive Board, Standardization thus ensuring a high quality of pur- Agreeing on the items for the group purchased items. chasing programme leads to standardiza- • Ensuring use of the same drugs manu- tion of the drugs used in the GCC states factured by the same company by all and allows quality supplies to be obtained in the GCC states. emergency situations. • Rapid processing and awarding of presented tenders. Labour reduction • Ensuring a continuously supply of Reductions in labour are attained through drugs, hospital supplies and equipment minimizing administrative and regulatory all year round through regular succes- burdens on the ministries of health and cut- sive deliveries. ting down the number of samples that have • Encouraging other health sectors, e.g. to be checked for quality control. specialized hospitals, to secure their needs through group purchasing. Enhancement of purchasing operations • Encouraging the policy of purchasing The key factors in the group purchasing from generic-registered companies to programme responsible for improved phar- obtain a greater financial surplus. maceutical supply in the GCC states are the reduction in the length of the procurement • Supporting the Gulf drug industry to process, the increased predictably of tim- achieve security of drug supplies in the ing and the simplification of procedures. Gulf countries.

27 Group purchasing.pmd 221 10/25/2005, 1:24 AM 222 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Information sharing • Provision of technical support to mem- Exchange and sharing of information ber states and the group purchasing among institutions, hospitals and countries committees through the consultative is very valuable. This information may be committees. clinical data, prescribing data, health edu- • Responding to any inquiries from the cation programmes, etc. states or the companies. • Collection of 0.5% of the total awarded Uniform types of drugs items as fees for each state. If there is uniformity in the drugs allowed for use in all the states, it is easier for peo- Commitments of member states ple to get the same drug in any other GCC country. • Each country should submit its needs at the time specified. Development of the drug policy in • Participation with no less than 60% of member countries items submitted and 20% of locally pro- The group purchasing programme has duced ones. contributed to establishing the following • The specifications of items and condi- progress in GCC states: tions of the tenders should be adhered • Central drug registration to. • Bioequivalence programmes • Authorizing representatives of the per- • GCC drug formulary manent committees to attend the meetings for preparing and finalizing • Good manufacturing practice awarding of tenders. • Accreditation of quality central labora- Each country has to pay the sum of tories • money indicated according to contracts • Support to the local pharmaceutical in- and purchase orders. dustry. • Notification of the Executive Board of the details of items that fail in the labora- Procedures of group purchasing tory analysis or which do not comply The mission towards group purchasing of with the required specifications. the Executive Board of the Council of Health Ministers for the GCC is as follows: • Member states have to prepare statistical studies indicating why they did not Inviting the relevant committees to • take part in the endorsed items in order meet, proposing the agenda and per- to determine the percentage of contri- forming the secretarial work. bution of each state in group purchas- • Submission of tenders from bidding ing. companies. • Notification of countries about awards Commitments of companies and sending samples of items awarded • Fulfilment of all conditions and tender to each country for delivery. documents. • Notification of the companies about • Submission of the following required items awarded and the total sum of de- bank guarantees: 1% bid bond of the livery.

27 Group purchasing.pmd 222 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 223

quotation through a local Saudi Arabian WTO Agreement on Trade-Related Aspects bank, which is returned after submis- of Intellectual Property Rights (TRIPS) as sion of the performance bond, and 5% the most far-reaching instruments that may performance bond of the awarded items influence pharmaceutical prices. Other for each state, which is returned after agreements such as the General Agreement completing delivery of items. on Tariffs and Trade (GATT), Agreement • Submission of samples for items in the on Technical Barriers to Trade (TBT), anti- tender. dumping laws, and government procurement agreements are also expected to Obstacles to group purchasing affect the programme. Several obstacles can be identified that may limit the development of the programmes. Potential effects on the programme The main concerns raised by the WTO • Rate of participation by each member agreements are the possibilities for increas- country. Some countries do not pur- es in the price of patented products, low chase all the quantities they require quality products, unfair competition for the through the programme which leads to local industry and price dumping. decreasing purchasing power. • Commitment on tender quantities. Higher prices of patented products Some members may decrease the The TRIPS agreement requires WTO amount of their original quantities sub- members to grant patent protection for a stantially after the tender award. minimum of 20 years for any invention in • Commitment on tender schedules. any branch of technology. Because of the Some members may delay their notifi- high prices of patented drugs, it is expected cation regarding their final quantities. that the total cost of drugs may gradually This results in delays in the tender increase in the next decade. award. • Lower prices at local tenders. Many Lower quality products companies offer reduced prices for The aim of the WTO agreements is to cre- their products in local tenders of some ate a liberal and open trading system in member countries. This action is aimed which business enterprises from member to keep these countries out of the pro- countries can trade with one another under gramme. the conditions of fair competition. It is ex- • Lack of participation in some tenders. pected that competition could force the Some member countries did not partici- competitors to lower the quality of their pate in all tenders offered by the pro- products in order to compete effectively. gramme. Unfair competition for local industry and price dumping Impact of globalization on the Local producers may face unfair competi- programme tion from foreign producers who can offer Several agreements of the World Trade Or- very cheap prices for their products. Usu- ganization (WTO) are expected to influence the group purchasing programme. ally foreign producers have advantages Since 1994, attention has focused on the over local producers that help them lower

27 Group purchasing.pmd 223 10/25/2005, 1:24 AM 224 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

their prices. These advantages include fence against pressures of various kinds. cheap labour and raw materials, a large lo- Purchases of this size and level are always cal market, subsidies by the government liable to interference from commercial and and strategic planning to supersede the other pressure groups. Several participants product market. in the process have stressed the importance of correct information and transparency. Potential advantages of the The involvement of many foreign partici- programme pants requires a high degree of trans- In contrast to the potential disadvantages, parency. the WTO agreements may have several advantages for the programme. The agree- Future of the group purchasing ments may lower prices, improve the programme quality of products, encourage growth in For the programme to be successful, the the local industry and improve the transpar- member countries must provide commit- ency of the system. ment and support to the programme philos- ophy. The following points must be Lower prices considered to enhance the programme. When the GCC countries fully implement 1. Strategic planning. To ensure success, the WTO agreements it is expected that member countries should employ stra- more foreign companies will compete in tegic planning processes and shift their the group-purchasing tenders. This could focus away from the acquisition cost of lead to a lower price for a large number of pharmaceutical products. products and huge saving for the countries. 2. Establishment of a common formulary. Without a strong formulary, system Improved product quality members cannot capitalize on the bene- It is expected that the price competition fits of the programme. will also lead to quality competition. Many 3. Information exchange. The information companies will invest in good quality prod- sharing and exchange is imperative in ucts with reasonable prices to compete in the process of the group purchasing the international market. The group pur- programme. chasing programme must achieve a balance 4. Purchasing volume. The aim of the pro- between quality and cost in order to serve gramme is to increase the volume of the GCC countries effectively. quantities purchased. Growth of local industry 5. Coordination. Coordination between In the short term, the local industry may pharmacy and therapeutic committees suffer some losses from foreign competi- and establishing common educational tion. However, a strong local industry can programmes. only grow in a competitive market with 6. Use of bioequivalence data. The policy carefully calculated protective measures. of generic interchange depends on the This will also lead to improved product availabilities of bioequivalence data that quality and encouragements to export. must be incorporated into the contract award decision-making process. Improved system transparency 7. Promoting pharmaceutical services and Although impressive results have been re- monitoring of drug use in member ported, the system has required a solid de- countries.

