Eastern Mediterranean Health Journal La Revue De Santé De La Méditerranée Orientale

Contents

Acknowledgement of the work of Dr Mohammad Haytham Khayat...... 103 Editorial

Child mortality in the Eastern Mediterranean Region: challenges and opportunities...... 104 Research articles

Factors affecting defaulting from DOTS therapy under the national programme of tuberculosis control in Alexandria, Egypt...... 107 Factors associated with patient delay in accessing pulmonary tuberculosis care, Gezira State, Sudan, 2009...... 114 Eastern Mediterranean Mycobacterium tuberculosis spoligotypes circulating in the Lebanese population: a retrospective study...... 119 Health education and peer leaders’ role in improving low vaccination coverage in Akre district, Health Journal Kurdistan region, Iraq...... 125 High incidence of hepatitis B infection after treatment for paediatric cancer at a teaching hospital in Baghdad...... 130 La Revue de Santé de Risk factors for human brucellosis in northern Jordan...... 135 Volume 19 Number 2 la Méditerranée orientale Public awareness of and support for smoke-free legislation in Turkey: a national survey using the lot quality sampling technique...... 141 Blood lead level among Palestinian schoolchildren: a pilot study...... 151 Birth weight and risk of childhood acute leukaemia...... 156 Mothers and children in the Region External cephalic version for breech presentation at term: predictors of success, and impact on Addressing the health of mothers and children is at the heart the rate of caesarean section...... 162 of the Millennium Development Goals (MDGs). With only February 2 years to go to reach the targets set down in MDG 4 and 5, Depression among a group of elders in Alexandria, Egypt...... 167 WHO, together with UNICEF AND UNFPA, recently held a

Frequency of impaired glucose tolerance and diabetes mellitus in subjects with fasting blood glucose 2013 high-level meeting to consider strategies to accelerate progress below 6.1 mmol/L (110 mg/dL)...... 175 towards achieving MDGs 4 and 5 in all countries of the Region. Interleukins 12 and 13 levels among beta-thalassaemia major patients...... 181 Reviews

Médicaments et allaitement maternel : évaluation du risque médicamenteux chez le nourrisson...... 186 Diagnostic et évaluation de l’hépatite virale C chez l’hémodialysé...... 192 Report

Evaluation of national tuberculosis surveillance system in Afghanistan...... 200

املجلد التاسع عرش / عدد Volume 19 / No. 2 2 شباط / فرباير February / Février 2013

Cover 19-2.indd 1 2/20/2013 2:31:21 PM Members of the WHO Regional Committee for the Eastern Mediterranean املجلة الصحية لرشق املتوسط Afghanistan . Bahrain . Djibouti . Egypt . Islamic Republic of Iran . Iraq . Jordan . Kuwait . Lebanon هى املجلة الرسمية التى تصدر عن املكتب اإلقليمى لرشق املتوسط بمنظمة الصحة العاملية. وهى منرب لتقديم Libya . Morocco . Oman . Pakistan . Palestine . Qatar . Saudi Arabia . Somalia . South Sudan السياسات واملبادرات اجلديدة ىف اخلدمات الصحية والرتويج هلا، ولتبادل اآلراء واملفاهيم واملعطيات الوبائية Sudan . Syrian Arab Republic . Tunisia . United Arab Emirates . Yemen ونتائج األبحاث وغري ذلك من املعلومات، وخاصة ما يتعلق منها بإقليم رشق املتوسط. وهى موجهة إىل كل أعضاء املهن الصحية، والكليات الطبية وسائر املعاهد التعليمية، وكذا املنظامت غري احلكومية املعنية، واملراكز املتعاونة مع منظمة الصحة العاملية واألفراد املهتمني بالصحة ىف اإلقليم وخارجه.

EASTERN MEDITERRANEAN HEALTH JOURNAL البلدان أعضاء اللجنة اإلقليمية ملنظمة الصحة العاملية لرشق املتوسط 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, con‑ cepts, epidemiological data, research findings and other information, with special reference to the Eastern Mediterranean Region. األردن . أفغانستان . اإلمارات العربية املتحدة . باكستان . البحرين . تونس . ليبيا . مجهورية إيران اإلسالمية It addresses all members of the health profession, medical and other health educational institutes, interested NGOs, WHO Col‑ ...... اجلمهورية العربية السورية اليمن جنوب السودان جيبويت السودان الصومال العراق عُ ام ن فلسطني قطر الكويت لبنان مرص .laborating Centres and individuals within and outside the Region املغرب . اململكة العربية السعودية LA REVUE DE SANTÉ DE LA MÉDITERRANÉE ORIENTALE EST une revue de santé officielle publiée par le Bureau régional de l’Organisation mondiale de la Santé pour la Méditerranée orientale. Elle offre une tribune pour la présentation et la promotion de nouvelles politiques et initiatives dans le domaine des ser‑vices de santé ainsi qu’à l’échange d’idées, de concepts, de données épidémiologiques, de résultats de recherches et d’autres Membres du Comité régional de l’OMS pour la Méditerranée orientale informations, se rapportant plus particulièrement à la Région de la Méditerranée orientale. Elle s’adresse à tous les professionnels de la santé, aux membres des instituts médicaux et autres instituts de formation médico-sanitaire, aux ONG, Centres collabora‑ Afghanistan . Arabie saoudite . Bahreïn . Djibouti . Égypte . Émirats arabes unis . République islamique d’Iran teurs de l’OMS et personnes concernés au sein et hors de la Région. Iraq . Libye . Jordanie . Koweït . Liban . Maroc . Oman . Pakistan . Palestine . Qatar . République arabe syrienne Somalie . Soudan . Soudan du Sud . Tunisie . Yémen

EMHJ is a trilingual, peer reviewed, open access journal and the full contents are freely available at its website: http://www/emro.who.int/emhj.htm

EMHJ is abstracted/indexed in the Index Medicus and MEDLINE (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, Scopus and the Index Medicus for the WHO Eastern Mediterranean Region (IMEMR). Correspondence

Editor-in-chief ©World Health Organization 2013 EMHJ All rights reserved WHO Regional Office for the Eastern Mediterranean P.O. Box 7608 Disclaimer The designations employed and the presentation of the material in this publication do not imply the expression of any opinion Nasr City, Cairo 11371 whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of Egypt its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border Tel: (+202) 2276 5000 lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products Fax: (+202) 2670 2492/(+202) 2670 2494 does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar Email: [email protected]/[email protected] nature that are not mentioned. All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either express or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. The named authors alone are responsible for the views expressed in this publication.

Subscriptions and Distribution Enquiries regarding subscriptions and distribution of the print edition of EMHJ should ISSN 1020-3397 be addressed to: Printing and Marketing of Publications at: email: [email protected]; tel: (+202) 2276 5000; fax: (+202) 2670 2492 or 2670 2494

Permissions Cover designed by Diana Tawadros Requests for permission to reproduce or translate articles, whether for sale or Internal layout designed by Emad Marji and Diana Tawadros non-commercial distribution should be addressed to Printed by WHO Regional Office for the Eastern Mediterranean EMHJ at: [email protected]

Cover 19-2.indd 2 2/20/2013 2:31:21 PM Eastern Mediterranean La Revue de Santé de Health Journal la Méditerranée orientale

املجلد التاسع عرش عدد Vol. 19 No. 2 • 2013 • 2 Contents

Acknowledgement of the work of Dr Mohammad Haytham Khayat...... 103 Editorial Child mortality in the Eastern Mediterranean Region: challenges and opportunities Zulfiqar A. Bhutta, MBBS, PhD, FRCP, FRCPCH, FCPS, FAAP ...... 104 Research articles Factors affecting defaulting from DOTS therapy under the national programme of tuberculosis control in Alexandria, Egypt M. Nour El-Din, T. Elhoseeny and A.M.M.A. Mohsen...... 107 Factors associated with patient delay in accessing pulmonary tuberculosis care, Gezira State, Sudan, 2009 E.Y. Mohamed, S.M. Abdalla, A.A. Khamis, A. Abdelbadea and M.A. Abdelgadir...... 114 Mycobacterium tuberculosis spoligotypes circulating in the Lebanese population: a retrospective study N. Bedrossian, M. Hamze, A.K. Rahmo, A. Jurjus, J. Saliba, F. Dabboussi and W. Karam...... 119 Health education and peer leaders’ role in improving low vaccination coverage in Akre district, Kurdistan region, Iraq M.A. Abdul Rahman, S.A. Al-Dabbagh and Q.S. Al-Habeeb...... 125 High incidence of hepatitis B infection after treatment for paediatric cancer at a teaching hospital in Baghdad M.F. Al-Jadiry, M. Al-Khafagi, A.F. Al-Darraji, R.M. Al-Saeed, S.F. Al-Badri and S.A. Al-Hadad...... 130 Risk factors for human brucellosis in northern Jordan M.N. Abo-Shehada and M. Abu-Halaweh...... 135 Public awareness of and support for smoke-free legislation in Turkey: a national survey using the lot quality sampling technique B. Cakir, T. Buzgan, S. Com, H. Irmak, E. Aydin and C. Arpad...... 141 Blood lead level among Palestinian schoolchildren: a pilot study A.F. Sawalha, R.O. Wright, D.C. Bellinger, C. Amarasiriwardean, A.S. Abu-Taha and W.M. Sweileh...... 151 Birth weight and risk of childhood acute leukaemia A. Gholami, S. Salarilak, S. Hejazi and H.R. Khalkhali...... 156 External cephalic version for breech presentation at term: predictors of success, and impact on the rate of caesarean section O.A. Hussin, M.A. Mahmoud and M.M. Abdel-Fattah...... 162 Depression among a group of elders in Alexandria, Egypt H.M. El Kady and H.K. Ibrahim...... 167 Frequency of impaired glucose tolerance and diabetes mellitus in subjects with fasting blood glucose below 6.1 mmol/L (110 mg/dL) S.H. Khan, A. Ijaz, S.A. Raza Bokhari, M.S. Hanif and N. Azam...... 175 Interleukins 12 and 13 levels among beta-thalassaemia major patients R.A. Hashad, N.A. Hamed, M.M. El Gharabawy, H.A. El Metwally and M.G. Morsi...... 181 Reviews Médicaments et allaitement maternel : évaluation du risque médicamenteux chez le nourrisson Y. Khabbal, S. Zaoui et Y. Cherrah...... 186 Diagnostic et évaluation de l’hépatite virale C chez l’hémodialysé A. Bahadi, O. Maoujoud, Y. Zejjari, A. Alayoud, K. Hassani, D. Elkabbaj et M. Benyahia...... 192 Report Evaluation of national tuberculosis surveillance system in Afghanistan K.M.I. Saeed, R. Bano and R.J. Asghar...... 200

Book 19-2.indb 103 2/21/2013 11:39:15 AM Dr Ala Alwan, Editor-in-chief Editorial Board Professor Zulfiqar Bhutta Professor Mahmoud Fahmy Fathalla Professor Rita Giacaman Dr Ziad Memish Dr Sameen Siddiqi Professor Huda Zurayk Editors Fiona Curlet, Guy Penet Eva Abdin, Alison Bichard, Marie-France Roux Graphics Suhaib Al Asbahi, Hany Mahrous, Diana Tawadros Administration Nadia Abu-Saleh, Yasmine El Sakhawy, Yasmeen Sedky, Heba ElBayoumi

Book 19-2.indb 720 2/21/2013 11:39:15 AM املجلد التاسع عرش املجلة الصحية لرشق املتوسط العدد الثاين

Acknowledgement of the work of Dr Mohammad Haytham Khayat

Following the appointment of a new editorial board and advisory panel for the Eastern Mediterranean Health Journal, I would like to pay tribute to Dr Mohammad Haytham Khayat, who served as Editor-in Chief for 19 years. Dr Khayat played a key role in creation of EMHJ in 1995 as a new peer-reviewed journal for the Eastern Mediterranean Region, through the amalgamation of two journals, the Health Services Journal and the EMRO Epidemiological Bulletin. Under his leadership, EMHJ has developed from just two issues a year, to 12 issues a year, with submissions increasing from just a few to over 700. It has evolved to become a leading journal in the Region, growing both in popular- ity and reputation, and becoming a flagship publication of the Regional Office. In recent years the journal has also been published online and is now one of the most frequently visited sections of the Regional Office web site.As of Volume 19, 2013, EMHJ has been included in Thomson Reuters’ Science Citation Index Expanded, the Social Science Citation Index, online at the Web of Knowledge. This will lead to the journal having an impact factor by 2015. All this could not have been achieved without Dr Khayat’s guidance and steadfast support. Dr Khayat is well known within and outside the World Health Organization Regional Office for the Eastern Mediterranean for his long and valuable contribution to public health in the Region, for his dedicated promotion of the use of Arabic and the dissemination of public health information in both English and Arabic. In the latter regard, supported by Dr Hussein A. Gezairy, former Regional Director, he founded the Regional Arabic Programme, which subsequently became the WHO Arabic Programme. He was a key figure in the incep- tion and production of many Regional Office publications in his 30-year career with the Organization. These include The unified medical dictionary, now in its 4th edition, which aims to make basic medical terms accessible to Arabic-speaking doctors, dentists and other medical and scientific professionals, and the educational series heT right path to health: health education through religion which, in a region where religion plays such an important role in people’s lives, aimed to draw on religion to help promote healthy lifestyles and practices. I would like to take this opportunity to thank Dr Khayat for his dedication to EMHJ in his 19 years as Editor-in-Chief. It is, and will continue to be, sincerely valued and appreciated.

Dr Ala Alwan WHO Regional Director for the Eastern Mediterranean

شكر وتقدير لعمل الدكتور محمد هيثم الخياط

البد يل بعد تعيني جملس جديد للتحرير وهيئة استشارية جديدة للمجلة الصحية لرشق املتوسط، أن َّأتقدم بخالص الشكر والتقدير للدكتور حممد هيثم اخلياط الذي شغل رئاسة حترير املجلة 19 ًعاما، وقد كان له دور رئييس يف إنشاء هذه املجلة عام 1995 لتكون جملة جديدة يف إقليم رشق املتوسط َّحمكمة بمراجعة الزمالء، بعد دمججملتني مها جملة اخلدمات الصحية والنرشة الوبائية للمكتب اإلقليمي لرشق املتوسط. وقد شهدت املجلة حتت قيادة األستاذ الدكتور اخلياط ًتطورا ًكبريا، فبعد 700 12 أن كان يصدر منها عددان ًسنويا، أصبح يصدر منها ًعددا ًسنويا، كام زاد عدد األوراق التي َّتقدم للنرش فيها من بضع آحاد لتزيد عن ، ومالبث أن َّتطورت أكثرلتصبح املجلة الرئيسية يف اإلقليم، فزادت شعبيتها وسمعتها، وأضحت من املطبوعات الذائعة الصيت يف املكتب اإلقليمي. ثم ما أن بدأ النرش اإللكرتوين للمجلة عىل صفحات اإلنرتنت، حتى أصبحت من أكثر املواقع ًاستقطابا للزوار ملوقع املكتب اإلقليمي. ومع صدور املجلد 19 يف هذا العام 2013 فإن املجلة ستدرج ضمن فهرس االقتباسات العلمية َّاملوسعلتوماس رويرت، وفهرس االقتباسات العلمية االجتامعية، واملعارف املنشورة عىل صفحات اإلنرتنت؛ ِّوسيؤدي ذلك إىل أن يسجل للمجلة عامل التأثري بحلول عام 2015،ومل يكن ألي من ذلك أن يتم لوال اجلهد الدؤوب واإلرشاد احلكيم لألستاذ الدكتور اخلياط.

والدكتور حممد هيثم اخلياط معروف ضمن املكتب اإلقليمي ملنظمة الصحة العاملية لرشق املتوسط وخارجه، ملا ُعرف عنه من إسهام ِّمقي ومديد يف الصحة العمومية يف اإلقليم، وإلخالصه العميق يف تعزيز استخدام اللغة العربية ونرش املعلومات يف جمال الصحة العمومية ٍبكل من اللغتني العربية واإلنكليزية. ويف هذا املجال، فقد َّأسس األستاذ الدكتور اخلياط، ٍبدعممن الدكتور حسني عبد الرزاق اجلزائري، املدير اإلقليمي السابق، الربنامج العريب اإلقليمي الذي رسعان ما َّحتولإىل الربنامج العريب ملنظمة الصحة العاملية. ولقد كان األستاذ الدكتور اخلياط الشخصية األكثر ً يفمتيزا إعداد وإنتاج الكثري من املنشورات واملطبوعات التي صدرت عن املكتب اإلقليمي خالل فرتة حياته املهنية يف املنظمة التي استمرت ثالثني ًعاما؛ ونذكر منها املعجم الطبي َّاملوحد الذي وصل اآلنإىل طبعته الرابعة، والذي يستهدف إتاحة املصطلحات الطبية والصحية األساسية لألطباء وألطباء األسنان ولغريهم من أرباب املهن الطبية والعلمية باللغة العربية، كام نذكر منها السلسلة التثقيفية اهلدي الصحي: التثقيف الصحي من خالل تعاليم الدين، وقد استهدفت هذه السلسلة االعتامد عىل الدين لتعزيز أنامط احلياة الصحية واملامرسات الصحية يف هذا اإلقليم الذي ِّيؤدي فيه الدين ًدورا ًهاما يف حياة الناس. 19 َوأ َو ُّ دأن أغتنم هذه املناسبة َّألتقدمإىل األستاذ الدكتور اخلياط بالشكر واالمتنان عىل تفانيه يف خدمة املجلة الصحية لرشق املتوسط طيلة ًعاما من خدمته هلا ًرئيسا للتحرير؛ فذلك اجلهد جدير ًحقا بالتقدير واالحرتام، وسيبقى كذلك عىل الدوام.

الدكتور عالء الدين العلوان املدير اإلقليمي ملنظمة الصحة العاملية لرشق املتوسط

103

Book 19-2.indb 103 2/26/2013 8:41:28 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Editorial Child mortality in the Eastern Mediterranean Region: challenges and opportunities Zulfiqar A. Bhutta, MBBS, PhD, FRCP, FRCPCH, FCPS, FAAP 1

In the year 2000, 189 heads of state by seven countries: Afghanistan, Iraq, under‐five population and mortality signed the United Nations (UN) Morocco, Pakistan, Somalia, Sudan rates in these countries, which represent Millennium Declaration consisting and Yemen. In terms of MDG5, five over half (54%) of the total population of eight Millennium Development countries are not on track to achieve its in the Region, that EMR is unlikely to Goals (MDGs) with a focus on an targets: Djibouti, Iraq, Pakistan, Soma- achieve MDG4. integrated global agenda for poverty lia and Tunisia [5]. We also know a fair bit about the reduction, health, education and hu- In EMR 12.2% of the population major causes of child mortality in the man development [1]. Although many comprises children under 5 years of Region and its determinants. Neonatal countries have made progress, corre- age, while 20% are women of child- deaths represent a substantial propor- sponding gains in targets for child sur- bearing age. An estimated 923 000 tion (43%) of under‐five mortality in vival (MDG4) and maternal survival children under 5 years die every year the Region and are clearly linked to the (MDG5) are slow. It is estimated that in the Region and under-five mortality poor status of maternal health among to achieve MDG4 targets, the annual has decreased by only 41% since 1990, the poorest sections of the population rate of decline in child mortality should from 99 deaths per 1000 live births to [7]. In four countries, namely Afghani- be around 4.4%, but the corresponding 58 per 1000 live births in 2011. EMR stan, Pakistan, Somalia and Yemen, less rate of decline since 1990 has only aver- ranks fifth among the six WHO regions than 50% of deliveries were attended aged 1.5% with wide variations between in terms of progress in reducing the by skilled health personnel in 2010 [8]. regions (Figure 1). Similarly, to achieve under‐five mortality, ahead only of the Across the Region, only 31% of mar- MDG5, the annual rate of maternal African Region [5]. ried women use modern contracep- mortality decline needed to be 5.5% Progress across the Region for child tives and 35% of women and newborns but global figures since 1990 indicate a survival remains uneven. Six countries are delivered without a skilled attendant decline of only 1.9% [2,3]. (Oman, United Arab Emirates, Bah- at childbirth. Beyond the neonatal pe- Around 39 000 women of child- rain, Lebanon, Qatar and Saudi Arabia) riod, four disorders – diarrhoea, pneu- bearing age still die every year in the in EMR have already achieved reduc- monia, malaria and measles – are the World Health Organization Eastern tions in the under-five mortality rate major causes of post-neonatal death Mediterranean Region (EMR) as a well beyond the targets of MDG4. The [9]. result of pregnancy-related complica- highest decrease in under‐five mortal- These findings reinforce the im- tions (sepsis, haemorrhage, eclampsia, ity in the world (72%) between 1990 perative for action. Without intensive obstructed labour and unsafe abor- and 2010 did occur in the Northern and accelerated action, especially in tions) [4]. Mortality rates are particu- African countries of the Region. Of the those countries contributing to the bulk larly high among pregnant adolescents total under-five deaths in EMR, 82% of under-five deaths, the Region will who account for a significant propor- occur in six countries (Afghanistan, not join the rank of developed coun- tion of first births. Overall the maternal Pakistan, Somalia, South Sudan, Sudan tries and those that have climbed the mortality ratio (MMR) in EMR fell by and Yemen). Of these, three countries, development ladder. In recent months, 53% between 1990 and 2010, and six Sudan, Afghanistan and Pakistan, are commendable progress has been countries have already achieved MMR among the 10 countries with highest made in raising awareness in the Re- beyond MDG5 targets and seven more child mortality in the world, a fact that gion and building the case for relevant are on track. Over 90% of the burden of has been recognized for almost a dec- evidence-based actions for change. A maternal deaths in the Region is shared ade [6]. It is largely because of the high high-level meeting of health ministers,

1Noordin Noormahomed Sharieff Professor & Founding Chair, Division of Women & Child Health, The Aga Khan University, Karachi 74800, Pakistan.

104

Book 19-2.indb 104 2/21/2013 11:39:16 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

2000–2010 Developing regions 1990–2000 Developed regions

Oceania

Latin America and the Carribbean

Northern Africa

Sub-Saharan Africa

South-Eastern Asia

Southern Asia

Western Asia

Eastern Asia

0 1 2 3 4 5 6 7 % decline

Figure 1 Trends in decline in under-5 mortality: Source: [1]

% decline technical experts, academia and lead- slums and marginalized on the basis with the clear recognition that regular ing UN agencies was held in Dubai of race or ethnicity. The monitoring monitoring of gaps and remedial meas- in January, 2013 and culminated in a process for coverage of essential inter- ures are critical to achieving equitable consensus “Dubai Declaration” and a ventions across the continuum of care coverage. There is clear evidence that renewal of pledges for accelerated ef- spearheaded by Countdown for 2015 greatest gains in equity for coverage are forts. Four pillars were identified as is one such initiative that could be used in countries that have targeted scaling critical to progress. at the country level to assess progress up universal coverage and deployed First, countries in the Region need and coverage [8]. Third, while interven- measures to reach the poor [12]. This to have reliable and robust informations exist, there is a need to focus on cannot be done without information tion for action [10]. Not only does delivery strategies and mechanisms for and targeted measures to address ser- this relate to mortality trends and scaling up coverage in the short term. vices in conflict-affected populations. overall status, but disaggregated data These include innovations for demand The challenge of wide differentials highlighting differentials and vulner- creation, removal of financial barri- in maternal and child survival and life able populations within countries is ers as well as service delivery through expectancy in EMR is real and a moral needed. Second, it was noted that community platforms and health work- imperative to act for a Region rich in evidence-based interventions existed ers who can reach the marginalized history, resources and a unifying faith but were failing to reach many who and poorest sectors of the population that places social justice and rights of needed them most. These include not [11]. Finally, none of the above can women and children at its core. We only the poorest sections of the popula- be achieved without robust measures need to rise to the challenge and achieve tion resident in difficult to access rural for monitoring and accountability. such gains in equity for future genera- outposts but also those living in urban This must be done transparently and tions, within a generation.

References

1. The Millennium Development Goals Report 2012. New York, United Nations Children’s Fund, 2012 (http://www.childinfo. United Nations, 2012. org/files/Child_Mortality_Report_2012.pdf, accessed 3 Feb- 2. Lozano R et al. Progress towards Millennium Development ruary 2013). Goals 4 and 5 on maternal and child mortality: an updated 4. Lozano R et al. Global and regional mortality from 235 causes systematic analysis. Lancet, 2011, 378:1139–1165. of death for 20 age groups in 1990 and 2010: a systematic 3. United Nations Inter-agency Group for Child Mortality Estima- analysis for the Global Burden of Disease Study 2010. Lancet, tion. Levels and trends in child mortality: Report 2012. New York, 2012, 380:2095–2128.

105

Book 19-2.indb 105 2/21/2013 11:39:17 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

5. Saving the lives of mothers and children: rising to the challenge. 9. Fischer-Walker C et al. Global burden of childhood diarrhoea Background document for the High Level Meeting on Saving and pneumonia: leading causes of child mortality. Lancet, 2013 the Lives of Mothers and Children: Accelerating Progress To- (in press). wards Achieving MDGs 4 and 5 in the Region, Dubai, United 10. Every woman, every child: from commitments to action: the first Arab Emirates, 29–30 January 2013 (http://applications.emro. report of the independent Expert Review Group (iERG) on Infor- who.int/docs/High_Level_Exp_Reg_doc_2013_EN_14811.pdf, mation and Accountability for Women’s and Children’s Health. accessed 3 February 2013). Geneva, World Health Organization, 2012. 6. Bhutta ZA et al. Child health and survival in the Eastern Medi- 11. Lassi ZS, Haider BA, Bhutta ZA. Community-based intervention terranean region. British Medical Journal, 2006, 333:839–842. packages for reducing maternal and neonatal morbidity and 7. Liu L et al. Global, regional, and national causes of child mor- mortality and improving neonatal outcomes. Cochrane Data- tality: an updated systematic analysis for 2010 with time trends base of Systematic Reviews, 2010, (11):CD007754. since 2000. Lancet, 2012, 379:2151–2161. 12. Victora CG et al. How changes in coverage affect equity in 8. Countdown to 2012 Maternal, Newborn and Child Survival. Build- maternal and child health interventions in 35 Countdown to ing a future for women and children. The 2012 Report. Geneva, 2015 countries: an analysis of national surveys. Lancet, 2012, World Health Organization and UNICEF, 2010. 380:1149–1156.

106

Book 19-2.indb 106 2/21/2013 11:39:17 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Factors affecting defaulting from DOTS therapy under the national programme of tuberculosis control in Alexandria, Egypt M. Nour El-Din,1 T. Elhoseeny 2 and A.M.M.A. Mohsen 3

العوامل ِّاملؤثرةعىل إخفاق املعاجلة القصرية َاألمدحتتاإلرشاف املبارش )دوتس( ضمن الربنامج الوطني ملكافحة السل يف اإلسكندرية، مرص مصطفى نور الدين حسن، تغاريد عباس احلسيني، عبد املحسن حممد عبد املحسن

اخلالصـة: هتدف هذه الدراسة للحاالت والشواهد غري املتوافقة إىل ُّالتعرف عىل العوامل ِّاملؤثرةيف اإلخفاق يف املعاجلة ضمن الربنامج الوطني 187 53 ملكافحة السل يف اإلسكندرية، مرص، وقد َ ى أجرالباحثون مراجعات للسجالت ومقابالت ّمنظمة مع من الذين أخفقت معاجلتهم، ومع 54 13 من الشواهد الذين اختاروهم ًعشوائيا. وأوضح التحليل الوحيد املتغ ِّير أن ًعامال من بني ًعامال استقصاها الباحثون قد ترابطت ُتراب ًطا ُي ْع َت ُّد به 0.16 ًإحصائيا مع اإلخفاق، وبعد إجراء التحوف اللوجستي التدرجيي، َبقيت مخسة عوامل يف النموذج، وهي: العمر األصغر )معدل األرجحية (، واإلقامة يف منطقة ريفية )معدل األرجحية (،12.9 وأوقات االنتظار الطويلة )معدل األرجحية (،5.81 وسوء التواصل بني املريض والطبيب )معدل 3.62 3.06 األرجحية (، واخلوف من ُّترسب املعلومات )معدل األرجحية (. أما األسباب التي ذكرها الذين أخفقت معاجلتهم فتشمل املسافة الطويلة للوصول إىل العيادة، وعدم مالءمة أوقات العيادات، وأوقات االنتظار الطويلة. أما العوامل الرئيسية لإلخفاق يف الربنامج الوطني ملكافحة السل يف اإلسكندرية، مرص فقد كانت عوامل تتعلق باخلدمة، وهي عوامل يمكن تصحيحها وحتسينها.

ABSTRACT This unmatched case–control study aimed to identify factors affecting default from therapy under the national programme of TB control in Alexandria, Egypt. Record reviews and structured interviews were made with 57defaulters and 187 randomly selected controls. Univariate analysis showed 13 out of 54 factors investigated were significantly associated with defaulting and, after stepwise logistic regression, 5 factors remained in the model: younger age (adjusted OR= 0.16), rural area of residence (OR = 12.9), long waiting times (OR = 5.81), poor physician–patient communication (OR = 3.06) and fear of information leakage (OR = 3.62). Reasons cited by defaulters included long distance to the clinic, unsuitable clinic times and long waiting times. The main factors associated with defaulting from the national programme of TB control in Alexandria, Egypt were service-related factors, which are amenable to improvement.

Facteurs liés à l'abandon du traitement DOTS dans le cadre du programme national de lutte antituberculeuse à Alexandrie (Égypte)

RÉSUMÉ La présente étude cas-témoins non appariés visait à identifier les facteurs influant sur l'abandon du traitement dans le cadre du programme national de lutte antituberculeuse à Alexandrie (Égypte). Les dossiers médicaux de 57 patients ayant abandonné le traitement et de 187 témoins sélectionnés aléatoirement ont été examinés puis des entretiens structurés ont été menés. Une analyse univariée a démontré que 13 facteurs étudiés sur 54 étaient significativement associés à un abandon et, après une analyse de régression logistique par étapes, cinq facteurs ont été dégagés à partir du modèle : un âge plus jeune (OR ajusté = 0,16), un lieu de résidence en milieu rural (OR = 12,9), un long temps d'attente (OR = 5,81), une mauvaise communication entre le médecin et le patient (OR = 3,06) et la crainte de la divulgation d'informations (OR = 3,62). Une longue distance pour atteindre l'établissement de soins, des horaires d'ouverture peu pratiques et de longs temps d'attente comptaient parmi les raisons citées par les patients ayant abandonné le traitement. Les principaux facteurs associés à un abandon du traitement dans le cadre du programme national de lutte antituberculeuse à Alexandrie (Égypte) relevaient du domaine des services se prêtant aux améliorations.

1e-School of Health and Environmental Studies, Hamdan Bin Mohammed e-University, Dubai, United Arab Emirates. 2Hospital Administration Division, Department of Health Administration and Behavioural Sciences, High Institute of Public Health, Alexandria, Egypt. 3National Programme of Tuberculosis Control, Egyptian Ministry of Health, Cairo, Egypt. Received: 01/10/11; accepted: 21/02/12

107

Book 19-2.indb 107 2/21/2013 11:39:17 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction infection) and service-related (long was required to demonstrate with 95% waiting times for a consultation, trans- certainty and with a power of 80% that Despite being a curable disease tuber- portation difficulties to the service) unemployment was a statistically sig- culosis (TB) remains a serious public [1,2,10,11]. Several studies have identi- nificant risk factor for default. To select health problem in Egypt, as it is world- fied that quality of communication be- controls, a systematic random sample wide [1–3]. Revised estimate of TB tween patients and health care workers was used to select patients from the incidence in Egypt published by the is also an important motivating factor register of TB patients who attend to World Health Organization (WHO) for completion of treatment [12–14]. collect their treatment. Every second in 2009 reported 21 cases per 100 000 The present study aimed to identify patient was selected to complete the population per annum. Of these cases factors affecting defaulting from DOTS sample size of controls (187 controls) 9 per 100 000 were sputum-positive, therapy among Egyptian TB patients [16]. which is the most dangerous source in Alexandria who started their anti-TB of infection [4]. After establishment of treatment under the NTP-Egypt. Data collection the national TB control programme in Data collection lasted from 1 February Egypt (NTP-Egypt) [5], the country to 30 April 2010. Data were collected has succeeded in meeting the World Methods using a structured interview in Arabic Health Organization’s global targets. In language and a record review. The inter- Study design and sample 2009 the case detection rate of positive view was developed by the researchers cases in Egypt was 72% (global target The study was conducted at 7 govern- based on a literature review of factors is 70%) and treatment success rate was ment-run chest disease dispensaries affecting defaulting among TB patients. 87% (global target is 85%) [4]. belonging to 6 districts in Alexandria, A pilot study was carried out and 2 items A major contributor to both treat- under the control of Egypt Health Af- were added based on its results. The ment failure and the rise of multidrug- fairs. At the start of the study in January total number of investigated factors resistant TB is inadequate and incom- 2010, the total number of patients on included in the final version of the in- plete treatment [6]. Default is defined the TB registries of these dispensaries terview schedule was 54. The following by the WHO as treatment interruption was 564. An unmatched case–control factors were included in the analysis: study design was used. Cases (default- of 2 consecutive months or more, and • Sociodemographic and economic factors. is often used synonymously with drop- ers) were defined as patients who had failed to collect medication for more age; sex; marital status; education; out from treatment before completion residence; crowding index; monthly than 2 consecutive months after the [7]. In 1996, the NTP-Egypt took income; household possessions; ow- date of the last attendance during the action to tackle low adherence to TB ing a private house. treatment and began implementation course of treatment [14]. Controls of the DOTS programme [8]. After (non-defaulters) were defined as pa- • History and habits. Period between this the situation improved; official data tients who continued their treatment symptoms and start of treatment; BCG vaccination; exposure to other published by NTP-Egypt showed that without defaulting during the past 3 TB patients; smoking; alcohol intake. the average rate of defaulting in Egypt months. in 2008 was 4% and that Alexandria All defaulters were recorded in a • Signs and symptoms. General weak- governorate had the same rate as the specially designed register. The total ness; loss of appetite; loss of weight; national average (4%) [9]. number of defaulters at the start of the low-grade fever; night sweating; Factors affecting the rate of default study was 57; all of them were included severe cough for 2 weeks or more; among TB patients have been studied in the study. Owing to the small number haemoptysis; chest pain; sputum ex- in countries other than Egypt and it of cases, an unequal sample size design amination result. seems that variables associated with was selected with a ratio of 3 cases to 10 • Patient’s knowledge. Disease signs and defaulting from TB programmes dif- controls. The sample size was calculated symptoms; methods of disease transfer according to the characteristics based on the assumption that unem- mission; duration of TB treatment. of the setting. Studies in Brazil, South ployment is an important risk factor for • Facility-related factors. Convenience of Africa, Russia and Thailand showed default. The sample size was calculated clinic hours; waiting time; conveni- that significant factors included socio- using the Fleiss formula with continu- ence of waiting place; getting medica- economic (unemployment, monetary ity correction factor [15]. Assuming an tions from different dispensary than resources, homelessness, history of odds ratio of 2.5, and unequal sample that of initial diagnosis; availability of imprisonment), clinical (side-effect, size of cases and controls (ratio of 1:3), a transportation to dispensary; trans- alcoholism, use of illegal drugs, HIV sample size of 57 cases and 187 controls portation cost; travelling time to dis-

108

Book 19-2.indb 108 2/21/2013 11:39:17 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

pensary; frequency of physician visits; At the end of the interview default- the association of different factors with physician tells patient about next visit ers were asked to list the most important defaulting from TB treatment. In case of date; rapport between clinic staff and reasons for defaulting from the pro- sparse data, the Fisher exact probability patients; respect and caring by physi- gramme using an open-ended question. was used as indicated. Unadjusted odds cian; any monetary payments at the Defaulters were approached by ratio (OR) and confidence intervals clinic; provision of initial medical one of the researchers and each social (CI) were presented to illustrate the examination on first contact; provi- worker at the 7 chest diseases dispensa- magnitude of effect of different factors sion of medical examination at each ries. Interviews were conducted at the on defaulting from TB treatment. Sig- encounter. patient’s home or one of the dispensa- nificant factors in univariate analysis • Physician–patient communication. ries. Controls were interviewed when were included in a stepwise logistic re- Physician listens to patient com- they attended to collect their treatment. gression analysis. The model fit was as- plaints; explains the current medi- Following patients’ interviews a review sessed using the omnibus test of model cal condition; explains the expected coefficients. It tests if the model with was made of medical records of both complications; explains the duration the predictors is significantly different cases and controls to collect clinical of treatment; explains the expected from the model with only the intercept. data such as date of starting treatment complications of treatment; explains Two-tailed P-value was reported and and initial symptoms. In addition, social the results of investigations. statistical significance was established worker records were reviewed to collect • Availability of home at P < 0.05. Other factors. sensitive sociodemographic data such care; availability of emotional sup- as income and housing condition. port; patient’s feeling of improvement with treatment; patient’s belief Consent was taken from each pa- Results that TB is not a curable disease; patient before participation. Patient iden- Univariate analysis tient being ashamed of the disease; tification data were kept confidential. patient’s belief that DOTS represents Out of 54 investigated factors, only Statistical analysis a burden of any kind; patient’s fear of those that had significant associations information leakage; patient’s accept- Statistical analysis was carried out with defaulting from NTP-Egypt are ance of medications (taste, size of using SPSS, version 16. Pearson chi- presented here. Table 1 shows that 2 tablets, number of tablets). squared test was carried out to assess sociodemographic factors (age and

Table 1 Univariate analysis of sociodemographic and clinical factors associated with defaulting from tuberculosis treatment Sociodemographic and clinical Defaulters Non-defaulters P-value OR (95% CI) factors (n = 57) (n = 187) No. % No. % Age (years) < 30a 13 22.8 31 16.6 30– 16 28.1 64 34.2 0.03 0.60 (0.26–1.39) 40– 20 35.1 39 20.9 1.22 (0.53–2.84) 50 + 8 14.0 53 28.3 0.36 (0.13–0.97) Area of residence Urbana 22 38.6 143 76.5 Squatter 24 42.1 41 21.9 < 0.001 3.80 (1.94–7.47) Rural 11 19.3 3 1.6 23.8 (6.15–92.2) Period between symptoms & treatment (weeks) 1–3a 27 4 7. 4 128 68.5 4+ 30 52.6 59 31.5 0.004 2.41 (1.32–4.41) Cough Mild or moderatea 1 1.8 23 12.2 Severe 56 98.2 164 87.8 0.02 7.85 (1.04–59.5)

aReference category. OR = odds ratio; CI = confidence interval.

109

Book 19-2.indb 109 2/21/2013 11:39:17 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

area of residence) were significantly to default (OR = 9.33 and OR = 4.33 factors measuring physician–patient different between defaulters and non- respectively). Three significant physi- communication [variance inflation fac- defaulters using univariate analysis. A cian–patient communication factors tor (VIF) > 2.5], these variables were higher percentage of defaulters than were identified Table( 2). Defaulting combined into a single variable (physi- non-defaulters were found among pa- from the national programme of TB cian–patient communication). Out of tients aged < 30 years (22.8% and 16.6% control was more likely to occur among 11 investigated variables, 5 variables respectively). The highest percentage of patients who reported “physicians don’t remained in the last step: age group, area defaulters (42.1%) lived in squatter ar- listen to complaints” (OR = 16.3), “phy- of residence, long waiting times, poor eas, while most of the controls (76.5%) sicians don’t explain TB complications” physician–patient communication and lived in urban areas. When compared (OR = 3.55) and “physicians don’t fear of information leakage (Table 4). with patients residing in the urban area, explain treatment side-effects” (OR = Patients aged 50+ years were 6 times less patients living in the rural area were 23.8 6.65). likely to default from TB treatment than times more likely to default while those Significant patient perception fac- patients aged < 30 years (adjusted OR = living in a squatter area were 3.8 times tors are illustrated in Table 3. Patients 0.16, 95% CI: 0.03–0.46). Patients who more likely to default than those living who did not feel better with treatment lived in rural areas showed the highest in urban areas. and who had a negative perception of likelihood of defaulting from TB treat- Significant clinical factors included the ability of TB treatment to comment (adjusted OR = 12.9). Patients the period between symptoms and pletely cure the disease defaulted more who feared information leakage and treatment and severity of cough. Pa- than those who answered positively to reported poor physician–patient com- tients with a long duration between both questions (ORs = 4.82 and 11.5 munication were 3 times more likely to symptoms and start of treatment were respectively). Patients feeling shame default from TB treatment (adjusted 2.41 times more likely to default than about TB and patients fearing informa- ORs = 3.62 and 3.06 respectively). patients with a short duration between tion leakage were more likely to default Defaulters’ reported reasons symptoms and start of treatment. Pa- than those who answered negatively to tients presented with severe cough were both questions (ORs = 0.45 and 0.39 Defaulters reported one or more rea- 7.85 times more likely to default from respectively). sons for defaulting from the TB DOTS treatment than those presenting with programme, amounting to a total of 135 mild or moderate cough. Multivariate regression reasons. Unsuitable opening time at the Factors related to the health ser- analysis clinic was the most frequently cited rea- vice, including facility-related factors Using univariate analysis significant fac- son for defaulting, accounting for over and physician–patient communication, tors associated with defaulting from TB a quarter of the reasons (28.2%). The were compared between defaulters and treatment (13 factors) were included next most common reasons were the non-defaulters. Patients who reported in a stepwise multivariate logistic re- distance of clinic from home, long wait- unsuitable clinic opening times and gression. Owing to the presence of ing times before examination and being long waiting times were more likely multicolinearity between 3 significant unaware of the need to complete the

Table 2 Univariate analysis of health service factors associated with defaulting from tuberculosis (TB) treatment Service factors Defaulters Non-defaulters P-value OR (95% CI) (n = 57) (n = 187) No. % No. % Facility Unsuitable clinic times 21 36.8 11 5.9 0.001 9.33 (4.14–21.0) Long waiting times 36 72.2 53 28.3 0.001 4.33 (2.32–8.10) Physician–patient communication Physicians don’t listen to complaints 15 26.3 4 2.1 < 0.001 16.3 (5.16–51.8) Physicians don’t explain TB complications 6 10.5 6 3.4 0.04 3.55 (1.10–11.5) Physicians don’t explain treatment side-effects 15 26.3 10 5.3 < 0.001 6.65 (2.65–15.1)

OR = odds ratio; CI = confidence interval.