27 Group purchasing.pmd 224 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 225

8. Vendor relationships. Health care ven- grammes established by the Executive dors are becoming involved in partner- Board of the Council of Health Ministers ships with ministries of health and for the GCC states. The programme will institutions. continue to develop and new tenders will be introduced in the future. It is expected that more institutions within the member coun- Conclusions tries will join the programme. The group purchasing programme represents one of the most successful pro-

The world medicines situation This second review of the world medicines situation (first published in 1988 as The world drug situation) presents the available evidence on global production, research and development, international trade and consumption of pharmaceuticals. In addition, it draws on the most recent surveys and studies in WHO Member States to examine the state of national medicines policy. The aim is to provide an easily accessible source of information on the pharmaceutical situation at global and national levels. This report covers a wide range of different products from multiple and varied sources, prescribed, purchased and consumed in very different domestic contexts. The report does not attempt to deal in a comprehensive way with a number of key policy issues in medicines policy, such as parallel trade, intellectual property rights, counterfeiting, or corpo- rate pricing strategy, around which vigorous debate continues at both the national and international level. Whilst WHO’s concerns and policy positions are made clear at relevant points in the text, our primary aim is to provide an up-to-date set of basic information on the global medicines situation and on the current status of national medicines policies. It is hoped that these data will serve as a useful set of reference material for analysts, researchers and others concerned with the global pharmaceutical situation. This document is available on line at: http://whqlibdoc.who.int/hq/2004/WHO_ EDM_PAR_2004.5.pdf

27 Group purchasing.pmd 225 10/25/2005, 1:24 AM 226 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Report Maternal and newborn care: practices and beliefs of traditional birth attendants in Sindh, Pakistan Z. Fatmi,1 A.Z. Gulzar 2 and A. Kazi 1

ABSTRACT Maternal mortality, infant mortality and neonatal mortality are high in Pakistan where maternal health services depend upon traditional birth attendants (TBAs). We examined the practices of TBAs in Dadu district in rural Sindh from September to November 1998 by interviewing and hosting focus group discussions with 17 TBAs. Health care personnel and other important members of the community were also interviewed. TBAs worked in areas demarcated by extended families, ethnicity or geographical access and a system of seniority was observed. Only one TBA was formally trained and antenatal and postnatal care concepts, cleanliness and equipment were inadequate. Communities trusted the TBAs and remunerated them according to factors particular to each birth. TBAs need training and to be linked with the formal health sector to effect change and to decrease maternal and neonatal mortality. Soins aux mères et aux nouveau-nés : pratiques et croyances des accoucheuses traditionnelles à Sindh (Pakistan) RÉSUMÉ La mortalité maternelle, infantile et néonatale est élevée au Pakistan où les services de santé maternelle dépendent des accoucheuses traditionnelles. Nous avons examiné les pratiques des accoucheuses traditionnelles dans le district de Dadu dans la province rurale de Sindh de septembre à novembre 1998 en organisant des entretiens et des groupes de discussion avec 17 accoucheuses traditionnelles. Le personnel de soins de santé et d’autres membres importants de la communauté ont également été interrogés. Les accoucheuses traditionnelles travaillaient dans des zones délimitées par les familles élargies, l’ethnicité ou l’accès géographique et un système d’ancienneté a été observé. Seule une ac- coucheuse traditionnelle avait été formée, et les notions de soins prénatals et postnatals ainsi que la propreté et le matériel étaient insuffisants. Les communautés avaient confiance dans les accoucheuses traditionnelles et les rémunéraient en fonction de facteurs spécifiques à chaque naissance. Les accoucheuses traditionnelles ont besoin d’une formation et doivent être reliées au secteur formel de la santé pour amener un changement et réduire la mortalité maternelle et néonatale.

1Division of Public Health Practice, 2School of Nursing; Department of Community Health Sciences, The Aga Khan University, Karachi, Pakistan (Correspondence to Z. Fatmi: [email protected]). Received: 11/09/03; accepted: 19/11/03

28 Maternal and newborn.pmd 226 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 227

Introduction Methods Traditional birth attendants (TBAs) assist in Union Council Jhangara is an area of 40 km 60%–80% of all deliveries globally and radius in the district Dadu, province of even more in the rural areas of developing Sind, Pakistan, that is dependent on farm- countries [1]. In South Asia, an estimated ing. Maternal mortality is the highest con- 80% of all deliveries were attended at home tributor to death among women and TBAs and mostly by TBAs [2]. The maternal conduct most deliveries (Z. Fatmi, S. Luby, mortality ratio, the infant mortality rate and unpublished report, 1998). Hakim, or tra- the neonatal mortality rate are high in Paki- ditional healers, who emphasize the impor- stan [3–5]. Maternal health services in Pa- tance of temperament and body fluids, are kistan also depend primarily on TBAs and the most common care providers. The approximately 85% of all births occur at hakim usually prescribe herbal medicines home [3–5]. and strict diets and often recommend life- A review of TBA training in more than style changes. In most cases, treatment in- 70 countries in the past 3 decades revealed volves the application of balms or herbal that, if unsupervised and unsupported by medicines [14]. skilled personnel, TBAs were often helpless In addition to hakim, government rural when delivery complications occurred [6]. health centres provided outpatient care for At the same time, the effect of trained general medical problems. To use any of TBAs on reducing maternal and infant mor- the maternal health services of the formal tality was not significant when proper re- health care sector, women must go to the ferral systems for essential obstetric and town of Sehwan to visit a government-run neonatal care were lacking [7,8]. Re- maternity centre. Only one road links the searchers have examined the knowledge Union Council Jhangara with Sehwan and and practices of untrained TBAs [9]. Other buses run during the daytime only. programmes have also tested various in- Data were gathered from September to structional methods and curricula for train- November 1998. We interviewed 17 TBAs, ing TBAs in Nigeria and selected countries 6 community members, 2 hakim and 4 [8,10]. staff members of rural health centres. We Characteristics and practices of TBAs then hosted focus group discussions with in Pakistan during pregnancy and newborn the TBAs [15]. After verbal consent, a fe- care have been briefly studied [11,12]. male physician and our researchers con- These studies suggest that to reduce mater- ducted the interviews in the local language. nal mortality, the major challenge in Paki- A guideline was used for interviewing and stan is to link TBAs with the formal health probing and modifications or reiterations sector in the short-term [13]. Therefore, were done as the interviews progressed. our study examined the milieu and practic- We explored aspects of antenatal, natal and es of TBAs during delivery and after birth postnatal maternal care, newborn care, fol- and identified reasons for establishing, as low-up procedures for the mother and the well as potential gaps in, a referral system newborn, cost of services and referral of TBAs with the formal health care sector chains. in rural Sindh, Pakistan.