110

Book 19-2.indb 110 2/21/2013 11:39:17 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 3 Univariate analysis of patient perception factors associated with defaulting from tuberculosis treatment Patient perceptions Defaulters Non-defaulters P-value OR (95% CI) (n = 57) (n = 187) No. % No. % Felt better with treatment Yes a 41 71.9 173 92.5 No 16 28.1 14 7. 5 < 0.001 4.82 (2.18–10.7) Treatment cured TB completely Yes a 48 84.2 184 98.4 No 9 15.8 3 1.6 < 0.001 11.5 (2.99–44.1) Feel shame about TB Yes a 43 75.4 108 57.8 No 14 24.6 79 42.2 0.02 0.45 (0.23–0.87) Fear of information leakage Yes a 35 61.4 72 48.5 No 22 38.6 115 61.5 0.002 0.39 (0.21–0.72)

aReference category. OR = odds ratio; CI = confidence interval.

treatment; each accounted for 12.6%. likelihood of defaulting, especially those Reasons cited for defaulting by our cases Feeling better after initial treatment residing in rural areas, who showed showed the importance of facility fac- ranked third (11.9%). the highest adjusted odds ratio (OR tors such as unsuitable clinic times and = 12.9). Profiles of high-burden coun- long waiting times before examination. tries showed that the public health care Problems with physician–patient com- Discussion systems, into which TB control is fully munication were also found to have a integrated, are constrained by a lack negative impact on continuation of TB Failure to complete the treatment regi- of human resources and difficulties treatment. Out of 6 factors related to men has historically been cited as one in providing outreach services. This is physician–patient communication, 3 of the most challenging problems in particularly the case in rural areas in communication factors were found to TB treatment. The aim of this study countries such as Ethiopia, Indonesia, be statistically significant: physicians was to investigate possible factors af- Nigeria and Pakistan [4]. Expansion not listening to complaints, physicians fecting patients’ defaulting from the of the network of general health care not explaining TB complications and NTP-Egypt DOTS programme at facilities will improve access to health physicians not explaining treatment government chest dispensaries in Al- care and ultimately help to achieve side-effects. Patients reporting poor exandria, Egypt. The study investigated targets for TB control [4]. The WHO physician communications were 3 times 54 factors that might affect defaulting. report published in 2006 on enhancing more likely to default from TB treatment Stepwise multivariate logistic regression the DOTS programme indicated that (adjusted OR = 3.06). Patients typically analysis revealed 5 significant factors, in- particular attention should be given to default when they experience improve- cluding younger age, living in a squatter the poorest and most vulnerable popu- ment of symptoms with treatment, or rural area, long waiting times, poor lation groups [18]. while unawareness of the side-effects physician–patient communication and Factors related to the health services of medication can be misinterpreted patients’ fear of information leakage. were found to be important variables by patients as drawbacks of treatment. Younger patients aged < 30 years associated with defaulting from DOTS. The relationship between medication were 6.25 times more likely to default Facility-related factors, namely unsuit- side-effects and treatment defaulting than patients aged 50+ years. A effect able clinic hours and long waiting times, has been frequently demonstrated in of age on defaulting was reported in a were significantly associated with de- other studies [3,10,19–21]. It has been study conducted in Russia where defaulting. Similar results were reported recommended that health care staff in faulting was associated with age < 45 in a study conducted in Brazil in 2004, TB programmes should direct health years [17]. which reported long waiting times for education towards expected side-effects Patients residing in rural and squat- a medical consultation as an associated of treatment. Nurses should teach pa- ter areas had a significantly higher factor for defaulting from treatment [1]. tients to recognize and manage severe

111

Book 19-2.indb 111 2/21/2013 11:39:17 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 4 Multivariate logistic regression analysis of factors associated with community. The stigma associated with defaulting from tuberculosis treatment TB has been shown to have a significant Factors Adjusted OR (95% CI) P-value effect on defaulting from treatment Age (years) [17,25]. TB treatment polices must < 30a address confidentiality of information 30– 0.71 (0.24–2.11) 0.53 about patients’ private data. Patients 40– 0.57 (0.20–1.66) 0.30 should be assured about the confidenti- 50 + 0.16 (0.03–0.46) 0.002 ality of their information to ensure their Area of residence compliance with attendance. Urbana There are a number of limitations in Squatter 2.74 (1.12–6.73) 0.027 the present study that affect the general- Rural 12.9 (2.53–65.7) 0.002 izability of the results. The small number Long waiting times 5.81 (1.80–18.8) 0.003 of cases increased the value of each case Poor physician–patient and may have exaggerated some factors. communication 3.06 (1.02–9.46) 0.049 Moreover, the study was conducted Fear of information leakage only in chest dispensaries of Alexan- 3.62 (1.54–8.53) 0.003 dria governorate. Despite similarities aReference category. between TB patients living in different OR = odds ratio; CI = confidence interval. governorates all over Egypt, we cannot generalize the results to the whole popu- medication side-effects and request with defaulting. These findings are in lation of Egyptian TB patients. patients to schedule an extra clinic visit agreement with a study conducted in In conclusion, the main factors if they experience side-effects in order to Russia in 2008, which indicated that associated with defaulting from the reduce treatment defaulting [10]. non-adherent patients did not believe national programme of TB control Several patient perception factors that they will fully recover and thus did in Alexandria, Egypt, were service- were investigated in the present study, not want to continue treatment [17]. An related factors, which are amenable to many of which were significant in uni- association between patients’ knowl- improvement. Expanding treatment variate analysis. Although multivariate edge about the duration of treatment outlets in rural settings, involving analysis revealed only 1 factor related and treatment adherence has also been providers who practice close to where to patient perceptions, it is important shown in other studies [3,17,20,22–24]. patients live, offering psychologi- to discuss these factors as they are ame- The present study showed that most of cal support and health education nable to improvement at a lower cost to the defaulters were ashamed of their programmes especially for younger the community. Lack of improvement disease and concerned about leakage patients, reduction of patient waiting with treatment and patients’ perception of information. Stigmatization of TB times and improvement of patient– that treatment cannot cure TB com- leads to a situation where many patients physician communication are recom- pletely were significantly associated are not treated well by members of the mended improvement strategies.

References

1. Salles CL et al. Defaulting from anti-tuberculosis treatment in a 6. Borgdorff MW, Floyd K, Broekmans JF. Interventions to reduce teaching hospital in Rio de Janeiro, Brazil. International Journal tuberculosis mortality and transmission in low- and middle- of Tuberculosis and Lung Disease, 2004, 8:318–322. income countries. Bulletin of the World Health Organization, 2. Holtz TH et al. Risk factors associated with default from multidrug- 2002, 80:217–227. resistant tuberculosis treatment, South Africa, 1999–2001. Interna- 7. Treatment of tuberculosis: guidelines for national programmes, tional Journal of Tuberculosis and Lung Disease, 2006, 10:649–655. 3rd ed. Geneva, World Health Organization, 2003. 3. Tekb B, Mariam DH, Ali A. Defaulting from DOTS and its deter- 8. Tuberculosis services in partnership. The case of Egypt. Cairo, Na- minants in three districts of Arsi zone in Ethiopia. International tional Tuberculosis Control Programme Egypt/World Health Journal Tubercle and Lung Disease, 2002, 6(7):673–679. Organization, Eastern Mediterranean Region/Royal Tropical 4. Global tuberculosis control: epidemiology, strategy, financing. Institute, 2004 (WHO-EM/TUB/243/E). Geneva, World Health Organization, 2009. 9. Tuberculosis profile—Egypt. World Health Organization [on- 5. National program of TB control. Social worker index, 2nd ed. line database] (http://www.who.int/tb/country/data/pro- Cairo, Ministry of Health and Population, 2004. files/en/index.html, accessed 24 December 2012).

112

Book 19-2.indb 112 2/21/2013 11:39:17 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

10. Jittimanee SX et al. Treatment default among urban tuberculo- 19. Quy HT et al. Treatment results among tuberculosis patients sis patients, Thailand. International Journal of Nursing Practice, treated by private lung specialists involved in a public–private 2007, 13:354–362. mix project in Vietnam. International Journal of Tuberculosis and 11. Jakubowiak WM et al. Impact of socio-psychological factors Lung Disease, 2003, 7:1139–1146. on treatment adherence of TB patients in Russia. Tuberculosis, 20. Okanurak K, Kitayaporn D, Akarasewi P. Factors contributing to 2008, 88:495–502. treatment success among tuberculosis patients: a prospective 12. Mishra P et al. Adherence is associated with the quality of cohort study in Bangkok. International Journal of Tuberculosis professional-patient interaction in Directly Observed Treat- and Lung Disease, 2008, 12:1160–1165. ment Short-course, DOTS. Patient Education and Counseling, 21. Gupta S, Gupta S, Behera D. Reasons for interruption of anti- 2006, 63:29–37. tubercular treatment as reported by patients with tuberculosis 13. Dimitrova B et al. Health service providers’ perceptions of bar- admitted in a tertiary care institute. Indian Journal of Tuberculo- riers to tuberculosis care in Russia. Health Policy and Planning, sis, 2011, 58:11–17. 2006, 21:265–274. 22. Comolet TM, Rakotomalala R, Rajaonarioa H. Factors deter- 14. Janakan N, Seneviratne R. Factors contributing to medication mining compliance with tuberculosis treatment in an urban noncompliance of newly diagnosed smear-positive pulmo- environment, Tamatave, Madagascar. International Journal of nary tuberculosis patients in the district of Colombo, Sri Lanka. Tuberculosis and Lung Disease, 1998, 2:891–897. Asia-Pacific Journal of Public Health, 2008, 20:214–223. 23. Liam CK et al. Attitudes and knowledge of newly diagnosed tu- 15. Fleiss JL et al. Statistical methods for rates and proportions, 3rd berculosis patients regarding the disease, and factors affecting ed. London, Wiley-Interscience, 2003. treatment compliance. International Journal of Tuberculosis and Lung Disease, 1999, 3:300–309. 16. Lwanga SK, Lemeshow S. Sample size determination in health studies: a practical manual. Geneva, World Health Organiza- 24. Hasker E et al. Why do tuberculosis patients default in Tashkent tion, 1991. City, Uzbekistan? A qualitative study. International Journal of Tuberculosis and Lung Disease, 2010, 14:1132–1139. 17. Jakubowiak WM et al. Risk factors associated with default among new pulmonary TB patients and social support in six 25. Dodor EA, Afenyadu GY. Factors associated with tuberculosis Russian regions. International Journal of Tuberculosis and Lung treatment default and completion at the Effia-Nkwanta Region- Disease, 2007, 11:46–53. al Hospital in Ghana. Transactions of the Royal Society of Tropical Medicine and Hygiene, 2005, 99:827–832. 18. The Stop TB strategy. Building on and enhancing DOTS to meet the TB-related Millennium Development Goals. Geneva, World Health Organization, 2006.

113

Book 19-2.indb 113 2/21/2013 11:39:17 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Factors associated with patient delay in accessing pulmonary tuberculosis care, Gezira State, Sudan, 2009 E.Y. Mohamed,1 S.M. Abdalla,2 A.A. Khamis,3 A. Abdelbadea 4 and M.A. Abdelgadir 5

2009 العوامل املرتبطة ُّ ر بتأخوصول املرىض إىل رعاية السل الرئوي، والية اجلزيرة، السودان. الصادق يوسف حممد، سوسن مصطفى عبد اهلل، عبد اهلل عيل مخيس، أمحد عبد البديع، حممد أمحد عبد القادر اإلمام

اخلالصـة: تعطي التحليالت لتأخر املرىض الذين يلتمسون املعاجلة للسل ِّبينات مفيدة للربامج الوطنية ملكافحة السل. وهتدف هذه الدراسة التي َأجريت يف والية اجلزيرة، السودان، إىل تقييم مدى ُّتأخر وصول املرىض إىل الرعاية للسل الرئوي والعوامل املرتبطة بذلك التأخري؛ وهي دراسة مستعرضة َأجريتاملرحلة األوىل منها للتعرف عىل طول فرتة التأخر، وأجريت املرحلة التالية لدراسة احلاالت والشواهد املتجمعة ملقارنة ُّتأخر املرىض 365–0 4 27.2 أكثر أو أقل من متوسط الوقت. ووجد الباحثون أن وسطي ُّتأخر املرىض ًيوما، وأن املتوسط أيام )يرتاوح بني (، ومل يكن هناك فرق ُي ْع َت ُّد به ًإحصائيابني جمموعات احلاالت والشواهد من حيث العمر واجلنس واحلالة الزواجية واملستوى التعليمي أو التدخني. إال أن أكثر احتامالت التأخر كانتلدى املرىض الذين يعيشون يف مناطق حرضية، من ذوي الدخل املنخفض، أو لدى زوجات ال يعملن خارج املنزل، أو العاطلني عن العمل، إىل جانب املرىض املخالطني ًسابقا ملريض مصاب بالسل، واملرىض الذين ُي َش ُّك بإصابتهم بالسل، أو لدهيم سوابق مرض رئوي مزمن ّساد.

ABSTRACT Analyses of patient delays in seeking treatment for tuberculosis (TB) provide useful evidence for national TB control programmes. The objectives of this study in Gezira State, Sudan were to estimate the extent of, and factors associated with, pulmonary TB patient delay in accessing care. A cross-sectional phase was conducted to determine the length of delay, followed by a nested case–control phase comparing patients delaying above or below the median time. The mean patient delay was 27.2 days, median 4 days (range 0–365 days). There were no significant differences between case and control groups in terms of age, sex, marital status, educational level or smoking status. However, patients living in urban areas, with low income status and who were housewives or unemployed were more likely to delay. Also patients with a history of contact with a TB patient, those who suspected TB and those with a history of chronic obstructive pulmonary disease were more likely to delay.

Facteurs associés au retard d'accès des patients au traitement de la tuberculose pulmonaire dans l'État de Gezira (Soudan) en 2009

RÉSUMÉ L'analyse des retards dans la recherche de traitement antituberculeux par les patients fournit des éléments utiles aux programmes nationaux de lutte antituberculeuse. La présente étude, menée dans l'État de Gezira (Soudan), visait à estimer l'importance du retard d'accès aux soins des patients atteints de tuberculose pulmonaire ainsi que les facteurs associés à ce retard. Une phase transversale a été menée afin de déterminer l'importance du retard d'accès, suivie par une phase cas-témoins emboîtée comparant les patients dont le retard était inférieur au retard médian avec ceux dont le retard était supérieur. Le retard d'accès moyen des patients était de 27,2 jours et le retard d'accès médian de 4 jours (extrêmes 0–365 jours). Aucune différence significative n'a été observée entre le groupe d'étude et le groupe de témoin en termes d'âge, de sexe, de situation matrimoniale, de niveau d'études ou de statut tabagique. Toutefois, les patients vivant en milieu urbain, disposant de revenus faibles et qui étaient soit femmes au foyer, soit sans emploi avaient davantage tendance à retarder leur recherche de soins. De même, les patients ayant des antécédents de contact avec un patient tuberculeux, ceux qui suspectaient être atteints de la maladie et ceux souffrant de bronchopneumopathie obstructive chronique risqueraient davantage de repousser leur recherche de soins.

1Department of Community Medicine, Faculty of Medicine, University of Khartoum, Khartoum, Sudan (Correspondence to E.Y. Mohamed: [email protected]). 2Department of Community Medicine; 3Department of Paediatrics; 4Department of Pathology; 5Department of Obstetrics and Gynaecology, Faculty of Medicine, National Ribat University, Khartoum, Sudan. Received: 01/08/10; accepted: 22/03/11

114

Book 19-2.indb 114 2/21/2013 11:39:18 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction Methods delay of the sample. Patients with delays greater than the median were catego- Tuberculosis (TB) is still a major cause Study area rized as cases, while those with delays of death in both developed and devel- Gezira State, located in the centre of less than or equal to the median were oping countries [1]. In 2005, 12 million Sudan, is one of the high TB burden considered as controls. new cases of TB were identified, an states in the country. The TB case detec- Data collection techniques almost 60% increase from the estimated tion rate was 37.9% in 2010 [15], which and tools number of cases in 1990 [2]. An effec- is far below the target of 70%, developed Data was collected during June and tive TB control programme requires by the Stop TB partnership and the December 2009 by health workers early diagnosis and immediate initiation World Health Assembly, and within the who underwent intense training on of treatment. Delay in diagnosis is seri- Millennium Development Goal frame- interviewing and probing techniques. ous because it not only adversely affects work [16]. TB care is provided through Patients were interviewed using a pre- disease prognosis at the individual level 41 TB management units (TBMUs) coded, pre-tested questionnaire. The but also promotes transmission within distributed throughout the State [17]. the community and enhances TB epi- questionnaire included information demics [3,4]. Study design about patients’ sociodemographic characteristics and selected factors that Delays in diagnosis and treatment of This was a cross-sectional and case– might affect delay in receiving TB care. TB may reflect patient delays in seeking control study to explore patient delay The laboratory register was examined care, health care provider delays in mak- in diagnosis and treatment of TB. A to confirm data regarding the time of ing diagnosis and starting treatment, cross-sectional phase was conducted to diagnosis. The time taken to initiate or both [5–11]. A delay in initiating determine the extent of delay, and then treatment was calculated after the treat- TB care is not always the responsibil- a nested case–control phase was carried ment was started. ity of the patient; sometimes patients out comparing patients delaying above contact the health services early but and below the median time. A written consent was taken from the physician in charge does not sus- the respondents. Ethical clearance was pect TB [12]. It has been found that Study population and obtained from the Ministry of Health. sampling patients become more contagious as Objectives, steps and expected outcomes the delay progresses. The longest delays The study population was newly diag- of the research were explained to the are associated with the highest numbers nosed (within 2 weeks) smear-positive participants as well as their right to with- of Mycobacterium tuberculosis bacilli on pulmonary TB cases aged 15 years and draw from the study at any time without sputum smears. Diagnostic delay may older who attended the selected TB- any consequences for their current care. result in more advanced and severe MUs during the study periods. Confidentiality of the data obtained were disease, higher mortality and sustained The sampling type was simple maintained before and during the study spread of M. tuberculosis in the com- random. All the 41 TBMUs in the 5 and will continue in the future. munity as untreated people continue to localities were considered and 10 of Data management and transmit the infection to others [13,14]. them were selected randomly (i.e. 2 out analysis The importance of delay is reflected in of every 8 on average). Based on the Data analysis was performed using SPSS estimated incidence of TB with a maxi- increasing costs and mortality due to for Windows, version 13. Descriptive mum allowed error of 10% and a 95% TB. The many factors causing delay in statistics were used, such as frequency, confidence interval (CI) the sample diagnosis and treatment must be clearly mean and standard deviation (SD), size was calculated as 282 and rounded identified and addressed locally in order median, minimum and maximum. up to 292 [18]. The sample was divided to improve the quality and effectiveness Comparisons between groups were across the 5 localities according to of national TB control programmes made using the chi-squared test. P value (NTP). Studies analysing delays and population size by dividing the locality < 0.05 was considered significant and all their determinants provide useful population over the total population of tests were 2-sided. evidence for assessment of case-finding the state multiplied by the sample size. success under DOTS. Within the TBMU, respondents were The objectives of this study in Sudan selected conveniently until the required Results were to estimate the extent of pulmo- sample was completed. The response nary TB patient delay in accessing care rate was 100%. The mean duration of patients’ delay in Gezira State and to determine some The cut-off point for long delays was 27.2 (95% CI: 21.0–33.2) days. of the factors associated with delays. was defined according to the median The median delay was 4 days (range

115

Book 19-2.indb 115 2/21/2013 11:39:18 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

0–365 days). There were 83 patients or unemployed (35.3%) were more versus those without (27.3%) (P = in the delayed group and 209 in the likely to delay than clerks and labour- 0.024). There was no significant differ- non-delayed group. ers (21.1%), and students (17.4%) (P ence in delay in care-seeking by smok- Table 1 shows the demographic = 0.014). Patients with debts (44.6%) ing status. characteristics of patients who delayed or no income (50.0%) were more seeking care for TB compared with likely to delay than those with income those who did not delay. There were equal to expenditure (17.2%) (P < Discussion no significant differences between 0.001). the delayed and not delayed groups Table 2 shows that patients with This study reported some extremely in terms of age, sex, marital status or a history of contact with a TB patient long delays between the onset of educational level. However, there were more likely to delay (38.7%) than symptoms until presentation to a were significant differences by area those who did not have contact (25.7%) health care provider, ranging from 0 of residence, occupation and income (P = 0.043). Patients who suspected TB to 365 days. The mean duration was level. Patients living in urban areas were also more likely to delay (44.7%) 27.2 days, which was lower than mean were more likely to delay (45.3%) than than those who had no suspicion of delays reported elsewhere in Sudan those in rural areas (23.7%) (P < infection (25.3%) (P = 0.007), as did in Khartoum (53 days) [Mustafa 0.001). Those who were housewives those with a history of COPD (60.0%) MH. unpublished document], the

Table 1 Relationship between patient delay in care-seeking for tuberculosis and socioeconomic factors Variable Total Delayed Not delayed χ2-test P-value (n = 292) (n = 83) (n = 209) No. No. % No. % Age (years) 15–35 160 46 28.8 114 71.3 0.16 0.899 > 35 132 37 28.0 95 72.0 Sex Male 179 45 25.1 135 75.4 2.24 0.134 Female 113 38 33.6 74 65.5 Education University and above 10 1 10.0 9 90.0 2.19 0.535 Basic and primary 150 43 28.7 108 72.0 Illiterate and khalwa 132 39 29.5 92 69.7 Residence 11.5 < 0.001 Urban 64 29 45.3 35 54.7 Rural 228 54 23.7 174 76.3 Occupation 14.3 0.014 Professional/ technical 9 3 33.3 6 66.7 Clerk/ labourer 109 23 21.1 86 78.9 Student 46 8 1 7. 4 38 82.6 Housewife/ unemployed 102 36 35.3 66 64.7 Other 26 13 50.0 13 50.0 Income 23.5 < 0.001 Indebted 65 29 44.6 36 55.4 No income 42 21 50.0 21 50.0 Income = expenses 180 31 17.2 149 82.8 Saving 5 2 40.0 3 60.0 Marital status 0.33 0.557 Married 193 57 29.5 136 70.5 Single 99 26 26.3 73 73.7 Total 292 83 28.4 209 71.6

116

Book 19-2.indb 116 2/21/2013 11:39:18 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 2 Relationship between patient delay in care-seeking for tuberculosis (TB) and patient clinical history Patient history Total Delayed Not delayed χ2-value P-value (n = 292) (n = 83) (n = 209) No. No. % No. % Contact with a TB patient Yes 62 24 38.7 38 61.3 4.09 0.043 No 230 59 25.7 171 74.3 History of smoking 0.03 0.858 Yes 44 13 29.5 31 70.5 No 248 70 28.2 178 71.8 Suspected TB 7.28 0.007 Yes 47 21 44.7 26 55.3 No 245 62 25.3 183 74.7 History of COPD 5.07 0.024 Yes 10 6 60.0 4 40.0 No 282 77 27.3 205 72.7 Total 292 83 28.4 209 71.6

COPD = chronic obstructive pulmonary disease.

Philippines and Ethiopia (30 and 30 a significant difference in care-seeking history of COPD also delayed signifi- days respectively) [19,20], but higher by occupational and income status. cantly more than those with a negative than in Egypt, Pakistan and Botswana Unemployed people and housewives history [26]. This may be because (24.3, 9.9 and 21 days respectively) were more likely to delay than other patients with a history of COPD as- [21,22]. occupational groups [28] (profession- sume that their symptoms are due to As regards the relation between als and technicians, clerks and labour- COPD and hence they do not seek patient delay and social factors [23], ers); students were the least likely help. the study revealed similar proportions to delay [29]. A study in California, Ou study found no relationship be- of younger and older age groups delay- United States, found that unemploy- tween smoking and delay in accessing ing seeking care. This finding was not ment was one the factors indepen- care. Another study in Kampala found in agreement with a study conducted dently predicting delay greater than 60 that one of the predictors of patient in Khartoum state where older age days [28]. Economic status seems to delay was smoking [26]. This may be groups delayed longer on average play an important role in patient delay due to the fact that cough appearing (50.2 versus 64.1 days) [Mustafa MH. to seek TB care, as those with no debts in TB patients who are smokers is of- unpublished document]. Slightly were less likely to delay than people ten attributed to smoking, resulting in more women in our study delayed with debts. These findings are in line delays in seeking help for TB-related seeking TB care than did men but the with reports from the Philippines and symptoms. difference was not significant. More Armenia [19,24]. There were some limitations to the women than men delayed in a study in This study found that the clinical study. Patients’ recall was needed to Yemen [21]; but not in reports from history of patients had effects on delay acquire data for the study and they Armenia and Brazil [24,25]. Patients in accessing TB care. Those who had may have had difficulty remembering living in rural areas were significantly had contact with TB patients delayed events that occurred some time ago. more likely to delay than those living significantly more than those who However, as in other studies, there was in urban areas [26]. had no contact [30]. Patients who no other way to obtain this type of data A study in China found that less suspected TB tended to delay more than patients’ recall. educated patients delayed longer than those who did not. This may be To summarize, the mean patient [27], perhaps because they lacked the due to stigma; patients who suspect delay was long (27.2 days). There were knowledge to help them seek care at TB may be afraid to consult a care no significant differences between the right time. In contrast, our study provider while those who think that the delayed and non-delayed groups found no significant differences by ed- they are free of TB more confidently in terms of age, sex, marital status or ucational level. However, we did find access care [26]. Patients who had educational level or smoking status.

117

Book 19-2.indb 117 2/21/2013 11:39:18 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

However, there were significant differ- Acknowledgements conduct this work. We would also ences by area of residence, occupation, like to extend our thanks to Gezira income level, history of contact with a We would like to acknowledge the Tuberculosis Control Programme TB patient, history of COPD and hav- National Tuberculosis Control Pro- and the staff at the localities who col- ing suspicion of pulmonary TB. gram for offering us the chance to lected the data.

References

1. Dolin PJ, Raviglione MC, Kochi A. Global tuberculosis inci- 17. TB control program. Annual report 2006. Khartoum, Federal dence and mortality during 1990–2000. Bulletin of the World Ministry of Health, 2006. Health Organization, 1994, 72:213–220. 18. Fleiss JL. Statistical methods for rates and proportions. New York, 2. Global tuberculosis control: surveillance, planning, financing. John Wiley and Sons, 1981. WHO report 2005. Geneva, World Health Organization, 2005 19. Auer C et al. Health seeking and perceived causes of tubercu- (WHO/HTM/TB/2005.349). losis among patients in Manila, Philippines. Tropical Medicine 3. Dye C et al. Consensus statement. Global burden of tubercu- and International Health, 2000, 5:648–656. losis: estimated incidence, prevalence and mortality by coun- 20. Yimer S, Bjune G, Alene G. Diagnostic and treatment delay try. WHO Global Surveillance and Monitoring Project. Journal among pulmonary tuberculosis patients in Ethiopia: a cross- of the American Medical Association, 1999, 282:677–686. sectional study. BMC Infectious Diseases, 2005, 5:112. 4. Bjune G. Tuberculosis in the 21st century: an emerging pan- 21. Diagnostic and treatment delay in tuberculosis: An in depth demic? Norsk Epidemiologi, 2005, 15:133–139. analysis of the health seeking behaviour of patients and health 5. Sherman LF et al. Patient and health care system delays in the system response in seven countries of the East Mediterra- diagnosis and treatment of tuberculosis. International Journal nean Region. Cairo, World Health Organization Regional of Tuberculosis and Lung Disease, 1999, 3:1088–1095. Office for the Eastern Mediterranean, 2006 (WHO-EM/ 6. Wandwalo ER, Morkve O. Delay in tuberculosis case-finding TDR/009/E). and treatment in Mwanza, Tanzania. International Journal of 22. Steen TW, Mazonde GW. Pulmonary tuberculosis in Kwen- Tuberculosis and Lung Disease, 2000, 4:133–138. ing district Botswana: delay in diagnosis in 212 smear positive 7. Demissie M, Lindtjorn B, Berhane Y. Patient and health care patients. International Journal of Tuberculosis and Lung Disease, service delay in the diagnosis of pulmonary tuberculosis in 1998, 2:627–634. Ethiopia. BMC Public Health, 2002, 25:23. 23. Schneider D et al. Reasons for delay in seeking care for tuber- 8. Golub JE et al. Delayed tuberculosis diagnosis and tuberculo- culosis, Republic of Armenia. Interdisciplinary Perspectives on sis transmission. International Journal of Tuberculosis and Lung Infectious Diseases, 2010:41262. Disease, 2006, 10(1):24–23 24. Schneider D et al. Reasons for delay in seeking care for tuber- 9. Ohmori M et al. Trends of delays in tuberculosis case finding culosis, Republic of Armenia, 2006–2007. Interdisciplinary in Japan and associated factors. International Journal of Tuber- Perspectives on Infectious Diseases, 2010:412624. culosis and Lung Disease, 2005, 9:999–1005. 25. Dos Santos MAPS et al. Risk factors for treatment delay in 10. Xu B et al. Diagnostic delays in access to tuberculosis care in pulmonary tuberculosis in Recife, Brazil. BMC Public Health, counties with or without the National Tuberculosis Control 2005, 5:25. Program in rural China. International Journal of Tuberculosis and 26. Kiwuwa MS, Charles K, Harriet MK. Patient and health service Lung Disease, 2005, 9:784–790. delay in pulmonary tuberculosis patients attending a referral 11. Rojpibulstit M, Kanjanakiritamrong J, Chongsuvivatwong V. hospital: a cross-sectional study. BMC Public Health 2005, Patient and health system delays in the diagnosis of tuberculo- 5:122. sis in Southern Thailand after health care reform. International 27. Xu B et al. Diagnostic delays in access to tuberculosis care in Journal of Tuberculosis and Lung Disease, 2006, 10:422–428. counties with or without the National Tuberculosis Control 12. Yimer S, Bjune G, Alene G. Diagnostic and treatment delay Program in rural China. International Journal of Tuberculosis and among pulmonary tuberculosis patients in Ethiopia: a cross- Lung Disease, 2005, 9:784–790. sectional study. BMC Infectious Diseases, 2005, 5:112. 28. Asch S et al. Why do symptomatic patients delay obtaining 13. Toman K. Tuberculosis case finding and chemotherapy questions care for tuberculosis? American Journal of Respiratory and Criti- and answers. Geneva, World Health Organization, 1979. cal Care Medicine, 1998, 157:1244–1248. 14. Styblo K. Epidemiology of tuberculosis. Selected papers. Volume 29. Okeibunor JC et al. Where do tuberculosis patients go for 24. The Hague, KNCV Tuberculosis Foundation, 1991. treatment before reporting to DOTS clinics in southern Nige- 15. National tuberculosis control programme Sudan. Progress report. ria? Tanzania Health Research Bulletin, 2007, 9(2):94–101. January–December 2010. Khartoum, Federal Ministry of Health, 30. Salaniponi FM, Harries AD, Banda HJ. Care seeking behavior 2010:35. and diagnostic processes in patients with smear positive pul- 16. Global tuberculosis control strategy 2009: epidemiology, strat- monary tuberculosis in Malawi. International Journal of Tuber- egy, financing. Geneva, World Health Organization, 2009. culosis and Lung Disease, 1991, 4:327–332.

118

Book 19-2.indb 118 2/21/2013 11:39:18 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Mycobacterium tuberculosis spoligotypes circulating in the Lebanese population: a retrospective study N. Bedrossian,1 M. Hamze,1,2 A.K. Rahmo,3 A. Jurjus,4 J. Saliba,4 F. Dabboussi 1,2 and W. Karam 5

أنامط عديدات النكليوتيد البينية للمتفطرات السلية التي ترسي بني السكان اللبنانيني: دراسة استعادية نوره بدروسيان، منذر محزة، عبد القادر رمحو، عبده جرجس، جيسيكا صليبا، فؤاد دبويس، وليد كرم قد يعوداخلالصـة: التنميط اجليني للمتفطرات السلية بالفوائد عىل املستوى الوطني يف لبنان من أجل رصد االستجابة العالجية للمعاجلة القصرية األمدحتت اإلرشاف املبارش )دوتس(. وهتدف هذه الدراسة إىل حتديد خصائص املستفردات الرسيرية من أنامط عديدات النكليوتيد البينية للمتفطرات 60 2005 2004 السلية التي مجعت من املرىض يف الفرتة من نيسان/أبريل وترشين األول/أكتوبر من مجيع املناطق اللبنانية. وبلغ عدد َاملستفردات ، وزرعت وتم ُّالتعرفعىل خصائصها الكيميائية احليوية، وتضخيم املستخلصات من الدنا بالتفاعل السلسيل للبوليمراز، وحتديد أنامطها اجلينية من 41.6 13 خالل التنميط التفصييل ألنامط عديدات النكليوتيد البينية. وتم ُّ فالتعر عىل ًنمطا من الذراري املعقدة لفصائل املتفطرات السلية: % منها من ذراري تنتمي إىل فصيلة T1، و25% تنتمي إىل فصيلة LAM9، و10% منها تنتمي إىل فصيلة هارمل 3، و3.3 %منها تنتمي إىل كل من الفصائل CAS، و LAM و BCG، والفصيلة 36، يف حني كانت 1.7% منها تنتمي إىل كل من الفصائل هارمل 1، و LAM 10، و S، واملتفطرات األفريقية، و X1، و .T3 ومل يكن الغياب امللحوظ لفصائل بيجني واهلندية الرشق أفريقية ً متامشيامع النامذج ّاملسجلة يف البلدان املجاورة؛ مما يدل عىل رضورة القيام بدراسة أكثر إحاطة وشموال ًللسكان اللبنانيني من أجل ُّالتعرف بدقة عىل الفصائل األكثر ً انتشارايف لبنان.

ABSTRACT Genotyping Mycobacterium tuberculosis in Lebanon on the national level may be beneficial for assessing patients and monitoring the therapeutic response to DOTS. This study aimed to characterize the spoligotypes of clinical isolates of M. tuberculosis patients collected between April 2004 and October 2005 from all Lebanese provinces. Isolates (n = 60) were cultured and identified by their biochemical characteristics. DNA extracts of these samples were amplified by PCR and genotyped by spoligotyping. Thirteen (13) patterns of M. tuberculosis complex family strains were identified: 41.6% of the strains belonged to the T 1 family, 25.0% to LAM 9, 10.0% to Haarlem 3, 3.3% to each of CAS, LAM 8, BCG and Family 36 and 1.7% to each of Haarlem 1, LAM 10, S, M. africanum, X 1 and T 3 families. The noticeable absence of Beijing and East African Indian families was not consistent with the patterns reported in neighbouring countries. A more inclusive study of the Lebanese population is necessary to accurately identify most of the prevailing families in the country.

Spoligotypes de Mycobacterium tuberculosis circulant dans la population libanaise : une étude rétrospective

RÉSUMÉ Le génotypage de Mycobacterium tuberculosis au Liban au niveau national peut être utile pour dépister les patients et surveiller la réponse au traitement de brève durée sous surveillance directe ou DOTS. L'objectif de l'étude visait à caractériser les spoligotypes des isolats cliniques de M. tuberculosis prélevés entre avril 2004 et octobre 2005 chez des patients de toutes les provinces libanaises. Les isolats (n = 60) ont été mis en culture puis identifiés en fonction de leurs caractéristiques biochimiques. Les extraits d'ADN de ces échantillons ont été amplifiés par PCR puis génotypés par spoligotypage. Treize (13) souches différentes de la famille du complexe M. tuberculosis ont été identifiées : 41,6 % des souches appartenaient à la famille T 1 ; 25,0 % à la famille LAM 9 ; 10,0 % à la famille Haarlem 3 ; 3,3 % chacune aux familles CAS, LAM 8, BCG et à la famille 36 et 1,7 % chacune aux familles Haarlem 1, LAM 10, S, M. africanum, X 1 et T 3. L'absence remarquable des familles Beijing, indiennes et d'Afrique de l'Est ne coïncidait pas avec les tendances rapportées dans les pays voisins. Une étude plus globale de la population libanaise est nécessaire pour identifier avec précision la plupart des familles dominantes dans notre pays.

1Faculty of Medicine; 5Department of Biology, Lebanese University, Hadath Campus, Hadath, Lebanon (Correspondence to M. Hamze: mhamze@ monzerhamze.com). 2AZM Center for Biotechnology, Lebanese University, Tripoli, Lebanon. 3National Commission of Biotechnology, Damascus, Syrian Arab Republic. 4Department of Human Morphology, American University of Beirut, Beirut, Lebanon. Received: 12/08/11; accepted: 05/02/12

119

Book 19-2.indb 119 2/21/2013 11:39:18 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction to DOTS. In a previous study, charac- 7H9 broth (BD BBL MGIT, Becton terization of M. tuberculosis in Lebanese Dickinson) supplemented with After HIV/AIDS, tuberculosis (TB) patients by double-repetitive-element PANTA (BBL MGIT PANTA, Becton is the second most common cause of polymerase chain reaction (DRE-PCR) Dickinson) and oleic acid, albumin, death due to an infectious disease, and indicated the presence of several geno- dextrose, and catalase (OADC) en- current trends suggest that TB will still types with evidence of certain groups richment. The 2 media were incubated be among the 10 leading causes of global specific to geographic areas, implicat- at 35–37 ºC, the LJ slant was exam- disease burden in the year 2020 [1]. ing separate evolution of M. tuberculosis ined for growth twice weekly. The BD Molecular tools have enhanced our strains [9]. This investigation reports on BBL MGIT tube was read daily with understanding of the epidemiology of the first nationwide study to character- ultraviolet light starting on the 2nd day TB by providing new insight into the ize by spoligotyping clinical isolates of of incubation using a positive control transmission, dynamics, source and M. tuberculosis from different regions of and a negative control. Fluorescence is spread of Mycobacterium tuberculosis Lebanon. It will further provide insight detected as a bright orange colour in the clones [2,3], which are often difficult to into the future dynamics of the disease bottom of the tube and also as an orange identify by traditional epidemiological in our country and even among neigh- reflection on the meniscus. A positive investigations alone [4]. International bouring countries. tube was subcultured and an acid-fast databases, such as the World Spo- smear prepared. A positive acid-fast smear result indicated the presumptive ligotyping Database, SpolDB3.0, have Methods revealed the clonal structure of M. tu- presence of viable microorganisms in the tube. berculosis isolates in different geographi- Sampling of patients cal settings. The SpolDB4.0 database Following an agreement with the na- Identification of strains further defines super-families specific to tional TB programme of the Lebanese Identity of strains was based on the certain locations [5]. The abundance of Ministry of Health, we obtained sputum following biochemical characteristics: polymorphism indicates that transposi- samples from newly detected pulmo- production of niacin, nitrate reductase, tion and homologous recombination nary TB cases (60 samples). Samples catalase at laboratory temperature, cata- are the major events contributing to the were collected between April 2004 lase at 68 ºC, hydrolysis of Tween 80 in diversity of M. tuberculosis strains [6]. and October 2005 from all Lebanese 10 days, urease in 18 hours, arylsulfatase Due to the importance and gravity provinces (muhafazat) and were stored in 3 days [10]. of this disease in Lebanon, the Ministry in the laboratory of the hospital at the of Public Health set up the national Middle East Health Centre in Bsalim/ DNA extraction TB programme in 1992, which started Metn in Lebanon. Relevant informa- A loopful of each culture was suspended the DOTS implementation/expansion tion (e.g. sex, age, location, new/old in animal-tissue lysis buffer (0.2 mL). in 1998. DOTS refers to a broad TB case) was also obtained for each of the DNA samples were extracted in a labo- control strategy outlined by the World samples; however no information about ratory free of all mycobacterial products Health Organization (WHO) which the HIV status was available. using QIAamp DNA blood mini kit aims to halt the spread of the infection As this was a retrospective study, (Qiagene). The accompanied proce- and of multidrug resistant TB strains and the samples tested consisted of dure was adhered to, except for the [7]. The total number of cases in Leba- stock bacterial cultures and did not incubation period which was extended non fell from 993 in 1993 to 195 in directly involve any human subjects, to 3 hours. DNA was eluted in 100 µL 2005, with pulmonary TB representing ethical approval was not required. PCR water (Gibco). 60% to 65% of all TB cases. At present, according to the WHO report, the in- Specimen preparation and PCR amplification cidence of TB per 100 000 people in culture All DNA samples were positive for Lebanon was reported as 14 in 2008, 15 Specimens were decontaminated by the IS6110 insertion element as in 2009 and 17 in 2010 (report released the 2% N-acetyl-L-cysteine NaOH determined by PCR amplification in 2011) [8]. method. After neutralization and cen- followed by detection by agarose gel Genotyping M. tuberculosis in Leba- trifugation, 0.2 mL of the concentrated electrophoresis using the method of non on the national level in parallel specimen was inoculated onto one slant Eisenach [11]. PCR set-up was also with the DOTS programme may prove of Lowenstein-Jensen (LJ) medium performed in a remote area separate beneficial for assessing patients and (Becton Dickinson), 0.5 mL was also from all subsequent procedures. Meas- monitoring the therapeutic response inoculated into modified Middlebrook ures to avoid cross- and carry-over PCR

120

Book 19-2.indb 120 2/21/2013 11:39:18 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

contaminations were observed: use of Table 1 Prevalence of the different Mycobacterium tuberculosis complex family aerosol barrier-fitted pipette tips and strains in the tested samples from Lebanese patients (n = 60) molecular biology-grade water and Family strains No. % reagents and frequent decontamina- T 1 25 41.6 tion of work areas with diluted bleach LAM 9 15 25.0 and short-ultraviolet surface performed Haarlem 3 6 10.0 regularly between experiments. CAS 2 3.3 LAM 8 2 3.3 Spoligotyping Family 36 2 3.3 Spoligotyping was performed for each BCG 2 3.3 sample in duplicate and was initially Haarlem 1 1 1.7 repeated twice to assess the reproduc- LAM 10 1 1.7 ibility of the method. Duplicate positive S family 1 1.7 and negative controls were included M. africanum 1 1.7 in each run. DNA was replaced with X 1 1 1.7 water in negative control reactions and T 3 1 1.7 DNA of M. tuberculosis strains H37Rv Total 60 100.0 and M. bovis BCG P3 supplied in the kit were used in positive control reactions. The procedure was performed using a reverse dot-blot spoligotyping 272 and lacked spacers 1–4 and 12–13. another as ST 418 (lacking 3–5, 31 kit with chemiluminescent detection Two isolates could not be classified un- and 33–36), and 4 others as ST 775, (Isogen Bioscience) as follows: 5 µL der any of the shared types: 1 having an 2 of which lacked 28–37 and 2 others of mycobacterial DNA was added to a octal designation 017711727760760 additionally lacking spacer 13. PCR amplification reaction in which and lacking spacers 1–5, 14–15, 21–22, There were 10 other family strains the forward direct repeat (DR) primer 32–36 and 42, and 1 with an octal identified. One sample (1.7%) belonged was biotinylated. Amplification was per- designation 7777377777760371 and to the T 3 family as ST 149, lacking formed for 25 cycles as recommended lacking spacers 2, 5, 13–15, 19, 32–35, spacers 33–36 and 10–19. One speci- by the manufacturer. 37–39 and 42. men (1.7%) belonged to Haarlem 1 as The second predominant family in ST 602 and lacked spacers 25–36. Two our results was the LAM 9 family in 15 samples (3.3%) were Asian (CAS) line- Results samples (25.0%). Most of these strains age; 1 belonged to ST 25 (lacking 4–7, (14/15) were prototyped as ST 41, 13 21–34 and 37–38 spacers) and 1 to ST The prevalence of the different M. tu- simultaneously lacking 20–24, 26–27 1199 (lacking 2, 4–7, 10, 13–15, 21–34, berculosis complex family strains in the and 33–36 spacers and 1 in addition 37–39 and 42 spacers). Two of the sam- tested samples are summarized in Table lacking 13 and 15. The other LAM 9 ples (3.3%) belonged to Family 36; 1 1. Table 2 shows a detailed analysis of strain belonged to ST 42, lacking spac- was ST 4 and lacked 1–24 and 33–36 the different M. tuberculosis complex ers 21–24 and 33–36. Another 3.3% of and the other was ST 125 and addition- family clades with the individual shared samples (n = 2) belonged to the LAM ally lacked spacers 40–41. One sample type (ST) designation, the number of 8 family: 1 was prototyped as ST 511, (1.7%) belonged to the X 1 family. With isolates and the corresponding lack of lacking simultaneously 13–17, 19–31, an octal format 017100777760760 and spacers for each. 33–36 and 39–42 spacers; the other the absence of spacers 1–5, 10, 12–15, Of the samples tested 25 (41.6%) with an octal format 777777400000371 21–22, 32–36 and 42, it could not be belonged to the T1 family. In this clade, (lacking 14–17, 19–31, 33–36 and 39– classified under any of the shared types. all isolates lacked spacers 33–36. Sepa- 42 spacers) and was unclassified. One One strain (1.7%) belonged to M. africa- rate STs exhibited additional absence sample belonged to LAM 10 (1.7%), num ST 536, with the absence of 4–19 of spacers: 17 strains were ST 53 and lacking 23–25 and 33–36, and was pro- and 39 spacers. Two samples (3.3%) lacked spacers 39–41; 1 was ST 154 totyped as ST 61. belonged to M. bovis BCG were ST 482, and lacked spacers 5; 1 was ST 751 In the Haarlem (H) lineage, 6 sam- lacking spacers 3, 9, 16 and 39–43. One and lacked spacers 1–3, 12–15, 21 and ples (10.0%) belonged to Haarlem 3. specimen (1.7%) was characterized as 39–42; 1 was ST 879 and lacked spac- One was prototyped as ST 35 (lack- the S family (ST 784) and lacked spacers 10, 21 and 39; and 3 isolates were ST ing 13, 29–31, 33–36 and 40 spacers), ers 9–10, 33–36 and 40.

121

Book 19-2.indb 121 2/21/2013 11:39:19 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 2 Summary of the different clades of Mycobacterium tuberculosis complex family strains identified, their shared type designations, the number of isolates and the absence of spacers for the specimen tested from Lebanese patients Clade Shared type No. of isolates Absent spacers No. of isolates T 1 53 17 33–36 13 33–36, 39–41 2 33–36, 40–41 1 33–36, 39–40 1 T 1 154 1 33–36, 5 1 T 1 751 1 33–36, 1–3, 12–15, 21, 39, 42 1 T 1 879 1 33–36, 10, 21, 39 1 T 1 272 3 33–36, 1–4, 12–13 3 T 1 Unclassified 2 1–5, 14–15, 21–22, 32–36, 42 2 T 3 149 1 33–36, 10–19 1 LAM 9 41 14 33–36, 20–24, 26–27 13 33–36, 20–24, 26–27, 13, 15 1 LAM 9 42 1 33–36, 21–24 1 LAM 8 511 1 33–36, 13–17, 19–31, 39–42 1 LAM 8 Unclassified 1 14–17, 19–31, 33–36, 39–42 1 LAM 10 61 1 33–36, 23–25 1 Haarlem 1 602 1 25–36 1 Haarlem 3 35 1 33–36, 13, 29–31, 40 1 Haarlem 3 418 1 33–36, 3–5, 31 1 Haarlem 3 775 4 28–37 2 28–37, 13 2 CAS 25 1 4–7, 21–34, 37–38 1 CAS 1199 1 4–7, 13–15, 21–34, 37–39, 2, 10, 42 1 BCG 482 2 39–43, 3, 9, 16 2 BCG (control) 683 1 39–43, 8–12, 3, 16 1 M. africanum 536 1 4–19, 39 1 Family 36 4 1 33–36, 1–24 1 Family 36 125 1 33–36, 1–24, 40–41 1 S 784 1 33–36, 9–10, 40 1 X 1 Unclassified 1 1–5, 10, 12–15, 21–22, 32–36, 42 1 H37Rv (control) 451 1 33–36, 20–21 1

Discussion [14], hence strong phylogeographical of Ethiopian women in the past few years clustering of TB bacilli populations has to work as housemaids in Lebanon. Many genetic loci within the M. tubercu- been reported. The second predominant family in losis complex genomes are polymorphic Our results found that the majority of our results was the LAM family: 25.0% and may be used for molecular evolu- the samples tested (41.6%) belonged to were LAM 9, 3.3% LAM 8 and 1.7% tionary studies [12]. Among these, the the T 1 family. This family is widely pre- LAM 10. In general, the LAM family DR locus, which consists of alternating sent in populations in all continents and has been found to be most common identical DRs and variable spacers, corresponds to about 30% of all entries in Venezuela, in the Mediterranean ba- can be genotyped using spoligotyping in the international database [15]. One sin and in the Caribbean region [16]. methods [13]. sample (1.7%) belonged to the T 3 fam- Lebanon’s geographical location in the Several studies have shown that ily, which has recently been sub-classified Mediterranean basin and the tendency there is an apparent stable association as T 3 Ethiopia [15]. The presence of of the Lebanese, since ancient times, to of TB bacilli populations with their this type with an Ethiopian origin can be migrate to Venezuela, Central America human hosts in various environments explained by the migration of thousands and to some other African countries,

122

Book 19-2.indb 122 2/21/2013 11:39:19 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

with intermittent visits to their mother USA, Australia, South Africa and the from different regions of Lebanon neces- country could explain our findings. Caribbean [25]. However, according to sitates additional studies by the simple In the Haarlem (H) lineage, 10.0% other investigations, this group of strains spoligotyping test in order to achieve ac- of our samples belonged to Haarlem is currently correlated with African- curate epidemiological tracing of specific 3 and 1.7% to Haarlem 1. In Europe, Americans, a fact that may not represent clusters and shared types which, in con- this lineage represents about 25.0% of the ancestry of this genotype [26]. junction with TB control programmes, the isolates [15]. Outside Europe, the Another sample (1.7%) belonged will help in prevention and treatment of Haarlem strains were mainly found in to M. africanum. It was first described TB. Determination of the sensitivity of Central America and the Caribbean in Senegal in 1968, in comparison to the identified spoligotypes to the various (about 25%), suggesting a link of Haar- half of the smear-positive pulmonary antituberculosis drugs is in progress. This lem to the post-Columbus European TB cases in West Africa, but has not will provide further evidence of the clini- colonization [17]. Today its widespread been found in other geographical ar- cal applicability of this work and quicker distribution in different geographical re- eas except among recent West African recognition of the long-feared virulent gions of the world such as Asia (except migrants [27]. There is a specificity of multidrug resistant strains. for India [18]), Europe and Africa has the M. africanum lineage for the African been documented [19]. The presence continent, with a decreasing gradient of of Haarlem may always be related to prevalence from west to east [27]. Conclusion European settlers [14] and in our areas, Finally, 1 sample (1.7%) was char- Our results fit fairly well the molecular ep- to the historical contacts that existed acterized as the S family. This lineage, idemiology of M. tuberculosis genotypes between Lebanon and many of the Eu- which is highly prevalent in Sicily and described in the international database. ropean communities. Sardinia, is thought to be identical to Four unclassified strains were detected: The Central Asian (CAS) lineage the F 28 in South Africa [15], although 2 strains belonging to the T 1 family covered 3.3% of the samples tested. its origin remains unknown. (octal formats 017711727760760 and Strains of this family have been reported Noticeably none of the samples 7777377777760371), 1 strain belong- in different countries of the Middle East tested belonged to the Beijing family ing to LAM 8 (777777400000371) (Islamic Republic of Iran, Pakistan, (characterized by the absence of spac- and 1 strain belonging to X 1 India and Afghanistan) [20], and to a ers 1–34), which is quite prevalent in (017100777760760). However, the lesser extent in several other regions countries of the Middle East and Far absence of Beijing and East African In- (Africa 5.3%, Central America 0.1%, East. Actually, they represent around dian is not consistent with the patterns Europe 3.3%, Far East Asia 0.4%, North 50% of the strains in Far East Asia, 16.5% reported in neighbouring countries, America 3.3% and Oceania 4.8%). in the Middle East and Central Asia, despite the fact that Lebanon is open to Another 3.3% of the samples be- 17.2% in Oceania [15], 10% in India workers from the Far and Middle East longed to Family 36. Until now, this [18], and 13% of isolates globally [15]. (e.g. Philippines, Bangladesh, Nepal, Sri family has been identified as solely of This genotype, which may have been Lanka, India, Pakistan, Islamic Republic USA origin [21]. endemic in China for a long time [28], of Iran, Russia, Iraq), from Africa (e.g. is emerging in some parts of the world, There were 3.3% of samples belong- Egypt, Sudan, Ethiopia), in addition especially in countries of the former ing to M. bovis. This strain is character- to the multinational forces lodging in Soviet Union, and to a lesser extent in ized by the presence of spacers 33 and south Lebanon. A more inclusive study the Western hemisphere [29]. Either 34 [22] and shares 99.95% identity with of the Lebanese population is necessary this genotype is actually absent in our M. tuberculosis [23]. Humans are rarely af- to more accurately identify most of the country or additional samples should be fected, but people in some occupations prevailing families in our country and tested in order to confirm this finding. such as veterinarians, farmers and abat- further extrapolate the benefits for pre- toir workers may be more at risk [23]. Another spoligotype that could not vention and possibly treatment of TB. There was 1 sample (1.7%) belong- be detected was the East African Indian ing to the X 1 family. The X super-family lineage. This super-family is highly prev- is subdivided into at least 3 distinct fami- alent in Far East Asia (33.8%), in the Acknowledgements lies (X1 to X3) [24]. Its prevalence in Middle East and Central Asia (24.3%) North America is 21.5%, and in Central and in Oceania (22.9%). This work was totally funded by the America 11.9%. It has been linked to The current study demonstrated the Lebanese University. TheSpolDB4 an Anglo-Saxon ancestry, since it has circulation of several M. tuberculosis spoli- database was kindly provided by Chris- been encountered in the UK and in gotypes in the Lebanese population. The topher Sola and was used as a basis for English-colonized areas such as the genetic diversity of M. tuberculosis isolates data analysis.

123

Book 19-2.indb 123 2/21/2013 11:39:19 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

References

1. Murray M. Determinants of cluster distribution in the mo- 15. Brudey K et al. Mycobacterium tuberculosis complex genetic di- lecular epidemiology of tuberculosis. Proceedings of the Na- versity: mining the fourth international spoligotyping database tional Academy of Sciences of the United States of America, 2002, (SpolDB4) for classification, population genetics and epidemi- 99:1538–1543. ology. BMC Microbiology, 2006, 6:23–50. 2. Gopaul KK et al. Progression toward an improved DNA ampli- 16. Sola C, Rastogi N. Molecular epidemiology and population fication-based typing technique in the study of Mycobacterium genetics of tuberculosis. Kuala Lumpur, Malaysia, Academy of tuberculosis epidemiology. Journal of Clinical Microbiology, Sciences of Malaysia, 2006. 2006, 44:2492–2498. 17. Duchêne V et al. Phylogenetic reconstruction of Mycobacte- 3. Mathema B et al. Molecular epidemiology of tuberculosis: cur- rium tuberculosis within four settings of the Caribbean region: rent insights. Clinical Microbiology Reviews, 2006, 19:658–685. tree comparative analyse and first appraisal on their phyloge- 4. Durmaz R et al. Population-based molecular epidemiological ography. Infection, Genetics and Evolution, 2004, 4:5–14. study of tuberculosis in Malatya, Turkey. Journal of Clinical 18. Kulkarni S et al. Spoligotyping of Mycobacterium tuberculosis Microbiology, 2007, 45:4027–4035. isolates from patients with pulmonary tuberculosis in Mumbai, 5. Sola C et al. Spoligotype database of Mycobacterium tubercu- India. Research in Microbiology, 2005, 156:588–596. losis: biogeographic distribution of shared types and epide- 19. Mardassi H et al. Tuberculosis due to resistant Haarlem strain, miologic and phylogenetic perspectives. Emerging Infectious Tunisia. Emerging Infectious Diseases, 2005, 11:957–961. Diseases, 2001, 7:390–396. 20. Farnia P et al. Prevalence of Haarlem I and Beijing types of My- 6. Sreevatsan S et al. Restricted structural gene polymorphism cobacterium tuberculosis strains in Iranian and Afghan MDR-TB in the Mycobacterium tuberculosis complex indicates evolu- patients. Journal of Infection, 2006, 53:331–336. tionarily recent global dissemination. Proceedings of the Na- 21. TB insight. Tuberculosis tracking and control [website] (http://tb tional Academy of Sciences of the United States of America, 1997, insight.cs.rpi.edu, accessed 28 November 2012). 94:9869–9874. 22. Singh UB et al. Predominant tuberculosis spoligotypes, Delhi, 7. The five elements of DOTS. World Health Organization [online India. Emerging Infectious Diseases, 2004, 10:1138–1142. factsheet]. (http://www.who.int/tb/dots/whatisdots/en/ 23. Taylor GM et al. Genotypic analysis of Mycobacterium tuber- index2.html, accessed 12 November 2012). culosis from medieval human remains. Microbiology, 1999, 8. Global tuberculosis control report 2012. Geneva, World Health 145:899–904. Organization, 2012. 24. Filliol I et al. Global distribution of Mycobacterium tuberculosis 9. Hamze M, Rahmo A, Saade M. Characterization of Mycobac- spoligotypes. Emerging Infectious Diseases, 2002, 8:1347–1349. terium tuberculosis of Lebanese patients by double-repetitive- 25. Sebban M et al. A data-mining approach to spacer oligonu- element polymerase chain reaction. Eastern Mediterranean cleotide typing of Mycobacterium tuberculosis. Bioinformatics, Health Journal, 2010, 16:812–819. 2002, 18:235–243. 10. Shinnick TM, Good RC. Mycobacterial taxonomy. European 26. Kempf MC et al. Long-term molecular analysis of tuberculosis Journal of Clinical Microbiology and Infectious Diseases, 1994, strains in Alabama, a state characterized by a largely indig- 13:884–901. enous, low-risk population. Journal of Clinical Microbiology, 11. Eisenach KD et al. Polymerase chain reaction amplification of 2005, 43:870–878. a repetitive DNA sequence specific for Mycobacterium tubercu- 27. De Jong BC et al. Use of spoligotyping and large sequence losis. Journal of Infectious Diseases, 1990, 161:977–981. polymorphisms to study the population structure of the Myco- 12. Mostowy S, Behr MA. The origin and evolution of Mycobacte- bacterium tuberculosis complex in a cohort study of consecu- rium tuberculosis. Clinics in Chest Medicine, 2005, 26:207–216, tive smear-positive tuberculosis cases in The Gambia. Journal v–vi. of Clinical Microbiology, 2009, 47:994–1001. 13. Groenen PM et al. Nature of DNA polymorphism in the direct 28. Qian L et al. Retrospective analysis of the Beijing family of repeat cluster of Mycobacterium tuberculosis; application for Mycobacterium tuberculosis in preserved lung tissues. Journal strain differentiation by a novel typing method. Molecular Mi- of Clinical Microbiology, 1999, 37:471–474. crobiology, 1993, 10:1057–1065. 29. Glynn JR et al. Worldwide occurrence of Beijing/W strains 14. Hirsh AE et al. Stable association between strains of Mycobac- of Mycobacterium tuberculosis: a systematic review. Emerging terium tuberculosis and their human host populations. Proceed- Infectious Diseases, 2002, 8:843–849. ings of the National Academy of Sciences of the United States of America, 2004, 101:4871–4876.

124

Book 19-2.indb 124 2/21/2013 11:39:19 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Health education and peer leaders’ role in improving low vaccination coverage in Akre district, Kurdistan region, Iraq M.A. Abdul Rahman,1 S.A. Al-Dabbagh 1 and Q.S. Al-Habeeb 1

التثقيف الصحي ودور القادة الروحيني يف حتسني املعدالت املنخفضة للتطعيم يف منطقة أكري، إقليم كردستان، العراق مسعود عبد الكريم عبد الرمحن، صميم أمحد الدباغ، قيرص صاحب حبيب

اخلالصـة:مل تتناول البحوث دور القادة الدينيني يف حتسني املعدالت املنخفضة للتطعيم عىل ٍنحو جيد. وقد َأجرى الباحثون هذه الدراسة التدخلية ُّللتعرفعىلدور محالت التثقيف الصحي والقادة الزمالء الروحيني يف حتسني معدالت التغطية بالتطعيم يف منطقة أكري يف إقليم كردستان العراق، وأجرى الباحثون محلة إعالمية شملت 30 قرية تنخفض فيها معدالت التطعيم، ودرسوا إسهام القادة الزمالء الروحيني يف 15 قرية من قبيلة سورجي املعروفة بانخفاض معدالت التطعيم فيها. ووجد الباحثون أن معدالت التغطية بالتطعيم باجلرعات األوىل والثانية والثالثة من اللقاح الثالثي ولقاح 95 95.5 2007 احلصبة خالل الفرتة التي تلت التدخل )كانون الثاين/يناير حتى حزيران/يونيو ( قد َّحتسنت ُّحتس ًنا ُي ْع َت ُّد به ًإحصائيا ) % للجرعة األوىل، % للجرعة الثانية، 84.4% للجرعة الثالثة، 80.3% للحصبة( مقارنة باملعدالت يف الفرتة قبل التدخل يف الفرتة من كانون الثاين/يناير حتى حزيران/يونيو 27.6 21.5 42.2 55.9 2006 ، ) % للجرعة األوىل، % للجرعة الثانية، % للجرعة الثالثة، %للحصبة(. كام َن َق َصت معدالت ُّالترسب من اللقاحات ًنقصا ًكبريا. وكانت نتيجة الدراسة أن التطعيم قد َّحتسن يف القرى التي ساهم فيها القادة الروحيون ُّحتس ًنا يزيد زيادة ُي ّعتد هبا ًإحصائيا عىل ما لدى القرى األخرى.

ABSTRACT The role of religious leaders in improving vaccination coverage has not been well researched. This intervention study investigated the role of a health education campaign and peer spiritual leaders in improving vaccination coverage rates in Akre district in Kurdistan region, Iraq. An information campaign was conducted in 30 villages with low vaccination coverage. The participation of peer spiritual leaders was sought in 15 villages of the Sorchi tribe known to have persistent low coverage rates. The vaccination coverage rates of DPT1, DPT2, DPT3 and measles vaccines during the post-intervention period (January to June 2007) were significantly improved (95.5%, 90.0%, 84.4% and 80.3% respectively) compared with the pre-intervention period (January to June 2006) (55.9%, 42.7%, 21.5% and 27.6% respectively). The dropout rates of those vaccines were also significantly decreased. Vaccination in villages where spiritual leaders were involved improved significantly more than other villages.

Éducation sanitaire et rôle des chefs de groupe dans l'amélioration de la faible couverture vaccinale dans le district d'Akre, région du Kurdistan (Iraq)

RÉSUMÉ Le rôle des chefs religieux dans l'amélioration de la couverture vaccinale n'a pas été bien étudié. La présente étude d'intervention a évalué le rôle d'une campagne d'éducation sanitaire et des chefs de groupe spirituels dans l'amélioration des taux de couverture vaccinale dans le district d'Akre, région du Kurdistan (Iraq). Une campagne d'information a été menée dans 30 villages où la couverture vaccinale était faible. La participation des chefs de groupe spirituels a été recherchée dans 15 villages de la tribu Sorchi, connue pour avoir en permanence des taux de couverture faibles. Les taux de couverture vaccinale pour les première, deuxième et troisième doses du vaccin antidiphtérique-anticoquelucheux-antitétanique et pour le vaccin contre la rougeole pendant la période suivant l'intervention (de janvier à juin 2007) ont nettement augmenté (95,5 %, 90,0 %, 84,4 % et 80,3 % respectivement) par rapport à la période précédant l'intervention, de janvier à juin 2006 (55,9 %, 42,7 %, 21 5 % et 27,6 % respectivement). Les taux des perdus de vue pour ces vaccins ont aussi beaucoup diminué. La vaccination dans les villages où les chefs spirituels étaient impliqués a davantage augmenté que dans les autres villages.

1Department of Family and Community Medicine, Duhok College of Medicine, University of Duhok, Duhok, Iraq (Correspondence to S.A. Al- Dabbagh: [email protected]). Received: 29/08/11; accepted: 05/02/12

125

Book 19-2.indb 125 2/21/2013 11:39:19 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction and can also reduce or remove barriers The sample for the study was by changing negative attitudes and be- achieved by selecting all the target Vaccination is considered to be among liefs about vaccination. Yet other studies population of selected villages. The vil- the most cost-effective public health have provided insufficient evidence to lages selected for inclusion in this study interventions programmes [1]. Yet the determine the effectiveness of client had to fulfil a DPT 3 coverage rate < global commitment to vaccination has education in increasing targeted vac- 20% and a target population (aged < 1 not been sustained in all developing cination coverage in adult populations year) between 15–24 infants. The latter countries. In some countries, less than at risk [9–12]. was adopted in order to have a reason- 1 in 3 children are immunized during In many developing countries, re- able size of target population for the their first year of life [2]. ligious leaders have a legitimacy that intervention. The procedure resulted In Iraq immunization coverage rates political leaders may not have in advo- in 30 villages scattered through all sub- in 2005 were less than that of neigh- cating for vaccination, but their role in districts, 15 of which belonged to the bouring countries. In the Kurdistan improving vaccination coverage has not Sorchi tribe. region the rates of fully immunized been fully documented. The aim of this Intervention infants with valid doses before their 1st work was to study the effect of health birthday were 54% in Suleimany and education and the role of peer spiritual Before starting the study, a visit was 34% in Erbil and Duhok. The dropout leaders in improving low vaccination carried out to the family of the local rates between Bacille Calmette–Gué- coverage in Akre district of Duhok gov- sheikh who was the most influential rin (BCG) and measles vaccinations ernorate in Iraq. spiritual leader for the Sorchi popula- were 30.5%, 25.0% and 20.6% in Erbil, tion, to request their help in improving Duhok and in Suleimany respectively. vaccination coverage among their tribe, Moreover, annual vaccination statistics Methods stressing the need for preventing the of Duhok in 2005 revealed coverage re-emergence of vaccine preventable rates for full immunization with oral Study setting diseases such as the pertussis epidemic polio vaccine (third dose), diphtheria- This intervention study was carried out which occurred in most Sorchi villages pertussis-tetanus (DPT) (third dose) in Akre district, which is one of the 6 during 2005–06. The sheikh’s family and measles of 64.8%, 66.1% and 60.5% districts of Duhok governorate, cover- promised support and offered help by in the governorate in comparison with ing 1800 km2 about 120 km south-east sending, in advance, a verbal message 40.3%, 40.1% and 44.3% in Akre district, of Dohuk city. Duhok governorate is to their recognized representatives in which represented the lowest cover- one of the 3 governorates of Iraqi Kurd- all tribal villages. The message was that age rates among all districts in Duhok istan region. Akre district has 5 sub- the vaccination programme is of great [3]. DPT3 coverage rates are the most districts (Bardarash, Girdasin, Dinarta, benefit for people in preventing many frequently used to monitor vaccination Bejiland and central Akre) with a total diseases and that they were requesting coverage, while dropout rates are used of 179 204 inhabitants (46 942 urban all families of the tribe to vaccinate their as indicators of a health system’s ability and 132 262 rural) belonging to 4 main male and female children, as they did to deliver services requiring multiple vis- tribes (Goran, Zebaar, Herky and Sor- with their family and relative’s children. its [4]. The Global Alliance for Vaccines chi). They also gave the research team per- and Immunization set the objective of mission to use that verbal message. reaching 90% coverage at national level Study sample For the health education stage each and 80% in each district in all countries The strategy for delivery of routine of the 30 selected villages was offered a by 2010 [2,5]. immunization via the expanded pro- visit. Two days before each scheduled In Duhok governorate about 50% of gramme on immunization (EPI) in visit, the research team met members the causes of vaccination failure or low Akre district is though 11 fixed units of the relevant local health unit or coverage rates were due to lack of infor- in 11 PHCs in the district, delivering with the peer leader in villages where mation and motivation [6,7]. Studies vaccines 1–3 times per week according no such units were available. During done in different parts of the world have to their targets, in addition to 6 mobile each visit the planned health education shown the important role of health edu- outreach teams, which visit villages programme was applied by one of the cation in improving vaccination cover- bimonthly according to a prepared researchers (M.A.) with 2 paramed- age rates [8]. Education interventions timetable. Based on DPT3 coverage ics from the vaccination unit. Invita- can help people to identify their risk rates during 2005, 260 villages did not tion for attendance was done through status, indications for specific vaccines achieve the target coverage rate of 80% using loudspeakers, sending children and the potential benefits of vaccination with 139 of them having a rate of < 20%. to nearby houses and interpersonal

126

Book 19-2.indb 126 2/21/2013 11:39:19 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

communication. The activities lasted for Table 1 Distribution of the study population aged < 1 year old by sub-district in the 3 hours and included health talks (lec- study years 2006 and 2007 tures), posters and a video film with the Sub-district No. of Estimated target population participation of local peer leaders. The villages 2006 2007 sessions were held at selected places Age (months) Age (months) traditionally known to the villagers as 6 12 6 12 collection sites during special occasions. Bardarash 12 119 238 123 246 Recognized representatives of the local Girdasin 11 101 202 105 210 sheikh’s family were involved with the Bejil 4 34 68 35 70 team in all 15 Sorchi villages. Akre centre 2 14 28 14 28 The post-intervention stage lasted Dinarta 1 11 22 12 24 6 months from January to June 2007. Total 30 279 558 289 578

All the procedures conducted, other Based on Directorate of Duhok Statistics, 2005. than health education, were exactly the same as in 2006, i.e. the routine vaccination programme usually conducted Comparison of pre- and post- (24.7%). This difference was highly sig- by the local Department of Health. intervention vaccination nificant (P < 0.001) .The DPT1–DPT3 The researchers kept a neutral attitude The DPT1, DPT2, DPT3 and measles dropout rate in Sorchi villages was also during the post-intervention stage by vaccination coverage rates during 2007 significant decreased (8.0%) compared neither intruding upon the usual vacci- in the 30 studied villages were highly with non-Sorchi villages (15.1%) (P < nation programme of the Directorate of significantly improved (95.5%, 90.0%, 0.05) (Table 5). Health nor advocating for an additional 84.4% and 80.3% respectively) com- vaccination campaign. The main role of pared with those of 2006 (55.9%, 42.7%, the research team was to coordinate the 21.5% and 27.6% respectively) (P < Discussion campaign in all aspects and maintain 0.001). During 2007 the highest cover- In Kurdistan region, significant efforts liaison with the sheikh’s family repre- age rate was that of DPT1 followed by have been focused on the vaccination sentatives in the 15 Sorchi villages. DPT2 and DPT3 (Table 2). programme since 1992. The target of Data collection and analysis The dropout rates between DPT1 the local Department of Health is that and DPT3 and between DPT1 and Data regarding the numbers of < 1 the coverage rate should exceed at least measles vaccinations both decreased year old infants vaccinated with DPT 80% of the target population. Villages highly significantly from 60.7% and 1, DPT2, DPT3 and measles were col- in Akre district, particularly those of lected. The coverage and dropout rates 49.6% respectively in 2006 to 11.5% and the Sorchi tribe, have been shown to for each village and sub-district were 15.9% respectively during (P < 0.001) have low vaccination coverage and high calculated in the same way as during the (Table 3). dropout rates, which were a stimulus to efforts to improve such unaccepted pre-intervention stage. Comparison of Sorchi and Z-test for 2 proportions was em- non-Sorchi tribes rates [3]. Several studies have shown that health education and motivation ployed to detect significant differences Generally, vaccination coverage rates between different indicators. AP -value can lead to significant improvement during 2007 were higher in the 15 Sor- in vaccination coverage [8]. There is < 0.05 was considered as statistically sig- chi tribe villages compared with the 15 nificant and < 0.01 as highly significant. little or no evidence of such trials being non-Sorchi villages. The greatest differ- conducted in Kurdistan region and or ence was that of the measles vaccination even in Iraq. Results coverage rate (P < 0.001), followed by It was decided to study villages with that of DPT2 and DPT3 coverage rates low vaccination coverage rates of < 20% Table 1 shows the distribution of the (P < 0.05), while the difference related in order to detect any benefit from the study population during 2006 and 2007. to DPT1 coverage was not significant programme and to provide the best For 2007, the study population was 578 (P = 0.148) (Table 4). assistance to these villages. Moreover infants for the whole year and as the The DPT1–measles vaccina- these villages constituted > 50% of all duration of post-intervention period tion dropout rate in the Sorchi tribe villages with < 80% coverage rate. Vil- was 6 months, the denominator of 289 villages during 2007 was much lower lages with a small target population represented half of the annual target. (8.0%) than that of non-Sorchi villages (< 14) were excluded because it was

127

Book 19-2.indb 127 2/21/2013 11:39:19 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 2 Coverage rates of first, second and third doses of diphtheria-pertussis-tetanus (DPT) and measles vaccinations among infants in the 30 studied villages in Akre district during 2006 and 2007 Vaccine Coverage rate 2006 2007 P-value No. of infants % No. of infants % vaccinated vaccinated DPT1 156 55.9 276 95.5 < 0.001 DPT2 119 42.7 260 90.0 < 0.001 DPT3 60 21.5 244 84.4 < 0.001 Measles 77 27.6 232 80.3 < 0.001

Table 3 Dropout rates between first and third doses of diphtheria-pertussis-tetanus (DPT) vaccination and between DPT1 and measles vaccination among infants in the 30 studied villages in Akre district during 2006 and 2007 Vaccine schedule 2006 2007 P-value Infants Infants Dropout Infants Infants Dropout vaccinated vaccinated rate vaccinated vaccinated rate with DPT1 with DPT3 with DPT1 with measles No. No. % No. No. % DPT1–DPT3 156 60 60.7 276 244 11.5 < 0.001 DPT1–measles 156 77 49.6 276 232 15.9 < 0.001

thought that the effects of such pro- after the implementation of the health The study also showed a signifi- gramme might be of limited value if very education programme. In an attempt cantly greater improvement in vacci- small numbers of mothers were avail- to reduce other possible confounding nation coverage post-intervention in able. Big villages were also excluded, factors, the post-intervention vaccina- villages belonging to the Sorchi tribe in as it would have been more difficult to tion programme followed the same pat- comparison with other villages. By com- implement health sessions and extra tern as the pre-intervention period as paring villages with and without a spir- time would have been needed to collect it was conducted for the same duration itual leader advocating for vaccination participants. and the same routine procedures were we were able to demonstrate a possible In this study the DPT1, DPT2, conducted, except for the implementa- association between the intervention of DPT3 and measles vaccination cover- tion of health education. It seems likely peer leaders and improvement in vac- age rates significantly improved in the therefore that the improvements were cination coverage rates. These results intervention villages in 2007, the year related to the educational activities, es- are consistent with a study conducted of the education campaign, compared pecially as 50% of causes of vaccination in Madagascar in 2003, in which com- with 2006, the year before the interven- failure or low coverage rates are believed munity leaders (local political, admin- tion. In addition, dropout rates for both to be due to lack of information and mo- istrative and religious leaders) played a DPT1–DPT3 and DPT1–measles tivation [7]. Similar results have been role in increasing vaccination coverage vaccinations decreased significantly reported in other countries [9,13–15]. above the national average (DPT3 from

Table 4 Coverage rates of first, second and third doses of diphtheria-pertussis-tetanus (DPT) and measles vaccinations among infants in the 15 Sorchi tribe villages and 15 other villages during 2007 Vaccine Coverage rate P-value Non-Sorchi villages Sorchi villages No. of infants % No. of infants % vaccinated vaccinated DPT1 139 94.0 137 96.5 0.148 DPT2 127 85.8 123 94.3 0.013 DPT3 118 79.7 126 89.3 0.018 Measles 106 71.6 126 89.3 < 0.001

128

Book 19-2.indb 128 2/21/2013 11:39:19 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 5 Dropout rates between first and third doses of diphtheria-pertussis-tetanus vaccination (DPT) and between DPT1 and measles vaccination among infants in the 15 Sorchi tribe villages and 15 other villages during 2007 Vaccine schedule Non-Sorchi villages Sorchi villages P-value Infants Infants Dropout rate Infants Infants Dropout rate vaccinated vaccinated vaccinated vaccinated with DPT1 with DPT3 with DPT1 with measles No. No. % No. No. % DPT1–DPT3 139 118 15.1 137 126 8.0 < 0.05 DPT1–measles 139 106 24.7 137 126 8.0 < 0.001

50% in 2002 to 95% in 2003) and in area, inhabiting more than 120 villages reported the sheikh’s family requested dropout reduction (below the national in Akre district. During 2005, the DPT3 their neighbours to remove their do- average of 12%) in 2 provinces [16]. It is coverage rate among them was less mestic birds and next day all such birds well known that interpersonal commu- than 10%, representing the lowest rate were slaughtered by all tribe fellows. nication activities with influential local among other tribes in the district. Sorchi Comprehensive health education leaders (religious, medical and politi- populations bear great respect, loyalty especially modulated to suit the dif- cal) can positively affect a community’s and obedience to the local sheikh’s fam- ferent educational levels with active trust and willingness to vaccinate their ily, regarding them as the reference spir- involvement of local tribes and spiritual children [10–12,17–20]. The Sorchi itual and religious authority. It has been leaders is recommended in plans to im- tribe is one of the largest tribes in the noted that when avian influenza was prove vaccination coverage rates.

References

1. Canada Communicable Diseases Report: measuring up results 12. Ndiaye SM et al. Interventions to improve influenza, pneumo- from the national immunization coverage survey in 2002. Ot- coccal polysaccharide, and hepatitis B vaccination coverage tawa, Ontario, Public Health Agency of Canada, 2004. among high-risk adults: a systematic review. American Journal 2. State of the world's vaccines and immunization. Geneva, World of Preventive Medicine, 2005, 28(Suppl.):248–279. Health Organization, 2002. 13. Zimicki S et al. Improving vaccination coverage in urban areas 3. Annual review of Duhok DOH: activities in 2005. Duhok, Iraq, through a health communication campaign: the 1990 Philip- Directorate of Health Dohuk, 2006. pine experience. Bulletin of the World Health Organization, 1994, 72:409–422. 4. Bos E, Batson A. Using immunization coverage rates for monitoring health sector performance. Measurements and interpretation is- 14. Jassim AK et al. Visceral leishmaniasis control in ThiQar Gov- sues. HNP Discussion Paper. Washington DC, World Bank, 2000. ernorate, Iraq. Eastern Mediterranean Health Journal, 2003, 5. Immunize every child: GAVI strategy for sustainable immuniza- 12:933–940. tion services. New York, Global Alliance for Vaccines and Im- 15. Anjum Q et al. Improving vaccination status of children under munization, 2000. five through health education. Journal of the Pakistan Medical 6. Ministry of Health, Kurdistan Region/UNICEF Iraq. Household Association, 2004, 54:610–613. immunization coverage survey, 2007. Duhok, Iraq, Hawar Press, 16. Shimp L. Strengthening immunization programs: the communi- 2007. cation component. Arlington, Virginia, Basic Support for Insti- 7. Immunization summary: the 2007 edition. A statistical reference tutionalizing Child Survival Project (BASICS II) for the United containing data through 2005. New York, United Nations Chil- States Agency for International Development, 2004. dren’s Fund, 2007. 17. Building trust in immunization: partnering with religious leaders 8. Shefer A et al. Improving immunization coverage rates: an ev- and groups. New York, United Nation Children's Fund, 2004. idence-based review of the literature. Epidemiologic Reviews, 18. Ilika AL, Emelumadu OF, Okonkwo IP. Religious leaders’ per- 1999, 21:96–142. ceptions and support of reproductive health of young people 9. Porter RW et al. Role of health communications in Russia’s in Anambra State, Nigeria. Nigerian Postgraduate Medical Jour- diphtheria immunization program. Journal of Infectious Dis- nal, 2006, 13:111–116. eases, 2000, 181(Suppl. 1):S220–S227. 19. Obregón R et al. Achieving polio eradication: a review of 10. Das J, Das S. Trust, learning, and vaccination: a case study health communication evidence and lessons learned in India of a North Indian village. Social Science and Medicine, 2003, and Pakistan. Bulletin of the World Health Organization, 2009, 57:97–112. 87:624–630. 11. Verma AK, Bansal RK, Pawar AB. Facilitating behavioral change 20. Somali religious leaders join the fight against polio. Press release, for acceptance of oral polio vaccine. Indian Pediatrics, 2004, 18 July 2007. UNICEF [online] (http://www.unicef.org/media/ 41:951–952. media_40360.html, accessed 11 November 2012).

129

Book 19-2.indb 129 2/21/2013 11:39:20 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

High incidence of hepatitis B infection after treatment for paediatric cancer at a teaching hospital in Baghdad M. F. Al-Jadiry,1 M. Al-Khafagi,2 A.F. Al-Darraji,2 R.M. Al-Saeed,2 S.F. Al-Badri 2 and S.A. Al-Hadad 1

َّ لمعد مرتفع حلدوث العدوى بالتهاب الكبد "يب" بعد معاجلة رسطانات األطفال يف مستشفى تعليمي يف بغداد مازن فيصل اجلادري، ميثاق اخلفاجي، عامر فاضل الدراجي، رغد ماجد السعيد، صفاء فرج البدري، سلمى عباس احلداد

اخلالصـة:هتدف هذه الدراسة لتقرير معدل حدوث التهاب الكبد الفريويس لدى األطفال عقب معاجلتهم من الرسطان، ُّوللتعرف عىل املتغريات التي قد ِّتؤثر عىل هذا املعدل، ولتقييم دور التطعيم ضد فريوس التهاب الكبد "يب" يف الوقاية من العدوى. ففي الفرتة بني أيلول/سبتمرب 2007 وحزيران/يونيو 2008، درس الباحثون دراسة استقبالية 256 ًطفاليف وحدة األورام الدموية يف مستشفى محاية األطفال التعليميم يف بغداد. وقد سجل الباحثون املعطيات الديمغرافية " " والرسيرية وسوابق التطعيم، وأجروا للمرىض اختبارات ّلتحري فريوس التهاب الكبد يب وقت التشخيص )حيث كانت مجيع النتائج سلبية(، وبعد البدء باملعاجلة الكيميائية. وعند قبول األطفال املرىض يف املستشفى، كان 231 منهم )90%( قد تلقى التطعيم مرة ثانية. أما عند إعادة التقييم بعد معاجلة الرسطان فقد كشف الباحثون العدوى بفريوس التهاب الكبد "يب" لدى 70 ًمريضا )27.3%(؛ وكانت املتغريات التي تزيد من خطر العدوى بفريوس التهاب الكبد "يب" زيادة ُي ْع َت ُّد هبا ًإحصائيا هي: تشخيص ابيضاض الدم، ِّوتلقي أكثر من ثالث وحدات من الدم. عىل أن زيادة أعداد التطعيامت املضادة لفريوس التهاب الكبد "يب" يف املستشفى قد أنقصت من خطر العدوى بفريوس التهاب الكبد "يب". ويشري َّاملعدل املرتفع الكتساب العدوى بفريوس التهاب الكبد "يب" يف هذه الدراسة إىل احلاجة إىل ّحتريات أفضل ملنتجات الدم، وإىل االلتزام بالتقنيات العقيمة عند معاجلة هذه املجموعة من املرىض.

ABSTRACT This study estimated the incidence of viral hepatitis in children treated for cancer, to identify variables that could affect this incidence and to assess the role of hepatitis B virus (HBV) vaccination in preventing infection. Between September 2007 and June 2008, 256 children in the haemato-oncology unit at the Children’s Welfare Teaching Hospital, Baghdad, were studied prospectively. Demographic and clinical data and vaccination history were recorded. Patients were tested for HBV at the time of diagnosis (all were negative) and after starting chemotherapy. On admission to the unit, 231 patients (90.2%) were revaccinated. At reassessment after treatment for cancer, HBV infection was found in 70 patients (27.3%). The variables that significantly increased the risk for HBV infection were a diagnosis of leukaemia and receiving more than 3 units of blood. A higher number of HBV vaccinations in hospital reduced the risk for HBV infection. The high rate of acquisition of HBV infection found in this study indicates the need for better screening of blood products and adherence to aseptic techniques in management of this group of patients.

Incidence élevée de l'infection par le virus de l'hépatite B après un traitement pour un cancer chez l'enfant à l'hôpital universitaire de Bagdad

RÉSUMÉ La présente étude a estimé l'incidence de l'hépatite virale chez les enfants traités pour un cancer, afin d'identifier les variables qui pourraient influer sur son incidence et d'évaluer le rôle de la vaccination contre le virus de l'hépatite B dans la prévention de l'infection. Entre septembre 2007 et juin 2008, 256 enfants admis au service d'hémato-oncologie de l'hôpital pédiatrique universitaire à vocation sociale de Bagdad ont fait l'objet d'une étude prospective. Les données démographiques et cliniques ainsi que les antécédents de vaccination ont été notés. Les patients ont passé un test de dépistage de l'infection par le virus de l'hépatite B à la pose du diagnostic (tous étaient négatifs) et après l'instauration de la chimiothérapie. Lors de leur admission dans le service, 231 patients (90,2 %) ont été revaccinés. Après le traitement pour le cancer, un nouveau dépistage a été conduit et une infection par le virus de l'hépatite B a été retrouvée chez 70 patients (27,3 %). Un diagnostic de leucémie et une transfusion de plus de trois unités de sang étaient les variables augmentant nettement le risque d'infection par le virus de l'hépatite B. Un nombre plus élevé de vaccinations contre le virus de l'hépatite B à l'hôpital réduisait le risque d'une telle infection. Le taux élevé de nouvelles infections par le virus de l'hépatite B observé dans cette étude met en évidence la nécessité d'un meilleur dépistage des produits sanguins et d'une meilleure observance des techniques d'asepsie dans la prise en charge de ce groupe de patients.