28 Maternal and newborn.pmd 227 10/25/2005, 1:24 AM 228 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

The goal of the focus group discussions however, a new blade is purchased for was to elaborate upon any discrepancies each birth to cut the umbilical cord. that arose during individual interviews and TBAs use medicines prepared and pro- also to identify factors responsible for non- vided by the hakim. Butreeh, literally “32” referral. The purpose of rural health centre in the Sindhi language, is a mixture of 32 staff member, hakim and community herbal medicines used in raw resin or syrup member interviews was to clarify concerns forms; it is used for all kinds of illnesses raised during interviews with TBAs and to during pregnancy, whether to induce con- complement and triangulate the informa- tractions, to relieve false labour pain or to tion. treat antepartum or postpartum haemorrhage. The mixture is bought from the pan- sari, the grocers who sell herbal medicines Results under the supervision of the hakim, and is All TBAs were female, married or wid- commonly known as the “drug of TBAs”. owed, with no formal schooling and of low Butreeh is applied in a crude ground form in socioeconomic status. All had learned to the vagina and sometimes is also given oral- deliver babies from senior TBAs. Their ly in its syrup form. When asked how they ages ranged from 30 to 80 years (mean age applied butreeh during labour, TBAs indi- = 49 years). Of 17 TBAs, 9 were Sindhi cated that when women had pain they put it (the major ethnic group in the area), 7 were in cotton, tied a thread around it and insert- Balochi (a minor ethnic group in Sindh that ed it in the vagina. If the baby is due (full- is in the majority in Balochistan) and 1 was term), the pain will increase; if not, the pain Saraiki (a minor ethnic group also found in will subside on its own. the Punjab, Balochistan and North-West TBAs also said that they did not cut the Frontier provinces). cord until the child and placenta were deliv- Each TBA attended to a defined area de- ered and that they cut the cord from the marcated either by her extended family, middle with the blade. ethnic group or subgroup, or geographical To remove the placenta, TBAs said they access. If more than one TBA worked in an applied pressure on the stomach; if then the area, they were either relatives or close baby still didn’t come out, they put their friends. hand inside (the womb) and pulled the baby out. When asked how they apply pressure, Maternal care they said, they applied pressure with their There was no concept of antenatal care and knee on the stomach pregnant women contact the TBAs only to In response to questions of cleanliness, confirm their pregnancies. the majority of TBAs agreed that after han- TBAs do not undertake any measures of dling the neonate they washed their hands, cleanliness for mothers, newborns or because as a human being it is dirty. They themselves. They do not wash their hands also tended to take bath and change or instruments or clean the perineal surface clothes. No mention was made regarding of the pregnant woman. When questioned, hand-washing before the delivery. one TBA responded that no specific treat- No specific postpartum care was pro- ment was practised. Any available rags in vided to the women. However, rubri, a lo- the home are used during the delivery; cally made concentrated milk product, was

28 Maternal and newborn.pmd 228 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 229

given for the first 3 days after delivery to TBAs apply oil and surma, a mixture of the mothers because of its nutritional value. ground lead and other ingredients, to the eyelid and around eyes to make the eyes Neonatal care appear big, beautiful and bright. They also TBAs also gave advice and offered help apply it to the umbilical cord of newborns. during the early neonatal period. They said Celebrations of the birth are held on the that most babies were usually all right; they newborn’s sixth day of life. Special treat- only needed massages after birth and again ment is given to the newborn’s bath and on the sixth day. They believed that no oth- dress by the TBA on that day. er special care was required for the new- TBAs stay at the mother’s home for a born. few days before and after the delivery to Practices regarding management of perform household chores such as clean- neonatal asphyxia varied; however, in most ing, cooking and washing clothes, as well cases the TBAs turned the child upside as taking care of the mother and the child. down and patted the baby’s back with the hand in the interscapular region. There was Charge for services some belief that women should not be giv- Besides cash, TBAs also accepted cattle, en too much water to drink so that the child foods like rice, wheat, sugar or ghee, does not sink in it (feel suffocated). One clothes, utensils, and in some circumstanc- TBA reported breathing into the child’s es, gold. One TBA reported that parents mouth to provide air. There was also the give about 500–700 rupees along with belief that it was God’s will that some ba- clothes and food to eat (US$ 1 = 45 ru- bies survive and some die. pees). If a son is born, then as a gesture of Almost all babies were breastfed and happiness, parents tend to give more. One hence received colostrum. Further breast- said that TBAs sometimes do not take mon- feeding was recommended by the TBAs ey at all and that whatever amount the vil- and some teach mothers how to breastfeed lagers gave them, they accepted humbly. the baby. They believed that the baby should receive only breast milk and no oth- Links to the formal health care er kind of milk, whether commercial prepa- sector ration or other animal. Some recommend TBAs were able to identify some compli- breastfeeding for more than 2 years. cated deliveries such as breech, obstructed In response to questions about which labour and haemorrhage. To manage com- prelacteals were given before breastfeeding plicated cases TBAs consult with senior and why, the TBAs said they gave give bits TBAs or take the mother to the hospital. of butter or honey. They also took eilichi When in the field, the immunization of- and misri (a sugar), ground it and filtered it ficer of the rural health centre collects in- in cloth and spoon-fed it to the newborn. formation from TBAs regarding the They believed it cleaned the intestine and number of deliveries expected in each made it soft so the baby could easily pass month. This information is passed to the the meconium without any problem. For rural health centre in charge of the area and treatment purposes, if a child had a is in turn submitted to the district health stomach-ache, they give them butter and officer of the area on a regular basis. Dur- mashed mint leaves. ing an interview, the director of a rural

28 Maternal and newborn.pmd 229 10/25/2005, 1:24 AM 230 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

health centre showed us supporting paper- any event, there is a need to increase the work and told us the number of pregnan- coverage of training of TBA. cies expected in two villages. In an effort TBAs had information regarding num- to encourage TBAs to refer complicated ber of pregnant women and deliveries in deliveries to the formal health care sector their areas and worked together as a net- providers, they receive money for referring work in these rural areas. One characteris- the cases to them. tic of this network is that TBAs tend to attend deliveries in populations defined by Community trust of TBAs ethnicity, extended families or geography. By living and serving in a specific commu- This tradition is important and should be nity throughout their entire lives, TBAs considered when the government plans have gained the trust and faith of the locals. training programme covering the different People listen to their advice and act accord- areas. There also seems to be a system of ingly. This was evident from their state- seniority with one TBA who is consulted in ments. One health care provider who lived times of complications and junior TBAs and worked in the area said that women who handle normal deliveries. There also were not usually taken to Sehwan or to the always seem to be consultations with se- doctor as TBAs were relied upon to make nior TBAs before families take women to the deliveries. They were generally only hospital for care. Therefore, identifying se- taken to the doctor when the TBA advised nior TBAs, giving them the responsibility this. Another health worker indicated that for early referral of complications and link- TBAs do not leave the women and go with ing them to the nearest maternity centre is them when they are taken to a female doc- important. Operational research should in- tor. vestigate options to accomplish this. TBAs primarily provide services during delivery and during the first week of the Discussion child’s life. Practices and beliefs surround- The sociodemographic profile of TBAs in ing maternal and newborn care are inade- our study was similar to studies of TBAs in quate. Practices are often based on other developing countries [16]. In Sindh, superstitious beliefs and do not conform all TBAs were female unlike in some coun- with recognized medical practice. There is tries where male TBAs also deliver babies no concept of antenatal care in the commu- [9]. Also in Sindh, most TBAs were unedu- nities and TBAs offer no services during cated and had some erroneous beliefs about this period, except to confirm the pregnan- pregnancy, delivery and newborn care. cy, often unreliably. TBAs count the gesta- As we expected, most TBAs were not tional period in lunar months (Muslim formally trained (16 of the 17 TBAs). calendar months); a full-term pregnancy Through the Family Health Project the gov- comprises 10 lunar months. This suggests ernment has trained TBAs in only 10 of the that TBA training programmes should refer 21 districts in Sindh, and the study area, to this system of counting for gestational the Dadu district, was not one of these. age. Our TBAs learned to deliver babies either The equipment used by TBAs during la- through their own experiences of childbirth bour is meagre and unsanitary. A common or from close relatives and senior TBAs. In herbal medicine used by TBAs is butreeh;

28 Maternal and newborn.pmd 230 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 231

however, the effects of this drug are un- fective link between the health facilities and known and need further investigation. Oth- the TBAs [18]. er herbal preparations are also used. The An appropriate referral system requires use of surma, the ground lead mixture, on a referral protocol specifying when and the umbilical cord and the baby’s eyes is where to refer. It also requires coordina- practised. There is no awareness that this tion, communication, confidence and un- could be harmful for the child. Indeed it is derstanding between the TBAs and their believed that the use of surma encourages supervisors, who are often midwives or the growth of eyebrows and increases the Lady Health Visitors, the basic health care child’s beauty. Honey and butter are also workers who provide outreach services in believed to give strength to the child and areas of Pakistan. Furthermore, requires are frequently fed to newborns. A strongly the availability of health centres and hospi- held belief is that this facilitates the easy tals with surgical facilities. passing of meconium. The TBAs have informal links to the TBAs do not understand the complica- formal health personnel of the area. When tions of pregnancy very well if at all. They TBAs refer a complicated pregnancy to the must be trained to identify complications formal health personnel, they receive mon- early so that referral can be made at the ap- etary incentives. The TBA accompanies the propriate time. woman from home and stays with her as One characteristic of the services of long as she remains in the health facility. TBAs is flexibility in service charges as Studies in other countries also suggest that they accepte any amount of money present monetary incentives for TBAs could im- in the household at the time of birth. This prove outreach services and lead to compli- might also be an important factor that leads ance by TBAs for referral of cases [19]. In women to use their services. TBAs are also fact, in 1995 it was recommended that willing to accept payment in instalments monetary compensation was needed to which is easier for poor households to pay make TBAs part of the mainstream health compared with the fixed and one-time system in Pakistan [20]. These monetary charges of the formal health-care sector. compensations could be formal fees-for- An evaluation of the TBA programme in referral. Pakistan indicated that it was unable to develop an effective link between health facil- Conclusions ities and the TBAs [17]. Past evaluations of TBA training in Pakistan identified the need The World Health Organization (WHO) for effective referrals, but no policy has yet recognizes the need at the time of birth for been formulated. For instance, the Pakistan skilled attendants such as midwives, nurses government started a training programme with midwifery education or physicians of TBAs in 1950 and abandoned it in 1970 with appropriate training and experience; because it did not result in a decrease in however, TBAs are not skilled birth atten- maternal mortality. In Sindh province, the dants [21]. Studies have shown that TBAs Sindh Dai Training Project (SDTP) was cannot prevent or treat most life-threaten- begun in 1982 and emphasized “safe and ing obstetric complications, and therefore, clean deliveries”. An SDTP evaluation some critics charge that training them is a showed that the main problem of the pro- waste of resources [22]. However, TBAs gramme was that it did not develop an ef- could be very effective if a formal link

28 Maternal and newborn.pmd 231 10/25/2005, 1:24 AM 232 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

could be made between them and the well- Operational research is needed to equipped formal health sector [23,24]. strengthen links between TBAs and formal Studies have also found that TBAs can be health care providers. In the immediate fu- trained to identify delivery complications ture, we need to train and link TBAs to the [25]. Most infant mortality is due to neona- formal health sector to decrease maternal tal deaths, and among neonatal deaths, as- and neonatal mortality in Sindh and other phyxia is the major cause [26]. Timely similar areas. In the long run, we need to intervention by a skilled attendant is vital to train midwives who will also need strong decrease neonatal deaths as well [27]. links to the communities and to the formal TBAs should be better trained to manage health sector to have any effect. neonates in this regard. Our study used a triangulation technique to complement and validate oral re- Acknowledgements ports from different sources, i.e. besides We have worked to establish a primary interviewing TBAs, we interviewed health health care system in Union Council Jhan- care personnel from the government health gara for more than 2 years prior to this facilities, hakim and other health workers study. We are indebted to primary health and community members. Furthermore, care workers Dr Nusrat Lakhiar, Ms Hami- we have been working in the area for the dah Bajar, Ms Shireen Channa, Ms Mujahi- establishment of a health care system for a da Jabbar and Mr Mehtab Hussain Jokhio couple of years prior to this study and our for their contributions to this study. We accumulated observations are also reflect- also thank the Jhangara community health ed in the results. Our paper provides quali- workers who helped lead the way in these tative research of the working scenario of remote villages. TBAs in an area of Sindh.