1Baghdad College of Medicine; 2Children Welfare Teaching Hospital, Paediatric Oncology Unit, Baghdad, Iraq (Correspondence to M.F. Al-Jadiry: [email protected]). Received: 08/07/11; accepted: 01/11/11

130

Book 19-2.indb 130 2/21/2013 11:39:20 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction of child, date of birth, sex, residence, exact test were used to compare groups. type of cancer, date of diagnosis, re- Statistical significance was set at P < Children living in intermediate- and sult of hepatitis screening at the time 0.05. high-endemicity areas for hepatitis B of diagnosis, duration of observation, virus (HBV) are at risk of getting the presumed number of hepatitis vac- infection, especially those children cinations given at home, number of Results with cancer who are obliged to receive HBV vaccinations given at the hos- intensive cytostatic chemotherapy, pital, result of hepatitis screening at The majority of the children (192/256, which requires multiple blood transfu- the time of reassessment, number of 75.0%) were diagnosed between 2006 sions and causes immunodeficiency. blood transfusions and hospital in and 2007; for the remainder, the diag- HBV vaccination of seronegative which they were done. The sources for nosis had been made between 2001 patients at diagnosis may be recom- this information were the parents, the and 2006. mended [1,2]. hospital records, and medical notes All patients had a negative screen Iraq is a developing country, where and documentation about the child's for HBV and HCV at the time of diag- HBV and hepatitis C virus (HCV) treatment held by the parents. Simple nosis (the result for HCV was missing infections are still prevalent, with an written consent was obtained from for 1 patient). At the time of reassess- HBV carrier rate of 2%–5%. Although the families although all procedures ment, 70/256 (27.3%) patients were Iraq includes HBV vaccination in its and investigations were within the positive (screening and confirma- Expanded Programme on Immuniza- normal range of our work as we do tory test) for HBV, and 19/244 (7.8%) tion, the coverage rate is less than 80% frequent screening for hepatitis (every were positive for HCV in the screening [3,4]. 3 months) because of high incidence test (12 patients were not screened for HCV because of lack of laboratory Hospital-acquired HBV and HCV in our patients. materials). infections continue to occur despite The factors which were considered increased awareness of the problem might have a statistically significant The median age of the children in the among the medical community [2]. impact on the incidence of hepatitis B study was 5 years and 5 months (range, Children with cancer need frequent among the children were age, type of 3 months to 15 years and 5 months); blood counts, invasive diagnostic pro- malignancy, residence, number of blood 168 (65.6%) were boys (male: female cedures (e.g. bone marrow aspiration transfusions and hospital in which they ratio, 1.9:1) (Table 1). The patients and biopsy), intravenous therapy and were done, vaccination at home in the were referred from various parts of Iraq, surgery, which may contribute to their national immunization programme and however, the majority (149, 58.2%) risk for acquiring HBV infection. vaccination at the hospital. were from Baghdad. We determined the risk for infection Blood samples were tested for hepa- The duration of observation from with hepatitis viruses among children titis B surface antigen (HBsAg) and the time of diagnosis of any cancer until being treated for cancer at our hospital, anti-hepatitis C antibody (anti-HCV) the last screen for hepatitis was 2–82 the Children’s Welfare Teaching Hos- with commercially available enzyme- months, with a median of 15.5 months. pital, Baghdad. We also assessed the linked immunosorbent assay (ELISA) The majority of patients (166, 64.8%) role of HBV vaccination in preventing diagnostic kits. Hepanostika HBsAg had leukaemia and 59 (35.5%) of these infection. ultra 576T (bioMèrieux) was used to were found to be HBsAg positive, the screen for HBV and Hepanostika HB- earliest at 3 months and the last at 80 sAg Ultra 25T for confirmation. Bio- months during the 2–82 months obser- Methods elisa HCV 480T (Biokit) was used to vation period (Table 1). screening for HVC, but no confirma- The number of units of blood trans- Between September 2007 and June tory tests were available. fused ranged from 0 to 30 units, with a 2008, all 256 children ≤ 15 years old Multi-transfused patients were de- median of 7 units; 158 children (of the (the upper limit for receiving children in fined as those who had received more 231 who received transfusions) were this hospital is 14 years) who were being than 3 units of blood. transfused exclusively in our hospital treated for cancer or who had finished (Table 1). treatment and were attending for follow- Statistical analysis Most children had received 3 up at the Children’s Welfare Teaching Statistical analysis was performed with doses of HBV vaccine at home accord- Hospital, Baghdad, were studied. GraphPad Instat 3 for Windows. Descrip- ing to the national Iraqi vaccination Demographic and clinical infor- tive statistics were reported. The chi- schedule (3 doses: birth, 2 months and mation was obtained, including name squared test for trend and the Fisher 6 months); 40 children had received

131

Book 19-2.indb 131 2/21/2013 11:39:20 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

fewer than 3, and 14 children had not Statistically significantly higher inci- Discussion received any (Table 1). It is the policy dences of hepatitis B were found among of our unit that all children receive HBV children with leukaemia (35.5%) A high incidence of HBV infection vaccination on admission: 231 children (P < 0.0001) and among those who (27%) was found in children with can- received a variable number of doses, had received more than 3 units of blood cer treated in our unit, especially among while 25 children received no vaccine (34.0%) (P < 0.0105); a greater number those with leukaemia and those who owing to a shortage of supply (Table of doses of hepatitis B vaccine reduced received more than 3 units of blood; 1). The rate of infection with HBV was the incidence (P < 0.0001). however, vaccination at the hospital had 28/152 (18.4%) among children re- Duration of follow-up did not a significant impact on preventing HBV ceiving 3 or 4 doses of HBV vaccine in show any statistically significant cor- infection. hospital and 42/103 (40.7%) receiving relation with incidence of hepatitis (P The prevalence of HBsAg at the < 3 doses. = 0.1275). time of diagnosis was zero, which is

Table 1 Patient characteristics and incidence of hepatitis B at time of reassessment 2–82 months after treatment at the Children’s Welfare Teaching Hospital (CWTH), Baghdad Characteristic Total HBsAg +ve P-value No. % No. % Age (years) 0.0680a 1–6 177 69.1 42 23.7 > 6 79 30.9 28 35.4 Sex 1.000a Male 168 65.6 46 27.3 Female 88 34.4 24 27.2 Site of cancer Leukaemia 166 64.8 59 35.5 Other 90 35.1 11 12.2 Residence 0.5710a Baghdad 149 58.2 43 28.8 Other 107 41.8 27 25.2 Blood transfusion (no. of units) 0.0105 0 25 9.8 1 4.0 1–3 131 51.2 35 26.7 > 3 100 39.0 34 34.0 Place of transfusion 0.8776a CWTH 158 68.4 48 30.3 Other 73 31.6 21 28.7 Vaccination at home (no. of doses)b 0.12 3 202 78.9 54 26.7 1–2 40 15.6 9 22.5 0 14 5.4 7 50.0 Vaccination at hospital (no. of doses)b 0.0001 4 15 5.9 2 13.3 3 137 53.5 26 18.9 2 44 17.2 17 38.6 1 34 13.3 8 23.5 0 25 9.8 17 68.0

aFisher’s exact test, otherwise χ-squared test. bPossible reasons for not receiving the full complement of vaccinations include: young age, lack of vaccine in the primary health clinic, ignorance, inability to go to the primary health clinic for security reasons. Also, each primary health clinic serves a certain residential area and if a child/family is displaced they cannot get the vaccine. HBsAg = hepatitis B surface antigen.

132

Book 19-2.indb 132 2/21/2013 11:39:20 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

lower than the estimated carrier rate vials or through contamination of A shortage of kits for measuring anti- in our country, perhaps reflecting the injectable medications or flush solu- bodies to HBsAg precluded us from success of the active Expanded Pro- tions, the lack of an association with assessing the response to vaccination. gramme for Immunization in the coun- length of follow-up suggests that long- Our results show, however, a statisti- try and the young age of the sample. The term exposure to medical and nursing cally significant association between prevalence of HBsAg among children procedures does not increase the risk the number of doses of vaccine given after treatment for cancer in this study if basic infection control measures are in hospital and a reduction in the inci- was, however, higher than that found respected. dence of hepatitis, which implies that in other studies in the region. Mostafa Our finding that children treated the strict immunization policy we use et al. [5] in Egypt showed that the inci- for leukaemia had a significantly higher in our oncology unit was successful. dence of HBsAg increased from 3.6% incidence of HBV infection might be re- This finding differs from findings in 2 at the time of diagnosis to 18.2% after lated to the duration of chemotherapy, studies on children with acute lympho- 6 months of therapy in 111 paediatric the degree of immunosuppression and blastic leukaemia treated at the Tata malignancies. Kebudi et al. [6] in Tur- the requirement for blood products. Memorial Hospital in India. In the first key studied the prevalence of HBV and The effect of a greater number of blood study, only 10.5% of 162 patients devel- HCV infections in 50 children with transfusions on the incidence of HBV oped protective levels of anti-HBsAg solid tumours at diagnosis who under- infection might point to inadequate after 3 double doses of HBV vaccine, went intensive chemotherapy and mul- screening in the national blood bank and 48.0% became infected with HBV tiple transfusions. At diagnosis, 4% were centre. Several factors can lead to hu- [10]. In the second study, children HBV-positive and 2% HCV-positive; man immunodeficiency virus (HIV), were given 5 primary doses at monthly these values rose significantly to 20% HCV or HBV infections through blood intervals, followed by a booster 1 year and 14%, respectively after therapy. donation: donation of blood during after the first dose. Serum antibodies The incidence of HCV infection in the infectious “window period” follow- were detected in only 30% of children our study was lower than that in other ing infection when the available tests who received all 6 doses of vaccine, and 43% became infected with HBV studies of children with and without cannot detect infection; false-negative [11]. Fioredda et al. in Italy, however, malignancies in developing countries results due to poor test sensitivity (less found that the addition of boosters to [7,8]. This finding might raise questions than 100%), and blood donations the regular vaccination schedule was about the accuracy of HCV screen- falsely labelled as virus-negative due to effective if they were given at least 6 ing in our study, which depends on errors in sampling, testing or recording months after the end of treatment [12]. the method and on the quality of the of test results. In addition, blood dona- Further study is needed to confirm available kits. The lack of polymerase tions collected from individuals with these observations. chain reaction (PCR) testing kits in Iraq fluctuating or waning levels of HBsAg precluded confirmation of the presence during the later stages of HBV carriage of HCV in our patients. can give false-negative results during Conclusions We found no effect of age on the donation [3]. incidence of HBV infection, in accord- Our study shows that prior vac- The high rate of acquisition of HBV ance with the study of Baytan, Gunes cination does not prevent infection, infection in children with malignancy and Gunay in Turkey [9]. perhaps raising the question of the effi- found in this study indicates the need Although there is no guarantee cacy of the vaccines given, the reliability for better screening of blood products that the practices of the paramedi- of histories taken from the parents and and rigorous adherence to aseptic tech- cal staff in our unit led to infection the degree of attenuation of the im- niques in the management of this group through the reuse of single medication mune system during chemotherapy. of patients.

References

1. Kebudi R, Agaoglu L, Badur S. The seroprevalance of HIV-1 3. Hepatitis B. Geneva, World Health Organization, 2008 and HBV infections in multitransfused pediatric hematology- (Factsheet No. 204) (http://www.who.int/mediacentre/ oncology patients in Istanbul. Journal of Pediatric Hematology/ factsheets/fs204/en/, accessed 28 December 2012). Oncology, 1992, 9:389–391. 4. Immunization service delivery. Hepatitis B. Countries using 2. Dumpis U et al. An outbreak of HBV and HCV infection in a hepatitis B vaccine. Geneva, World Health Organization, paediatric oncology ward: epidemiological investigations and 2009 (http://www.who.int/immunization_delivery/new_ prevention of further spread. Journal of Medical Virology, 2003, vaccines/hepb/en/index4.html, accessed 28 December 69(3):331–338. 2012).

133

Book 19-2.indb 133 2/21/2013 11:39:20 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

5. Mostafa A et al. Seroprevalence of hepatitis B and C in pediatric 9. Baytan B, Gunes AM, Gunay U. Efficacy of primary hepatitis B malignancies. Journal of the Egyptian National Cancer Institute, immunization in children with acute lymphoblastic leukemia. 2003, 15:33–42. Indian Pediatrics, 2008, 45:265–270. 6. Kebudi R et al. Seroprevalence of hepatitis B, hepatitis C and 10. Goyal S et al. Hepatitis B vaccination in acute lymphoblastic human Immunodeficiency virus infection with cancer at di- leukemia. Leukaemia Research, 1998, 22:193–195 agnosis and following therapy in Turkey. Medical and Pediatric 11. Somjee S et al. Hepatitis B vaccination in children with acute Oncology, 2000, 34:102–105. lymphoblastic leukemia: results of an intensified immunization 7. Mollah AH et al. Common transfusion-transmitted infectious schedule. Leukaemia Research, 1999, 23:365–367. agents among thalassaemic children in Bangladesh. Journal of 12. Fioredda F et al. Re-immunisation schedule in leukaemic Health, Population and Nutrition, 2003, 21:67–71. children after intensive chemotherapy: a possible strategy. 8. De Paula EV et al. Transfusion-transmitted infections among European Journal of Haematology, 2005, 74:3–20. multi-transfused patients in Brazil. Journal of Clinical Virology, 2005, 34(Suppl. 2):S27–S32.

Hepatitis B and C in the Eastern Mediterranean Region

Despite the availability of effective prevention strategies, hepatitis B virus (HBV) and hepatitis C virus (HCV) transmission occur throughout the WHO Eastern Mediterranean Region. In fact it is estimated that around 4.3 million persons are infected with HBV and 800 000 persons with HCV in the Region each year. The cost to treat patients with chronic HBV or HCV infection far outweighs the cost of implementing prevention programmes. Strategies recommended by the World Health Organization to reduce HBV and HBC infection include: sustainable hepatitis B vaccination of all infants, including provision of a first dose of vaccine within 24 hours of birth; vaccination and education of all persons occupationally exposed to blood and also health care students, and education and quality assurance in health care to ensure patient safety, injection safety and safe dental care. Source: The growing threats of hepatitis B and C in the Eastern Mediterranean Region: a call for action (http:// applications.emro.who.int/docs/EM_RC56_3_en.pdf)

134

Book 19-2.indb 134 2/21/2013 11:39:20 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Risk factors for human brucellosis in northern Jordan M.N. Abo-Shehada 1 and M. Abu-Halaweh 2

عوامل اختطار داء الربوسيالت البرشي يف شامل األردن حممود أبو شحادة، مروان أبو حالوة اخلالصـة:إن املعلومات حول عوامل اختطار داء الربوسيالت يف األردن قليلة. وقد أجرى الباحثان دراسة للحاالت والشواهد شملت 56 أردنيا 247 عوجلوا إلصابتهم بداء الربوسيالت مع ثالثة شواهد عىل األقل يف مقابل كل حالة، فبلغ عدد الشواهد ، وكانت ُ هأوج التقابل هي اجلنس والعمر واملوقع )يف نفس القرية(، واملعيار االجتامعي واالقتصادي. وباملجمل درس الباحثان 17 ًعامالمن عوامل االختطار تتعلق باملخالطة ملختلف املوايش، واستهالك اللبن احلليب ومنتجات احلليب، ومعاجلة مياه الرشب والوعي باملرض. وكانت أكثر املتغريات ترتبط بداء الربوسيالت يف التحليل الوحيد 3.5 املتغري، إال أن النموذج اللوجستي النهائي َّيتضمن أربعة متغريات فقط، هي: َح ْلب النعاج والعنـزات )معدل األرجحية (، واستهالك جبن أبيض )فيتا( غري مطبوخ مصنوع من ألبان النعاج والعنـزات )معدل األرجحية 2.8(، واستهالك حليب البقر )معدل األرجحية (،0.4 واستهالك اجلبن 0.4 األبيض )فيتا( املطبوخ )معدل األرجحية (. ويرى الباحثان رضورة تدريب الفئة القليلة املتبقية من املزارعني حول املامرسات األكثر ًأمانا َللح ْلب وإلجراءات إعدادات اجلبنة )فيتا( البيضاء.

ABSTRACT Little is known about the risk factors of human brucellosis in Jordan. A case–control study was conducted involving 56 Jordanians who had been treated for brucellosis and at least 3 matched controls for each case (n = 247). Matching was for sex, age, locality (the same village) and socioeconomic standard. Univariate and multivariate logistic regression analyses were used. In all, 17 risk factors were examined related to: contact with various livestock, milk and milk product consumption, drinking-water treatment and disease awareness. Most variables were associated with brucellosis in the univariate analysis but the final logistic model included only 4: milking sheep and goats (OR 3.5), consumption of raw feta cheese made from sheep and goat milk (OR 2.8) and consumption of cows’ milk (OR 0.4) and boiled feta cheese (OR 0.4). Small ruminant farmers need to be trained in safer milking practices and feta cheese making procedures.

Facteurs de risque de la brucellose humaine dans le nord de la Jordanie

RÉSUMÉ Les données sur les facteurs de risque de la brucellose humaine en Jordanie sont rares. Une étude cas- témoins a été menée auprès de 56 Jordaniens traités pour une brucellose et d'au moins trois témoins appariés pour chaque cas (n = 247). L'appariement concernait le sexe, l'âge, la localité (le même village) et le statut socioéconomique. Des analyses de régression logistique univariées et multivariées ont été utilisées. En tout, 17 facteurs de risque liés aux domaines suivants ont été examinés : contact avec du bétail divers, consommation de lait et de produits laitiers, traitement de l'eau de boisson et connaissance de la maladie. La plupart des variables étaient associées à la brucellose dans l'analyse univariée mais le modèle de régression logistique final n'en a conservé que quatre : la traite des brebis et des chèvres (OR = 3,5), la consommation de fromage feta cru fait à partir de lait de brebis et de chèvre (OR = 2,8), la consommation de lait de vache (OR = 0,4) et de fromage feta bouilli (OR = 0,4). Les éleveurs de petits ruminants doivent être formés à des pratiques de traite et à des méthodes de fabrication du fromage feta plus sûres.

1Department of Basic Veterinary Medical Sciences, Faculty of Veterinary Medicine, Jordan University of Science and Technology, Irbid, Jordan (Correspondence to M.N. Abo-Shehada: [email protected]). 2Department of Biotechnology and Genetic Engineering, Faculty of Science, Philadelphia University, Jerash, Jordan. Received: 24/05/09; accepted: 20/07/09

135

Book 19-2.indb 135 2/21/2013 11:39:20 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction laboratory records of 2 teaching hospi- Potential exposure to livestock was ob- tals (Princess Basma and King Abdulla) tained by recording the number of the Brucellosis is a zoonosis of importance were examined for cases with Brucella- different livestock animals kept or noted in most countries of the Middle East positive isolation and/or were seroposi- in the immediate area of the residence and South America and half a million tive with the Rose Bengal test. A hospital and whether these animals were ever new human cases each year are reported nurse approached each eligible patient, allowed in the house/yard at the time worldwide [1]. The annual incidence described the purpose of the study and of the visit. Because of the potential of human brucellosis ranges from less asked for participation in the study. The indirect transmission of Brucella spp. than 1 to 78 cases per 100 000 popula- consent of the patient/patient’s legal through drinking-water, contact with tion in the Middle East [2]. Reported guardian was sought where appropri- ruminants and their manure, consump- numbers greatly underestimate the true ate. Out of 65 confirmed brucellosis tion of milk and milk products, related incidence of human brucellosis because patients asked to participate, 56 (86%) questions were asked. of variability in the clinical picture [3], (42 males and 14 females) agreed to To reduce recall-bias especially on misdiagnosis and under reporting to lo- participate and signed a consent form. diet, further questions were asked on cal authorities. Brucella abortus, B. canis, All cases selected had received treat- the source of milk products, i.e. home- B. suis and especially B. melitensis are able ment during the period January to July made or where the item(s) were usually to cause human infection [2]. The only 2007. The inclusion criteria for cases bought, and children’s mothers were species isolated in Jordan is B. melitensis were: brucellosis positive with a history asked for confirmation. [4]. of fever, and/or sweating and/or joint Data analysis In Jordan, a higher seroprevalence pain and Brucella spp. isolation (n = of 8% was reported among high-risk 8) or seropositive (n = 56) with Rose Data were stored and analysed using people compared to only 0.5% in the Bengal test and confirmed with ELISA; Epi-Info, version 6 according to the control sample [5], and an even higher resident in northern Jordan; and agree- case–control design and compared in seroprevalence of 55% was reported ment to participate in the study. relation to the exposure to the follow- among veterinarians [6]. Brucellosis Each case was matched with at least ing potential risk factors: livestock near is usually transmitted to man by the 3 controls on age, sex, locality (same house; contact with sheep; contact with ingestion of unpasteurized dairy prod- village) and socioeconomic status. goats; contact with cattle; slaughtering ucts or by direct contact with infected Their consent was also obtained. The animals; contact with manure; help in animals. The results of previous studies inclusion criteria for the controls were: animal delivery; milking small rumi- on high-risk people emphasized the no history of brucellosis and seronega- nants; consumption of sheep’s milk, importance of contact infection and the tive by the Rose Bengal test, resident in goats’ milk, cows’ milk, boiled milk, occupational nature of the disease in northern Jordan and agreement to par- feta cheese and boiled feta cheese; boil Jordan [5,6]. ticipate. water; filter water; and have knowledge To date, few community-based The total sample recruited was 303 of brucellosis. studies have been carried out in Jordan people. The study had 80% power at the Uni- and multivariate analyses were and little is known about the risk factors 5% significance level to detect an odds performed using the chi-squared test of human brucellosis in the Middle East. ratio (OR) ≥ 2 for risk factors present in and logistic regression. Seventeen di- We conducted a case–control study 50% of controls, and an OR ≥ 3 for those chotomous variables were studied in to investigate possible risk factors for present in 20% of controls. the univariate analysis. A multivariate human brucellosis in northern Jordan, analysis was then conducted separately Data collection namely contact with livestock, milk and for each group of factors and the OR milk products consumption, drinking- Patients and controls were visited and and its 95% confidence intervals (CIs) water treatment and disease awareness. interviewed at their homes. Data were were calculated, starting with the factors collected using purposely designed, that were statistically significant in the pretested and validated questionnaire; univariate analysis (P ≤ 0.05) or an OR Methods completion was interviewer-assisted by ≤ 0.3 or ≥ 3.0. a trained nurse and it was conducted in A forward stepwise simple logistic Sample Arabic. The interviewer was supervised regression analysis was done includ- A matched case–control study was for the first 10 interviews. Information ing the significant variables to control undertaken of indigenous Jordanians collected included; name, age, gender, for confounding and get a final logistic who had recently been treated for bru- place of residence, medication or regression model. Only those factors cellosis in Irbid, northern Jordan. The health history and telephone number. that remained statistically significant in

136

Book 19-2.indb 136 2/21/2013 11:39:20 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

the final model are presented. A P-value house, contact with sheep, contact consumption of raw feta cheese were of < 0.05 was considered statistically with goats, contact with manure, and positively associated with brucellosis, significant. milking animals (Table 1). Five factors while the consumption of cows’ milk of milk and milk product consump- and boiled feta cheese decreased the Ethical considerations tion, including consumption of sheep risk of brucellosis (Table 4). The ap- The study protocol was approved by and/or cows’ milk, boiled milk, feta parent significant association of the the concerned committees at Jordan cheese and boiled feta cheese, were also other factors found in the univariate University of Science and Technol- significantly associated with brucel- analysis disappeared after adjustment ogy. Informed consent was given and losis in the univariate analysis (Table for confounding. all identifying information was kept 2). On univariate analysis, boiling confidential. drinking-water was significantly associ- Discussion ated with brucellosis (Table 3). These Results 11 significant factors were entered in In our study, although 11 statistically multivariate logistic regression analysis significant risk factors for brucellosis were Of the 17 examined factors; 11 were using forward stepwise approach. In found in the univariate analysis, only 4 associated with brucellosis in the uni- the final model after controlling for all factors remained associated with infec- variate analysis at P < 0.05. Of these, the variables only 4 factors remained tion after adjustment for confounding in 5 involved contact with animals, significantly associated with brucel- the logistic regression analysis: milking namely: keeping livestock near the losis: milking small ruminants and the small ruminants, the consumption of raw

Table 1 Animal contact risk factors for human brucellosis in northern Jordan: univariate analysis Variable Controls Cases OR (95% CI) P-value (n = 247) (n = 56) No. No. Livestock near house No 219 38 3.7 (1.9–7.3) 0.01 Yes 28 18 Contact with sheep No 211 34 3.7 (2.0–7.2) 0.01 Yes 36 22 Contact with goats No 215 34 4.4 (2.3–8.5) 0.01 Yes 32 22 Contact with cattle No 239 52 2.3 (0.7–7.9) 0.30 Yes 8 4 Slaughtering animals No 228 47 2.3 (0.98–5.4) 0.08 Yes 19 9 Contact with manure No 213 38 3.0 (1.5–5.8) 0.01 Yes 34 18 Help in animal delivery No 229 49 1.8 (0.7–4.6) 0.30 Yes 18 7 Practise milking small ruminants No 232 41 5.7 (2.6–12.5) 0.01 Yes 15 15

OR = odds ratio; CI = confidence interval.

137

Book 19-2.indb 137 2/21/2013 11:39:21 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 2 Milk and milk product consumption as risk factors for human brucellosis in northern Jordan: univariate analysis Milk product consumption Controls Cases OR (95% CI) P-value (n = 247) (n = 56) No. No. Sheep’s milk No 83 9 2.6 (1.2–5.7) 0.01 Yes 164 47 Goats’ milk No 161 30 1.6 (0.9–2.9) 0.10 Yes 86 26 Cows’ milk No 88 36 0.3 (0.2–0.6) 0.01 Yes 159 20 Boiled milk No 30 17 0.3 (0.2–0.6) 0.01 Yes 217 39 Feta cheese No 127 18 2.2 (1.2–4.1) 0.01 Yes 120 38 Boiled feta cheese No 69 28 0.4 (0.2–0.7) 0.01 Yes 178 28

OR = odds ratio; CI = confidence interval.

feta cheese, and consumption of cows’ in the uterine fluids, as well as in milk In northern Jordan, the flock-level milk and boiled feta cheese. The associa- [2]. During miscarriage the fetal fluids prevalence of brucellosis among sheep tion with small ruminants accords with wet the animal’s back and udder. This is 56% [8] and 54% in goats [9] and the findings of Al Sekait [7] from Saudi Brucella rich fluid will contaminate the represents a major source of infection Arabia, where similar small ruminant environment and the rear parts of the ani- to other animals [10] and humans di- husbandry is practised. Infected animals mal exposing animal handlers, especially rectly through contact and indirectly secrete Brucella spp. in large numbers milkers, to a high risk of infection. through contact with contaminated

Table 3 Univariate analysis of drinking-water treatment and disease awareness as risk factors for human brucellosis in northern Jordan Variable Controls Cases OR (95% CI) P-value (n = 247) (n = 56) No. No. Water treatment by: Boiling No 215 55 0.1 (0.02–0.9) 0.02 Yes 32 1 Filtering No 172 45 0.6 (0.3–1.1) 0.20 Yes 75 11 Knowledge of brucellosis No 101 16 1.7 (0.9–3.3) 0.10 Yes 146 40

OR = odds ratio; CI = confidence interval.

138

Book 19-2.indb 138 2/21/2013 11:39:21 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 4 Final logistic regression model for brucellosis in humans in northern to bias, especially selection, recall and in- Jordan ter-interviewer bias and measures were Variable OR (95% CI) P-value taken to minimize this as described. Contact with animals In addition, the use of simple regres- Milking, yes 3.5 (1.5–8.4) 0.02 sion analysis in a matched case–control Consumption of milk and milk products study results in larger odds ratios than Cows’ milk, yes 0.4 (0.2–0.8) 0.01 conditional regression analysis. How- Feta cheese, yes 2.8 (1.4–5.6) 0.01 ever, we kept the stratum sizes large to Boiled feta cheese, yes 0.4 (0.2–0.8) 0.01 minimize this effect.

OR = odds ratio; CI = confidence interval. Conclusions objects and consumption of animal found that greatest risk was associated products [4,5]. with the consumption of milk and la- This study showed association with In Jordan, the only isolated Brucella ban (buttermilk), as opposed to cheese both the practice of manual milking of species is B. melitensis [4]. Although, B. [14]. The consumption of homemade sheep and goats and the consumption melitensis infects cattle with high mor- milk products has been reported to be of raw feta cheese, and human brucel- bidity in Jordan (M.N. Abo-Shehada, associated with B. melitensis infection losis. Also, the consumption of boiled unpublished observation), the signifi- [15], which concurs with our finding of feta cheese of sheep and goat origin and cance of this infection and its role in an increased risk associated with con- cows’ milk was associated with reduced human infection has never been evalu- sumption of raw feta cheese. Locally, odds of brucellosis. ated. The current results showed bru- at home and in small dairy businesses, Our findings show that contaminat- cellosis to have an inverse association feta cheese is made from heat untreated animals and milk products remain (OR = 0.4) with cows’ milk consump- ed sheep and goat milk. As the local sources of brucellosis in Jordanians. tion. This protective effect of cow’s milk production of sheep and goat milk is Health education efforts should target can be explained because in northern seasonal, the cheese is preserved using small ruminant farmers and small rumi- Jordan cow’s milk only is available as a boiling and pickling in brine and boiled nant milk product consumers regarding pasteurised or heat-treated product in cheese is used outside the season. brucellosis symptoms and transmis- shops and supermarkets. On the other The results showed boiled feta cheese sion. Small ruminant farmers need to hand, small ruminant milk is mostly consumption to be associated with be trained in safer milking practices and purchased from farmers and con- decreased odds (0.4) of brucellosis feta cheese making procedures. These sumed directly without heat treatment contrary to that when consuming the may include washing and cleaning the as most of these farmers are nomads raw feta cheese. This association may udder before milking and the milkers’ with no milk heat-treatment facilities. explain the seasonality of brucellosis in hands after milking and heat treating the Brucellosis had an OR of 2.8 among Jordan and the Middle East, where the milk used in making feta cheese. subjects consuming raw feta cheese peak of infections occurs during the of sheep and goat origin, confirming lambing season and small ruminant earlier findings in central Greece [11], milking (raw feta cheese production) Acknowledgement Khuestan Province in Islamic Republic season and drops afterwards, when the of Iran [12] and the Gaza Strip [13]. boiled cheese is consumed. This work received financial support However, this is contrary to the find- There are some limitations to the from the Deanship of Research, Jordan ings of a study in Saudi Arabia which study. Case–control studies are prone University of Science and Technology.

References

1. Joint FAO/WHO Expert Committee. The development of new 5. Abo-Shehada MN et al. Seroprevalence of brucellosis among improved brucellosis vaccine. Geneva, World Health Organiza- high risk people in northern Jordan. International Journal of tion, 1997. Epidemiology, 1996, 25:450–454. 2. Joint FAO/WHO Expert Committee on Brucellosis. Geneva, 6. Abo-Shehada MN, Rabi AZ, Abuharfeil N. The prevalence World Health Organization, 1986. of brucellosis among veterinarians in Jordan. Annals of Saudi 3. Young E. An overreview of human brucellosis. Clinical Infec- Medicine, 1991, 11:356–357. tious Diseases, 1995, 21:283–289. 7. Al-Sekait MA. Seroepidemiology survey of brucellosis an- 4. Refai M. Incidence and control of brucellosis in the Near East tibodies in Saudi Arabia. Annals of Saudi Medicine, 1999, region. Veterinary Microbiology, 2002, 90:81–110. 19:219–222.

139

Book 19-2.indb 139 2/21/2013 11:39:21 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

8. Al-Talafhah AH, Lafi SQ, Al-Tarazi Y. Epidemiology of ovine 13. Awad R. Human brucellosis in the Gaza Strip, Palestine. Eastern brucellosis in Awassi sheep in Northern Jordan. Preventive Mediterranean Health Journal, 1998, 4:225–233. Veterinary Medicine, 2003, 60:297–306. 14. Cooper CW. Risk factors in transmission of brucellosis from 9. Al-Majali A. Seroepidemiology of caprine brucellosis in Jor- animals to humans in Saudi Arabia. Transactions of the Royal dan. Small Ruminant Research, 2005, 58:13–18. Society of Tropical Medicine and Hygiene, 1992, 86:206–209. 10. Abo-Shehada MN. Seroprevalence of Brucella species in 15. Kolar J. Control of Brucella melitensis brucellosis in develop- equids in Jordan. Veterinary Record, 2009, 165(9):267–268. ing countries. Annales de l'Institut Pasteur. Microbiology, 1987, 11. Minas M et al. Epidemiological and clinical aspects of human 138(1):122–126. brucellosis in Central Greece. Japanese Journal of Infectious Diseases, 2007, 60:362–366. 12. Alavi S, Rafiei A, Nikkhooi A. The effect of lifestyle on brucellosis among nomads in Khuzestan province of Iran. Pakistan Journal of Medical Sciences, 2007, 23:358–360.

Research priorities for zoonoses and marginalized infections: Technical Report of the TDR Disease Reference Group on Zoonoses and Marginalized Infectious Diseases of Poverty

The above-mentioned report provides a review and analysis of the research on zoonoses and marginalized infections which affect poor populations, and a list of research priorities to support disease control. The work is the output of the disease reference group on zoonoses and marginalized infectious diseases (DRG6), which is part of an independent "think tank" of international experts, established and funded by the Special Programme for Research and Training in Tropical Diseases (TDR), to identify key research priorities through the review of research evidence and input from stakeholder consultations.

The report covers a diverse range of diseases including zoonotic helminth protozoa, viral and bacterial infections considered to be neglected and associated with poverty. Disease-specific research issues are elaborated under individual disease sections and many common priorities are readily identified among the disease such as need for new and/or improved drugs and regimens, diagnostics and, where appropriate, vaccines. The disease specific priorities are described as micro priorities compared with the macro level priorities which will drive such policies as the need for improved surveillance; the need for intersectoral interaction between health, livestock, agriculture, natural resources and wildlife in tackling the zoonotic diseases; and the need for a true assessment of the burden of the zoonoses.

Further information about this and other WHO publications is available at: http://apps.who.int/bookorders/anglais/ home1.jsp

140

Book 19-2.indb 140 2/21/2013 11:39:21 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Public awareness of and support for smoke-free legislation in Turkey: a national survey using the lot quality sampling technique B. Cakir,1 T. Buzgan,2 S. Com,2 H. Irmak,2 E. Aydin 2 and C. Arpad 1

وعي َّة عامالناس ودعمهم للترشيعات حول ّاخللو من الدخان يف تركيا: مسح وطني باستخدام تقنية االعتيان جلودة التشغيلة باتو تشاكري، طوران بوزقان، رساج الدين تشوم، حسن إرماق، إيربو آيدن، جيالن آرباد اخلالصـة:بعد مرور ستة أشهر عىل إصدار الترشيع اجلديد يف تركيا حلظر التدخني يف مجيع األماكن العامة، أجرى الباحثون ًمسحا لتقييم مدى التنفيذ. ِّوتقدم هذه الورقة ًتلخيصا للنتائج الرئيسية ملدى الوعي بني َّعامة الناس والدعم الذي يقدمونه هلذا القانون اجلديد. وقد َأجرى الباحثون مقابالت شملت 32 972 من السكان البالغني الذين يمثلون السكان البالغني األتراك، ووجدوا أن نسبة مرتفعة من ٍكل من غري املدخنني )91.4%( ومن املدخنني يف الوقت احلايل )67.2(تدعم القانون اجلديد بقوة. إال أن املعارف حول املخاطر الصحية للتدخني السلبي ودعم القانون كانت أخفض لدى املدخنني مما لدى غري املدخنني. وبعد التصحيح اخلاص بالوضع من حيث التدخني، كان احلاصلون عىل معارف أفضل حول التدخني السلبي أكثر ًدعام للقانون بمقدار يقرب منمخسة أضعاف؛ مما يعني أنه ينبغي تصميم أنشطة مكافحة التدخني لتناسب االحتياجات املحلية، مع الرتكيز عىل زيادة الوعي حول املخاطر الصحية للتدخني السلبي. وقد عرضت الدراسة ًأيضا ًمثالعىل تطبيق تقنية االعتيان جلودة التشغيلة لرصد أنشطة مكافحة التبغ يف املسح الوطني.

ABSTRACT Six months after new legislation in Turkey banning smoking in all public places, a national survey was carried out to assess its implementation. This paper summarizes the main findings on the public’s awareness of and support for the new law. In a household interview survey of 32 972 adults representative of the Turkish adult population, a high proportion of both non-smokers and current smokers (91.4% and 67.2% respectively) strongly supported the new law. Knowledge about the health hazards of passive smoking and support for the law, however, were relatively lower among smokers than non-smokers. After controlling for smoking status, people with better knowledge about passive smoking were over 5 times more likely to support the new law. Tobacco control activities should be tailored to local needs, with a focus on increasing awareness about the health hazards of passive smoking. The study also demonstrated application of the lot quality sampling technique (LQT) for monitoring tobacco control activities in a national survey.

Sensibilisation et soutien du public à la législation antitabac en Turquie : enquête nationale recourant à l'échantillonnage par lots pour l'assurance qualité

RÉSUMÉ Six mois après la nouvelle législation en Turquie interdisant le tabac dans tous les lieux publics, une enquête nationale a été menée pour évaluer sa mise en œuvre. Le présent article résume les principaux résultats sur la sensibilisation du public à la nouvelle législation antitabac et son soutien en la matière. Dans une enquête par entretien auprès des ménages portant sur 32 972 adultes représentatifs de la population adulte turque, une proportion élevée de non-fumeurs et de fumeurs (91,4 % et 67,2 % respectivement) soutenait fortement la nouvelle législation. Les connaissances sur les risques sanitaires du tabagisme passif et le soutien à la loi étaient toutefois plus faibles chez les fumeurs que chez les non-fumeurs. Après vérification du statut tabagique, les adultes ayant davantage de connaissances sur le tabagisme passif étaient plus de cinq fois plus susceptibles de soutenir la nouvelle loi. Il faut adapter les activités de lutte antitabac aux besoins locaux en faisant davantage de sensibilisation aux risques sanitaires du tabagisme passif. L'étude a aussi permis de mettre en application la technique d'échantillonnage par lots pour l'assurance qualité dans le cadre suivi du activités de lutte antitabac dans une enquête nationale.

1Department of Public Health, Faculty of Medicine, Hacettepe University, Ankara, Turkey (Correspondence to B. Cakir: [email protected]). 2General Directorate of Primary Health Care Services, Ministry of Health, Ankara, Turkey. Received: 24/10/10; accepted: 18/04/11

141

Book 19-2.indb 141 2/21/2013 11:39:21 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction small geographical or population-based of 400 was divided by the number of lots areas (“lots”) using small sample sizes in determined for that city; for each lot, the In January 2008 Turkey passed new a cost-efficient way. To our knowledge, required minimum number calculated legislation on tobacco control, to be this is the first study using LQT in a was rounded to the upper integer. Deci- implemented in 2 phases. The first tobacco-related survey. Details of the sion values were calculated based on the phase prohibited smoking inside all national survey and full set of analytical topic of interest and the lot sample size public places and workplaces in Turkey tables are presented elsewhere [5]. This in that particular city. and went into effect in May 2008. The paper summarizes the main findings on The most commonly used sampling second phase, in July 2009, was to fur- the public’s awareness of and support frames were household registry records ther prevent smoking inside all public for the new smoke-free legislation and of the MoH, which are updated an- places, including inside all restaurants, demonstrates application of the LQT nually by local health officials. In cities bar and cafes and their open spaces. The for monitoring tobacco control ordi- where family medicine practice was in 18-month period between the 2 phases nances in a national survey. use (about 25 cities), family medicine was expected to help in the readjust- listings were used as the sampling frame. ment process of smokers, as an adapta- In small rural areas where no listing was tion period. Even a complete tobacco Methods available (< 5% of the whole adult popu- ban, however, may not be effective in lation), maps were used to identify areas Study design and sampling tobacco control unless it is well-known strategy for interviews. Given that individuals and supported by the public. Observa- reached at houses were often females tions and general agreement over the 6 The target population in this national, (even after 3 visits), city supervisors months after the enactment of the law cross-sectional survey was all Turkish were requested to interview a randomly seemed to indicate that there had been a adults (aged 15 years or over) residing in selected male when the selected male dramatic decrease in smoking in public all 81 cities, corresponding to about 51 could not be reached after 3 house vis- and workplaces. However, almost no million individuals. The confidence level its or to substitute a randomly chosen scientific evidence was available about was set at 95% and the accuracy level at female for an unreached female; the the impact of the new law [1] and, given ± 5%. The threshold levels to identify spe- complementary (second) lists were that the second phase of the law would cial subpopulations within each city with prepared preceding the field study ac- bring in more restrictions on smok- lower than acceptable rates were deter- cordingly. It is noteworthy that there ing behaviour, it was felt essential to mined by the Ministry of Health (MoH) was no requirement for a gender bal- investigate the public’s knowledge about as < 85% for the prevalence of awareness ance in selection of the master lists. For the new law, their support for it, the ef- of the new smoke-free legislation and/ each incomplete interview, the reason fectiveness of earlier bans on tobacco or passive smoking and < 70% for the for missing data was recorded together advertising, promotion and sponsorship prevalence of support for the new law. with the sex and age of the originally and the public’s compliance with the to- Correspondingly, 384 fully-completed selected individual and this informa- bacco control ordinances in general. The surveys in each of the 81 provinces were tion was further used to compare and ultimate goal was to be fully prepared for determined as the minimum sample contrast the characteristics of missing potential obstacles in implementation of size required [6]. The minimum num- and completed interviews. the second phase of the law. ber of questionnaires to be completed Data collection A national survey was therefore in each province was rounded to 400, carried out 6 months after the imple- corresponding to a minimum of 32 400 Questionnaire mentation of the first phase of the 100% interviews nationwide. Data were collected via face-to-face smoke-free law in Turkey. Of several In each city, the lots were deter- interviews in households from each available sampling methodologies, the mined by the provincial health directo- randomly selected individual using a authors chose to use the lot quality sam- rate as the official/administrative units standard questionnaire. This question- pling technique (LQT), which is an through which they provided educa- naire included a total of 43 questions easy, quick, low-cost method developed tional activities and/or interventions. on sociodemographic characteristics, by the World Health Organization Lots were chosen either as districts, knowledge about the health hazards (WHO) [2–4]. While traditional sur- community health centres, family medi- of passive smoking, awareness about vey methods are preferable for providing cine service areas and/or district health the new smoke-free legislation, level of information on prevalence rates at the directorate service areas, as appropriate support for the new legislation, expo- regional and national levels, this method for the health care structure in the city. sure to any anti-smoking media mes- makes it possible to conduct surveys in In addition, in each city, this sample size sages related to the publicizing of the

142

Book 19-2.indb 142 2/21/2013 11:39:21 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

new regulations and frequency of use Statistical analysis more than 25% of the answers for 43 of smoke-free eating places after the The major dependent variables in the questions, with inconsistencies in con- legislation. Inquiries were based both on study were knowledge about, support trol questions and missing information open-ended questions and on opinions for and exposure to media messages on sex and age were excluded in statisti- provided for given statements and/ about the new law. Other covariates cal analyses (n = 215 questionnaires). or conditions. For example, study par- included sociodemographic character- The results were therefore based on a ticipants rated their approval level (on istics, smoking history, exposure to pas- total of 32 972 questionnaires, com- a 5-level Likert scale) for statements, sive smoking, thoughts and behaviours pleted by adults residing in all 81 cities such as “The new legislation prohibiting on given statements (on tobacco-use in throughout Turkey. The number of smoking in workplaces will never work the public, exposure to media messages, lots studied in each city ranged from in Turkey—people will continue to etc.), (specific enquiries on) level of 4 to 28. smoke wherever they want” or “The approval/support for prohibition of Prevalence of smoking new legislation prohibiting smoking smoking in eateries in indoor places in workplaces is a good idea because but not in open areas and experience The prevalence of current smoking was exposure to secondhand smoke is a with violation of the new law over the found to be 33.8% (SE 0.07%) in the direct threat to our health”. preceding 6 months (if any). Turkish adult population, corresponding to 46.9% (SE 0.01%) of all adult Fieldwork The data analyses included frequency and percentage distributions and calcula- males (n = 16 064) and 20.9% (SD To train provincial team leaders, pro- 0.08%) of all adult females (n = 16 908). vincial supervisors, regional supervisors tion of prevalence rates with relevant standard errors (SE). Logistic regres- Of all quitters (4812 males and 2835 fe- and central supervisors a set of 3 con- males), 9.6% (SE 1.0%) of male quitters secutive “training of trainers” sessions of sion modelling was used for multivariate analysis: modelling was conducted for 5 and 14.7% (SE 1.6%) of female quitters 2–3 days were held at 3 separate loca- reported that they stopped smoking tions (to guarantee full attendance of outcomes of interest: current smoking status; having knowledge on the new within 6 months prior to the survey, the trainees). These trainings provided meaning that they stopped smoking technical instructions on determination law; support for the new law; having knowledge on health hazards of passive after the enactment of the new smoke- of lots, sample selection and other city- free legislation. specific preparations based on docu- smoking; and exposure to at least one an- ments brought from 81 provinces and ti-smoking media message. Models 5 and Awareness of and support for introduced the field manuals. 6 additionally included current smoking the new smoke-free law status of the respondents, while model 3 Prior to the field study, a pretest The rates of awareness about the new was controlled for current smoking status was conducted in the Ankara area on a law, knowledge about health hazards and the knowledge on the new law in ad- convenience sample of rural and urban of passive smoking and the level of sup- dition to the covariates in the first 3 mod- residents, aged 15 years or above. Su- port for the new law were studied by els. Based on the sampling characteristics, pervisors travelled to the sites prior to sex and current smoking status. Overall all analyses were weighted, where weights initiation of the fieldwork, assisted with 57.7% of smokers and 47.7% of non- were calculated as inverses of the sam- city-specific preparations, supervised smokers had heard/read a great deal pling fractions. Statistical significance of the local interviewers, field control- about the smoke-free law and 34.3% the differences was discussed based on lers and provincial supervisors (from of smokers and 42.5% of non-smokers 95% confidence intervals (CI) and SE, provincial health directorates) and also knew that passive smoking was a very but chi-squared test P-values were also helped them with managing problems serious risk to people’s health (Table provided, with a pre-set alpha of 0.05. during data collection and/or transfer 1). More non-smokers (91.4%) than All analyses were conducted using the of the questionnaires to Ankara. smokers (67.2%) were strongly in fa- statistical software package SPSS, version Field interviewers were chosen 15, complex samples module. vour of the new law; an additional 5.3% from among local nurses and midwives of non-smokers and 19.8% of smokers (as interviewers) and physicians (as were somewhat in favour of the new controllers or provincial supervisors). Results law (Table 1). Local interviewers, controllers and su- Respondents were asked about their pervisors were trained on how to select A total of 33 187 questionnaires were experience of seeing/reading/hearing a household and they could request returned from the field and optically messages about the new law in the me- assistance from the study centre (by scanned. Questionnaires with unac- dia and/or in workplaces/indoor places phone) if required. ceptable and unreliable entries, missing where smoking was prohibited over the

143

Book 19-2.indb 143 2/21/2013 11:39:21 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale 22009) % (SE) 1.5 (0.3) 1.5 5.8 (0.5) 5.8 (0.3) 5.1 5.3 (0.4) 5.3 2.4 (0.2) 2.4 2.0 (0.3) 0.8 (0.1) (0.0) 0.1 0.4 (0.1) 0.4 0.2 (0.0) 0.7 (0.1) 0.9 (0.1) 0.2 (0.0) 47.7 (1.2) 47.7 91.4 (0.7) 91.4 52.5 (1.1) 42.5 (1.1) 40.5 (1.1) n = ( Non-smoker Total < 0.001 < 0.001 10608) % (SE) 1.1 (0.2) 1.1 1.0 (0.3) 1.0 7.2 (0.6) 7.2 3.4 (0.6) 3.4 2.3 (0.3) 2.4 (0.5) 2.4 4.6 (0.4) 9.0 (0.8) 9.0 0.8 (0.3) (0.1) 0.1 0.4 (0.1) 0.4 0.2 (0.1) Smoker 57.7 (1.4) 57.7 19.8 (1.0) 19.8 67.2 (1.1) 67.2 35.7 (1.3) 35.7 52.9 (1.3) 34.3 (1.3) n = < 0.001 ( 13844) % (SE) 1.1 (0.2) 1.1 1.3 (0.2) 1.3 (0.2) 1.3 7.3 (0.8) 7.3 (0.4) 7.0 2.4 (0.3) 2.4 (0.3) 2.4 4.5 (0.6) 0.2 (0.0) 0.3 (0.1) 0.3 (0.1) 0.6 (0.2) 0.2 (0.0) 41.5 (1.3) 41.5 (1.3) 41.2 53.4 (1.3) 53.4 n = 42.9 (1.1) 92.0 (0.7) ( Non-smoker < 0.001 < 0.001 Females 2868) % (SE) 1.6 (0.8) 1.6 3.2 (0.5) 3.2 3.9 (0.6) 3.9 2.2 (0.8) 4.3 (1.1) 6.4 (1.3) 6.4 6.6 (0.9) 0.9 (0.7) 0.1 (0.1) 0.1 (0.1) 0.4 0.6 (0.2) 0.3 (0.2) Smoker 37.1 (1.9) 37.1 57.5 (1.8) 57.5 n = 18.5 (1.6) 53.3 (2.1) 53.3 34.0 (2.1) 69.3 (1.8) 69.3 ( < 0.001 8165) 1.4 (0.3) 1.4 1.7 (0.4) 1.7 % (SE) 3.4 (0.4) 3.4 2.3 (0.3) 2.4 (0.3) 2.4 6.4 (1.0) 6.4 0.3 (0.1) 0.1 (0.0) 0.1 0.5 (0.1) 0.2 (0.0) 0.9 (0.2) 0.3 (0.1) (0.0) 0.1 51.1 (1.5) 51.1 57.1 (1.7) 57.1 n = 44.5 (1.6) 36.7 (1.7) ( 90.5 (1.1) Non-smoker 0.283 Males < 0.001 7740) % (SE) 1.9 (0.3) 1.9 (0.0) 1.0 1.3 (0.2) 1.3 7.4 (0.6) 7.4 5.0 (0.5) 5.0 2.9 (0.5) 2.5 (0.5) 0.8 (0.4) 0.4 (0.1) 0.4 0.7 (0.2) 0.2 (0.1) Smoker 57.8 (1.8) 57.8 n = 10.1 (0.9) 10.1 33.1 (1.4) 33.1 52.7 (1.4) 36.5 (1.5) ( 20.4 (1.1) 20.4 66.3 (1.4) < 0.001 a a a Public awareness about the new smoke-free law, knowledge about health about the awareness of law, hazards for thePublic new smoke-free smoking and support passive legislation, smoke-free smoking and current by sex -value -value -value A great deal A great Somewhat little Very Nothing No idea No answer P serious risk Very Quite serious risk risk A little No risk No idea No answer P Somewhat oppose Strongly favour Strongly Somewhat favour oppose Strongly No idea No answer P Item about new law Heard/read healthBelieve health smoking to people’s of passive hazards Support for new law Table 1 Table status SE = standard error. SE = standard

144

Book 19-2.indb 144 2/21/2013 11:39:21 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

30 days preceding the survey (Table Respondents’ experiences included in models because these vari- 2). While 96.1% of smokers and 88.9% after the ban ables were either primary risk factors for of non-smokers had been exposed to Two Likert-type questions were used smoking and/or might affect individu- such messages from at least 1 source, to enquire about respondents’ experi- als’ smoking habits/smoking-related only 25.2% and 21.0% respectively had ences of visiting restaurants, bars and perceptions directly. Sex, as done in been exposed to such messages in all 4 teahouses in shopping malls after the other analysis, was considered as an of these sources. smoking ban. About 4 out of 5 smok- effect modifier, and multivariate models Awareness about the new smoke- ers (78.7%) reported that they found were done separately for males and free law (heard/read about the new such places more enjoyable or found females (Tables 4 and 5). The principal law a great deal or somewhat versus no difference than before the law Ta( - finding was that current smoking and other categories) was found to be as- ble 3). Smokers reported so in 70.9% awareness of and support for the new sociated with support for the new law of the cases. About one-fifth (20.4%) law were all significantly associated with (strongly/somewhat in favour versus of non-smokers and 9.5% of smokers individuals’ knowledge about the health other categories). In males, rates for reported that they visited restaurants, hazards of passive smoking. Those who support for the law were 92.8% (SE bars and cafes inside shopping malls had knowledge about passive smok- 0.5%) and 81.1% (SE 2.5%) for those more often after the new law. Alto- ing were more likely to be aware of the who were and were not aware about the gether, the study findings revealed that new law (OR = 2.89 and OR = 2.60 for law, respectively and in females the cor- the claims about a significant decrease males and females respectively) and to responding rates were 95.9% (SE 0.4%) in use of shopping malls and/or res- support the law (OR = 5.02 and 5.39 for and 87.7% (SE 1.5%) (OR = 3.29) taurants inside such malls due to the males and females respectively), while respectively. The association between smoking ban regulations were not current smokers were less likely to know awareness of the law and support for the well-grounded. about the risks of passive smoking (OR law was even stronger among current = 3.73 and 1.94 for males and females non-smokers. Examining awareness Risk factors respectively). and support as binary variables as de- Sociodemographic characteristics and scribed above, among current smokers potential risk factors gathered from the the rate of awareness about the new present data were simultaneously stud- Discussion law and support for it were 87.7% (SE ied to model status of awareness about 0.8) and 71.1% (SE 0.4) respectively the new law (present versus absent), Use of the LQT in this survey for males (OR = 2.90) and 88.9% (SE knowledge of health hazards of passive The study did not include a specific 1.1) and 75.1% (SE 7.6) respectively smoking (present versus absent) and methodological component to inves- for females (OR = 2.67). Among cur- support for the new law (support versus tigate the robustness of prevalence rent non-smokers the rates of awareness no support), using logistic regression estimates obtained from this study. about and support for the new law were modelling. In all models, the same 5 However, the Global Adult Tobacco 97.9% (SE 0.3) and 89.4% (SE 1.4) sociodemographic characteristics (age, Survey (GATS), an international respectively (OR = 4.49) for males and educational attainment, marital status, study [7], was being conducted in Tur- 97.9% (SE 0.5) and 89.3% (SE 2.1) number of children and presence of key at the same time as our field study, respectively for females (OR = 5.44). any paid job outside the home) were with special emphasis on tobacco use

Table 2 Public exposure to messages about the new smoke-free law from 4 sources (signs in public institution/workplace, signs in public transport vehicle/terminal, read about in newspaper/journal or heard about on radio) over the previous 30 days, by sex and current smoking status Exposed to messages about new law Males Females Total Smoker Non-smoker Smoker Non-smoker Smoker Non-smoker (n = 7740) (n = 8165) (n = 2868) (n = 13844) (n = 10608) (n = 22009) % (SE) % (SE) % (SE) % (SE) % (SE) % (SE) From at least 1 source 96.2 (0.4) 94.1 (0.6) 95.8 (0.5) 85.4 (0.7) 96.1 (0.3) 88.9 (0.5) P-valuea < 0.005 < 0.001 < 0.001 From all 4 of these sources 24.7 (1.3) 23.3 (1.0) 25.6 (1.9) 17.3 (1.0) 25.2 (1.1) 21.0 (0.8) P-valuea 0.348 < 0.001 < 0.001

aChi squared test. SE = standard error.

145

Book 19-2.indb 145 2/21/2013 11:39:21 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

among Turkish adults. Both studies used the same set of questions for sociodemographic characteristics and 22 009) % (SE) 3.7 (0.3) 3.7 5.0 (0.4) 5.0 2.1 (0.2) 2.1 2.9 (0.3) 31.2 (1.0) 31.2 15.5 (0.6) 15.5 19.6 (1.1) 19.6 59.1 (1.1) 59.1 20.4 (0.7) 20.4 40.6 (1.2) smoking history. GATS used a multi- n = ( Non-smoker staged stratified systematic cluster

Total sampling technique and the estimated < 0.001 < 0.001 prevalence of tobacco use. In GATS, on a total of 9030 adults, the preva- 10 608) % (SE) 3.6 (0.5) 3.6 2.9 (0.5) 9.6 (0.6) 9.6 (0.8) 9.5 Smoker 31.8 (1.2) 31.8 16.6 (1.0) 16.9 (1.0) 39.1 (1.2) 39.1 49.8 (1.2) 49.8 20.3 (0.8) lence of current smoking was found n = ( to be 31.1% (95% CI: 29.9%–32.4%) and the corresponding prevalence in our survey was 33.8% (SE 0.07%) for the same age group. The closeness of 13 844) 1.7 (0.3) 1.7 % (SE) 5.6 (0.5) 5.6 2.4 (0.4) 2.4 4.5 (0.4) 18.9 (0.7) 18.2 (0.8) 18.6 (1.2) 35.1 (0.9) 35.1 38.7 (1.2) 56.4 (1.4) 56.4

n = the 2 prevalence rates suggests that the ( Non-smoker LQT method can be considered as a robust method, besides its simplicity < 0.001 < 0.001 Females and ease of use.

2868) The relatively ease of the method, % (SE) 2.8 (0.7) 4.0 (0.8) 9.1 (1.0) 9.1 (1.3) 9.0 Smoker 31.6 (2.1) 31.6 n = 37.5 (2.1) 37.5 18.9 (1.8) 18.3 (2.1) 18.6 (1.3) 50.1 (2.3) 50.1 ( together with no absolute need for complete sampling frames, makes LQT-based sampling very suitable for conducting tobacco-related surveys

8165) nationwide. The LQT enabled us % (SE) 3.7 (0.4) 3.7 2.5 (0.3) 2.8 (0.4) 4.1 (0.4) 4.1 n = 21.2 (1.2) 21.2 10.4 (0.9) 10.4 23.7 (1.1) 23.7 25.2 (1.6) 25.2 43.3 (1.7) 43.3 ( 63.1 (1.4) 63.1 to identify subpopulations in which Non-smoker awareness of and support for the new Males

< 0.001 < 0.001 law were above or below the threshold of expected levels and hopefully

7740) this will allow local personnel to focus % (SE) 3.0 (0.6) 3.0 (0.5) 3.4 9.8 (0.7) 9.8 (0.8) 9.7 15.5 (1.1) 15.5 21.1 (1.1) Smoker 16.2 (1.0) 32.0 (1.4) n = 39.8 (1.4) 39.8 49.6 (1.6) 49.6 ( their future anti-tobacco efforts and resources in the most cost-efficient way. The survey was also important in terms of national capacity development for tobacco-related surveillance activities. Both the core team of the tobacco control unit in the General Directorate of Primary Care at the MoH and heads of tobacco control units in provincial health directorates had the opportunity to take an active part in planning and implementation of a large survey at the national level.

Prevalence of smoking The baseline data of this study confirmed that, despite all efforts, current smoking is still common in Turkey (46.9% of adult males and 20.9% of adult females). Quitting rates over the 6 months following the first phase of a a

Public opinions about visiting thePublic inside malls after shopping malls and restaurants new legislation prohibiting smoking and current smoking in indoor places, by sex the new smoke-free law (9.6% of males

-value -value and 14.7% of females) were about twice Don’t go to these places No answer P often More No difference often Less Don’t go to these places No answer P More enjoyable More No difference enjoyable Less Frequency of visitingFrequency new law? inside shopping malls since bars and cafes restaurants, Item enjoyable than enjoyable or less Is going to shopping malls more new law? before Chi-squared test. Chi-squared the comparable rates before [1,8]. This a error. SE = standard Table 3 Table status

146

Book 19-2.indb 146 2/21/2013 11:39:22 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين value 0.001 0.116 0.111 0.055 P-

< 0.001 32 027) n = ( a Model 6: Exposed to ≥ 1 media message OR (95% CI) 1.22 (0.59–2.52) 1.22 (0.44–3.51) 1.25 1.34 (0.52–3.47) 1.34 1.88 (0.83–4.26) 1.88 (0.30–4.22) 1.13 (0.47–3.35) 1.25 1.51 (0.99-2.30) 1.51 1.0 3.60 (2.07–6.24) 3.60 5.54 (2.97–10.4) 5.54 6.17 (3.98–9.57) 6.17 8.24 (4.59–14.8) 0.80 (0.33–1.91) 0.97 (0.36–2.60) 0.52 (0.18–1.50) 0.26 (0.09–0.76) 0.32 (0.71–1.45) 0.54 (0.26–1.12) 0.62 (0.30–1.29) 0.66 (0.28–1.55) (0.19–0.99) 0.44 41.0 (20.1–83.7) 41.0 value 0.864 0.042 0.181 0.398

P- < 0.001 32 423) n = (

a Model 5: Know about passive Know about passive smoking OR (95% CI) 1.17 (0.70–1.96) 1.17 (0.69–1.91) 1.15 (0.68–2.23) 1.24 1.49 (0.70–3.14) 1.49 (0.60–3.08) 1.36 1.75 (0.83–3.68) 1.75 (0.50–3.54) 1.33 (0.99–2.57) 1.59 (0.74–2.22) 1.28 (0.95–2.87) 1.65 (0.83–3.06) 1.60 (0.51–2.24) 1.07 (0.53–4.82) 1.60 (0.74–3.23) 1.54 (0.47–2.20) 1.02 1.00 (0.42–2.38) 1.00 1.0 1.13 (0.85–1.52) 1.13 0.76 (0.42–1.37) 0.25 (0.12–0.55) (0.16–1.05) 0.41 0.93 (0.41–2.13) value 0.017 0.019 0.237 0.360

P- < 0.001 32 423) n = Model 4: Know about passive Know about passive smoking ( OR (95% CI) 1.10 (0.51–2.38) 1.10 1.28 (0.56–2.89) 1.28 (0.90–3.89) 1.87 (0.63–4.30) 1.65 1.70 (1.12–2.58) 1.70 (0.85–2.25) 1.38 (1.15–3.10) 1.89 (1.10–3.63) 1.99 (0.52–2.28) 1.09 (0.49–4.00) 1.39 (0.70–3.20) 1.50 (0.46–2.22) 1.01 (0.41–2.54) 1.02 1.0 (0.86–1.53) 1.15 0.74 (0.44–1.24) 0.74 0.69 (0.40–1.17) 0.77 (0.41–1.42) 0.82 (0.46–1.47) 0.28 (0.12–0.62) (0.15–1.09) 0.41 0.94 (0.40–2.21)

a,b value 0.212 0.014 0.254 0.353 0.314

P- 32 384) Model 3: n = ( Support new law Support OR (95% CI) 0.95 (0.46–1.96) 1.04 (0.40–2.73) 1.04 (0.39–2.58) 1.00 (0.43–3.07) 1.14 1.10 (0.42–2.89) 1.10 2.25 (0.80–6.31) 0.66 (0.21–2.04) 1.42 (0.69–2.92) 1.42 (1.05–2.73) 1.69 (0.74–2.36) 1.32 0.96 (0.57–1.61) (0.62–1.78) 1.05 0.92 (0.45–1.90) 0.31 (0.01–1.00) 0.65 (0.13–3.36) 0.83 (0.34–1.99) (0.76–3.29) 1.59 (0.54–1.97) 1.03 (0.56–2.43) 1.16 1.16 (0.54–2.49) 1.16 1.0 (0.86–1.57) 1.17 Reference category = no children. category Reference f a value 0.018 0.012 0.166 0.949

P- < 0.001 32 407) Model 2: n = ( Aware ofAware new law OR (95% CI) 1.49 (0.80–2.78) 1.49 (0.83–2.86) 1.54 (0.81–3.21) 1.61 1.14 (0.50–2.64) 1.14 (0.44–2.78) 1.11 1.93 (0.82–4.52) 1.93 (0.78–3.65) 1.69 1.0 (0.75–1.36) 1.01 3.29 (2.09–5.17) 3.29 2.34 (0.89–6.14) 2.32 (1.08–4.98) 4.12 (2.83–5.99) 4.12 6.52 (3.76–11.3) 8.98 (4.53–17.8) 0.95 (0.56–1.63) 0.48 (0.19–1.21) 0.48 0.70 (0.30–1.63) 0.99 (0.31–3.12) 0.25 (0.06–1.02) 0.35 (0.13–0.92) 12.4 (4.84–31.7) 12.4 Reference category = Never married; category Reference e value 0.002 0.178 0.244 0.822 P-

< 0.001 32 423) Model 1: n = ( OR (95% CI) 1.34 (0.82–2.18) 1.34 1.84 (0.90–3.76) 1.84 (0.61–1.96) 1.10 (0.77–1.64) 1.12 1.0 (0.87–1.20) 1.02 Current smoking status Current 5.87 (3.92–8.81) 5.87 (3.46–8.07) 5.29 Reference category = illiterate; = illiterate; category Reference 2.74 (1.80–4.16) 2.74 4.58 (3.20–6.56) 0.83 (0.50–1.37) 0.47 (0.23–0.97) 0.47 0.69 (0.45–1.03) 0.79 (0.54–1.15) 0.80 (0.53–1.19) 0.64 (0.43–0.96) (0.32–0.75) 0.48 0.99 (0.70–1.41) 0.67 (0.36–1.25) 0.99 (0.69–1.41) 0.86 (0.58–1.28) 0.88 (0.61–1.26) d d 15–19 years; 15–19 years; f e c Logistic models for the regression 5 outcomes studied: males 20–29 30–39 40–49 50–59 60–69 70–79 80+ Literate Primary Secondary High school University Married Separated Divorced Widowed 1 2 3 4/+ Yes No Covariates in theCovariates model Age (years) Educational attainment Marital status No. of children Paid job outside home Paid This model is additionallyThis smoking status of the for current controlled (non-smokers versus smokers) and the respondent versus other). deal knowledge on the a great new law (heard/read These models also included current smoking status of the models also included current These (non-smokers versus smokers) as a potential respondent confounder. Reference category= category= Reference Table 4 Table a b c OR = odds ratio; interval. CI = confidence

147

Book 19-2.indb 147 2/21/2013 11:39:22 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale a -value 0.798 0.556 0.809 P < 0.001 < 0.001 32 027) (n = media message OR (95% CI) 1.15 (0.60–2.23) 1.15 1.21 (0.57–2.56) 1.21 1.16 (0.66–2.03) 1.16 1.23 (0.69–2.20) 1.23 1.00 Model 6: Exposed to ≥ 1 7.63 (6.02–9.67) 7.63 1.05 (0.63–1.77) 1.05 23.2 (13.3–.40.4) 23.2 0.81 (0.37–1.79)

74.6 (31.9–174) 74.6 0.43 (0.25–0.74) 0.43 2.70 (1.96–3.71) 0.93 (0.58–1.48) 0.63 (0.16–2.53) 0.95 (0.55–1.62) 8.40 (4.62–15.3) 8.40 0.82 (0.52–1.27) 0.92 (0.54–1.59) 0.29 (0.16–0.53) 0.26 (0.12–0.56) 0.69 (0.42–1.16) 0.90 (0.37–2.19) a 0.171 0.036 0.392 0.260 -value P < 0.001 32 423)

n = ( passive smoking passive OR (95% CI) Model 5: Know about 1.37 (0.68–2.75) 1.37 1.32 (0.76–2.31) 1.32 1.42 (0.75–2.72) 1.42 1.00 1.24 (0.54–2.86) 1.24 1.48 (0.99–2.20) 1.48 1.07 (0.38–3.05) 1.07 1.20 (0.59–2.43) 1.20 1.88 (1.04–3.42) 1.88 (1.25–3.17) 1.99 1.04 (0.51–2.13) 1.04 1.09 (0.59–2.00) 1.09 0.55 (0.23–1.32) 2.40 (1.81–3.17) 2.40 0.74 (0.42–1.29) 0.74 (0.39–1.40) 0.74 0.72 (0.36–1.42) 0.72 (0.41–1.27) 0.44 (0.12–1.56) 0.44 0.65 (0.28–1.50) 0.92 (0.45–1.89) 0.86 (0.25–2.93) 0.041 0.015 0.491 0.375 -value < 0.001 P 32 423) n = Model 4: Know about passive Know about passive smoking ( OR (95% CI) 1.71 (1.05–2.79) 1.71 1.01 (0.49–2.11) 1.01 1.35 (0.70–2.60) 1.35 1.43 (0.97–2.12) 1.43 1.00 1.36 (0.78–2.35) 1.36 1.65 (0.95–2.88) 1.65 1.26 (0.53–2.96) 1.26 1.30 (0.64–2.65) 1.30 1.06 (0.50–2.27) 1.06 (0.57–1.98) 1.06 0.61 (0.30–1.25) 2.24 (1.71–2.94) 0.53 (0.22–1.30) 0.77 (0.43–1.37) 0.63 (0.32–1.23) 0.38 (0.12–1.25) 0.83 (0.23–2.95) 0.65 (0.28–1.50) 0.82 (0.39–1.75) 0.66 (0.37–1.16) 0.98 (0.34–2.83)

a,b 0.055 0.014 0.075 0.622 0.951 -value

P 32 384) Model 3: n = ( Support new law Support OR (95% CI) 1.19 (0.77–1.83) 1.19 1.33 (0.75–2.38) 1.33 1.00 1.79 (0.69–4.59) 1.79 1.29 (0.46–3.63) 1.29 1.02 (0.63–1.64) 1.02 0.41 (0.15–1.15) 0.41 0.43 (0.18–1.02) 0.43 0.57 (0.22–1.47) 0.23 (0.06–0.69) 0.49 (0.20–1.22) 0.49 0.76 (0.34–1.70) 0.78 (0.34–1.78) 0.76 (0.24–2.43) 0.76 (0.24–2.37) 0.97 (0.33–2.82) 0.62 (0.33–1.17) 0.62 (0.23–1.66) 0.62 (0.32–1.19) 0.82 (0.26–2.57) 0.40 (0.16–0.99) 0.40 0.89 (0.37–2.11) Reference category = no children. category Reference f a -value 0.030 0.075 0.622 0.976 P < 0.001 32 407) Model 2: n = ( Aware ofAware new law OR (95% CI) 1.19 (0.61–2.35) 1.19 1.01 (0.57–1.80) 1.01 (0.56–1.83) 1.01 1.35 (0.87–2.09) 1.35 1.00 2.71 (1.51–4.84) 1.46 (0.62–3.44) 1.46 1.29 (0.68–2.44) 1.29 (0.68–2.42) 1.28 (8.29–56.4) 21.6 (0.45–3.70) 1.29 1.50 (0.80–2.84) 1.50 0.51 (0.23–1.13) 2.75 (2.04–3.70) 2.25 (1.57–3.25) 0.81 (0.44–1.48) 0.77 (0.39–1.53) 6.50 (4.09–10.4) 0.67 (0.31–1.45) 0.92 (0.54–1.58) 0.84 (0.45–1.57) 0.99 (0.56–1.76) Reference category = Never married; category Reference e 0.439 0.004 -value < 0.001 < 0.001 < 0.001 P 32 423) Model 1: n = ( 1.51 (0.87–2.62) 1.51 1.19 (0.55–2.57) 1.19 OR (95% CI) 1.00 1.46 (0.91–2.35) 1.46 (0.93–2.41) 1.49 1.29 (0.71–2.34) 1.29 4.33 (2.22–8.44) 3.58 (1.96–6.53) 3.58 1.37 (0.75–2.51) 1.37 1.23 (0.60–2.54) 1.23 3.16 (1.47–6.81) 3.16 Current smoking status Current 0.67 (0.51–0.88) 1.96 (0.87–4.42) 1.96 2.42 (0.97–6.09) 2.42 3.63 (2.31–5.72) 3.63 4.32 (2.09–8.94) 3.03 (1.83–5.03) 3.03 3.20 (1.81–5.68) 3.20 0.74 (0.43–1.27) 0.74 2.20 (1.50–3.24) 0.26 (0.09–0.75) 0.66 (0.14–2.99) Reference category = illiterate; = illiterate; category Reference d 15–19 years; 15–19 years; Logistic models for the regression 5 outcomes studied: females 20–29 30–39 40–49 50–59 60–69 70–79 80+ Literate Primary Secondary High school University Married, Separated Divorced Widowed 1 2 3 4/+ Yes No Covariates in theCovariates model Age (years) Educational attainment Marital status No. of children job outside home Paid This model is additionallyThis smoking status of the for current controlled (non-smokers versus smokers) and the respondent versus other). deal knowledge on the a great new law (heard/read These models also included current smoking status of the models also included current These (non-smokers versus smokers) as a potential respondent confounder. Reference category= category= Reference a b c OR = odds ratio; interval. CI = confidence Table 5 Table

148

Book 19-2.indb 148 2/21/2013 11:39:22 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

phase of the programme included ef- about 20.3% of smokers did not go to first informed about the new law, with forts to increase the public awareness shopping malls at all. Thus studies con- a standard sentence explaining it. Later, about the health hazards of smoking ducted to those visiting shopping malls, each individual was asked whether s/ and its economic burden and an in- as is usually the case when monitor- he was in favour of the law or opposed crease in the number of smoking cessa- ing tobacco control activities, cannot it. Therefore, no “differential” misclas- tion clinics. Further studies are needed capture about 20%–30% of the general sification bias (if any) was expected due to investigate whether smoking bans, adult population and inferences based to a difference (if any) in knowledge especially those in the workplace, affect on data from this method will therefore on the new law and/or current smok- quit rates directly. be biased. ing status. Besides a general statement Regardless of sex, there was a posi- about the support for the new 100% Awareness of and attitudes tive and dose–response association smoke-free law, several statements re- to the new national tobacco control laws between educational attainment and garding smoking bans were presented awareness about the new law and to the interviewed individuals and their The main findings of this survey re- about the risks of passive smoking opinions were obtained. vealed that the rate of awareness about to smoking-related media messages, the new smoke-free law was fairly high matching our expectation that bet- among the general public (47.7% of ter educated people may have more Conclusion non-smokers and 57.7% of smokers access to the media than their coun- knew about the law) and the law was terparts with lower educational attain- This study has provided important supported by a high proportion of ment. The content and transmission findings about the awareness of the Turkish adults, even among current of anti-smoking media messages may general public in Turkey about the smokers (91.4% of non-smokers and need to be improved for better use by new 100% smoke-free law and the 67.2% of smokers strongly favoured the the population with low educational level of compliance with the law. Even new law). Furthermore, it is noteworthy attainment. more important for the international that, controlling for smoking status, one Last but not least, the study re- literature, we demonstrate the use of of the strongest, yet modifiable, risk fac- vealed, at a national level, that indi- the LQT in a tobacco-related survey in tors among the studied predictors of viduals did not change their frequency a developing country. We believe that awareness and attitudes was the general of visits either to shopping malls or to conducting similar surveys would be public’s knowledge about the health the eateries inside such malls after the of great benefit in developing countries hazards of passive smoking. Those ban. Even more importantly, 59.1% not only for tobacco control surveil- who knew about the dangers of pas- of non-smokers and 31.8% of smok- lance, but to develop national capac- sive smoking were nearly 3 times more ers agreed that they found smoke-free ity to provide evidence on a variety of likely to know about the law and were shopping malls “more enjoyable than other health topics of interest at the over 5 times more likely to support before” and 20.4% of non-smokers and local level. the new law. Current smokers were about 9.5% of smokers even reported significantly less likely to have knowl- increasing their visits to smoke-free edge about the health effects of passive restaurants in shopping malls after Acknowledgements smoking than were smokers. This high- May 2008. The results can be seen lights the importance of increasing the as important in putting an end to the This survey was supported technically awareness of the general public about discussion of unfair business in the and/or financially by the World Health the health threats of passive smoking hospitality sector after the smoking Organization- Tobacco Free Initiative, in future interventional activities. Iden- ban in May 2008, in favour of the eater- World Health Organization Regional tification of subpopulations with less ies outside the indoor areas. Office for Europe, World Health -Or knowledge and/or support for the law ganization Country Office (Turkey) will lead to effective targeted plans for Limitations of the study and the US Centers for Disease Control intervention. It is important to note that the findings and Prevention; and the Turkish Min- The new 100% smoke-free law about the awareness of the new smoke- istry of Health, Primary Health Care banned smoking in closed areas in free law were based on self-reports; General Directorate, Tobacco Control Turkey such as shopping malls. It is evaluation of the actual content of this Unit, within the scope of Bloomberg important to note that 31.2% of non- knowledge was beyond the scope of this Initiative to Reduce Tobacco Use. smokers reported that they did not go study. In evaluating the level of support The authors would like to thank to eating places in shopping malls and for the new law all participants were Cevdet Erdol, Jo Bickmayer, Nichole

149

Book 19-2.indb 149 2/21/2013 11:39:22 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Veatch, Stephen Hamill, Wick C. Ayfer Aytemur. Hayati Baykan, Hü- midwives, health technicians and Warren, Juliette Lee, Veronica Lee, seyin İlter, Serdar Acar, Derya Özkaya, drivers who took part and all other Meltem Sengelen, Meliksah Ertem, Pi- Irmak Arpad, Gokturk Demirel, Toker colleagues at the provincial health de- nar Pazarli, Osman Elbek and Zeynep Erguder; and all physicians, nurses, partments.

References

1. Sferrazza L et al. Using polling for smoke-free implementa- 5. Buzgan T et al., eds. Initial 6-month evaluation of the effects of tion and enforcement in Turkey. Paper presented that the the 100 percent-smoke-free legislation and health hazards of 14th World Conference on Tobacco or Health, Mumbai, 8–12 passive smoking: and evaluation of the coverage of the related March 2009. mass media campaign. Ankara, Turkey, Ministry of Health, 2. Robertson S. Monitoring immunization services using the lot xxxx . quality technique. Geneva, World Health Organization, 1996 6. Sample LQ software package. Brixton Health [website] (http:// (WHO/VRD/TRAM/96.01). www.brixtonhealth.com/index.html, accessed 22 July 2012). 3. Lemeshow S, Taber S. Lot quality assurance sampling: single- 7. Global Adult Tobacco Survey.Turkey report—2010. Anka- and double-sampling plans. World Health Statistics Quarterly, ra, General Directorate of Primary Health Care, Ministry of 1991, 44:115–132. Health, 2010 (MoH Publication No. 803). 4. Robertson SE, Valadez JJ. Global review of health care surveys 8. Bilir N et al. Tobacco Control in Turkey Copenhagen, Den- using lot quality assurance sampling (LQAS), 1984–2004. Social mark, World Health Organization Regional Office for Europe, Science and Medicine, 2006, 63:1648–1660. 2009.

150

Book 19-2.indb 150 2/21/2013 11:39:22 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Blood lead level among Palestinian schoolchildren: a pilot study A.F. Sawalha,1 R.O. Wright,2 D.C. Bellinger,2 C. Amarasiriwardean,2 A.S. Abu-Taha 3 and W.M. Sweileh 3

مستوى رصاص الدم بني أطفال املدارس الفلسطينيني: دراسة ارتيادية أنسام صواحلة، روبرت رايت، ديفيد بيلينجر، تشيرتا اماراسرييواردين، أدهم ابو طه، وليد صويلح

اخلالصـة: مل َ ْيح َظ ُّالتعرض املزمن للرصاص يف فلسطني بالدراسة الكافية باعتباره أحد املشكالت لدى األطفال. وقد أجرى الباحثون ًمسحا ملستويات الرصاص يف الدم يف ثالث مدارس يف مدينة نابلس، ومجعوا املعطيات الديموغرافية والرسيرية لتقييم ُّتعرض أطفال املدارس الفلسطينيني للرصاص. فجمعوا عينات الدم من 178 ًطفال )140 فتى و 38فتاة(، وهم من جمموعة عمرية ترتاوح بني 6 و8 سنوات. ووجدوا أن املستوى الوسطي اإلمجايل للرصاص يف الدم 3.2 مكغ/دييس لرت ± 2.4، وأن 4.5% من األطفال كان لدهيم مستويات تزيد عىل 10 مكغ/دييس لرت. وقد كانت مستويات الرصاص يف الدم أعىل بمقدار ُي ْع َت ُّد به ًإحصائيالدى األطفال الذين يعيشون يف خميامت النازحني القريبة من مناطق صناعية أو مكتظة باحلركة املرورية، ولدى األطفال الذين يعيشون يف مناطق سكنية ضمن املدينة. وقد كانت مستويات الرصاص يف الدم تـرتابط ُتراب ًطا ً إجيابيامع حجم األرسة r =0.18 r = 0.15 ) ( ُوتراب ًطا ًسلبيا مع مساحة املسكن ) (. وكانت مستويات الرصاص يف الدم لدى أطفال املدارس الفلسطينية أعىل مما هي عليه يف البلدان األخرى التي ُ ظريح فيها الغازولني َاملعالج بالرصاص، ويبدو أنه أعىل لدى األطفال املحرومني ًاقتصاديا.

ABSTRACT In Palestine, chronic exposure to lead has not been adequately addressed as a problem for children. To assess the exposure of Palestinian schoolchildren, we surveyed blood lead levels in 3 schools in Nablus city and collected demographic and clinical data. Blood samples were collected from 178 children (140 boys, 38 girls), age range 6–8 years. The overall mean blood lead level was 3.2 (SD 2.4) µg/dL, and 4.5% of children had levels above 10 µg/dL. Blood lead levels were significantly higher among children living in refugee camps near industrial/high traffic regions than among children living in residential areas of the city. Blood lead levels were positively correlated with family size (r = 0.15) and negatively correlated with household area (r = –0.18). Blood lead levels among these Palestinian schoolchildren were higher than those of other countries where leaded gasoline has been banned and seemed to be higher in more economically deprived children.

Concentration sanguine de plomb chez des écoliers palestiniens : une étude pilote

RÉSUMÉ En Palestine, l'exposition chronique au plomb n'a pas été abordée adéquatement comme un problème chez l'enfant. Pour évaluer l'exposition des écoliers palestiniens, nous avons recherché les taux de plomb sanguins chez des élèves de trois écoles de la ville de Naplouse et avons recueilli les données démographiques et cliniques. Des échantillons de sang ont été recueillis auprès de 178 enfants (140 garçons, 38 filles), âgés de six à huit ans. La concentration de plomb moyenne globale dans le sang était de 3,2 µg/dl (E.T. 2,4), et 4,5 % des enfants présentaient des niveaux supérieurs à 10 µg/dl. Les taux de plomb sanguins étaient nettement plus élevés chez les enfants vivant dans des camps de réfugiés près de zones industrielles/de circulation intense que chez les enfants vivant dans des zones résidentielles de la ville. Les taux de plomb dans le sang étaient positivement corrélés à la taille de la famille (r = 0,15) et négativement corrélés au nombre de mètres carrés du logement familial (r = –0,18). Les taux de plomb dans le sang chez ces écoliers palestiniens étaient supérieurs à ceux d'autres pays où le carburant contenant plomb a été interdit. Ils semblaient aussi supérieurs chez les enfants plus faibles économiquement.

1Poison Control and Drug Information Centre; 3College of Pharmacy, An-Najah National University, Nablus, Palestine (Correspondence to A.F. Sawalha: [email protected]). 2Harvard Medical School and School of Public Health, Boston, Massachusetts, United States of America. Received: 22/11/11; accepted: 05/02/12

151

Book 19-2.indb 151 2/21/2013 11:39:22 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction Furthermore, baseline data about BLL estimated 10 private schools, which was in children are important to inform located at the middle of Nablus city Lead is a toxic heavy metal that is ubiq- advocacy efforts to phase out leaded away from the main highway and which uitous in the environment as a result gasoline in Palestine. Such data need to therefore might have lower exposure of industrialization. Exposure occurs take into consideration the social and to environmental lead. Private schools primarily through ingestion and inhala- economic differences between children are attended by children from families tion [1,2]. In the Middle East, the major living in different environments. This living outside the refugee camps, who reported sources of lead exposure are study therefore aimed to investigate generally have higher economic and industrial, including smelters, battery BLL among schoolchildren in Pales- parental education levels. The Al-Talaa factories and radiator repair shops; tine, with emphasis on the comparison private school has approximately 1000 flour from traditional stone mills; and between those living in a refugee camp students who come from a number of the occasional burning of wastes [3,4]. environment and those living in the city. different residential areas in Nablus city. In Palestine, leaded gasoline used by automobiles remains the major source Sample of environmental lead pollution. These Methods We calculated the sample size with a exposures can be prevented by increas- 99% confidence limit with an interval Study settings and schools ing public health awareness and by of ± 2 and a standard deviation of 5. implementing measures to phase out This cross-sectional survey was carried The expected sample size was 166. The sources of lead from the environment. out in the spring of 2009 in Nablus city, final sample included 103 first graders Lead poisoning frequently goes un- one of the largest cities in the West Bank from around 300 in total at the UN- recognized. Elevated blood lead levels of Palestine. The total population of the RWA schools and 75 first graders from (BLL) can adversely affect many organ West Bank is approximately 2.5 mil- around 150 at the private school, which systems including mental development lion, 20% of whom have been living in represented approximately 45% of chil- [5], haemoglobin level [6], kidney refugee camps since 1948 [12]. dren who were invited to participate in function [7], cardiovascular function in The sample was taken from among the study and 6% (178/3000) of the adults and reproduction in women [8]. 1000 first-grade students attending 3 total number of 6–8-year-old children It has been reported that even slightly schools in different areas of Nablus. The in Nablus city. elevated levels can result in reduced IQ, schools were selected to represent dif- Data collection learning disabilities and behavioural ferent geographic and socioeconomic problems [9,10]. The United States strata in the city. We choose 2 of the 3 On the first day of the study, all children Centers for Disease Control and Pre- refugee camp schools run by the United in the 3 selected schools were given a vention (CDC) recommends a BLL Nations Relief and Works Agency for written consent form and a brochure of 10 µg/dL or above as the level of Palestine Refugees in the Near East explaining the study to be handed to concern at which public health action (UNRWA). The choice of UNRWA their parents. Of 400 consent forms needs to be initiated. The CDC consider schools was based on the fact that the distributed, 178 were signed by the low-dose exposure to lead poisoning in populations of refugee camps are char- parents and returned to the researchers children to be a preventable paediatric acterized by low socioeconomic status a week later. health problem and have emphasized and poor health standards. The camps Before venous blood samples were the need for primary prevention [11]. have concrete houses built before 1950, collected, parents completed a ques- Therefore screening for elevated BLL with narrow streets and high popula- tionnaire to collect data about sociode- and prevention measures are recom- tion density and are located close to the mographic parameters, child’s general mended [11], including universal main streets with heavy traffic which health status (i.e. presence of any eating screening, if the prevalence of elevated might to greater lead exposure for chil- or sleeping problems, growth problems, BLL in the community is unknown. dren. Al-Ain UNRWA school is located any diseases suffered, vaccinations taken In Palestine, chronic exposure to at the northern part of Nablus city and and potential sources of lead exposure in lead has not been adequately addressed serves boys living at Al-Ain refugee the home environment). The question- as a problem for schoolchildren of differ- camp. Asker UNRWA school is located naire was written in Arabic. The ques- ent socioeconomic strata. Health care in the eastern part of Nablus city, close tionnaires were filled by parents without providers do not routinely screen chil- to an industrial area, and serves solely assistance from the authors. Only children for lead exposure, and so no data boys, who have a similar background dren who brought back the completely are available to evaluate BLL and the ex- to children who attend Al-Ain UN- filled questionnaire along with a signed tent of lead exposure in schoolchildren. RWA school. We also chose 1 of the consent form were included in the study.