References

1. Alisjahbana A. How to improve training 3. Fikree FF et al. Maternal mortality in dif- of the traditional birth attendants. In: ferent Pakistani sites: ratios, clinical Kessel E, Wiknjosastro GH, Alisjahbana causes and determinants. Acta obstetri- A, eds. Maternal and infant mortality – cia and gynecologica scandinavica, closing the gap between prenatal health 1997, 76(7):637–45. services. Proceedings: Fourth Interna- 4. United Nations Children’s Fund. The tional Congress for Maternal and Neo- State of the World’s Children, 2000. New natal Health, Bandung, Indonesia. North York, Oxford University Press, 2000. Carolina, Chappel Hill Southern Print- ing, 1991:75–82. 5. Pakistan Demographic and Health Sur- vey, 1990–1991. Islamabad, Pakistan 2. Mathur HN, Sharma PN, Tain TP. Tradi- and Columbia, Maryland, National Insti- tional ways of maternal and child health tute of Population Studies and IRD/ care in Rajhastan, India. In: Aviado DM, Macro International, 1992. Del Mundo F, Cuyegkeng EI, eds. Pri- mary maternal and neonatal health: A 6. Bergstrom S, Goodburn E. The role of tra- global concern. New York, Plenun, 1983: ditional birth attendants in the reduction 473–81. of maternal mortality. In: De Brouwere V,

28 Maternal and newborn.pmd 232 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 233

Van Lerberghe W, eds. Safe motherhood 15. Verderese M, Turball L. The traditional strategies: a review of the evidence. birth attendant: A guide to her training Antwerp, Belgium, ITG Press, 2001:77– and utilization. Geneva, World Health 96. Organization, 1975 (WHO offset publication, No. 18). 7. Sibley L, Sipe TA, Koblinsky M. Does traditional birth attendant training improve 16. Egullion C. Training traditional midwives referral of women with obstetric compli- in Manicaland, Zimbabwe. International cations: a review of the evidence. Social journal of gynaecology and obstetrics, science and medicine, 2004, 59(8): 1985, 23(4):287–90. 1757–68. 17. Kamal IT. Impact of training on the prac- 8. Itina SM. Characteristics of traditional tices of traditional birth attendants birth attendants and their beliefs and (TBAs). Journal of the Pakistan Medical practices in the Offot Clan, Nigeria. Bul- Association, 2001, 51(7):239–40. letin of the World Health Organization, 18. Tinker AG. Improving women’s health in 1997, 75(6):563–7. Pakistan. Washington DC, The World 9. Matthews MK et al. Training traditional Bank, Human Development Network, birth attendants in Nigeria—the pictorial health, nutrition and population series, method. World health forum, 1995, 16(4): 1998. 409–13. 19. Utilization of trained traditional birth at- 10. Sibley L, Armbruster D. Obstetric first aid tendants. Bangladesh, International in the community—partners in safe Centre for Diarrhoeal Disease Re- motherhood. A strategy for reducing ma- search, 1994, 13(131):5–6. ternal mortality. Journal of nurse mid- 20. Greene P. Discussion: TBA training wifery, 1997, 42(2):117–21. programme. International journal of 11. Jafarey SN. Characteristics and prac- gynaecology and obstetrics, 1995, tices of traditional birth attendants (Dais) 50(suppl. 2):141–3. a preliminary survey. Journal of the Paki- 21. Revised 1990 estimates of maternal stan Medical Association, 1981, 31(12): mortality: a new approach by WHO and 288–91. UNICEF. Geneva, World Health Organi- 12. Islam A, Malik FA. Role of traditional birth zation, 1990 (Document WHO/ FRH/ attendants in improving reproductive MSM/96.11; UNICEF/PLN/96.1). health: lessons from the family health 22. Rosario S. Traditional birth attendants in project, Sindh. Journal of the Pakistan Bangladeshi villages: Cultural and so- Medical Association, 2001, 51(6):218– ciological factors. International journal of 22. gynaecology and obstetrics, 1995, 13. Kamal IT. The traditional birth attendant: 50(suppl. 2):145–52. a reality and a challenge. International 23. Leedam E. Traditional birth attendants. journal of gynaecology and obstetrics, International journal of gynaecology and 1998, 63(suppl. 1):S43–52. obstetrics, 1985, 23(4):249–74. 14. Zaman R. Hakims and Vaids the most 24. Marshall MA, Buffington ST. Life saving popular doctors in Bangladesh, but they skills manual for midwives, 3rd ed. know little about AIDS. Aids analysis Washington DC, American College of Asia, 1996, 2(1):12–3. Nurse Midwives, 1998.

28 Maternal and newborn.pmd 233 10/25/2005, 1:24 AM 234 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

25. Minden M, Levitt MJ. The right to know: based neonatal care. Journal of women and their traditional birth atten- perinatology, 2005, 25(suppl. 1):S92– dants. In: Murray SF, eds. Midwives and 107. safer motherhood. London, Mosby, 27. Kumar R et al. Reproductive health 1996. behaviour of rural women. Journal of the 26. Bang AT et al. Neonatal and infant mor- Indian Medical Association, 1995, 93(4): tality in the ten years (1993 to 2003) of 129–31. the gadchiroli field trial: effect of home-

Health service planning and policy-making: a toolkit for nurses and midwives The purpose of this toolkit, consisting of 7 booklets, is to provide nurses and midwives with tools to effectively participate in and influence health care planning and policy-making. The tool-kit has been designed for use by any nurse or midwife who has an interest in advocating for change in their work environment. This includes chief nurses and midwives who are in a position to impact policy decision-making at a country level, as well as clinical nurses and midwives advocating for community development or other programmes at the operational level. Each module in the tool-kit has objectives and reference lists and provides definitions, examples, exercises and accompanying guiding exercise sheets. Additional supportive materials are provided in Information and Skills Sheets. Further information on this publication can be obtained from WHO Press:http://www.who.int/bookorders/anglais/home1. jsp?sesslan=1

28 Maternal and newborn.pmd 234 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 235

Short communication Injuries caused by falls from trees in Tehran, Islamic Republic of Iran M. Zargar,1 A. Khaji1 and M. Karbakhsh1

ABSTRACT Although falls from trees are rare, the consequences can be severe. Over 13 months in 6 hospitals in Tehran 49 (0.57%) of 8500 registered trauma patients had fallen from trees. The mean age was 31 years (range 6 to 74 years) and 21 (42.9%) were aged < 20 years. Home and recreational/sports grounds were the most common sites of trauma. Most falls occurred during leisure time; only 3 (6.1%) were occupation-related. Eleven patients (22.4%) sustained a fracture of the spinal column; 2 of them became permanently paraplegic and 2 others who had severe thoracic vertebral fractures and paraplegia on first observation were transferred to other hospitals for treatment. Due to the severity of this type of injury, especially among the young, effective preventive efforts may be necessary.

Les traumatismes causés par une chute d’un arbre à Téhéran (République islamique d’Iran) RÉSUMÉ Bien qu’il soit rare qu’une personne tombe d’un arbre, les conséquences peuvent être graves. Sur une période de 13 mois, 49 (0,57 %) des 8500 patients traumatisés enregistrés dans 6 hôpitaux à Téhéran étaient tombés d’un arbre. L’âge moyen était de 31 ans (extrêmes : 6-74 ans) et 21 (42,9 %) avaient moins de 20 ans. La maison et les terrains de jeu/de sport étaient les lieux où les traumatismes survenaient le plus fréquemment. La plupart des chutes survenaient pendant les loisirs ; seules 3 chutes (6,1 %) étaient liées au travail. Onze patients (22,4 %) ont subi une fracture de la colonne vertébrale ; 2 d’entre eux ont été atteints de paraplégie motrice permanente et 2 autres qui présentaient des fractures des vertèbres dorsales et une paraplégie lors de la première observation ont été transférés dans d’autres hôpitaux pour traitement. En raison de la gravité de ce type de traumatismes, notamment chez les jeunes, des efforts efficaces de prévention peuvent être nécessaires.

1Sina Trauma and Surgery Research Centre, Sina Hospital, Tehran, Islamic Republic of Iran (Correspondence to M. Zargar: [email protected]). Received: 29/06/03; accepted: 18/02/04

29 Injuries caused.pmd 235 10/25/2005, 1:24 AM 236 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Introduction tients in emergency rooms and wards and filled in the questionnaire. The collected Falls are a common cause of accidents in data included: patients’ demographic de- the world [1–4]. Yet falls from trees are un- tails; how and where the injury occurred; common, and so little is known about the length of hospital stay and source of reim- characteristics of this type of injury. It is an bursement for care; clinical signs on ad- occupational injury among traditional farm- mission; anatomical site of injury; Injury ers in developing countries [5,6] and most Severity Score and Glasgow Coma Scale reports of injuries due to falling from trees score; prehospital care and procedures per- are from the tropics [7]. formed during hospitalization. This kind of trauma can cause severe Descriptive analysis was performed us- injuries leading to paraplegia and quadriple- ing SPSS software, version 10.0 for Win- gia. For example, in parts of the eastern dows. Injuries were grouped into standard forest belt of southern Nigeria, falls from categories based on ICD-9 [8]. trees account for more spinal cord injuries than injuries due to road traffic accidents [7]. The present study aimed to describe Results the characteristics and clinical conse- During the 13 months, 8500 trauma pa- quences of falls from trees among patients tients were admitted to the study hospitals; admitted to 6 hospitals in Tehran, the capi- 3100 had been injured due to a fall. Falling tal city of the Islamic Republic of Iran. from trees accounted for 49 (1.6%) of all injuries from falls. Methods All the falls from trees were reported to be accidental. Most of these patients sus- The study group were patients admitted to tained injuries during leisure time; only 3 emergency rooms in 6 general hospitals in injuries (6.1%) were occupation-related. Tehran from 23 August 1999 to 21 Septem- Homes (40.8% of cases) and recreational/ ber 2000. These hospitals are situated in sports grounds (32.7%) were the most different regions of the city: Imam Hossain common places where trauma occurred. hospital in the western part, Moayeri and The height of falling could not be estab- Sina hospitals in the central part, Hafteteer lished accurately, but the maximum falling hospital in the southern part, and Imam height was estimated to be about 7 metres. Khomeini and Shariatti hospitals in the east- Most falls had occurred in the 3rd month of ern part. Trauma patients who were hospi- the summer (23 August–21 September) talized for more than 24 hours and had and on Tuesdays with 12 (24.0%) cases sustained injuries within 7 days prior to ad- occurring on this day. The type of tree mission were included in the study. Patients could not always be identified, but it was who suffered from burns and poisonings most often a fruit tree such as mulberry or were excluded, since there are other spe- apple. cialized referral centres for these patients in The mean age of patients was 31 years Tehran. (range 6 to 74 years) and 21 (42.9%) were The data was collected using a ques- under 20 years of age (Table 1). Further- tionnaire, designed and validated at the Sina more, 17 patients (34.7%) were students. Trauma Research Centre. During the Most patients (46, 93.9%) were male. study, trained physicians visited trauma pa-

29 Injuries caused.pmd 236 10/25/2005, 1:24 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 237

Table 1 Distribution of patients falling from Table 2 Distribution of patients falling from trees by age and sex trees according to anatomical site of injury

Age (years) Males Females Site of injury No. of No. No. patients

< 10 10 0 Head and neck Mandibular fracture 2 11–20 11 0 Skull vault fracture 1 21–30 8 0 Dental fracture 1 31–40 5 0 Total 4 41–50 3 1 Spinal column, thorax and pelvic 51–60 2 0 Spinal fracture with neurological deficit 4 61–70 4 2 Spinal fracture with no neurological 71–80 3 0 deficit 7 80+ 0 0 Ilium fracture 1 Acromioclavicular dislocation 1 Rib fracture with pneumothorax 1 Total 14 Upper limbs All patients had a normal Glasgow Humerus fracture (lower end) 3 Coma Scale score of 15 except for 1 pa- Colle’s fracture 8 tient with a reduced score of 14. The ma- Colle’s fracture with ulnar and jority of the patients had suffered only mild radius fracture 1 or moderate injuries; the Injury Severity Ulnar and radius fracture (shaft) 3 Score was rated mild for 31 cases Ulnar and radius fracture (lower end) 1 (63.3%), moderate for 15 (30.6%) and se- Radius fracture (body) & wrist dislocation 1 vere for 3 (6.1%). No internal organ inju- Thumb fracture 1 ries were detected among the patients. The Total 18 most common anatomical site of injury Lower limbs was the upper limbs, followed by the lower Femur fracture (neck) 3 limbs and spinal column, respectively (Ta- Femur fracture (pretrochanteric) 4 ble 2). Eleven patients (22.4%) sustained a Femur fracture (body) 1 fracture of the spinal column, 4 of them Femur fracture (lower end) 2 became permanently paraplegic (2 who had Patellar fracture 4 severe thoracic vertebral fractures were Tibia fracture 2 transferred to other hospitals for treat- Calcaneus fracture 2 ment). One patient had Colle’s fracture in Toe fracture 1 both arms. Total 19 The average length of hospitalization All sitesa 55

was 6.1 days (range 1 to 37 days), and no aSome patients had more than 1 injury. patients were admitted to the intensive care unit. Only 20 patients (40.8%) had a source