152

Book 19-2.indb 152 2/21/2013 11:39:23 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

For each child, weight, height and arm between continuous variables were as- (SD 2.4) µg/dL. However, 8 children circumference were measured before sessed using Pearson correlation. The (4.5%) had BLL above 10 µg/dL blood sample collection. Blood sample significance level was set atP < 0.05. (maximum was 13.9 µg/dL), which is collection was carried out by well trained the level of concern according to CDC nursing staff supervised by a medical guidelines. All 8 children with a BLL > laboratory expert. Results 10 µg/dL were from Asker UNRWA school. Further analysis showed that Blood samples were analysed for Background demographic and lead at the trace metals laboratory at clinical data there were significant differences in BLL Harvard School of Public Health in among children from the 3 schools (P < Boston, Massachusetts. For the lead A total of 178 children were included 0.01, F = 41). The lowest mean blood assay blood samples were first weighed in the study (75 from Al-Tala private lead levels were at the private school (~1 g) and digested for 24 hours in 2 school, 31 from Al-Ain UNRWA school [1.9 (SD 0.6) µg/dL, range 0.7–4.3 mL of concentrated nitric acid. These and 72 from Asker UNRWA school). µg/dL], followed by Al-Ain UNRWA samples were then treated with 1 mL In our sample, there were 140 boys school [2.7 (SD 0.9), range 1.2–4.4 of 30% hydrogen peroxide per 1 g of (78.7%) and 38 (21.3%) girls; all the µg/dL], while Asker UNRWA school blood and left overnight. Samples were girls were from Al-Talaa private school. had the highest level [4.9 (SD 0.4) µg/ subsequently diluted to 10 mL with The mean age of the children was 6.4 dL, range 0.8–13.9 µg/dL]. Post hoc deionized water. Lead concentrations (SD 0.5) years, range 6–8, median 6 analysis using the Tukey test showed were measured using a dynamic reac- years. that children at Asker UNRWA school tion cell-inductively coupled plasma The mean family size, measured as had a significantly higher mean BLL mass spectrometer (DRC-ICP-MS, number of family members, was 4.3 than children at Al-Tala private school DRC II, Perkin Elmer). Analyses were (SD 1.7) and the mean household area (P < 0.01) or at Al-Ain UNRWA school 2 performed using an external calibration was 126 (SD 66) m . There was a sig- (P < 0.01). method, with lutetium as the internal nificant difference among children in standard for lead. the 3 schools with regard to family size. Risk factors for high lead levels Quality control measures included Children at Al-Talaa private school had Pearson correlation analysis showed analysis of initial and continuous cali- the smallest family size [3.6 (SD 1.4)] that age was not correlated with BLL bration verification standards [National and Al-Ain UNRWA school had the for the whole sample or at any of the 3 Institute of Standards and Technology highest [5.2 (SD 1.9)] (Table 1). There schools (P > 0.5). The only school with standard reference material for trace was also a significant difference among data on both boys and girls was Al-Talaa elements in water (NIST 1643e)], 1 children in the 3 schools with regard private school. The mean BLL for boys ppb lead standard, procedural blanks, to household area, with children at Al- at was 2.0 (SD 0.7) µg/dL and for girls duplicate samples, spiked samples and Talaa having the largest household area at the same school was 1.9 (SD 0.6) µg/ 2 certified reference material (NIST 955b 152 (SD 45) m and Al-Ain UNRWA dL. This difference was not statistically 2 in bovine blood for lead) to monitor for school the lowest [104 (SD 75) m ] significant P( > 0.05). contamination, accuracy and recovery (Table 1). None of the other health param- rates. Recovery rates for lead in quality Their mean haemoglobin level of eters measured (haemoglobin level, control and spiked samples were 90%– the children was 12.3 (SD 0.8) g/dL. weight or arm circumference) were 115%, and precision was measured as The mean weight of the children was significantly correlated with BLL. Fam- % relative standard deviation (SD) and 23.5 (SD 3.9) kg, and the mean arm ily size was positively correlated with it was less than 5% for lead. The limits circumference was 17.5 (SD 1.7) cm. children’s BLL (r = 0.153, P = 0.048) of detection of lead were 0.2 µg/dL. For Al-Ain UNRWA school, the mean and household size was negatively cor- Results from ICP-MS analyses are the weight was 23.3 (SD5.7) kg, for Al- related with BLL (r = –0.18, P = 0 .031) average of 5 replicate measurements. Talaa it was 24.1 (SD 3.5), kg and for (Table 1). Asker it was 22.9 (SD 3.1) kg. As for the Data analysis arm circumference, it was 17.2 (SD 2.4) Data obtained from questionnaires and cm for Al-Ain, 17.9 (SD 1.5) cm for Al- Discussion blood sample analyses were analysed Talaa and 17.2 (SD 1.3) cm for Asker. using SPSS, version 16 for Windows. Lead is considered a serious neuro- Comparisons among the 3 schools were Blood lead levels toxicant, and research suggests that made using robust ANOVA test with The mean BLL among all children children's intellectual functioning is Tukey post hoc analysis. Correlations when the 3 sites were pooled was 3.2 impaired by BLL concentrations even

153

Book 19-2.indb 153 2/21/2013 11:39:23 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 1 Demographic and clinical variables and blood concentrations of lead among children in the 3 schools in different areas of Nablus city, West Bank of Palestine Variable Al-Talaa private school Al-Ain UNRWA school Asker UNRWA school P-value (n = 75) (n = 31) (n = 72) No. % No. % No. % Sex Male 37 49 31 100 72 100 Female 38 51 0 0 0 0 Mean (SD) Mean (SD) Mean (SD) Age (years) 6.4 (0.5) 6.4 (0.7) 6.5 (0.5) Family members (no.) 3.6 (1.4) 5.2 (1.9) 4.4 (1.7) < 0.01 Household area (m2) 152 (45.0) 104 (75) 105 (73) < 0.01 Arm circumference (cm) 17.9 (3.5) 17.2 (2.4) 17.2 (1.3) NS Weight (kg) 24.1 (3.5) 23.3 (5.7) 22.9 (3.1) NS Haemoglobin level (mg/dL) 12.3 (0.7) 12.2 (0.9) 12.5 (0.8) NS Blood lead level (µg/dL) 1.9 (0.6) 2.7 (0.9) 4.9 (0.4) < 0.01

UNRWA = United Nations Relief and Works Agency for Palestine Refugees in the Near East; SD = standard deviation; NS = not significant.

below 10 µg/dL [9,10]. This suggests however, typically have higher BLL to develop and evaluate intervention that BLL in children should be reduced than school-age children due to their policies [24]. International experience as much as possible [13]. In our study, hand-to-mouth behaviours and 2–4 shows that it might take years to see the mean BLL among school-age chil- years of age are the peak ages for lead substantial reductions in lead levels in dren was 3.2 (SD 2.4) µg/dL. This is poisoning [5]. Age as a risk factor for the environment after lead is removed higher than that the average level found elevated BLL among children 4–12 from gasoline [24]. in US schoolchildren (1.9 µg/dL) years of age has not generally been Our study had a number of limi- [14–16]. The mean BLL of children reported. In a study in Russia, little tations. First, the surveyed schools 6–11 years of age in the US was 1.9 µg/ variation in BLL by age was found were not randomly selected and re- dL in the early 1990s and the rate of [18]. Studies from the United States cruitment of children within schools BLL >10 µg/dL was 2.0% in the early and Australia showed that BLL were depended on attendance and paren- 1990s and is continuing to drop [16]. It highest in 1–2-year-old children and tal consent. However, the selected is encouraging that the mean level was declined at older ages [19,20]. schools represented a spectrum of not even higher, considering the harsh In our study, the mean BLL were geographic and socioeconomic strata, economic situation, environmental pol- highest among the children living in a and the true mean geographic BLL lution and the continued use of leaded refugee camp near an industrial area levels of first-grade children was likely gasoline in Palestine. with high traffic and lowest among to be within the range found. Our Our results are similar to those children living in residential areas away sample size may not have allowed de- reported in a previous study in Pales- from high traffic. A study carried out tection of small statistical associations tine by Safi et al. who reported that the in Cairo, Egypt among children 3–15 between some potentially important mean BLL in Palestinian pre-school years of age found that the mean BLL exposure sources and elevated BLL children (2–6 years of age) was 4.2 was 4.82 (SD 2.97) µg/dL with a range levels. Finally, there was a biased dis- µg/dL and 5.2% had BLL > 10 µg/dL 1.1–14.3 µg/dL, and that children liv- tribution of the sexes from the differ- [17]. In the same study, the mean BLL ing in high traffic areas had the highest ent schools because in the UNRWA among pre-school children in Jordan mean levels [21]. The mean BLL in our schools, only males students agreed and Israel was found to be 3.2 µg/dL. study was lower than levels reported to participate, whereas in the private The children in our sample were older from countries in which leaded gaso- school, both males and females par- (6–8 years) than Safi et al.’s sample line is still used [22,23]. At the time ticipated. However, since our study and it might be expected that BLL for of this study, leaded gasoline was still was a pilot and exploratory study, this children in our study would be higher sold at gasoline stations in Palestine. limitation can be overcome in the due to increased exposure to sources In countries that plan to phase out later studies to be carried out at the of lead over time. Younger children, leaded gasoline, baseline BLL is critical national level.

154

Book 19-2.indb 154 2/21/2013 11:39:23 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Conclusions the mean BLL in children was below recommend public health action be ini- the CDC level of concern and was tiated. BLL were higher in children living Our study examined BLL among first- similar to those observed in children in refugee camps near industrial regions grade schoolchildren living in differ- from neighbouring countries, although with high traffic and significantly higher ent parts of Nablus city, West Bank, 4.5% of children had BLL above 10 µg/ in children from larger families and fami- Palestine. The major findings were that dL, the level at which CDC guidelines lies with smaller household area.

References

1. Brown LM et al. Blood lead levels and risk factors for lead poi- 13. Järup L. Hazards of heavy metal contamination. British Medical soning in children and caregivers in Chuuk State, Micronesia. Bulletin, 2003, 68:167–182. International Journal of Hygiene and Environmental Health, 14. Iqbal S et al. Estimated burden of blood lead levels 5 microg/dl 2005, 208:231–236. in 1999–2002 and declines from 1988 to 1994. Environmental 2. Singh AK, Singh M. Lead decline in the Indian environment Research, 2008, 107:305–311. resulting from the petrol-lead phase-out programme. Science 15. Pirkle JL et al. Exposure of the U.S. population to lead, 1991– of the Total Environment, 2006, 368:686–694. 1994. Environmental Health Perspectives, 1998, 106:745–750. 3. Hershko C et al. Lead poisoning by contaminated flour.Re - 16. Centers for Disease Control and Prevention (CDC). Mortality views on Environmental Health, 1989, 8:17–23. Morbidity Weekly Report Blood lead levels in young children 4. El Sharif N et al. Re-emergence of lead poisoning from con- and selected sites, 1996–1999. Mortality Morbidity Weekly Re- taminated flour in a West Bank Palestinian village. Interna- port, 2000, 49:1133–1137. tional Journal of Occupational and Environmental Health, 2000, 17. Safi J et al. Childhood lead exposure in the palestinian author- 75:183–186. ity, Israel, and Jordan: results from the Middle Eastern regional 5. Bellinger DC, Stiles KM, Needleman HL. Low-level lead ex- cooperation project, 1996–2000. Environmental Health Per- posure, intelligence and academic achievement: a long-term spectives, 2006, 114:917–922. follow-up study. Pediatrics, 1992, 90:855–861. 18. Rubin CH et al. Childhood lead poisoning in Russia: a site- 6. Ahamed M et al. Environmental exposure to lead and its corre- specific pediatric blood lead evaluation.International Journal lation with biochemical indices in children. Science of the Total of Occupational and Environmental Health, 1997, 3:241–248. Environment, 2005, 346:48–55. 19. Brody DJ et al. Blood lead levels in the US population. Phase 1 7. Fadrowski JJ et al. Blood lead level and kidney function in US of the Third National Health and Nutrition Examination Survey adolescents: The Third National Health and Nutrition Exami- (NHANES III, 1988 to 1991). Journal of the American Medical As- nation Survey. Archives of Internal Medicine, 2010, 170:75–82. sociation, 1994, 272:277–283. 8. Vahter M et al. Metals and women’s health. Environmental 20. Baghurst PA et al. Determinants of blood lead concentrations Research, 2002, 88:145–155. to age 5 years in a birth cohort study of children living in the 9. Jusko TA et al. Blood lead concentrations < 10 microg/dL and lead smelting city of Port Pirie and surrounding areas. Archives child intelligence at 6 years of age. Environmental Health Per- of Environmental Health, 1992, 47:203–210. spectives, 2008, 116:243–248. 21. Sharaf NE et al. Evaluation of children's blood lead level in 10. Lanphear BP et al. Low-level environmental lead exposure Cairo, Egypt. American–Eurasian Journal of Agricultural and and children’s intellectual function: an international pooled Environmental Sciences, 2008, 3:414–419. analysis. Environmental Health Perspectives, 2005, 113:894–899. 22. Kaiser R et al. Blood lead levels of primary school children in 11. Centers for Disease Control and Prevention. Advisory commit- Dhaka, Bangladesh. Environmental Health Perspectives, 2001, tee on childhood lead poisoning prevention Interpreting and 109:563–566. managing blood lead levels < 10 microg/dL in children and 23. Lovei M. Eliminating a silent threat: World Bank support for reducing childhood exposures to lead: recommendations of the global phase out of lead from gasoline. In: George AM, CDC's Advisory Committee on Childhood Lead Poisoning ed. Lead poisoning prevention and treatment: implementing a Prevention. Mortality Morbidity Weekly Report, 2007, 56(RR- national program in developing countries. Bangalore, India, The 8):1–16. George Foundation, 1999:169–180. 12. Palestinian Central Bureau of Statistics. Palestinian National 24. Tong S, von Schirnding YE, Prapamontol T. Environmental lead Authority [online database] (http://www.pcbs.gov.ps/Desk- exposure: a public health problem of global dimensions. Bul- topDefault.aspx?lang=en, accessed 20 November 2012). letin of the World Health Organization, 2000, 78:1068–1077.

155

Book 19-2.indb 155 2/21/2013 11:39:23 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Birth weight and risk of childhood acute leukaemia A. Gholami,1 S. Salarilak,2 S. Hejazi 3 and H.R. Khalkhali 4

الوزن عند الوالدة واختطار ابيضاض الدم احلاد يف الطفولة عيل غالمي، شاكر ساالري لك، ساسان حجازي، محيد رضا خلخايل اخلالصـة: ًنظراألن الدراسات حول عوامل اختطار ابيضاض الدم احلاد مازالت غري جازمة فقد أجرى الباحثون هذه الدراسات للحاالت والشواهد يف والية غرب أذربيجان يف مجهورية إيران اإلسالمية ُّللتعرفعىل العالقة بني الوزن عند الوالدة وابيضاض الدم احلاد لدى األطفال دون سن اخلامسة عرشة. وقد اختار الباحثون اثنني من الشواهد مقابل كل مريض يوافقانه بالعمر وباجلنس من السكان يف املجتمع ويف املستشفى. وقد شملت الدراسة 130 ًمريضا تم تشخيصهم خالل الفرتة 2003–2009، وكان لدى 108 منهم )83.1%( نمط األرومات اللمفاوية، ولدى 22 منهم )16.9%( نمط األرومات النقوية. 44.6 55.4 واتضح من حيث االعتداد اإلحصائي أن الذكور ) %( أكثر من اإلناث ) (. وعند إجراء النموذج ُّالتحويف املتعدد املتغ ِّيرات اتضح أن املتغريات 2.25 التي تـرتابط ُتراب ًطا ُي ْع َت ُّد به ًإحصائيا مع ابيضاض الدم احلاد هي: الوزن عند الوالدة )نسبة األرجحية (، الرتتيب بني املولودين )نسبة األرجحية 3.23 0.46 7.93 2.25 (، مكان الوالدة )نسبة األرجحية (، قصة إصابة سابقة باحلُامق )نسبة األرجحية (، درجة ُّتعلم األمهات )نسبة األرجحية (. واستنتج الباحثون أن خطر االبيضاض احلاد يزداد زيادة ُي ْع َت ُّد هبا ًإحصائيا بازدياد الوزن عند الوالدة يف جممل املجموعة وبني الفتيات ال بني الفتيان.

ABSTRACT Studies of risk factors for acute leukaemia are inconclusive. This case–control study was done in West Azerbaijan province, Islamic Republic of Iran, to determine the relationship between birth weight and acute leukaemia in children aged under 15 years. For every patient 2 age- and sex-matched controls were selected from hospital and community populations. Of 130 cases diagnosed over the period 2003–2009, 108 (83.1%) had lymphoblastic and 22 (16.9%) myloblastic type. Significantly more of them were male than female (55.4% versus 44.6%). In a multivariate logistic regression model variables significantly associated with acute leukaemia were: birth weight (OR = 2.25), birth order (OR = 2.25), birth place (OR = 7.93), history of chickenpox (OR = 0.46) and mothers’ education (OR = 3.23). The risk of acute leukaemia increased significantly with increasing birth weight in the total group and among girls, but not among boys.

Poids de naissance et risque de leucémie aiguë de l'enfant

RÉSUMÉ Les études des facteurs de risques d'une leucémie aiguë n'ont pas permis de tirer de conclusions. La présente étude cas-témoins a été menée dans la province d'Azerbaïdjan de l'ouest (République islamique d'Iran), pour déterminer la relation entre le poids de naissance et une leucémie aiguë chez l'enfant de moins de 15 ans. Pour chaque patient, deux témoins appariés pour l'âge et le sexe ont été sélectionnés dans des populations hospitalières ou communautaires. Sur 130 cas diagnostiqués entre 2003 et 2009, 108 (83,1 %) étaient porteurs du type lymphoblastique et 22 (16,9 %) du type myéloblastique. Les patients de sexe masculin étaient nettement plus nombreux que les patients de sexe féminin (55,4 % contre 44,6 %). Dans un modèle de régression logistique multivariée, les variables fortement associées à une leucémie aiguë étaient les suivantes : poids de naissance (OR = 2,25), rang de naissance (OR = 2,25), lieu de naissance (OR = 7,93), antécédents de varicelle (OR = 0,46) et niveau d'études de la mère (OR = 3,23). Le risque d'une leucémie aiguë augmentait nettement avec un poids de naissance accru dans l'ensemble du groupe et chez les filles, mais pas chez les garçons.

1School of Nursing, Neyshabur Faculty of Medical Sciences, Neyshabur, Islamic Republic of Iran. 2Department of Public Health, Islamic Azad University (Tabriz Branch), Tabriz, Islamic Republic of Iran (Correspondence to S. Salarilak: salarilak@ yahoo.com). 3Motahari Teaching Hospital, Children’s Blood Section; 4Department of Biostatistics and Epidemiology, Urmia University of Medical Sciences, Urmia, Islamic Republic of Iran. Received: 21/11/11; accepted: 21/02/12

156

Book 19-2.indb 156 2/21/2013 11:39:23 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction groups: a hospital control and a com- relationship between birth weight and munity control. Cases and controls acute leukaemia and also to control Leukaemia is a malignant progressive were matched according to age group for confounding factors and investigate cancer which is caused by incomplete (< 5, 5–9 and 10–14 years old) and the possible effects on acute leukaemia production of white blood cells and sex. Hospital controls were selected risk of the risk factors investigated. We their precursors in blood and bone from the children’s ward and the chil- report odds ratios (OR), 95% confi- marrow. The disease is classified into dren’s clinic centre in Motahari hos- dence interval (CI) and P-values for all acute and chronic types according to pital in the city of Urmia. Community variables. The chi-squared test for trend the speed of progression of cancer and controls were selected from children was used to determine the trend of into lymphoblastic or myeloblastic ac- who came to Urmia health centres for birth weight in the developing of acute cording to the types of white blood cell routine health care. Inclusion criteria leukaemia. affected [1–3]. Leukaemia is the most for controls were not having acute prevalent childhood cancer and the leukaemia or any other blood disease, most common type in children is acute age less than 15 years at the time of Results lymphoblastic leukaemia (ALL). ALL is data collection and residing in West a very rapid cancer which is usually seen Azerbaijan province. A total of 138 children affected by acute in ages between 2 and 6 years [2]. ALL leukaemia were diagnosed in west Azer- Data collection has a high mortality rate, great costs for baijan province over the study period treatment and a long period of hospi- A specially designed questionnaire was (113 with ALL and 25 with AML); 8 talization, with associated psychological used to collect data on the child’s demo- patients dropped out of the study, 1 trauma for patients and their families graphic characteristics, type of cancer because the parents did not agree to [4]. (for cases only), child’s risk factors (birth participation and 7 because of a change of address. So the study was based on In view of the social and economic order, birth place, history of chickenpox, 130 patients, of whom 108 (83.1%) burden of this disease, a number of stud- history of icterus, history of mumps were affected by ALL and 22 (16.9%) ies have been done in different coun- infection, history of X-ray exposure, by AML. Of these, 72 patients (55.4%) tries to study the relationship between breastfeeding, formula-milk feeding) were boys and 58 (44.6%) girls. At the genetic and environmental factors and and parents’ demographic characteris- time of diagnosis, 52.3% of their families the risk of acute leukaemia. One of tics (mother’s age, father’s age, mother’s lived in urban areas and 47.7% in rural these factors is birth weight and in some education and father’s education at the areas. Most cases (43.1%) were aged studies a statistically significant relation- birth of the child). The questionnaire under 5 years (Figure 1). There were ship has been reported between birth design was based on a questionnaire 260 age and sex-match controls: 144 weight and leukaemia [5–21], while provided by the Washington Health boys (55.4%) and 116 girls (44.6%). other studies have found no relation- Department [26] and questionnaires ship [22–25]. In view of the conflicting from previous studies in the Islamic The mean birth weight of children findings on this subject, we decided to Republic of Iran and with input from affected by acute leukaemia [3400 (SD study the relationship between birth local professors of epidemiology and 650) g] was significantly higher than weight and acute leukaemia in children oncology. A pilot study was done to test that the control group [3240 (SD 592) in the province of West Azerbaijan, Is- the validity of the questionnaire. Ques- g] (P = 0.02). lamic Republic of Iran. tionnaires were completed using data A univariate logistic regression files and face-to-face interviews with the model was used to evaluate the contri- mothers of patients and controls. bution of each variable and its relation Methods To prevent recall bias in the birth to leukaemia. Variables with a signifi- weight variable, we used the birth weight cant association were as follows: birth Study sample recorded in the household folders kept weight (< 4000 g versus ≥ 4000 g), This was a case–control study. The by the health system and in vaccination birth order (1 versus ≥ 2), birth place case group were patients with acute cards; the birth weight of 90.8% of the (home versus hospital), history of leukaemia, diagnosed from 20 March case patients and 95.0% of controls were chickenpox, mother’s and father’s age 2003 and 20 March 2009, age less recorded. (< 35 versus ≥ 35 years) and mother’s than 15 years at the time of diagnosis and father’s education (below high and residing in West Azerbaijan prov- Statistical analysis school versus high school and above) ince at the time of diagnosis. For each A logistic regression model in SPSS, (Table 1). Variables without any sig- patient we selected 2 controls from 2 version 16 was used to investigate the nificant relationship according to the

157

Book 19-2.indb 157 2/21/2013 11:39:23 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

univariate logistic regression model ALL AML were as follows: history of icterus, his- 60 tory of mumps, history of X-ray expo- 52 sure, breastfeeding (≥ 6 months and 50 s e

< 6) and formula milk feeding (only s 40 a 35 c

formula feeding) (Table 1). f 30 o

. Using a multivariate logistic regres- o 21 N 20 sion model with a forward method, 10 10 8 we evaluated different variables in 4 0 relation to acute leukaemia. Variables < 5 years 5–9 years 10–14 years that remained significantly associated with acute leukaemia were as follows: Figure 1 Frequency of childhood acute leukaemia by age groups. ALL = acute birth weight (≥ 4000 versus < 4000 lymphoblastic leukaemia; AML = acute myloblastic leukaemia g) (OR = 2.25; 95% CI: 1.09–4.64), birth order (≥ 2 versus 1) (OR = 2.25; 95% CI: 1.33–3.83), birth place (home versus hospital) (OR = 7.93; that more than 80% of acute leukaemia case–control studies. In addition, the 95% CI: 4.07–15.5), positive history cases were ALL type. use of vital records eliminated the of chickenpox (OR = 0.46; 95% CI: Slightly more of the cases were problem of recall bias (in the birth 0.25–0.96) and mothers’ education among males than females (55.4% weight variable) because the data were (below high school versus high school versus 44.6%). This agrees with the re- not self-reported after a diagnosis of or above) (OR = 3.23; 95% CI: 1.77– sults of other research. For example, leukaemia. 5.89) (Table 2). Hjalgrim et al. found 52.9% of patients We observed a strong relationship As seen in Table 3, the risk of acute were males and 47.1% were females between increasing birth weight and leukaemia increased with increasing [5], Rosenbaum et al., studying patients acute leukaemia risk. Several studies birth weight category, from < 3000 with ALL only, found 57% were males have suggested a possible relation- g (OR = 0.65; 95% CI: 0.34–1.24), and 43% were females [29], while ship between acute leukaemia risk through 3000–3499 (reference Mallol-Mesnard et al. reported that in children and particularly high group), 3500–3999 g (OR = 1.30; 54.4% of patients were males and 45.6% birth weight [5–21,23,32–36]. For 95% CI: 0.76–2.20) to ≥ 4000 g (OR were females [30]. Most of the cases example in Hjalgrim et al.’s study a = 2.20; 95% CI: 1.09–4.50). This trend in our study were under 5 years of age significant relationship was seen be- was statistically significant in the total (43.1%), which is similar to the results tween birth weight and ALL but not group (P = 0.016) and among girls of Westergard et al. [7] Hjalgrim et al. between birth weight and AML [5]. In (P = 0.006), but not among boys (P = [5] and Zolala et al. [22]. In our study Caughey and Michel’s meta-analysis 0.133). the number of patients living in urban [6] and a meta-analysis by Hjalgrim areas (52.3%) was greater more than in et al [36], a significant relationship rural areas (47.7%). In Auvinen et al.’s was seen between weight ≥ 4000 g at Discussion study on ALL patients 24% of patients the time of birth and acute leukaemia lived in rural areas, 32% in urban areas (AML and ALL) [6,27]. Some stud- In this study we examined 130 children and 44% in suburban areas [31]. Zolala ies did not observe any relationship with acute leukaemia who were diag- et al. reported 63.5% of patients living between high birth weight and leukae- nosed in West Azerbaijan province. In in urban areas and 36.5% in rural areas mia [22–25]. our sample of children diagnosed with [22]. We found a relationship not only acute leukaemia, 83.1% were affected by One of the major advantages of for those with high birth weight but ALL and 16.9% by AML. In MacArthur this study was that the subjects in the also a steady increase in leukaemia risk et al.’s study in Canada 88% of patients control group were randomly selected with increasing birth weight category. were affected by ALL, 10% by AML and from the children’s ward and from the This relationship was statistically 2% by acute unknown lymphoblastic children’s clinic centre in Motahari significant for girls but not for boys. leukaemia [27]. In Milne et al.’s study hospital in Urmia city and from chil- This agrees with Westergard et al., in Western Australia 86.8% of children dren who came to Urmia health cen- who observed that for each kilogram were affected by ALL and 13.2% by tres for routine care rather than from increase in birth weight the relative risk AML [28]. In all these studies, we see recruited volunteers as in most other increased by 1.46 in ALL and 2.14 in

158

Book 19-2.indb 158 2/21/2013 11:39:23 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 1 Univariate logistic regression model of factors associate with acute leukaemia compared with healthy controls Variables Patients Controls OR (95% CI) P-value (n = 130) (n = 260) No. No. Child variables Birth weight (g) ≥ 4000 21 22 2.21 (1.16–4.21) 0.014 < 4000 97 225 missing 12 13 Birth order ≥ 2 87 112 2.67 (1.72–4.15) > 0.001 1 43 148 Birth place Home 55 21 8.35 (4.74–14.7) > 0.001 Hospital 75 239 History of chickenpox Yes 15 59 0.44 (0.24–0.82) 0.008 No 115 201 History of icterus Yes 46 119 0.65 (0.42–1.00) 0.05 No 84 141 History of mumps infection Yes 6 5 2.47 (0.74–8.24) 0.13 No 124 255 History of X-ray exposure Yes 33 61 1.11 (0.68–1.81) 0.68 No 97 199 History of breastfeeding Yes a 118 247 0.52 (0.23–1.17) 0.11 Nob 12 13 History of formula milk feeding Yes a 31 67 0.90 (0.55–1.47) 0.68 Nob 99 193 Parent variables Mother’s age (years) ≥ 35 24 21 2.58 (1.37–4.83) 0.002 < 35 106 239 Father’s age (years) ≥ 35 48 65 1.76 (1.12–2.76) 0.014 < 35 82 195 Mother’s education Below high school 111 140 5.01 (2.91–8.63) < 0.001 High school and above 19 120 Father’s education Below high school 94 143 2.14 (1.36–2.37) < 0.001 High school and above 36 117

a ≥ 6 months; b < 6 months or not used. OR = odds ratio; CI = confidence interval.

159

Book 19-2.indb 159 2/21/2013 11:39:24 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 2 Multivariate logistic regression model showing significant risk factors associated with childhood acute leukaemia Variable OR (95% CI) P-value Birth weight (< 4000 g vs ≥ 4000 g) 2.25 (1.09–4.64) 0.028 Birth order (1 vs ≥ 2) 2.25 (1.33–3.83) 0.003 Birth place (home vs hospital) 7.93 (4.07–15.5) < 0.001 History of chickenpox (yes vs no) 0.46 (0.25–0.96) 0.038 Mother’s education (< high school vs ≥ high school) 3.23 (1.77–5.89) < 0.001

OR = odds ratio; CI = confidence interval.

Table 3 Risk of developing acute leukaemia according to birth weight and sex of children with acute leukaemia compared with healthy controls Sex/birth weight (g) Patients Controls OR (95% CI) P-value (n = 130) (n = 260) Both sexes < 3000 18 64 0.65 (0.34–1.24) 3000–3499 39 90 1 (Ref.) 0.016 3500–3999 40 71 1.30 (0.76–2.20) ≥ 4000 21 22 2.20 (1.09–4.50) Missing data 12 13 Girls < 3000 10 32 0.84 (0.32–2.17) 3000–3499 12 46 1 (Ref.) 0.006 3500–3999 20 27 2.37 (0.95–5.92) ≥ 4000 7 4 5.60 (1.36–23.1) Missing data 9 7 Boys < 3000 8 32 2.46 (0.99–6.10) 3000–3499 27 44 1 (Ref.) 0.133 3500–3999 20 44 1.82 (0.71–4.65) ≥ 4000 14 18 3.11 (1.10–8.83) Missing data 3 6

OR = odds ratio; CI = confidence interval; Ref. = reference group.

AML [7]. Robert et al. also observed excess risk associated with high birth Further studies in Islamic Republic of that increasing birth weight increased weight [16]. It has been suggested that Iran and other countries are needed the risk of acute leukaemia in children high birth weight may result from high to evaluate the risk factors for child- and the association was statistically levels of growth factors in uterus, and hood leukaemia, including high birth significant [8]. Our findings might these growth factors might increase weight. imply that factors that are influential the risk of ALL by inducing prolif- on fetal growth play an important role erative stress on the bone marrow in the development of childhood leu- Acknowledgements [37,38]. kaemia. It was previously suggested that children with a high birth weight In conclusion, our study adds We would like to extend best regards may be more likely to be exposed to to the evidence that birth weight is and thanks to the staff of the haema- diagnostic radiation in utero or in the related to an increased risk of child- tology ward in Motahari teaching neonatal period and that this radiation hood acute leukaemia and this may be hospital and also to the patient and exposure might explain some of the useful for diagnosis and prevention. control groups’ families.

160

Book 19-2.indb 160 2/21/2013 11:39:24 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

References

1. Acute lymphocytic leukemia. Leukemia and Lymphoma Society 21. Cnattingius S et al. Prenatal and neonatal risk factors for child- [online factsheet] (http://www.lls.org/#/diseaseinformation/ hood lymphatic leukemia. Journal of the National Cancer Insti- leukemia/acutelymphoblasticleukemia/ 2006, accessed 26 tute, 1995, 87:908–914. November 2012). 22. Zolala F et al. Determination the inducing factors of acute 2. Henry JB. Clinical diagnosis and management by laboratory lymphoblastic leukemia in children under 15 years old in Fars methods, 20th ed. Philadelphia, Saunders, 2001. province (Iran) in the year 2001. Rafsanjan of Medical Sciences 3. Acute myelogenous leukemia. Leukemia and Lymphoma So- and Health Services, 2004, 3:267–275. ciety [online factsheet] (http://www.lls.org/diseaseinforma- 23. Eisenberg DE, Sorahan T. Birth weight and childhood cancer tion/leukemia/acutemyeloidleukemia/2006, accessed 26 deaths. Journal of the National Cancer Institute, 1987, 78:1095– November 2012). 1100. 4. Pizzo P. Poplack D. Principles and practice of pediatric oncology, 24. Shu XO et al. Breast-feeding and risk of childhood acute leuke- 3rd ed. Philadelphia, Pennsylvania, Lippincott Williams and mia. Journal of the National Cancer Institute, 1999, 91:1765–1772. Wilkins, 2001:409–419. 25. Berger R. Acute lymphoblastic leukemia and chromosome 21. 5. Hjalgrim LL et al. Birth weight and risk for childhood leukemia Cancer Genetics and Cytogenetics, 1997, 94:8–12. in Denmark, Sweden, Norway, and Iceland. Journal of the Na- 26. A survey related to childhood acute lymphocytic leukaemia. tional Cancer Institute, 2004, 96:1549–1556. Whatcom County, Washington, Washington State Depart- 6. Caughey RW, Michels KB. Birth weight and childhood leu- ment of Health, 1999 (www.co.whatcom.wa.us/health/pdf/ kemia: a meta-analysis and review of the current evidence. leukemia_survey.PDF, accessed 10 December 2012). International Journal of Cancer, 2009, 124:2658–2670. 27. MacArthur AC et al. Risk of childhood leukemia associated 7. Westergaard T et al. Birth characteristics, sibling patterns, with parental smoking and alcohol consumption prior to con- and acute leukemia risk in childhood: a population-based ception and during pregnancy: the cross-Canada childhood cohort study. Journal of the National Cancer Institute, 1997, leukemia study. Cancer Causes and Control, 2008, 19:283–295. 89:939–947. 28. Milne E et al. Fetal growth and acute childhood leukemia: 8. Caughey RW, Michels KB. Birth weight and childhood leu- looking beyond birth weight. American Journal of Epidemiol- kemia: a meta-analysis and review of the current evidence. ogy, 2007, 166:151–159. International Journal of Cancer, 2009, 124:2658–2670. 29. Rosenbaum PF, Buck GM, Brecher ML. Early child-care and 9. Smith A et al.; UKCCS investigators. Birth weight, sex and child- preschool experiences and the risk of childhood acute lymph- hood cancer: A report from the United Kingdom Childhood oblastic leukemia. American Journal of Epidemiology, 2000, Cancer Study. Cancer Epidemiology, 2009, 33:363–367. 152:1136–1144. 10. Samuelsen SO et al. Birth weight and childhood cancer. Epide- 30. Mallol-Mesnard N et al. Vaccination and the risk of childhood miology, 2009, 20:484–487. acute leukaemia: the ESCALE study (SFCE). International Jour- 11. McLaughlin CC et al. Birth weight, maternal weight and child-� nal of Epidemiology, 2007, 36:110–116. hood leukaemia. British Journal of Cancer, 2006, 94:1738–1744. 31. Auvinen A et al. Extremely low-frequency magnetic fields 12. Schüz J, Forman MR. Birthweight by gestational age and child- and childhood acute lymphoblastic leukemia: an exploratory hood cancer. Cancer Causes and Control, 2007, 18:655–663. analysis of alternative exposure metrics. American Journal of Epidemiology, 2000, 152:20–31. 13. Kaye SA et al. Maternal reproductive history and birth characteristics in childhood acute lymphoblastic leukemia. Cancer, 32. Fasal E, Jackson EW, Klauber MR. Birth characteristics and 1991, 68:1351–1355. leukemia in childhood. Journal of the National Cancer Institute, 1971, 47:501–509. 14. Petridou E et al. The risk profile of childhood leukaemia in Greece: a nationwide case–control study. British Journal of 33. Shu XO et al. A population-based case–control study of child- Cancer, 1997, 76:1241–1247. hood leukemia in Shanghai. Cancer, 1988, 62:635–644. 15. Robison LL et al. Birth weight as a risk factor for childhood 34. Cnattingius S et al. Prenatal and neonatal risk factors for child- acute lymphoblastic leukemia. Pediatric Hematology and On- hood lymphatic leukemia. Journal of the National Cancer Insti- cology, 1987, 4:63–72. tute, 1995, 87:908–914. 16. Daling JR et al. Birth weight and the incidence of childhood 35. Hirayama T. Descriptive and analytical epidemiology of cancer. Journal of the National Cancer Institute, 1984, 72:1039– childhood malignancy in Japan. In: Kobayashi N, ed. Recent 1041. advances in management of children with cancer. Tokyo, The Children’s Cancer Association of Japan, 1980:27–43. 17. Yeazel MW et al. High birth weight and risk of specific childhood cancers: a report from the Children’s Cancer Group. 36. Hjalgrim LL et al. Birth weight as a risk factor for childhood leu- Journal of Pediatrics, 1997, 131:671–677. kemia: a meta-analysis of 18 epidemiologic studies. American Journal of Epidemiology, 2003, 158:724–735. 18. Schüz J et al. Association of childhood cancer with factors related to pregnancy and birth. International Journal of Epide- 37. Ross JA et al. Big babies and infant leukemia: a role for in- miology, 1999, 28:631–639. sulin-like growth factor-1? Cancer Causes and Control, 1996, 7:553–559. 19. Ross JA et al. Evaluating the relationships among maternal reproductive history, birth characteristics, and infant leukemia: a 38. Albanes D, Winick M. Are cell number and cell proliferation report from the Children’s Cancer Group. Annals of Epidemiol- risk factors for cancer? Journal of the National Cancer Institute, ogy, 1997, 7:172–179. 1988, 80:772–775. 20. Wertelecki W, Mantel N. Increased birth weight in leukemia. Pediatric Research, 1973, 7:132–138.

161

Book 19-2.indb 161 2/21/2013 11:39:24 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

External cephalic version for breech presentation at term: predictors of success, and impact on the rate of caesarean section O.A. Hussin,1 M.A. Mahmoud 1 and M.M. Abdel-Fattah 2

التحويل اخلارجي للرأس يف املجيء املقعدي يف متام احلمل: منبئات النجاح وأثرها عىل معدالت إجراء العملية القيرصية أميمة حسني، حممود حممود، معتز عبد الفتاح اخلالصـة:لقد زادت معدالت إجراء العمليات القيرصية يف حاالت املجيء املقعدي زيادة ملحوظة خالل العقدين املاضيني. وقد أجرى الباحثون دراسة استقبالية تدخلية أترابية حول معدالت نجاح التحويل اخلارجي للرأس ومنبئات النجاح وأثرها عىل معدالت إجراء العمليات القيرصية ًمنسوبةإىل الوالدة املقعدية عرب املهبل. وشملت الدراسة 128 امرأة أدخلن قسم التوليد خالل فرتة الدراسة يف مستشفى عسكري للرعاية الثالثية يف الطائف، يف اململكة العربية السعودية، ولدهين جميء مقعدي يف متام احلمل، دون اعتبار العمر وعدد مرات الوالدات السابقة، وقد َقبِ ْل َ ن ًمجيعا إجراء التحويل اخلارجي للرأس لدهين. واتضح للباحثني أن التحويل اخلارجي للرأس قد نجح يف 53.9 %من النساء، وأن معظم من نجح لدهين التحويل قد 8.5 14.5 84.1 َو َل ْد َن والدة طبيعية ) %(، وأن % فقط منهن َو َل ْد َن بالعملية القيرصية، وباملقابل فقد حدثت الوالدة الطبيعية لدى %ممن حدث لدهين حتويل تلقائي من بني من أخفق لدهين التحويل اخلارجي للرأس، أما ما يقرب من الثلثني منهن فقد أجريت هلن العملية القيرصية )62.7%(. ويرى الباحثون أن تطبيق التحويل اخلارجي للرأس ُينقص معدالت إجراء العملية القيرصية والوالدة املقعدية لدى احلوامل الالئي نجح لدهين ذلك التحويل.

ABSTRACT The incidence of caesarean section for breech presentation has increased markedly in the last 20 years. A prospective, interventional cohort study was carried out of the success rate of external cephalic version (ECV) and its predictors of as well as its impact on the rate of caesarean section for vaginal breech delivery. All 128 women admitted during the study period to the obstetrics department of a tertiary care military hospital in Taif, Saudi Arabia with breech presentation at term, regardless of age and parity, who accepted ECV were recruited. ECV was successful in 53.9% of the women. Most of the women with successful ECV delivered normally (84.1%) and only 14.5% of them delivered by caesarean section. Conversely, normal vaginal delivery was reported among 8.5% of those who had spontaneous version with failed ECV and approximately two-thirds of them delivered by caesarean section (62.7%). Successful ECV reduced the breech and caesarean section rate.

Version céphalique externe en cas de présentation par le siège à terme : facteurs prédictifs de succès et impact sur le taux de césariennes

RÉSUMÉ L'incidence des césariennes en cas de présentation par le siège a beaucoup augmenté au cours des 20 dernières années. Une étude interventionnelle d'une cohorte prospective a été menée pour connaître le taux de succès des versions céphaliques externes et ses facteurs prédictifs, ainsi que l'impact sur le taux de césariennes par rapport à l'accouchement du siège par voie vaginale. Les 128 femmes qui ont été admises pendant l'étude au service d'obstétrique de l'hôpital militaire de soins de santé tertiaires à Taif (Arabie saoudite) pour une présentation par le siège à terme, et qui ont accepté une version céphalique externe ont toutes été recrutées indépendamment de leur âge et du nombre de leurs accouchements. La version céphalique externe a été un succès chez 53,9 % de ces femmes. La plupart des femmes pour qui la version céphalique externe a été un succès ont accouché par voie basse (84,1 %) et seules 14,5 % d'entre elles ont accouché par césarienne. À l'inverse, un accouchement par voie basse a été rapporté chez 8,5 % des femmes ayant eu une version spontanée après un échec de la version céphalique externe et environ deux tiers d'entre elles ont accouché par césarienne (62,7 %). Le succès dans la version céphalique externe permettait de réduire le taux de présentation par le siège et le taux de césariennes chez les femmes.

1Department of Gynaecology and Obstetrics; 2Department of Preventive Medicine (Research Unit), Al-Hada Armed Forces Hospital, Taif, Saudi Arabia (Correspondence to M.M. Abdel-Fattah: [email protected]; [email protected]). Received: 31/10/11; accepted: 21/02/12

162

Book 19-2.indb 162 2/21/2013 11:39:24 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction The current study aimed to study ultrasound was done to confirm the the success rate of ECV and its predic- presentation and a cardiotography was Breech presentation complicates tors in a tertiary care setting as well as its repeated. Patients were discharged with 3%–4% of all term deliveries [1]. It has impact on reducing the rate of caesar- a documented plan of delivery. been widely recognized that there is ean section for breech presentation and The rates of successful ECV, normal higher perinatal morbidity with breech vaginal breech delivery. spontaneous vaginal delivery after suc- presentation, due principally to prema- cessful ECV, caesarean section after turity, congenital malformations, birth attempted ECV and ECV-related fetal asphyxia and birth trauma [1]. Breech Methods complications (bradycardia, emergency presentation, whatever the mode of Study setting and sample caesarean section or spontaneous rup- delivery, is a signal for potential fetal ture membrane during the procedure) handicap and this should be commu- In a prospective, interventional cohort were computed as outcome measures. nicated to the mother during antenatal, study all women admitted to the ob- The variables included in the study intrapartum and neonatal management stetric department of Al-Hada Armed were: success of ECV, complications, [2]. The incidence of caesarean section Forces Hospital (a tertiary care hospital presentation in labour, mode of deliv- for breech presentation has increased with 360 beds) in Taif, Saudi Arabia, ery, fetal outcome (Apgar score < 7 at markedly in the last 20 years [3]. The during the period June 2008 to June 5 minutes) and length of post-delivery Term Breech Trial concluded that, 2010, regardless of age and parity, with hospital stay. at least for mortality and markers of breech presentation at term were in- intermediate term morbidity, elective cluded in the study. Inclusion criteria Statistical analysis caesarean section was safer for the fetus were women with gestation of 37 weeks Data were analysed using SPSS, version and of similar safety to the mother when and onwards with breech presentation 13 and Epi-info, version 3.3.2. The chi- compared with intention to deliver vagi- regardless of age and parity, who ac- squared test was used to compare 2 or nally [4–6]. This means that measures cepted ECV. The exclusion criteria were more qualitative variables. Failed ECV to reduce the incidence of breech pres- refusal of ECV, multiple pregnancy, was treated as the dependent variable entation have become more important severe intrauterine growth retardation, in both bivariate and logistic regres- and that the effect of any such measure severe amniotic fluid abnormalities, sion analysis. Age, parity, position of the on the incidence of caesarean section placenta previa, significant third tri- placenta, fetal weight, maximum pool will be more evident [1]. mester bleeding, uterine malformation, depth of amniotic fluid and tocolysis External cephalic version (ECV) uncontrolled hypertensive disorders, were treated as independent categorical at or near to term is a safe procedure major fetal anomalies, non-reassuring variables. Bivariate data analysis was which could effectively reduce the in- cardiotocography and any indication performed and expressed as crude odds cidence of caesarean section in breech for caesarean section. ratio (OR) and their 95% confidence pregnancies [1]. ECV has been sub- Data collection intervals (CI). Multiple associations jected to rigorous scientific appraisal were evaluated in a multiple logistic re- in more than 6 randomized controlled Each patient was instructed to present gression model based on the backward studies. There is a significant reduction to the labour ward in a fasting state, and stepwise selection, where significant in the incidence of caesarean section in an initial cardiotocography was done. variables from the univariate analysis women where there is an intention to A clinical assessment as well as bedside were included. The adjusted measure undertake ECV without any increased ultrasound scan was performed by the of association between risk factors and risk to the baby [7]. Routine use of operator to confirm the fetal presenta- failed ECV was expressed as the OR ECV could reduce the rate of caesarean tion. The degree of fetal flexion and the with 95% CI. Adjusted or crude ORs delivery by about two-thirds [8]. It is position of the fetal back were noted. with 95% CI that did not include 1.0 recommended that all women with Tocolysis was given to patients with were considered significant. an uncomplicated breech pregnancy irritable uterus and the procedure was at term (37–42 weeks) should be of- done 1 hour later. fered ECV [1,9]. Although primar- The ECV procedure included Results ily intended for uncomplicated breech dislodging the fetal breech from the pregnancies at term, ECV has been pelvis while holding the fetus in a flexed For the 128 women recruited to the carried out successfully in previous position with both hands, turning the study the mean age was 30.1 (SD 6.7) caesarean sections [10,11] and during breech away from the pelvis. After years with a minimum of 17 years and a early labour [7]. completion of the procedure another maximum of 46 years.

163

Book 19-2.indb 163 2/21/2013 11:39:24 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Mode of delivery and duration Variables associated with None of the studied fetal complica- of hospitalization failed ECV tions—bradycardia, emergency cae- ECV succeeded in more than half of Multivariate logistic regression analy- sarean section, spontaneous rupture of cases (69, 53.9%). Most of the women sis was performed to predict independ- membranes and low Apgar score (< 5 at with successful ECV delivered nor- ent variables associated with failed 5 minutes)—were reported. mally (84.1%), while only 14.5% of ECV. Age (25–35, < 25 versus < 35 them delivered by caesarean section. years old), parity (1–3, 0 versus > 3), Conversely, normal vaginal delivery placental position (fundal, anterior Discussion was reported among only 8.5% of versus posterior), fetal weight (> 3 kg those who had spontaneous version versus ≤ 3 kg), maximum pool depth of The overall success rate of ECV in stud- with failed ECV and nearly two-thirds amniotic fluid (< 4.6 mm versus ≥ 4.6 ies published in the United Kingdom of them delivered by caesarean sec- mm) and tocolysis (used versus not was 46% [12,13]. Among published tion (62.7%). Vaginal breech delivery used) were included in the bivariate studies in the Untied States it reached was reported in 27.1% of women with and multivariate regression analyses. 65% [1] and 69.5% in a recent study [8]. In a study conducted at a tertiary teach- failed ECV compared with none of In the multivariate analysis, primi- ing hospital in Malaysia, involving 41 those with successful ECV (as all of gravida were significantly more likely them remained with cephalic fetal to be associated with failed ECV as op- pregnant women with malpresentation presentation until delivery). The dif- posed to those with parity > 3 (adjusted at term, ECV was successful in 63% of ference between the 2 groups was OR = 3.17, 95% CI: 1.08–7.01). On the women [14]. In our study of women statistically significant P( < 0.001) other hand, women presenting with an- with breech presentation at term, the (Table 1). terior placenta were significantly more success rate was 53.9%. Among those with success- likely to have failed ECV than those Most of the reports in the medical ful ECV, almost two-thirds stayed presenting with posterior placenta (ad- literature that were predictive of the for only 1 day in the hospital post- justed OR= 2.10, 95% CI: 1.01–5.17). outcome of ECV were limited by small delivery (63.8%) as compared with Regarding the amount of amniotic fluid, sample sizes, with conflicting results. 28.8% among those with failed ECV. patients with maximum pool depth < Univariate analysis predominated in On the other hand, only 13.0% of 4.6 cm had almost double the risk for these reports, but this does not exclude women with successful ECV stayed failed ECV compared with those having the possibility of confounding effect > 3 days in the hospital post-delivery maximum pool depth ≥ 4.6 cm (ad- between the variables [15–18]. compared with 37.3% among those justed OR= 1.99, 95% CI: 1.04–4.58). A variety of factors have been as- with failed ECV. This difference was However, maternal age, fetal weight and sociated with ECV success in the statistically significant P( < 0.001) tocolysis showed no significant associa- literature. Newman et al. [19], using (Table 1). tion with failed ECV (Table 2). linear regression analysis, found that

Table 1 Association of outcome of external cephalic version with mode of delivery and duration of hospitalization Variables External cephalic version P-valuea Successful Failed (n = 69) (n = 59) No. % No. % Mode of delivery Normal vaginal 58 84.1 5 8.5 Instrumental 1 1.4 1 1.7 Breech 0 0.0 16 2 7. 1 Caesarean 10 14.5 37 62.7 < 0.001 Duration of post-delivery hospitalization (days) 1 44 63.8 17 28.8 2–3 16 23.2 20 33.9 > 3 9 13.0 22 37.3 < 0.001

aChi-squared test.

164

Book 19-2.indb 164 2/21/2013 11:39:24 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 2 Risk factors for failed external cephalic version for breech presentation at term: results of bivariate and multivariate logistic regression analyses Variables External cephalic version Crude OR (95% CI) Adjusted OR (95% CI) Successful Failed (n = 69) (n = 59) No. % No. % Age (years) > 35a 20 29.0 13 22.0 1 –b 25–35 33 47.8 29 49.2 1.35 (0.53–3.49) < 25 16 23.2 17 28.8 1.63 (0.55–4.89) Parity > 3a 29 42.0 19 32.2 1 1 1–3 28 40.6 21 35.6 1.14 (0.47–2.79) 1.56 (0.62–3.12) 0 12 1 7. 4 19 32.2 2.42 (0.87–6.80) 3.17 (1.08–7.01) Placental position Posteriora 36 52.2 21 35.6 1 Fundal 10 14.5 7 11.9 1.20 (0.35–4.12) 0.85 (0.41–4.06) Anterior 23 33.3 31 52.5 2.31 (1.01–5.33) 2.10 (1.01–5.17) Fetal weight (kg) ≤ 3a 53 76.8 44 74.6 1 –b > 3 16 23.2 15 25.4 1.13 (0.47–2.74) Amniotic fluid MPD (cm)c ≥ 4.6a 41 59.4 23 39.0 1 1 < 4.6 28 40.6 36 61.0 2.29 (1.06–4.97) 1.99 (1.04–4.58) Tocolysis Not useda 56 81.2 43 72.9 1 –b Used 13 18.8 16 2 7. 1 1.60 (0.65–4.00)

aReference category; bRemoved from final model;c Based on median cut-off value. OR = odds ratio; CI = confidence interval;MPD = maximum pool depth.

parity, cervical dilatation, estimated fetal caesarean section rates is therefore very section rate in women with successful weight, breech station and placental labour intensive. ECV. Nevertheless, if ECV is to make implantation site were the most useful It has been reported that the use of an impact on breech vaginal deliveries predictors of ECV success. Maternal tocolytic agents increased the success rate and caesarean sections, efforts must be weight, gestational age, type of breech, of ECV, both when used routinely and made to expand the suitability criteria amniotic fluid volume and cervical when used selectively [7,13]. In our study, for ECV and increase the success rate of effacement did not have a significant the use of tocolytic agents was not sig- ECV, without increasing the morbidity effect. The present study found that the nificantly associated with successful ECV, and mortality (both maternal and peri- amount of amniotic fluid had an almost perhaps due to the small sample size. Two natal) associated with the procedure. significant but non-linear effect on suc- other small prospective trials showed no cess rate, which may explain why linear benefit of tocolysis in ECV [20,21]. regression analysis excluded it in New- Acknowledgements man’s study. In our population, a 48.2% Around 3 to 5 days in the hospital is lower caesarean section rate and 27.1% the common length of stay following cae- We would like to thank all physicians, lower vaginal breech delivery rate was sarean birth, whereas it is less than 1–3 midwives and nurses in the department seen in cases of successful ECV of term days for a vaginal birth [22]. In accord- of obstetrics and gynaecology who par- breeches. Each attempted ECV took ance with these data, our study revealed ticipated in diagnosing, referring and from 5–10 minutes of operator time, that successful ECV was significantly counselling of patients. We would also preceded and followed by cardiotocog- associated with shorter hospital stays. thank members of research unit for their raphy for 40 minutes. Use of ECV as a Offering an ECV service reduced advice and help throughout preparation method for reduction of term breech the breech delivery rate and caesarean of this work.

165

Book 19-2.indb 165 2/21/2013 11:39:24 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

References

1. External cephalic version (ECV) and reducing the incidence of 12. Bewley S et al. The introduction of external cephalic version breech presentation (Green-top Guideline 20a). London, Royal at term into routine clinical practice. European Journal of Ob- College of Obstetricians and Gynaecologists, 2006. stetrics, Gynecology, and Reproductive Biology, 1993, 52:89–93. 2. Danielian PJ, Wang J, Hall MH. Long term outcome by method 13. Breech presentation, management (Green-top Guideline 20b). of delivery of fetuses in breech presentation at term: popula- London, Royal College of Obstetricians and Gynaecologists, tion based follow up. British Medical Journal, 1996, 312:1451– 2001. 1453. 14. Yong SPY. Introducing external cephalic version in a Malaysian 3. Rietberg CC, Elferink-Stinkens PM, Visser GH. The effect of setting. Hong Kong Medical Journal, 2007, 13:40–45. the Term Breech Trial on medical intervention behaviour and 15. Brocks V, Philipsen T, Secher NJ. A randomized trial of external neonatal outcome in The Netherlands: an analysis of 35,453 cephalic version with tocolysis in late pregnancy. British Jour- term breech infants. British Journal of Obstetrics and Gynaecol- nal of Obstetrics and Gynaecology, 1984, 91:653–656. ogy, 2005, 112:205–209. 16. Donald WL, Barton JJ. Ultrasonography and external cephalic 4. Hannah ME et al.; Term Breech Trial Collaborative Group. version at term. American Journal of Obstetrics and Gynecology, Planned caesarean section versus planned vaginal birth for 1990, 162:1542–1545. breech presentation at term: a randomised multicentre trial. 17. Shalev E et al. External cephalic version at term—using to- Lancet, 2000, 356:1375–1383. colysis. Acta Obstetricia et Gynecologica Scandinavica, 1993, 5. Hofmeyr GJ, Hannah ME. Planned caesarean section for term 72:455–457. breech delivery. Cochrane Database of Systematic Reviews, 18. Klatt TE, Cruikshank DP. Breech, other malpresentations, 2003, (3):CD000166. and umbilical cord complications. In: Gibbs RS, et al., eds. 6. Cunningham FG et al. Breech presentation and delivery. Danforth's Obstetrics and Gynecology, 10th ed. Philadelphia, In: Williams’ obstetrics, 23rd ed. New York, McGraw-Hill, Lippincott Williams and Wilkins, 2008:400–416. 2010:527–543. 19. Newman RB et al. Predicting success of external cephalic 7. Hofmeyr GJ, Kulier R. External cephalic version for breech version. American Journal of Obstetrics and Gynecology, 1993, presentation at term. Cochrane Database of Systematic Reviews, 169:245–249. 2000, (2):CD000083. 20. Robertson AW et al. External cephalic version at term: is a toco- 8. Zhang J, Bowes EA, Fortney JA. Efficacy of external cephalic lytic necessary? Obstetrics and Gynecology, 1987, 70:896–899. version: a review. Obstetrics and Gynecology, 1993, 82:306–312. 21. Tan GWT et al. A prospective randomised controlled trial of 9. Lau TK et al. Predictors of successful external cephalic version external cephalic version comparing two methods of uterine at term: a prospective study. British Journal of Obstetrics and tocolysis with a non-tocolysis group. Singapore Medical Jour- Gynaecology, 1997, 104:798–802. nal, 1989, 30:155–158. 10. Shalev E, Battino S, Giladi Y. External cephalic version at term 22. Santrock JW, Brown WC. Child development, 6th ed. Madison, using tocolysis. Acta Obstetricia et Gynecologica Scandinavica, Wisconsin, Brown and Benchmark, 1994. 1993, 72:455–457. 11. Flamm BL et al. External cephalic version after caesarean section. American Journal of Obstetrics and Gynecology, 1991, 165:370–372.

166

Book 19-2.indb 166 2/21/2013 11:39:24 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Depression among a group of elders in Alexandria, Egypt H.M. El Kady 1 and H.K. Ibrahim 1

االكتئاب لدى جمموعة من املسنني ّيف اإلسكندرية، مرص هبة حممود القايض، هالة قدري إبراهيم اخلالصـة: َّيقدرأن االكتئاب سيكون هو السبب الرئييس للعبء املريض لدى املسننيّ يف عام 2020. وهتدف هذه الدراسة لتقييم معدل انتشار االكتئاب بني جمموعة من املسننييف ّاإلسكندرية، مرص، ومقارنة خصائص املسنني يف ثالثة مواقع خمتلفة، وهي دراسة وصفية مستعرضة شملت مئة شخص ممن تزيد أعامرهم عىل ستني ًعاما، يقيمون يف أحد النزل احلكومية و ىف أحد أجنحة قسم األمراض الباطنة باملستشفى الرئيسى اجلامعى، ويف جناح الطب الباطني ويف أحد املناطق العشوائية يف اإلسكندرية. َواستخد َمت الباحثتان سلم مقياس االكتئاب للمسنني )بشكله املخترص(، فوجدتا أنأعىل معدالت االنتشار التي تشري أو توحي باالكتئاب كانت لدى املودعني باملستشفى الذين ُأدخلوا املستشفيات )79%(، أو الذين لدهيم مرض 83.3 64.9 80 85.7 عيني ) %(، أو أورام ) %(، أو لدهيم أكثر من َمرض ْين مزمنَ ْين ) %(، أو الذين يتناولون أكثر من أربعة أدوية ًيوميا ) %(، أو الذين 90.9 يعتمدون ًبدنياعىل غريهم أكثر من سواهم، أو الذين أدخلوا املستشفى ألكثر من مرتني خالل السنوات الثالث املنرصمة ) %(. و َتمَ ُّس احلاجة للنظر يف العوامل القابلة للتعديل والتي ِّتؤثرعىل االكتئاب البتكار تدخالت تستهدف حتسني األحوال املزاجية لدى املسنني.

ABSTRACT Depression is projected to be the leading cause of disease burden in older populations by the year 2020. The aim of this study was to assess the prevalence depression among a group of elders in Alexandria, Egypt and compare the characteristic of elders in 3 different settings. A descriptive, cross-sectional study was carried out among 100 people aged 60+ years in a government elderly home, an internal medicine ward in Alexandria Main University Hospital and a slum area in Alexandria. Based on the Geriatric Depression Scale (short form) the highest prevalence of scores that were suggestive or indicative of depression was among elders who were hospitalized (79.0%), had ophthalmic diseases (85.7%) or tumours (80.0%), had 2+ chronic morbidities (64.9%), were taking 4+ medications daily (83,3%), were more physically dependent and had had 2+ hospital admissions in the last 3 years (90.9%). Modifiable factors that impact on depression need to be considered in developing interventions for improving mood states in the elderly.

Dépression dans un groupe d'anciens à Alexandrie (Égypte)

RÉSUMÉ Il a été estimé que la dépression représenterait la cause principale de la charge morbide dans les populations de personnes âgées d'ici 2020. L'objectif de la présente étude était d'évaluer la prévalence de la dépression dans un groupe d'anciens à Alexandrie (Égypte), et de comparer les caractéristiques de ces derniers dans trois milieux différents. Une étude transversale descriptive a été menée auprès de 100 personnes âgées de plus de 60 ans dans une maison de retraite du gouvernement, dans un service de médecine interne et dans un bidonville à Alexandrie. À partir de la version abrégée de l'échelle gériatrique de dépression, les scores indiquant ou suggérant une dépression étaient plus élevés dans les populations suivantes : les anciens hospitalisés (79,0 %), les personnes âgées souffrant de maladies ophtalmiques (85,7 %) ou de tumeurs (80,0 %), atteintes de plus deux comorbidités chroniques (64,9 %), ou prenant plus de quatre traitements quotidiens (83,3 %), ou davantage dépendants physiquement et admis plus de deux fois à l'hôpital au cours des trois années précédentes (90,9 %). Les facteurs modifiables ayant un impact sur la dépression doivent être pris en compte dans l'élaboration d'interventions visant à lutter contre la dépression chez les anciens.

1Department of Family Health, High Institute of Public Health, University of Alexandria, Alexandria, Egypt (Correspondence to H.K. Ibrahim: [email protected]). Received: 15/02/12; accepted: 21/02/12

167

Book 19-2.indb 167 2/21/2013 11:39:25 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction of self-actualization, reliance on others The sample size was calculated using and a general lack of mastery or sense Epi-info 2002 software. We randomly Depression is an important problem of control. Institutionalized persons are selected 33 elders from institutions, 33 that may highly influence the quality of typically those with no spouse or living who were in hospital at the time of the life of elderly people in different settings. children or with offspring who are not study and 34 who were currently living The prevalence of depression in people available to provide care [10]. In many in the community. aged 65 years and older may be as high as cases, these individuals are considered Data collection 40% in hospitalized and 30% in nursing abandoned by children who have not home patients and 8%–15% in commu- fulfilled their filial obligations. Conse- A structured pre-coded interview nity settings. Studies of depressed elders quently, the institutionalized are socially schedule was used to collect data about indicate that they have comparable or disadvantaged and depend on support sociodemographic characteristics and poorer physical functioning than elders from limited subsidies from local gov- medical history (the type and number with chronic medical conditions such ernment or collectives [9]. Depression of chronic diseases from which the as heart and lung diseases, arthritis and can also have adverse health effects for patient suffered, the number of medica- diabetes [1,2]. Depression can increase many hospitalized older adults. High tions he/she was taking and the number perceptions of poor health, the utiliza- levels of depressive symptoms are asso- of hospital admissions in the 3 years tion of medical services and health care ciated with poor treatment adherence, before the study). costs. Conversely, a diagnosis of certain longer lengths of hospital stays, increase The Katz scale for ADLs was used chronic diseases may trigger symptoms in hospital readmission and reduced to assess degree of dependency in of depression [3]. functional status [11]. performing the basic ADLs including There are many factors related to Most of the research on this subject bathing, dressing, toileting, transfer, depression among elders. It has been at- is based on high-income countries and urinary and faecal continence and feed- tributed to the increasing stressors with little is known about elders in developing ing [13]. The 6 functions are measured age resulting from declining health and countries. The aim of this study was to as- and scored according to the individual’s dwindling social relationships [4]. Low sess the prevalence of depression among actual performance of these functions. socioeconomic status was found to be elders in Alexandria, Egypt, and compare They are categorized into 3 levels of significantly associated with depression the characteristics of those from 3 differ- dependency: each item is scored from after adjustment for age, illness, sex and ent settings: hospitalized, institutional- 0–2, where 2 = full independence self-rated health [5]. Presence of co ized or community dwelling. without need for human assistance, 1 morbidity may correlate with depression = partial dependence with the need for among the elderly. Also, several clinical some help and 0 = total dependence trials reported a positive correlation of de- Methods with an inability to perform the task pression with functional decline, impair- even with assistance. The total score of ment in activities of daily living (ADL) Study setting and sample the scale is from 0–12. According to and a steep decline in strength [6]. A descriptive, cross-sectional study the scale patients are classified into 3 Research suggests that the liv- was carried out among elders aged 60+ categories: totally independent (score ing environment of older adults is an years in a government-run home for 9–12), partially dependent (score 5–8) important determinant of health and the elderly (Dar El Saada), an internal or totally dependent (score 0–4). The longevity. Existing evidence indicates medicine ward in the Main Univer- Katz scale was translated in to Arabic that the health care needs and costs sity Hospital and a slum area (Ezbet language and tested for reliability using of institutionalized persons are much Sekeena) in Alexandria city. test–retest ], which showed a Pearson higher than those of elders in other set- The sample included all elders who correlation coefficient of 0.8. tings [7]. Significant differences have were able to communicate and agreed The short form of the Geriatric De- been found between elders residing to participate in the study. Using a pression Scale (GDS) [14] was used in institutions or in the community in power of 80% to detect a difference of to assess the presence or absence of terms of mortality risk, daily function- 37% [4,12] at the depression prevalence depression among the elders. The re- ing, disease prevalence, life satisfaction, among hospitalized and community sponses to each item are yes/no (scored depression and quality of life [8,9]. For dwelling elders as the largest recorded 1 or 0) and the total scores (15 items) many older adults, especially in devel- difference in comparison to institution- are classified as follows: no depression oped countries, institutionalization rep- alized elders, α-error = 0.05, the minimal (score 0–4), suggestive of depression resents a loss of independence, marked required sample size was found to be (score 5–9), almost always indicative by feelings of limited usefulness, a loss 28 (increased to 33) for each group. of depression (score > 9). The scale was

168

Book 19-2.indb 168 2/21/2013 11:39:25 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

translated into Arabic, and validated were used as tests of significance. There Figure 1 shows the distribution of by the juries including 6 experts in ad- were no missing data. the study sample according to the GDS ministration and geriatric health. The scores. The majority of community required corrections and modifications dwelling and institutionalized elders were carried out accordingly. Reliability Results had no depression (77.0% and 64.0% was tested using internal consistency respectively), but the majority of hos- and a reliability coefficient (Cronbach The mean age for hospitalized elders pitalized elders had scores that were alpha) of 0.88 was found for the 15 was 70.0 (SD 5.3) years, for community suggestive or indicative of depression items of the scale. dwelling elders was 68.7 (SD 5.2) years (58.0% and 21.0% respectively). The The study was approved by the eth- and for institutionalized elders was 69.0 data also showed that 18.0% of com- ics committee of the High Institute of (SD 4.3) years. Among hospitalized munity dwelling and 24.0% of institu- Public Health, University of Alexandria and institutionalized elders the high- tionalized elders had scores that were before the start of the study and written est percentage were widowed (54.5% suggestive or indicative of depression approvals from the elders were taken and 69.7% respectively) whereas more respectively. before interviewing. Each elder was than two-thirds of those living in the Table 1 shows the relation between interviewed individually after explana- community were married (67.6%). As GDS score and demographic variables. tion of the purpose of the study and regards education, the majority of hos- The percentage of elders with scores assurances of confidentiality were given. pitalized elders (78.8%) were illiterate that were suggestive or indicative of Each individual interview took about 20 or could only read and write, while the depression were higher among those minutes. Data were collected during a majority of community dwelling and in- aged 75+ years (75.0%) compared with period of 5 months from the beginning stitutionalized elders (82.4% and 60.6% those aged 60–< 65 years (20.0%); of March 2011 to August 2011. respectively) were secondary school or among females (57.1%) compared with university graduates. Just over half of the males (35.3%); in widowed or single Statistical analysis hospitalized (51.5%) and the majority (90.0% and 80.0%) than married elders Data management and computations of community dwelling elders (67.6%) (31.7%); among elders living with their of descriptive statistics and prevalence considered there income was sufficient, offspring (68.4%) compared with those were performed using SPSS, version 16. but for the institutionalized elders more who lived in elderly homes (36.4%). Differences with P-value < 0.05 were than one-third (36.4%) reported it as The rate of depression was also higher considered statistically significant. The insufficient and the income of about half among those who were illiterate/semi- Pearson chi-squared, Monte Carlo of them (48.5%) was not sufficient and literate (66.7%) than those having sec- exact and Fisher exact probability tests they had to borrow money. ondary/university education (24.5%);

No depression Suggestive of depression Indicative of depression

80 77.0

70 64.0

s 58.0

t 60 n e d

n 50 o p s

e 40 r

f o 30 % 24.0 21.0 21.0 20 18.0 12.0 10 6.0

0 Hospitalized (n = 33) Community dwelling (n = 34) Institutionalized (n = 33)

Figure 1 Distribution of the study sample of elders according to depression scores on the Geriatric Depression Scale (short form)

169

Book 19-2.indb 169 2/21/2013 11:39:25 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 1 Relationship between depression on the Geriatric Depression Scale (short form) and the studied variables for the study sample of elders (n = 100) Variable Total No depression Depressiona χ2 P-value No. % % Age (years) 60– 20 80.0 20.0 65– 28 71.4 28.6 0.61 0.895 70– 36 55.6 44.4 75+ 16 25.0 75.0 Sex Male 51 64.7 35.3 4.8 0.028 Female 49 42.9 5 7. 1 Marital status Married 41 68.3 31.7 Widowed 50 10.0 90.0 7. 8 0.040b Divorced 4 25.0 75.0 Single 5 20.0 80.0 Educational level Illiterate/read and write 36 33.3 66.7 Primary 13 38.5 61.5 23.4 0.001b Preparatory 2 100.0 0.0 Secondary/university 49 75.5 24.5 Work status Housewife 32 31.3 68.8 13.3 0.009 Private employment 9 55.6 44.4 On pension/left work 59 66.1 33.9 Living condition 7. 5 0.186 Alone 16 62.5 37.5 With spouse only 15 46.7 53.3 With spouse and offspring 17 62.5 37.5 With offspring only 19 31.6 68.4 Elderly home 33 63.6 36.4 Source of income 17.8 0.001b Pension 65 76.9 23.1 Salary/wage 1 100.0 0.0 Family support 18 16.7 83.3 Social support 16 0.0 100.0 Income 14.3 0.006 Sufficient 45 88.9 11.1 Sufficient and saves 8 50.0 50.0 Insufficient 24 41.7 58.3 Insufficient and borrows 23 0.0 100.0 Presence of chronic disease(s) 15.0 0.001 No 15 100.0 0.0 Yes 85 45.9 54.1 No. of chronic diseases 23.2 0.001 0 15 100.0 0.0 1 28 67.9 32.1 2+ 57 35.1 64.9

170

Book 19-2.indb 170 2/21/2013 11:39:25 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 1 Relationship between depression on the Geriatric Depression Scale (short form) and the studied variables for the study sample of elders (n = 100) (concluded) Variable Total No depression Depressiona χ2 P-value No. % % No. of medications received 33.5 0.001 0 19 100.0 0.0 1– 51 58.8 41.2 4+ 30 16.7 83.3 No. of times hospitalized in last 3 years 15.1 0.001 0 76 64.5 35.5 1 13 30.8 69.2 2+ 11 9.1 90.9

aScores suggestive (score 5–9) or indicative (score > 9) of depression. bP-value based on Monte-Carlo exact test.

in those who were housewives (68.8%) ill elders, with the highest rates among in the literature can be attributed to versus those receiving a pension or had elders who had ophthalmic diseases differences in the methodology and left work (33.9%); and those receiving (85.7%), tumours (80.0%) and renal diagnostic criteria used. social support (100%) rather than a diseases (100.0%). The present study also found that pension (23.1%) and with insufficient depression was more prevalent among income/borrowing money (100%) females. This agrees with numerous rather than a sufficient income (11.1%). Discussion epidemiological studies showing that Table 1 also shows that more than women have a higher prevalence of half of the elders who had chronic dis- Depression is a major contributor to depression than men across genera- eases (54.1%) had scores that were sug- health care costs associated with older tions and cultures [17,18] and that sex gestive or indicative of depression and populations, and is projected to be the differences in the prevalence of depres- more elders who had 2+ diseases had leading cause of disease burden in older sion, which first become apparent in depression scores (64.9%) than those populations by the year 2020 [2]. Sev- early adolescence, continue into the who had only 1 disease (32.1%) or no eral factors are related to the causes of 60+ years age group. In addition, most diseases (0%). Elders were also more depression. Some researchers have fo- studies have shown that women are likely to have depression scores if they cused on socioeconomic variables such more exposed to risk factors for depres- were taking 4+ medications (83.3%) as advanced age, low education, poor sion, such as financial difficulties, -wid than 1–3 medications (41.2%) and economic status, manual occupation owhood, social isolation, low education were hospitalized 2+ times in the last 3 and current living situation as causes and functional disability. years (90.9%) than those who had no of depression, and demonstrated that The present study revealed that de- previous hospitalizations (35.5%). these variables had a relationship with pression was more prevalent among The correlation between the GDS depression [8]. widowed and single elders, who may scores and Katz scale scores showed a The present study showed that the suffer a sense of loneliness and bereave- significant inverse intermediate relation highest rate of depression was among ment that aggravates feelings of sadness between score of Katz scale for ADL those aged 75+ years (75.0%) com- among such elders. This is in accordance and level of depression as measured by pared with younger age groups. This is with a study in Korea which reported the GDS among hospitalized (–0.43, in accordance with other studies, which that widowhood significantly influenced P < 0.05), community dwelling (–0.4, reported similar findings [1,12]. This depression among older adults [19]. P < 0.05) and institutionalized (–0.5, may be attributed to the fact that with Socioeconomic variables, such as P < 0.01) elders and among the whole increasing old age people experience a low education, poor economic status sample (–0.5, P < 0.01). greater loss of physiological, psychologi- and current living situation, have been Table 2 shows the relationship be- cal and social functioning and become reported as causes of depression among tween GDS scores and chronic mor- increasingly prone to depression. How- elders in previous studies [1,8]. The bidity. The highest percentage of GDS ever, some other studies indicated a present study revealed that depression scores that were suggestive or indicative lower prevalence of depression among was more prevalent among elders of of depression was among chronically the elderly [15,16]. Such inconsistency lower educational status, which may

171

Book 19-2.indb 171 2/21/2013 11:39:25 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Table 2 Relationship between depression on the Geriatric Depression Scale (short form) and presence of chronic morbidities for the study sample of elders (n = 100) Chronic disease Total No depression Depressiona χ2 P-value No. % % Locomotor disease No 70 62.9 3 7. 1 7. 4 0.007 Yes 30 33.3 66.7 Cardiac disease No 74 59.5 40.5 3.4 0.065 Yes 26 38.5 61.5 Hypertension No 52 65.4 34.6 5.7 0.017 Yes 48 41.7 58.3 Respiratory disease No 98 54.1 45.9 – 0.995b Yes 2 50.0 50.0 Ophthalmic disease No 93 57.0 43.0 – 0.046b Yes 7 14.3 85.7 Diabetes No 62 62.9 3 7. 1 5.2 0.023 Yes 38 39.5 60.5 Tumour No 95 55.8 44.2 – 0.177b Yes 5 20.0 80.0 Psychiatric disease No 98 54.1 45.9 – 0.995b Yes 2 50.0 50.0 Renal disease No 94 5 7. 4 42.6 – 0.008b Yes 6 0.0 100.0 ENT disease No 99 53.5 46.5 – 0.997b Yes 1 100.0 0.0 GIT disease No 86 57.0 43.0 2.2 0.139 Yes 14 35.7 64.3 aScores suggestive (scores 5–9) or indicative (score > 9) of depression. bP-value based on Fisher exact probability. ENT = ear, nose the throat; GIT = gastrointestinal tract.

reflect their lower socioeconomic sta- The present study revealed that hypertension [2,21]. Several factors tus, which in turn may mean that they depression was more prevalent among associated with depressive symptoms, are exposed to more life stressors. The chronically ill elders, especially those including obesity-promoting health present study also revealed that depres- with 2 or more diseases, who received behaviours (e.g., physical inactivity, sion was more prevalent among elders 4 or more medications daily and those hypercaloric diets) and activation of whose source of income was social sup- who had ophthalmic diseases, tumours the neuroendocrine and inflammatory port and among those who reported and renal diseases. This is in accord- responses (resulting in increased cor- their income as insufficient. This agrees ance with other studies which found tisol, catecholamines and cytokines), with a study in Japan which found that that depressive symptoms were higher can induce the development of these low socioeconomic status was signifi- among chronically ill elders, especially chronic diseases. These associations cantly associated with depression, after those who had diabetes, stroke, car- may be also related to increased risk adjustment for age and sex [20]. diac disease, chronic lung disease and of depressive symptoms in individuals

172

Book 19-2.indb 172 2/21/2013 11:39:25 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

with these chronic diseases, increased been found between institutionalized level among the elderly population can risk of these chronic diseases in indi- and community-residing individuals in pave the way to developing effective viduals with depressive symptoms or mortality risk, daily functioning, disease interventions for decreasing geriatric both [2]. prevalence, life satisfaction, depression depression. Any interventions should In the present study, depression and quality of life [11]. In addition, be focused on factors that are modifia- was more prevalent among elders who depression can have adverse health ef- ble or on the behavioural characteristics had 2 or more previous hospital admis- fects for many hospitalized older adults. which contribute to depression. sions. This agrees with a study in the High levels of depressive symptoms are United States, which showed that high associated with poor treatment adher- rates of depressive symptoms were as- ence, longer lengths of stay, increase in Conclusion sociated with more frequent hospital hospital readmission and reduced func- admissions [11]. The present study tional status [24]. Compounding this In conclusion, depression rates were also revealed that there was an inverse relationship is the fact that hospitaliza- higher among hospitalized and insti- relationship between scores on the Katz tion can increase depressive symptoms tutionalized elders, among women, scale for ADL and level of depression in many older adults. The prevalence widowed and single elders, among as measured by the GDS among the of depressive symptoms was as high as housewives, among those living with study sample. This is in accordance 50% among hospitalized older adults in their offspring only and among more with several clinical trials that reported a a study in Australia, yet it was reported physically dependent elders. Depres- positive correlation of depression with to be much lower in persons 3–12 sion was also more prevalent among functional decline and impairment of months after hospital discharge [25]. elders whose source of income was ADL [22,23]. This is due to the fact The present study revealed similar find- social support and among those whose that physical dependence brings about ings; depression was more prevalent income was insufficient. It was more marked feelings of limited usefulness, among hospitalized and institutional- prevalent among elders having 2 or loss of self-actualization, reliance on ized elders than among those living in more chronic morbidities, those taking others and a general lack of mastery or the community. 4 or more medications daily and those sense of control. Most of the previous studies on geri- who had 2 or more hospital admissions Research suggests that the living atric depression have focused on its high in the last 3 years. It is recommended environment of older adults is an im- prevalence. They have not attempted to that modifiable factors such as com- portant determinant of their health and reduce the occurrence of depression petence in ADL, presence of chronic longevity. Existing evidence indicates among elders. The way in which depres- morbidities and degree of social and that the health care needs of institu- sion among the elderly is related with financial support should be considered tionalized persons are much higher their residence and competence of ADL as variables in developing interven- than those of elders in other settings has rarely been studied. The identifica- tions for improving mood states in the [9]. Significant differences have also tion of factors to predict the depression elderly.

References

1. Huang CQ et al. Chronic diseases and risk for depression in old 6. Rantanen T et al. Depressed mood and body mass index as age: a meta-analysis of published literature. Ageing Research predictors of muscle strength decline in old men. Journal of the Reviews, 2010, 9:131–141. American Geriatrics Society, 2000, 48:613–617. 2. Von Korff M et al. Disability and depression among high utiliz- 7. Federal Interagency Forum on Aging Related Statistics. Older ers of health care. A longitudinal analysis. Archives of General Americans 2004: key indicators of well-being. Washington DC, Psychiatry, 1992, 49:91–100. US Government Printing Office, 2004. 3. Carvalhais SM et al. The influence of socio-economic condi- 8. Van Dijk PT et al. Comorbidity and 1-year mortality risks in tions on the prevalence of depressive symptoms and its covari- nursing home residents. Journal of the American Geriatrics Soci- ates in an elderly population with slight income differences: ety, 2005, 53:660–665. the Bambuí Health and Aging Study (BHAS). International Jour- 9. Harris Y, Cooper JK. Depressive symptoms in older people nal of Social Psychiatry, 2008, 54:447–456. predict nursing home admission. Journal of the American Geri- 4. Borson S et al. Geriatric mental health services research: strate- atrics Society, 2006, 54:593–597. gic plan for an aging population. American Journal of Geriatric 10. Davis MA et al. Living arrangements, changes in living arrange- Psychiatry, 2001, 9:191–204. ments, and survival among community dwelling older adults. 5. Demura S, Sato S. Relationships between depression, lifestyle American Journal of Public Health, 1997, 87:371–377. and quality of life in the community dwelling elderly: a com- 11. Carrie A et al. Patterns and correlates of depression in hospital- parison between gender and age groups. Journal of Physiologi- ized older adults. Archives of Gerontology and Geriatrics, 2011, cal Anthropology and Applied Human Science, 2003, 22:159–166. 30:33–36.

173

Book 19-2.indb 173 2/21/2013 11:39:26 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

12. Jeon HS, Dunkle RE. Stress and depression among the oldest- 19. Shin KR et al. Depression among community-dwelling older old: a longitudinal analysis. Research on Aging, 2009, 31:661– adults in Korea: a prediction model of depression. Archives of 687. Psychiatric Nursing, 2009, 23:50–57. 13. Katz S et al. Studies of illness in the aged: the index of ADL: 20. Murata C et al. Association between depression and socio- a standardized measure of biological and psychological economic status among community-dwelling elderly in Japan: functions. Journal of the American Medical Association, 1963, the Aichi Gerontological Evaluation Study (AGES). Health and 185:914–919. Place, 2008, 14:406–414. 14. Sheikh JI, Yesavage JA. Geriatric Depression Scale: recent evi- 21. Nicholson A, Kuper H, Hemingway H. Depression as an dence and development of a shorter version. In: Brink TL, ed. aetiologic and prognostic factor in coronary heart disease: Clinical gerontology. New York, Haworth Press, 1986:112. a meta-analysis of 6362 events among 146 538 participants 15. Copeland JR et al. Depression in Europe. Geographical distri- in 54 observational studies. European Heart Journal, 2006, bution among older people. British Journal of Psychiatry, 1999, 27:2763–2774. 174:312–321. 22. Unützer J et al. Depressive symptoms and mortality in a pro- 16. Mara Cristina F et al. Depressive symptoms and cognitive spective study of 2,558 older adults. American Journal of Geriat- performance of the elderly: relationship between institution- ric Psychiatry, 2002, 10:521–530. alization and activity programs. Revista Brasileira de Psiquiatria, 23. Onishi J et al. The relationship between functional disability 2006, 28:118–121. and depressive mood in Japanese older adult inpatients. Jour- 17. Wang JK, Su TP, Chou P. Sex differences in prevalence and risk nal of Geriatric Psychiatry and Neurology, 2004, 17:93–98. indicators of geriatric depression: the Shih-Pai community- 24. Cullum S et al. Does depression predict adverse outcomes for based survey. Journal of the Formosan Medical Association, older medical inpatients? A prospective cohort study of indi- 2010, 109:345–353. viduals screened for a trial. Age and Ageing, 2008, 37:690–695. 18. Zunzunegui MV et al.; CLESA Working Group. Gender differ- 25. Brodaty H et al. Rates of depression at 3 and 15 months post- ences in depressive symptoms among older adults: a cross- stroke and their relationship with cognitive decline: the Sydney national comparison: the CLESA project. Social Psychiatry and Stroke Study. American Journal of Geriatric Psychiatry, 2007, Psychiatric Epidemiology, 2007, 42:198–207. 15:477–486.

Development of regional survey on ageing and health

The World Health Organization Regional Office for the Eastern Mediterranean (WHO/EMRO) is currently developing a regional comprehensive survey on ageing and health in Member States to collect a range of data on older persons. Data will be collected on demographic, socioeconomic, health, age-friendly health services and relevant legislation indicators in order to create an up-to-date database to inform evidence-based decision-making regarding comprehensive care for the elderly at the country and regional levels.

This survey will seek to include data on monitoring and evaluation of national capacities, numbers of primary health care centres which meet the criteria of age-friendly primary health care and changes in the environment which promote the development of age-friendly cities, and in turn, age-friendly countries

Further information about the work of WHO/EMRO on the health of the elderly is available at: http://www.emro. who.int/fr/entity/elderly-health/

174

Book 19-2.indb 174 2/21/2013 11:39:26 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Frequency of impaired glucose tolerance and diabetes mellitus in subjects with fasting blood glucose below 6.1 mmol/L (110 mg/dL) S.H. Khan,1 A. Ijaz,2 S.A. Raza Bokhari,1 M.S. Hanif 1 and N. Azam 1

تواتر خلل حتمل الغلوكوز ّريوالسك لدى من يقل غلوكوز الدم عىل الريق لدهيم عن 6.1 مييل مول/لرت )110 مييل غرام/دييس لرت( إسكندر حياة خان، عامر إعجاز، سيد عون رضا شاه يبخار، حممد ِشهزاد حنيف، نائلة أعظم اخلالصـة: ال يوجد اتفاق عىل تشخيص ّالسكري وفق املعايري املتاحة، ويعتمد إىل ّحد كبري عىل نتائج غلوكوز الدم الصيامي )عىل الريق(. وقد 127 2011 2010 َأجرى الباحثون هذه الدراسة يف َعام ْي - ، وهتدف لقياس تواتر خلل حتمل الغلوكوز ّوالسكري لدى ًشخصا يقل لدهيم غلوكوز الدم الصيامي عن 7مييل مول/لرت، وكذلك لقياس التوافق بني معايري التشخيص القياسية املختلفة. ووضع الباحثون املراجعني للمخترب من أجل حتليل غلوكوز الدم الصيامي أمام ٍحتد بالغلوكوز الفموي مقداره 75 ًغراما عىل مدى ساعتني، من أجل استبعاد تشخيص السكري؛ فوجدوا أن جممل 40.6%من املراجعني الذين كان مستوى غلوكوز الدم الصيامي لدهيم يصل إىل 5.6 - 6.0 مييل مول/لرت كان لدهيم تنظيم غري سوي للغلوكوز ًوفقا للمعيار الذهبي للتحدي بالغلوكوز. وكان التوافق بني معايري اجلمعية األمريكية ّللسكري ومنظمة الصحة العاملية مقبوالً )كابا = 0.32(. ويرى الباحثون أن حاالت اضطراب استقالب الغلوكوز بام فيها خلل حتمل الغلوكوز، ّوالسكري يمكن أن ُت َفصاد ضمن جمموعة نتائج غلوكوز الدم الصيامي التي تقل عن 6.1مييل مول/لرت ) 110مييل غرام/دييس لرت(.