29 Injuries caused.pmd 237 10/25/2005, 1:24 AM 238 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

of reimbursement for hospital care. Among umn injuries. Among them, 25 (70%) had the patients, 28 (57.1%) had received pre- fallen from trees. The majority of the cases hospital care. were young male farmers and injuries to the thoracolumbar area were prominent [6]. Discussion Compared with other reports, a few Disability resulting from trauma is a major points of interest have emerged from this problem for the health services especially study. Children and adolescents were when it involves the young [9]. Meanwhile, prominent (42.9% of cases). The most patients with spinal cord injuries carry a common sites of accidents were home and heavy burden, as they will be afflicted for recreational/sports grounds (40.8% and years, sometimes for the rest of their lives. 32.7% of cases respectively) and most falls In addition to the acute morbidity, people occurred during leisure time. Only in 3 cas- with spinal cord injuries will suffer from es (6.1%) was the injury considered occu- chronic problems as they age, such as pational. Spinal column factures were the chronic bladder infections and postural most serious injuries among our patients problems [10,11]. (22.4% of cases). As mentioned earlier, there are few pub- Due to the accidental nature of these in- lished reports about falling from trees. Eb- juries and high morbidity rates among ong reported 60 patients who had fallen young people, especially for severe spinal from trees in Nigeria [5]. Males were cord injuries, effective education pro- prominent; 78% of cases were farmers and grammes in the Islamic Republic of Iran only 22% were schoolchildren and the for high risk groups seems necessary. The most important injuries were vertebral most common reason for climbing trees fractures. In addition, all 6 patients who among children and adults was to pick the had sustained fracture dislocation of the fruit, so restrictions on planting fruit trees cervical spine suffered from permanent in public places might reduce the tempta- quadriplegia [5]. Okonkwo described 72 tion to climb trees. Planting fruit trees patients who were treated for spinal cord which do not grow tall, such as cherry and injuries, among whom the most common black cherry, or covering the ground below cause of injury was falling from trees (41% trees with soft ground or grass may reduce of the patients). The 40–49 year age group the severity of injuries in those who fall was at the greatest risk [7]. Toe reported from trees. 36 patients who had sustained spinal col-

References 1. Ferrera PC et al. Trauma management: 3. Injury prevention and control: an epide- an emergency medicine approach. St miological study of injuries in the area of Louis, Missouri, Mosby, 2001. Municipal Corporation of Delhi. World Health Organization Regional Office for 2. Kobusingye O, Guwatudde D, Lett R. In- South-East Asia, 2003. jury patterns in rural and urban Ungada. Injury prevention, 2001, 7(1):46–50.

29 Injuries caused.pmd 238 10/25/2005, 1:25 AM Eastern Mediterranean Health Journal, Vol. 11, Nos 1/2, 2005 239

4. Zargar M, Modaghegh MH, Rezaishiraz (ICD-9). Geneva, World Health Organi- H. Urban injuries in Tehran: demography zation, 1978. of trauma patients and evaluation of 9. Greaves I, Porter K, Ryan J. Trauma care trauma care. Injury, 2001, 32(8):613–7. manual. New York, Oxford University 5. Ebong WW. Falls from trees. Tropical and Press, 2001. geographical medicine, 1978, 30:63–7. 10. Cavigelli A, Fischer R, Dietz V. Socio- 6. Toe T. Spinal injuries in Rangoon, Burma. economic outcome of paraplegia com- Paraplegia, 1978, 16:118–20. pared to lower limb amputation. Spinal cord, 2002, 40:174–7. 7. Okonkwo CA. Spinal cord injuries in Enugu, Nigeria—preventable accidents. 11. McColl MA et al. International differ- Paraplegia, 1988, 26:12–8. ences in ageing and spinal cord injury. Spinal cord, 2002, 40:128–36. 8. International classification of diseases and related health problems, version 9

Accidents and injuries Each year unintentional injuries account for more than 400 000 deaths globally, the majority in children and adolescents. Most of these occur in low- and middle-income countries. Many of those who survive these injuries suffer life-long disabling health consequences. Information on WHO’s work in the prevention of accidents and injuries is available at: http://www.who.int/ceh/risks/cehinjuries/ en/

29 Injuries caused.pmd 239 10/25/2005, 1:25 AM 240 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

Short communication Characterization of Leishmania isolated from unhealed lesions caused by leishmanization S.M.H. Hosseini1, G.R. Hatam 2 and S. Ardehali 2

ABSTRACT Leishmania parasites were isolated after 13 and 8 years from the unhealed lesions of 2 soldiers who had been immunized against leishmaniasis during the war between Iraq and the Islamic Republic of Iran. Isoenzyme characterization on these isolates using 11 enzyme systems was carried out and the results were compared with the enzyme profiles of the original isolates of L. major used for leishmanization. Minor enzymatic differences in glucose-6-phosphate dehyrognase and phosphoglucomutase were observed but otherwise the strains appeared unchanged.

Caractérisation des Leishmania isolés sur des lésions non cicatrisées causées par la leishmanisation RÉSUMÉ Des parasites Leishmania ont été isolés après 13 et 8 années sur des lésions non cicatrisées de deux soldats qui avaient été vaccinés contre la leishmaniose durant la guerre entre l’Iraq et la République islamique d’Iran. On a procédé à la caractérisation isoenzymatique de ces isolats en utilisant 11 systèmes enzymatiques et comparé les résultats avec les profils enzymatiques des isolats originaux de L. major utilisés pour la leishmanisation. Des différences enzymatiques mineures ont été observées pour la glucose- 6-phosphate-déshydrogénase et la phosphoglucomutase, sinon les souches paraissaient inchangées.

1Department of Immunology, Razi Vaccine and Serum Research Institute, Shiraz, Islamic Republic of Iran. 2Department of Parasitology and Mycology, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Islamic Republic of Iran (Correspondence to G.R. Hatam: [email protected]). Received: 23/06/03; accepted: 17/02/04

Very sadly, Professor Ardehali passed away shortly after preparation of this report.

30 Characterization of Leishmania.pmd 240 10/25/2005, 1:25 AM Eastern Mediterranean Health Journal, Vol. 11 Nos 1/2, 2005 241

Introduction The second case was a 30-year-old man from Karaj, a suburb of Tehran, who Leishmanization with live virulent promas- had also resided in a war area (Khouzestan) tigotes has at times been carried out to pro- for 1 year after immunization. He had detect high-risk people against cutaneous veloped a lesion at the site of inoculation 8 leishmaniasis in endemic areas. During the years before this study, which had not war between Iraq and the Islamic Republic healed even after various treatments. The of Iran, immunization against leishmaniasis patient was apparently healthy on physical was carried out for more than 2 million Ira- examination carried out by a physician. In nian soldiers over several years. Soldiers direct smear preparation, no Leishmania were inoculated in the left deltoid muscle amastigotes were found, but promastigotes with 2–3 × 105 live Leishmania major pro- were obtained in NNN culture media 3 days mastigotes. Follow-up studies on individu- after inoculation in the medium. als who had received the immunization The isolates recovered from NNN me- indicated that 2%–3% had a lesion for over dium were further propagated in RPMI a year, and half of those did not heal after 1640 (Sigma, United Kingdom) supple- many years even with treatment [1,2]. mented with 15% fetal calf serum (Gibco- During a countrywide cutaneous leish- RBL, United Kingdom), penicillin/ maniasis survey in the Islamic Republic of streptomycin and L-glutamine. The meth- Iran from 1998 to 1999 on 407 patients ods and techniques for further cultivation, [3,4], we had opportunity to characterize preparation of cell extracts and enzyme isolates (isoenzymatically and with mono- staining were those described by Evans clonal antibodies) from the long-term le- [5]. Electrophoresis on polyacrylamide gel sions of 2 individuals who had been was carried out according to the methods immunized during the war. We were inter- described by Davis [6]. ested to find out the probable phenotypic The World Health Organization refer- changes in the parasite during its long presence strains were L. major (MHOM/SU/ ence in a human host. 73/5ASKH), L. tropica (MHOM/SU/71/ K27), L. donovani (MHOM/IN/80/DD8), Methods and L. infantum (MHOM/TN/80/IPTI). L. major (MHOM/IR/72/Nadim5), the strain The first case from whom Leishmania was used for leishmanization, was also applied. isolated was a 30-year-old man from the In order to determine if the original north of the country who had resided in the strain was still pathogenic, the organisms war region (Khouzestan) for 1 year after recovered from cryopreservation were had he had been immunized. He had a single propagated in RPMI 1640 medium supple- lesion at the inoculation site which had not mented with 15% fetal calf serum and L- healed in spite of extensive treatment for 13 glutamine. During the stationary growth years; otherwise he was an apparently phase, the parasites were harvested. Then healthy individual. No Leishmania amastig- 2 million promastigotes were inoculated otes were found in the direct smear of the into Balb/c mice at the base tail. Three lesion, but Leishmania promastigotes were weeks after inoculation, nodules appeared recovered in NNN medium, 3 days after which ulcerated a few days later indicating inoculation in the medium. the continued pathogenicity of the original

30 Characterization of Leishmania.pmd 241 10/25/2005, 1:25 AM 242 La Revue de Santé de la Méditerranée orientale, Vol. 11, No 1/2, 2005

strain. Amastigotes from the lesions that an Rf value of 0.62 was absent in the isolate developed were cultured on NNN medium compared with the original strain. In and further propagation was performed in ELISA, this isolate also only reacted with RPMI 1640 as above. monoclonal antibody specific for L. major, The enzymes studied were malate dehy- i.e. XLVI – 5B80 (T-1). drogenase (EC 1.1.1.37), malic enzyme (EC 1.1.1.40), phosphoglucomutase (EC 4.5.2.2), glucose-6-phosphate dehydroge- Discussion nase (G6PD) (EC 1.1.1.49), glucose phos- To our knowledge this is the first study at- phate isomerase (EC 5.3.1.9), nucleoside tempting to characterize the isolates from hydrolase (EC 3.2.2.1), isocitrate dehydro- individuals who have been immunized genase (EC 1.1.1.42), superoxide dismu- against leishmaniasis. The results obtained tase (EC 1.1.5.1.1), estrase (EC 3.1.1.1) demonstrated the following points. and alanine aminotransferase (EC 2.6.1.2). • The lesion that persisted in the 2 immu- The monoclonal antibodies used were nized cases was caused by the original XLVI-5B80 (T-1) anti L. major, IS2-2B4- strain used for leishmanization and was All anti L. tropica and LXXV III-2ES-A8 not due to concomitant infection with (D-2) anti L. infantum which were kindly another wild type of parasite. supplied to us by Dr F. Modabber from WHO, Geneva. Freshly prepared whole • The zymodeme of the original strain promastigotes of Leishmania isolates were used for leishmanization had not used as the source of antigen in the changed greatly in spite of residing in enzyme-linked immunosorbent assay host tissues for a long period of time. (ELISA) test system [7,8]. • All immunized individuals received the same dose of the same parasites under similar conditions, but in some cases Results the lesion that developed tended to per- sist. This therefore suggests that host The zymodeme of the case 1 isolate was factors play a role in the development the same as original strain used for leish- and establishment of lesions. However, manization, except for the G6PD enzyme more studies are necessary to investi- system for which the isolate had an extra gate the effect of host factors. band with an Rf value of 0.36. The zymo- gram of this enzyme system for the strain used for leishmanization showed 3 bands Acknowledgements with Rf 0.06, 0.12 and 0.3 respectively. In ELISA this isolate reacted only with mono- The authors would like to thank the United clonal antibody specific for L. major, i.e. Nations Development Programme/World XLVI – 5B80 (T-1). Bank/WHO Special Programme for Re- The zymodeme of the case 2 isolate search and Training in Tropical Diseases, was also the same as the original strain the Centre for Development of Clinical Re- used for leishmanization, except for the search of Nemazi Hospital, and Dr Davood PGM enzyme system in which a band with Mehrabani for editorial assistance.

30 Characterization of Leishmania.pmd 242 10/25/2005, 1:25 AM Eastern Mediterranean Health Journal, Vol. 11 Nos 1/2, 2005 243

References

1. Nadim A et al. Control of zoonotic 5. Evans DA et al., eds. Handbook on isola- cutaneous leishmaniasis by mass tion, characterization and cryopre- leishmanization in hyperendemic area servation of Leishmania. Geneva, United of Isfahan, Iran. Medical journal of the Is- Nations Development Programme/ lamic Republic of Iran, 1988, 2(2):113– World Bank/World Health Organization, 4. 1989. 2. Tropical disease research: progress 6. Davis BJ. Disc electrophoresis. II. Method 1975–94, highlights 1993–94, twelfth and application to human serum protein. programme report of the UNDP/World Annals of the New York Academy of Sci- Bank/WHO Special Programme for Re- ences, 1964, 121:404–27. search and Training in Tropical Diseases 7. Jaffe C, McMahon-Pratt D. Monoclonal (TDR). Geneva, World Health Organiza- antibodies specific for Leishmania tion, 1995:135–46. tropica. Characterization of antigens as- 3. Ardehali S. Characterization of Leishma- sociated with stage and species-specific nia isolated in Iran. I. Serotyping with determinant. Journal of immunology, specific monoclonal antibodies. Acta 1983; 131:1987–93. tropica, 2000, 75:301–7. 8. Jaffe C et al. Production and character- 4. Hatam GR, Hosseini SMH, Ardehali S. ization of species-specific monoclonal Evaluation of two biphasic culture media antibodies against Leishmania dono- for isolation of Leishmania parasites vani for immunodiagnosis. Journal of im- from cutaneous ulcers in Iranian pa- munology, 1989, 133:440–7. tients. Medical journal of the Islamic Re- public of Iran, 1997, 11(2):178–9.

30 Characterization of Leishmania.pmd 243 10/25/2005, 1:25 AM