ABSTRACT The diagnosis of diabetes mellitus by the available criteria is controversial and relies heavily on fasting glucose results. This cross-sectional study in 2010–2011 aimed to measure the frequency of impaired glucose tolerance and diabetes mellitus in 127 subjects having fasting blood glucose < 7.0 mmol/L and to measure the agreement between different standard diagnostic criteria. Subjects presenting to a laboratory for analysis of fasting blood glucose for excluding diabetes mellitus underwent a 2-hour 75 g oral glucose challenge. A total of 40.6% of subjects with fasting blood glucose from 5.6–6.0 mmol/L had abnormal glucose regulation on the basis of the gold standard glucose challenge. Agreement between American Diabetes Association and World Health Organization diagnostic criteria was only fair (kappa = 0.32). Abnormalities of glucose metabolism including impaired glucose tolerance and diabetes mellitus can exist at fasting blood glucose results < 6.1 mmol/L (110 mg/dL).

Fréquence de la diminution de la tolérance au glucose et du diabète chez des sujets ayant une glycémie à jeun inférieure à 6,1 mmol/l (110 mg/dl)

RÉSUMÉ Le diagnostic du diabète selon les critères disponibles est controversé et repose principalement sur les résultats de la glycémie à jeun. La présente étude transversale menée en 2010 et 2011 visait à mesurer la fréquence de la diminution de la tolérance au glucose et du diabète chez 127 sujets présentant une glycémie à jeun inférieure à 7,0 mmol/l et à mesurer la concordance entre différents critères diagnostiques standard. Les sujets se présentant au laboratoire pour une analyse de la glycémie à jeun visant à éliminer un diagnostic de diabète ont passé une épreuve d'hyperglycémie provoquée par voie orale deux heures après l'ingestion de 75 g de glucose. Au total, 40,6 % des sujets ayant une glycémie à jeun entre 5,6–6,0 mmol/l présentaient une régulation anormale de la glycémie selon les critères de référence de l'épreuve d'hyperglycémie. La concordance entre les critères diagnostiques de l'American Diabetes Association et de l'Organisation mondiale de la Santé était seulement passable (kappa = 0,32). Les anomalies du métabolisme du glucose telles que la diminution de la tolérance au glucose et le diabète sont possibles avec des résultats de la glycémie à jeun inférieurs à 6,1 mmol/l (110 mg/dl).

1Department of Pathology, PNS Rahat Hospital, Karachi, Pakistan (Correspondence to S.H. Khan: [email protected]). 2Department of Pathology, PNS Shifa Hospital, Karachi, Pakistan. Received: 01/11/11; accepted: 05/02/12

175

Book 19-2.indb 175 2/21/2013 11:39:26 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction Finally, older subjects may be normo- Subjects were grouped into 5 glycaemic, but can become glucose groups according to their fasting blood Diabetes mellitus (DM) is one of the intolerant on exposure to a glucose load glucose results: < 5.0 mmol/L (< 90 leading global causes of mortality and is [12]. mg/dL); 5.0–5.5 mmol/L (90–99 a growing epidemic [1]. While the caus- This study in Karachi, Pakistan was mg/dL); 5.6–6.0 mmol/L (100–108 es of the disease, i.e. sedentary lifestyles, therefore designed to measure the fre- mg/dL); 6.1–6.9 mmol/L (110–125 higher intake of refined carbohydrates quency of impaired glucose tolerance mg/dL); or > 6.9 mmol/L (125 mg/ and saturated fatty acids, synergize to (IGT) and DM in subjects having fast- dL). add to the disease burden, the situation ing blood glucose < 7.0 mmol/L and Subjects were defined from 2-hour is aggravated by the unavailability of to measure the agreement between the OGTT results as: normal glucose toler- any curative treatments [2]. Added to ADA and the WHO diagnostic criteria ance < 7.8 mmol/L (< 140 mg/dL); this, delays in diagnosis make subjects for DM. IGT 7.8–11.0 mmol/L (140–199 mg/ vulnerable to various complications of dL); or DM > 11.0 mmol/L (> 199 the disease [3]. mg/dL). The diagnosis of DM relies on the Methods demonstration of raised glucose levels Statistical analysis Study setting and sample in the patient’s blood. However, this All data were entered into SPSS, is more complex than it seems due to This cross-sectional study was carried version 15. Measures of central ten- existence of multiple diagnostic criteria. out at the department of pathology at dency and dispersion i.e., mean and Traditionally an oral glucose tolerance PNS Rahat and Shifa hospitals, Karachi, standard deviation (SD) was calcu- test (OGTT) was used to diagnose DM from March 2010 to June 2011. lated for age. Frequencies by sex were [4]. The latest World Health Organiza- All subjects who presented to the also calculated. Frequency of normal tion (WHO) criteria, however, recom- laboratory for analysis of fasting blood glucose tolerance, IGT and DM based mended a 2-step strategy, i.e. an initial glucose for excluding DM were eligible upon the 2-hour OGTT readings were fasting glucose test and, if levels are be- for the study. Subjects were initially measured at various levels of fasting tween 6.1–6.9 mmol/L, followed by a interviewed for clinical details (e.g. his- blood glucose defined cut-offs by 2-hour OGTT [5]. The 2003 American tory of hypertension, DM and ischae- utilizing descriptive statistics. The dif- Diabetic Association (ADA) guidelines mic heart disease). Subjects who gave ferences in 2-hour post-75-g glucose proposed a fasting glucose result with a history of intake of medicines, were challenge based diagnosis, i.e. normal, an upper cut-off of 5.6 mmol/L [6]. known diabetics, were not observing IGT and DM across various levels Moreover, the published data dealing proper medical fasting, were inpatients, (defined groups), were compared by with the labelling of hyperglycaemia were pregnant or had some physical or 1-way ANOVA. Level of agreement suffer from the problems of different mental stress were excluded from the between WHO and ADA criteria was diagnostic cut-offs and prevalence rates study. measured by the kappa statistic. for DM [7,8]. Apart from the lack of consensus Data collection regarding the diagnosis of DM, other After explanation of the study prereq- Results controversies exist. First, clinical prac- uisites and obtaining a signed written tice suggests that subjects having fasting consent, 172 subjects were sampled The mean age of our study population glucose levels < 6.1 mmol/L or even < for fasting blood glucose. After the was 43.7 (SD 10.8) years. Out of the 5.6 mmol/L may receive a diagnosis blood samples were taken subjects evaluated subjects, 139 were male and of DM in an OGTT and vice versa [9]. were requested to undergo a complete 33 were females. Secondly, the evidence also suggests a 2-hour 75-g glucose challenge. They Participants with IGT and DM a definite lag in the diagnosis of DM, were asked to drink the glucose solution based on the 2-hour OGTT results resulting in subjects being at risk of vari- (300 mL) over 5 minutes. Starting from spanned all the ranges of fasting blood ous DM-related complications from the time of intake of glucose, subjects glucose results (Figure 1). An increas- the outset [10]. Thirdly, the literature were advised to return for the 2-hour ing frequency of diagnosis of IGT and has highlighted the risks associated with sample for glucose. All samples were DM was seen after the fasting blood a decision based on a single reading analysed within 3 (± 1) hours. Glucose glucose results crossed the level of 5.6 of fasting blood glucose on account of was analysed by the hexokinase method mmol/L. patient non-compliance with fasting or on a random access clinical chemistry Table 1 shows that the mean 2-hour laboratory-related inaccuracies [11]. analyser (Hitachi-902). OGTT results crossed the WHO and

176

Book 19-2.indb 176 2/21/2013 11:39:26 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

41 40 Normal study (2 h OGTT result <7.8 mmol/L, i.e. 140 mg/dL)

30 Impaired glucose tolerance (2 h OGTT result 7.8-11.0 mmol/L, i.e. 24 140-199 mg/dL) 20 18 18 20 No. Diabetes mellitus (2 h OGTT result >11.0 mmol/L, i.e. 199 mg/dL) 12 13 10 8 6 5 3 0 1 1 1 1 Normal glucose Normal glucose Fasting blood Impaired fasting Newly diagnosed tolerance (<5.0 tolerance (5.0-5.5 glucose levels from glucose (6.1-6.9 diabetes mellius> mmol/L/<90 mmol/L/90-99 (5.6- 6.0mmol/L/ mmol/L/110-125 (6.9 mmol/L/125 mg/dL) mg/dL) (100-108 mg/dL) mg/dL) mg/dl)

Range of fasting blood glucose

Figure 1 Frequency distribution of normal glucose tolerance, impaired glucose tolerance and diabetes mellitus based on the 2-hour oral glucose oral glucose tolerance test results across fasting blood glucose groups (n = 172 patients)

ADA defined cut-off of 7.8 mmol/L in Discussion miss as many as 40% of cases with ab- subjects having fasting blood glucose normal glucose metabolism as depicted results > 5.6 mmol/L. Our study has shown that subjects by a 2-hour OGTT reading, i.e. IGT and The relative percentage of diagno- having fasting blood glucose < 6.1 DM. Similar results have been shown by sis rates of IGT and DM based upon mmol/L can have biochemically ab- studies in which subjects with normal 2-hour OGTT increased with age (Fig- normal response in terms of glucose glucose levels and abnormal 2-hour ure 2). intolerance after subjecting them to a glucose load readings [14]. Richard et A total of 40.6% of subjects (29.7% glucose load. A review of the literature al. showed no reliable cut-off for fasting IGT and 10.9% DM) with fasting blood revealed other studies that agree with glucose was sensitive enough to rule out glucose between 5.6–6.0 mmol/L our findings [11,13]. We also showed DM, and concluded that OGTT testing had abnormal glucose regulation on that even subjects in the normal range must be maintained for the diagnosis of the basis of the gold standard 2-hour of fasting glycaemia, i.e. fasting blood DM [8]. OGTT results (Figure 3). The agree- glucose < 5.6 mmol/L, can have glucose The possible reasons for such find- ment between the ADA and the WHO dysregulation demonstrated during a ings could be multifold. First, fasting diagnostic criteria was only fair (kappa = 2-hour glucose challenge test. Therefore and post-glucose challenge results 0.32, P < 0.001). the current diagnostic approach may demonstrate 2 different processes; the

Table 1 Differences of mean 2-hour oral glucose tolerance test (OGTT) blood glucose results across fasting blood glucose groups Fasting blood glucose level (mmol/L) No. of patients 2-hour OGTT blood glucose level (mmol/L) Mean (95% CI) < 5.0 (normal glucose tolerance) 22 6.7 (6.0–7.5) 5.0–5.5 (normal glucose tolerance) 31 7.2 (6.6–7.8) 5.6–6.0 (fasting blood glucose) 67 8.0 (7.6–8.5) 6.1–6.9 (impaired fasting glucose) 30 10.5 (9.4–11.5) > 6.9 (newly diagnosed diabetes mellitus ) 22 17.3 (15.1–19.6)

P < 0.001, 1-way ANOVA. CI = confidence interval.

177

Book 19-2.indb 177 2/21/2013 11:39:26 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

100

90 Diabetes mellitus 80 IGT 70 NGT 60

% 50 40 30 20 10 0 < 31 31–44 45–59 < 59 Age group (years)

Figure 2 Percentage distribution of normal glucose tolerance (NGT), impaired glucose tolerance (IGT) and diabetes mellitus based on the 2-hour oral glucose tolerance test results across various age groups (n = 172 patients )

former may reflect the glycaemic base- [18]. Our results suggest that sole re- of having DM? The quality and cost line being influenced by the amount of liance on fasting blood glucose will added by early diagnosis of DM can be fasting and associated physical and psy- compromise sensitivity due to the high weighed against the cost of late initia- chological state, while the later reflects number of false negative cases. Once tion of treatment. Already the literature a stimulated response from the beta a 2-hour OGTT reading is used, it not recommends the use of OGTT in the cells of pancreas after glucose loading only adds strength to the diagnosis but detection of undiagnosed DM in cer- [15]. Traditionally, stimulated or dy- also adds to the sensitivity by identify- tain disease categories such as stroke namic glucose testing has been referred ing subjects with glucose intolerance and myocardial infarction [21,22]. as the gold standard. So both glucose who have underlying atherosclerosis- Philips et al. have recommended a results represent different dimensions related complications [19]. Thirdly, glucose challenge test to improve the to the diabetogenic processes [15– 2-hour readings have been shown to diagnostic sensitivity for undiagnosed 17]. Secondly, the trade-off between be better correlated with markers of DM [23]. The question arises as to sensitivity and specificity of the test glycation including glycosylated hae- why wait for the development of com- is the underlying consideration once moglobin and fructosamine [20]. So plications whose treatment cost greatly decisions are required for screening or why not incorporate a single 2-hour exceeds the screening cost? Finally, a devising a confirmatory methodology reading to screen subjects suspected review of certain regional literature on

10.9 Normal glucose tolerance (2 h results < 7.8 mmol/L, i.e. 140 mg/dL)

Impaired glucose tolerance (2 h results 29.7 between 7.8–11.0 mmol/L, i.e. 140–199 59.4 mg/dL)

Diabetes mellitus (2 h results > 11.0 mmol/L, i.e. 199 mg/dL)

Figure 3 Percentage distribution of normal glucose tolerance, impaired glucose tolerance and diabetes mellitus based on the 2-hour oral glucose tolerance test results in subjects having fasting blood glucose between 5.6–6.0 mmol/L (n = 67 patients )

178

Book 19-2.indb 178 2/21/2013 11:39:27 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

diabetes screening highlights the vary- be an argument for further lowering for DM. At present clinical practice ing recommendations about screening the cut-offs for our population, but suggests labelling of individuals as in our population in contrast to their prospective clinical trials must address normoglycaemic based on a fasting non-Asian counterparts [12,24]. More this issue [18]. blood glucose result. However, in the regional literature do exist recom- We have also observed that ageing is light of our findings it can be suggested mending a lower threshold of screen- related to post-glucose load intolerance. that a patient can have DM when the ing for DM in our population [24–26]. Choi et al. reported similar findings [12]. fasting blood glucose result is normal. Thus racial and regional differences The reason could be the age-related Interpretation of fasting blood glucose in the pattern of diabetes could be an- deterioration in the functioning of beta results must include consideration of a other reason for subjects demonstrat- cells of the pancreas or the development patient’s age and clinical information. ing glucose intolerance during glucose of insulin resistance which appears later Moreover, a 2-hour OGTT reading loading. in life [12,28]. should also be considered as a screen- Comparing WHO criteria for di- There are some limitations to our ing test after a necessary cost–benefit agnosis of diabetes with ADA criteria study which should be noted. First, analysis and consensus among the au- shows minimal agreement between stress hyperglycaemia is a known en- thorities. the 2 definitions of similar disease. The tity. Attempts were made to exclude available literature review does sug- subjects with any degree of physical gest comparable results [7,8,27]. In ailments, but a psychological stress Conclusion the opinion of the authors, the lower evaluation was not carried out among cut-offs may be more compatible with our subjects. Secondly, the results may Abnormalities of glucose metabolism the screening concept for diagnosing have been confounded by the presence including impaired glucose tolerance DM, while subjects having a slightly of obesity, hypertension or ischaemic and DM can exist at fasting blood glu- higher degree of clinical suspicion heart disease. cose results < 6.1 mmol/L (110 mg/ may be further confirmed by adopting This study was clinically important dL). WHO and ADA criteria showed the 2-hour OGTT test as per WHO because it highlighted controversies poor agreement between each other for recommendations. There may also associated with the diagnostic criteria the diagnosis of DM.

References

1. Farag YM, Gaballa MR. Diabesity: an overview of a ris- 10. Haffner SM et al. Cardiovascular risk factors in confirmed pre- ing epidemic. Nephrology, Dialysis, Transplantation, 2011, diabetic individuals. Does the clock for coronary heart disease 26:28–35. start ticking before the onset of clinical diabetes? Journal of the 2. Chang K. Comorbidities, quality of life and patients' willing- American Medical Association, 1990, 263:2893–2898. ness to pay for a cure for type 2 diabetes in Taiwan. Public 11. Balion CM et al. Reproducibility of impaired glucose toler- Health, 2010, 124:284–294. ance (IGT) and impaired fasting glucose (IFG) classification: a systematic review. Clinical Chemistry and Laboratory Medicine, 3. Baldé NM et al. Frequency of diabetic microangiopathy in 2007, 45:1180–1185. newly diagnosed diabetes mellitus in Conakry: late diagnosis and lack of screening. Dakar Med., 2007, 52:165–170. 12. Choi KM et al. Comparison of ADA and WHO criteria for the diagnosis of diabetes in elderly Koreans. Diabetic Medicine, 4. National Diabetes Data Group. Classification and diagnosis of 2002, 19:853–857. diabetes mellitus and other categories of glucose intolerance. Diabetes, 1979, 28:1039–1057. 13. Kuzuya T et al. Committee of the Japan Diabetes Society on the diagnostic criteria of diabetes mellitus. Report of the Commit- 5. Definition and diagnosis of diabetes mellitus and intermediate tee on the classification and diagnostic criteria of diabetes mel- hyperglycaemia. Report of a WHO/IDF consultation. Geneva, litus. Diabetes Research and Clinical Practice, 2002, 55:65–85. World Health Organization, 2006. 14. De Vegt F et al. Relation of impaired fasting and post load 6. Summary of revisions for the 2010 clinical practice recommen- glucose with incident type 2 diabetes in a Dutch population: dations. Diabetes Care, 2010, 33(Suppl. 1):S3. The Hoorn Study. Journal of the American Medical Association, 7. Herdzik E et al. Comparison of ADA and WHO diagnostic 2001, 285:2109–2113. criteria for diabetes diagnosis and other categories of glucose 15. Li HY et al. The performance of risk scores and hemoglobin A1c intolerance. Polski Merkuriusz Lekarski, 2002, 13:316–320. to find undiagnosed diabetes with isolated post load hypergly- 8. Richard JL et al. Diagnosis of diabetes mellitus and intermedi- cemia. Endocrine Journal, 2011, 58:441–448. ate glucose abnormalities in obese subjects based on ADA 16. Temelkova-Kurktschiev TS, Hanefeld M. Oral glucose toler- (1997) and WHO (1985) criteria. Diabetic Medicine, 2002, ance test: to be or not to be performed? Clinical Laboratory, 19:292–299. 2002, 48:143–152. 9. Tirosh A et al. Normal fasting plasma glucose levels and type 17. Silverman RA et al. Hemoglobin A1c as a screen for previously 2 diabetes in young men. New England Journal of Medicine, undiagnosed prediabetes and diabetes in an acute-care set- 2005, 353:1454–1462. ting. Diabetes Care, 2011, 34(9):1908–1912.

179

Book 19-2.indb 179 2/21/2013 11:39:27 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

18. Waugh N et al. Screening for type 2 diabetes: literature review 24. Noda M et al. Fasting plasma glucose and 5-year incidence of and economic modelling. Health Technology Assessment, 2007, diabetes in the JPHC diabetes study—suggestion for the thresh- 11(17):iii–iv, ix–xi, 1–125. old for impaired fasting glucose among Japanese. Endocrine 19. Hanefeld M et al. Post-challenge hyperglycaemia relates more Journal, 2010, 57:629–637. strongly than fasting hyperglycaemia with carotid intima-media 25. Ryu S et al. Should the lower limit of impaired fasting glucose thickness: the RIAD Study. Risk Factors in Impaired Glucose be reduced from 110 mg/dL in Korea? Metabolism: Clinical and Tolerance for Atherosclerosis and Diabetes. Diabetic Medicine, Experimental, 2006, 55:489–493. 2000, 17:835–840. 26. Nang EE et al. Is there a clear threshold for fasting plasma 20. Rosediani M, Azidah AK, Mafauzy M. Correlation between glucose that differentiates between those with and without fasting plasma glucose, post prandial glucose and glycated neuropathy and chronic kidney disease? The Singapore Pro- haemoglobin and fructosamine. Medical Journal of Malaysia, spective Study Program. American Journal of Epidemiology, 2006, 61:67–71. 2009, 169:1454–1462. 21. Lindsberg PJ, Tuomi T, Kaste M. Oral glucose tolerance test 27. Park KS et al. Comparison of glucose tolerance categories in should be performed after stroke and transient ischemic at- the Korean population according to World Health Organiza- tack. International Journal of Stroke, 2011, 6:317–320. tion and American Diabetes Association diagnostic criteria. 22. Kitada S et al. Post-load hyperglycemia as an important predic- Korean Journal of Internal Medicine, 2000, 15:37–41. tor of long-term adverse cardiac events after acute myocardial 28. Hasegawa G. Decreased senescence marker protein-30 could infarction: a scientific study.Cardiovascular Diabetology, 2010, be a factor that contributes to the worsening of glucose toler- 9:75. ance in normal aging. Islets, 2010, 2:258–260. 23. Phillips LS et al. Glucose challenge test screening for prediabetes and undiagnosed diabetes. Diabetologia, 2009, 52:1798–1807.

Note from the Editor

We wish to draw the kind attention of our potential authors to the importance of applying the editorial requirements of the EMHJ when preparing their manuscripts for submission for publication. These provisions can be seen in the Guidelines for Authors, which are published at the end of every issue of the Journal. We regret that we are unable to accept papers that do not conform to the editorial requirements.

180

Book 19-2.indb 180 2/21/2013 11:39:27 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Interleukins 12 and 13 levels among beta- thalassaemia major patients R.A. Hashad,1 N.A. Hamed,2 M.M. El Gharabawy,1 H.A. El Metwally 1 and M.G. Morsi 1

مستويات اإلنرتلوكينات 12 و 13لدى مرىض الثالسيميا-بيتا الكربى روال أمحد حشاد، هنلة عبد املنعم حامد، مها منري الغرباوي، هالة عبد الرؤوف املتويل، منى مجال الدين مريس اخلالصـة: اليزال الغموض يشوب دور مركبات السيتوكني االلتهابية يف الفيزيولوجيا املرضية للثالسيميا-بيتا. وقد قاست الباحثات يف هذه الدراسة مستويات اإلنرتلوكينات 12 و13 مستخدمات املقايسة التجارية للممتز املرتبط باإلنزيم )إليزا( يف السوائل الطافية عىل ٍمزرعة خلاليا الدم املحيطي 30 الوحيدة النواة، َّهةواملنب بامدة ّحمرضة لالنقسام الفتييل، مستمدة من حالة ثالسيميا-بيتا مل جير هلا استئصال للطحال، ويزيد لدهيا عبء احلديد، مع 12 20 حالة ُتماثل مع تلك احلاالت من حيث العمر واجلنس، وتتمتع بالصحة. ووجدت الباحثات أن مستويات اإلنرتلوكني أخفض بمقدار ُي ْع َت ُّد به ًإحصائيا لدى احلاالت )91.4 بيكوغرام/مييل لرت( مما هي عليه لدى الشواهد )154.6 بيكوغرام/مييل لرت(، وأن مستويات اإلنرتلوكني 13 كانت أعىل 5.7 42.5 بمقدار ُي ْع َت ُّد به ًإحصائيا لدى احلاالت ) بيكوغرام/مييل لرت( مما هي عليه لدى الشواهد ) بيكو غرام/مييل لرت(. كام وجدت الباحثات ًترابطا r = –0.42 13 12 ًسلبيا ُي ْع َت ُّد به ًإحصائيا بني مستويات اإلنرتلوكني و لدى حاالت الثالسيميا-بيتا الكربى ) (. وكان لدى املرىض بالثالسيميا-بيتا لوحدها مستويات أعىل من اإلنرتلوكني 12 )140بيكو غرام/مييل لرت( مما لدى املرىض بالثالسيميا-بيتا اإلجيابيني ًمصليا للعدوى املزمنة بفريوس االلتهاب الكبدي "يب" و"يس" )50 بيكوغرام/مييل لرت(، أما مستويات اإلنرتلوكني 13 فقد كانت أخفض بقليل لدى مرىض الثالسيميا-بيتا الكربى لوحدها )65 بيكوغرام/مييل لرت( مما لدى مرىض الثالسيميا-بيتا الكربى اإلجيابيني ًمصليا للعدوى املزمنة بفريوس االلتهاب الكبدي "يب" و"يس" )67 بيكوغرام/مييل لرت(. وترى الباحثات أن اختالل التوازن يف حمور اإلنرتلوكينات 12 و13 قد يكون له دور يف الفيزيولوجيا املرضية للثالسيميا-بيتا.

ABSTRACT The role of inflammatory cytokines in the pathophysiology of beta-thalassaemia is still unclear. In this study production levels of interleukins (IL)-12 and IL-13 were measured by commercial ELISA in culture supernatants of mitogen-stimulated peripheral blood mononuclear cells from 30 non-splenectomized beta- thalassaemia cases with iron overload and 20 age- and sex-matched healthy individuals. IL-12 levels were significantly lower among cases compared with controls (91.4 pg/mL versus 154.6 pg/mL), while IL-13 levels were significantly higher (42.5 pg/mL versus 5.7 pg/mL). There was a significant negative correlation between IL- 12 and IL-13 levels among beta-thalassaemia cases (r = –0.42). Patients with beta-thalassaemia alone had higher IL-12 levels than beta-thalassaemia patients who were seropositive for chronic hepatitis B or C virus infection (140 pg/mL versus 50 pg/mL); IL-13 levels were slightly lower (65 pg/mL versus 67 pg/mL). An imbalance in the IL-12/IL-13 axis may be relevant to the pathophysiology of beta-thalassaemia.

Taux des interleukines 12 et 13 chez des patients atteints de bêta-thalassémie majeure

RÉSUMÉ Le rôle des cytokines inflammatoires dans la physiopathologie de la bêta-thalassémie reste à élucider. Dans la présente étude, les taux de production des interleukines 12 et 13 (IL-12 et IL-13) ont été mesurés à l'aide de la méthode ELISA dans les surnageants de culture de cellules mononucléées de sang périphérique stimulées par des mitogènes chez 30 cas de bêta-thalassémie non splénectomisés présentant une surcharge martiale ainsi que chez 20 individus en bonne santé appariés pour l'âge et le sexe. Les taux d'IL-12 étaient nettement inférieurs chez les malades par rapport aux témoins (91,4 pg/ml contre 154,6 pg/ml), alors que les taux d'IL-13 étaient significativement plus élevés (42,5 pg/ml contre 5,7 pg/ml). Une corrélation négative importante existait entre les taux d'IL-12 et d'IL-13 chez les patients atteints de beta-thalassémie (r = –0,42). Les patients uniquement porteurs d'une bêta-thalassémie avaient des taux d'IL-12 plus élevés que les patients atteints d'une bêta-thalassémie et également chroniquement infectés par le virus de l'hépatite B ou C (140 pg/ml contre 50 pg/ml) ; les taux de l'IL-13 étaient légèrement inférieurs (65 pg/ml contre 67 pg/ml). Un défaut de l'axe IL-12/IL-13 peut être révélateur de la physiopathologie de la bêta-thalassémie.

1Department of Medical Microbiology and Immunology; 2Haematology Unit; Department of Internal Medicine, Faculty of Medicine, University of Alexandria, Alexandria, Egypt (Correspondence to M.G. Morsi: [email protected]). Received: 31/10/11; accepted: 09/02/12

181

Book 19-2.indb 181 2/21/2013 11:39:27 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Introduction and IL-13 may also be relevant in the Incubation of cultures was performed pathophysiology of beta-thalassaemia. at 37 °C in a humidified atmosphere

Thalassaemias are a worldwide prob- To evaluate this hypothesis, IL-12 and of 5% CO2. After 2 days of culture, su- lem [1]. Beta-thalassaemia is the most IL-13 concentrations were investigated pernates were collected from each tube common type and has the highest in beta-thalassaemia patients and con- and stored at –20 °C to be assayed using frequency in the Mediterranean area. trols and in relation to their HBV and commercial ELISA kits (RayBio) were Heterozygotes have thalassaemia mi- HCV status. followed according to manufacturers’ nor, a condition in which there is usu- instructions [7,8]. ally mild anaemia and little or no clinical HBsAg and HCV Abs were de- disability. Homozygotes (thalassaemia Methods tected by ELISA HBsAg test system 2 major) are either unable to synthesize Sample and HCV version 3.0 ELISA test system haemoglobin A or at best produce very (Ortho-Clinical Diagnostics) [9,10]. little and, after the first 4 months of This case–control study was done on life, develop a profound hypochromic 30 patients [18 males and 12 females, Data analysis anaemia [2]. Blood transfusions may be mean age 21.6 (SD 9.3) years] admit- SPSS, version 15.0 for Windows was required every 4–6 weeks [1]. However, ted to the department of haematology used for analysis of the data. The 5% one of the most common side-effects at Alexandria Main University hospital, level of significance was taken to test of blood transfusion is infection of the Egypt, during the year 2010 with a final the significance of the results obtained; recipient with hepatitis B virus (HBV) diagnosis of beta-thalassaemia major. P ≤ 0.05 was considered significant. and/or C virus (HCV). HCV is the The inclusion criteria were all cases Data were shown as mean and standard most common transfusion-transmitted who were non-splenectomized, HIV deviation (SD). infection [2]. negative and having iron overload. Several immunological defects can They were diagnosed from complete be found in patients with beta-thalas- blood count, haemoglobin electro- Results saemia, among which the impairment phoresis and high serum ferritin, with of neutrophils and macrophage phago- bone marrow confirmation. Age- and Table 1 shows the IL-12 and IL-13 lev- cytic and killing functions and the sex-matched controls were 20 healthy els among both studied groups. In the production of some cytokines are the volunteers [8 males and 12 females, control group IL-12 levels ranged from most important. The plasma levels of mean age 21.3 (SD 11.9)] from among 46.2–243.9 (pg/mL), with a mean of these cytokines may be relevant in the hospital staff (doctors, technicians and 154.6 (SD 61.1) pg/mL, while patients’ pathophysiology of beta-thalassaemia laboratory workers) who were HBV, levels ranged from 40.4–135.4 pg/mL [3]. Interleukin (IL)-12 and IL-13 are 2 HCV and HIV negative. Informed with a mean of 91.9 (SD 28.5) pg/ recently characterized cytokines which consent was taken from all subjects mL, a highly significant difference (P play an important role in the induction participating in the study. All those en- < 0.001). IL-13 levels among controls of T-helper cell type 1 (Th1-) and Th2- rolled were subjected to the following: ranged from 2.7–8.5 pg/mL, with a like cells, respectively [4]. IL-12 has history taking, full clinical assessment, mean of 5.7 (SD 1.8) pg/mL, while the potential for use either as a single complete blood count and liver and patients’ values ranged from 9.9–65.0 immunotherapeutic agent, in combi- renal function tests. pg/mL with a mean of 42.5 (SD 18.0) nation with other chemotherapeutic pg/mL, also a highly significant differ- Data collection agents, or as a vaccine adjuvant [5]. IL- ence (P < 0.001). 13 is a cytokine secreted by many cell Estimation of IL-12 and IL-13 levels There were 4 beta-thalassaemic types especially Th2 cells. It displays was done by isolation of peripheral patients who were seropositive for homology with IL-4 and shares some blood mononuclear cells (PBMCs) HBV and HCV infection. There was no of its biological functions, especially from heparinized venous blood by correlation between IL-12 and IL-13 with regards to changes induced on Ficoll–Hypaque density layer centrifu- among controls, but there was a signifi- haematopoietic cells. It acts as an im- gation (Sigma) and cultured at 2 × 105 cant negative correlation between IL-12 munoregulatory and effector cytokine cells per 500 µL Rosell Park Memo- and IL-13 levels, i.e. a decrease in IL-12 [6]. Several studies have shown that rial Institute (RPMI) 1640 medium with increasing IL-13, among cases of an imbalance in the IL-3/IL-7 and supplemented with antibiotics and 5% beta-thalassaemia (n = 26) (r = –0.42, IL-6/IL-8 axes may contribute to the fetal calf serum (FCS). For stimula- P = 0.003). Figure 1a and 1b shows the development of beta-thalassaemia mation, 5 μg/mL phytohaemagglutinin IL-12 and IL-13 levels of controls and jor [3,4]. The plasma levels of IL-12 mitogen (PHA, Wellcome) was used. patients with thalassaemia major and

182

Book 19-2.indb 182 2/21/2013 11:39:27 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Table 1 Comparison of mean interleukin (IL)-12 and IL-13 levels among patients with beta-thalassaemia and controls Variable Controls Cases t-test P-value (n = 20) (n = 30) Mean (SD) Range Mean (SD) Range IL-12 level (pg/mL) 154.6 (61.1) 46.2–243.9 91.9 (28.5) 40.4–135.4 14.7 < 0.001 IL-13 level (pg/mL) 5.7 (1.8) 2.1–8.5 42.5 (18.0) 9.9–65.0 21.5 < 0.001

SD = standard deviation.

those co-infected with HBV or HCV. [50 (SD 5.5) pg/mL]. In contrast, IL- Discussion Patients with beta-thalassaemic alone 13 levels were higher among chronic had higher IL-12 levels [140 (SD 10.7) viral hepatitis[(67 (SD 3.6) pg/mL] Recent evidence suggests that levels pg/mL] compared with those co- cases compared with beta-thalassaemic of IL-8, tumour necrosis factor and infected with chronic HBV and HCV patients [65 (5.7) pg/mL]. a soluble receptor of IL-2 (sCD25)

300 )

L 250 m / g ( p

n 200 o i t r a t n e

c 150 n o c

2 - 1 L I

100 m r u S e 50

0 Control β-thalasseamia With HB With HCV

70 )

L 60 m / g ( p 50 s n o i t r a

t 40 n e c n o

c 30 3 - 1 L I

m 20 r u S e 10

0 Control β-thalasseamia With HB With HCV

Figure 1 Comparison of interleukin (IL) levels in controls (n = 20), patients with beta-thalassaemia alone (n = 26) and patients

with beta-thalassaemia plus hepatitis B virus (HBV) (n = 4) or hepatitis C virus (HCV) infection (n = 4)

183

Book 19-2.indb 183 2/21/2013 11:39:28 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

and imbalances in the IL-3/IL-7 and Substantial evidence suggests that hepatitis B or C patients (50 pg/mL). IL-6/IL-8 axes may have biological an imbalance in production of these 2 On the other hand, serum IL-13 con- and clinical importance in thalassaemic critical immunoregulatory cytokines centrations in patients with chronic patients [11]. Our data demonstrates (i.e. overproduction of IL-13 and im- viral hepatitis (67 pg/mL) were slightly that IL-12 production was significantly pairment of IL-12 production) may be higher compared with those of thalas- suppressed in thalassaemic patients associated with cellular immunological saemia patients (65 pg/mL).These (91.9 pg/mL) compared with controls alterations in beta- thalassaemia. Al- findings clearly agree with the results (154.6 pg/mL), indicating the involve- though it is not entirely clear what tips of Uguccioni et al. [15]. Similar val- ment of this cytokine in the suppres- the balance between these cytokines, ues were also reported by Di Marco sion of erythropoiesis in thalassaemic the potential role of IL-12 and IL-13 et al. [16]. Several other studies have patients, similar to the involvement of and the interaction between them in investigated the role of cytokines and IL-3 in erythropoiesis of thalassaemic thalassaemic patients require further interleukins in the pathophysiology of patients [12] and it suggests that IL-12 investigation. These findings clearly beta-thalassaemia [17–22]. is a cytokine involved in the cellular agree with the results of Willskarp who Our data and those from the lit- immunological alterations in beta-tha- suggested that an imbalance in the erature strongly support the hypothesis lassaemia. Meanwhile patients with production of IL-12 and IL-13 occurs that beta-thalassaemia major is associat- beta-thalassaemia were found to have in the lung of atopic and asthmatic ed with a downregulation of IL-12 and significantly higher IL-13 concentra- individuals [14]. In our study a signifi- an upregulation of IL-13 which leads to tions (42.5 pg/mL) than normal con- cant negative correlation was found an imbalance in IL-12/IL-13 axis which trols (5.7 pg/mL). Our data suggests between IL-12 and IL-13 levels (r = may be relevant in the pathophysiology that there could be an intrinsic cause –0.42). Meanwhile, there was no cor- of beta-thalassaemia. HBV and HCV, for the IL-13 increase in thalassaemia relation between these 2 inflammatory probably resulting from repeated blood major. This increase downregulates cytokines in normal controls. From transfusion, may aggravate the condi- macrophage activity thereby inhibit- these observations we can conclude tion. These cytokines deserve more in- ing production of proinflammatory that an imbalance in the IL-12/IL-13 vestigation on larger samples as well as cytokines and chemokines [13]. IL-13 axis may be relevant in the pathophysi- in vivo trials in experimental animals, as needs to be studied in detail in order ology of beta-thalassaemia. they may be a target for immunotherapy to learn more about the involvement Beta-thalassaemic patients had to clear hepatitis virus infection (result- of this cytokine in erythropoiesis of higher serum IL-12 levels (140 pg/ ing from repeated transfusion) and raise thalassaemic patients. mL) compared with those of chronic patients’ immunity.

References

1. Edwards C, Bouchier I. Davidson’s principles and practice of 9. Valentine-Thon E et al. European proficiency testing program medicine, 16th ed. Edinburgh, Churchill Livingstone 1991. for molecular detection and quantitation of hepatitis B virus 2. Shang G et al. Residual risk of transfusion-transmitted viral in- DNA. Journal of Clinical Microbiology, 2001, 39:4407–4412. fections in Shenzhen, China, 2001 through 2004. Transfusion, 10. Tobler LH et al. Impact of HCV 3.0 EIA relative to HCV 2.0 EIA 2007, 47:529–539. on blood-donor screening. Transfusion, 2003, 43:1452–1459. 3. Oztürk O et al. Increased plasma levels of interleukin-6 and 11. Rund D, Rachmilewitz E. Beta-thalassemia. New England Jour- interleukin-8 in beta-thalassaemia major. Haematologia, 2001, nal of Medicine, 2005, 353:1135–1146. 31:237–244. 12. Kutukculer N et al. Plasma interleukin-3 (IL-3) and IL-7 con- 4. Kutukculer N et al. Plasma interleukin-3 (IL-3) and IL-7 concentrations in children with homozygous beta-thalassemia. centrations in children with homozygous beta-thalassemia. Journal of Tropical Pediatrics, 1997, 43:366–367. Journal of Tropical Pediatrics, 1997, 43:366–367. 13. Sakamoto O et al. Interleukin-13 selectively suppresses the 5. Gately MK, Mulqueen MJ. Interleukin-12: potential clinical growth of human macrophage progenitors at the late stage. application in the treatment and prevention of infectious dis- Blood, 1995, 85:3487–3493. eases. Drugs, 1996, 52(Suppl. 2):18–25. 14. Willskarp M. IL-12/IL-13 axis in allergic asthma. Journal of Al- 6. Wynn TA. IL-13 effector functions. Annual Review of Immunol- lergy and Clinical Immunology, 2001, 107(7):9–18. ogy, 2003, 21:425–456. 15. Uguccioni M et al. Elevated interleukin-8 serum concentra- 7. O’Neill LAJ, Bowie A eds. Interleukin protocols (methods in mo- tions in beta-thalassemia and graft-versus-host disease. Blood, lecular medicine). Totawa, New Jersey, Humana Press, 2001. 1993, 81:2252–2256. 8. Morsi MG, Atta HY, Rabie AK. Comparative study of different 16. Di Marco V et al. Alpha interferon treatment of chronic hepa- test systems for measuring cytokine production among tuber- titis C in beta-thalassaemia. Gut, 1993, 34(2 Suppl.):S142–S143. culous andhealthy persons. Egyptian Journal of Immunology, 17. Lombardi G et al. Serum levels of cytokines and soluble 1998, 5:153–162. antigens in polytransfused patients with beta-thalassemia

184

Book 19-2.indb 184 2/21/2013 11:39:28 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

major: relationship to immune status. Haematologica, 1994, 21. Di Marco V et al. Liver disease in chelated transfusion depend- 79:406–412. ent thalassemics: the role of iron overload and chronic hepati- 18. Dore F et al. Serum interleukin-8 levels in thalassemia interme- tis C. Haematologica, 2008, 93:1243–1246. dia. Haematologica, 1995, 80:431–433. 22. Angelucci E et al. Italian Society of Hematology practice guide- 19. Kutukculer N et al. Plasma interleukin-3 (IL-3) and IL-7 con- lines for the management of iron overload in thalassemia major centrations in children with homozygous beta-thalassemia. and related disorders. Haematologica, 2008, 93:741–752. Journal of Tropical Pediatrics, 1997, 43:366–367. 20. Oztürk O et al. Increased plasma levels of interleukin-6 and interleukin-8 in beta-thalassaemia major. Haematologia, 2001, 31:237–244.

Community genetics services: report of a WHO consultation on community genetics in low- and middle- income countries

The objective of the above-mentioned Consultation was to develop an evidence-based report on community genetics services to provide guidance to low- and middle-income countries (LMIC) in accordance with the 2008–2013 Action plan for the global strategy for the prevention and control of noncommunicable diseases (NCDs). The high rates of congenital disorders in LMIC may result in part from: low availability of public health measures for the care and prevention of these disorders; high frequency of haemoglobinopathies in Africa, the Middle-East and South-East Asia; high rates of consanguineous marriage in the Eastern Mediterranean and South-East Asia regions that can increase the occurrence of recessively inherited diseases; advanced maternal age at conception in many such countries which increases the risk of chromosomal trisomies; and large family size that may increase the number of affected children in families with autosomal recessive conditions. While noting that prevention programmes have been successfully implemented in some LMIC (e.g. Bahrain, Cyprus and Iran), the Group emphasized the need to upgrade community genetics services in LMIC and the need for education in genetics to be provided to all health professionals, policy-makers and the general public. This also included sensitization to ethical, legal and social issues which are of key concern in the context of congenital disorders and genetic diseases.

This document is available at:http://whqlibdoc.who.int/publications/2011/9789241501149_eng.pdf

185

Book 19-2.indb 185 2/21/2013 11:39:28 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Revue Médicaments et allaitement maternel : évaluation du risque médicamenteux chez le nourrisson Y. Khabbal,1 S. Zaoui 2 et Y. Cherrah 3

األدوية واإلرضاع الطبيعي من الثدي: تقييم خطر األدوية عىل َّالرضع يوسف خبال، سناء الزاوي، حيي الرشاح

اخلالصـة: ُي َع ُّداللبن احلليب من العنارص اهلامة يف تنمية األطفال، إضافة إىل منافعه التي ال ُتنْ َكر يف التغذية ويف املناعة. وقد متنع التحذيرات، التي قد يكون ً مبالغافيها، األمهات واألطفال من ُّع التمتبمنافع كل من اإلرضاع الطبيعي من الثدي ومن تناول األدوية. ويعود تربير تلك األساليب إىل الغياب التام للدراسات حول إفراغ األدوية يف حليب الثدي من خالل دراسات رسيرية. وقد َّم قيالباحثون األخطار احلقيقية من خالل تصنيفات عديدة، ومن خالل تقدير عبور األدوية يف احلليب، ومن خالل تقدير مدى ُّتعرض َّالرضع لألدوية، وكلها ممكنة، ويمكنها توفري مؤرش يعتمد عليه للحصول عىل أحكام ْأص َوب حول األخطار التي هتدد َّ ع.وعىل الرضوجه اإلمجال فإننا بمواجهة حالتني خمتلفتني؛ إما معاجلة جديدة خالل فرتة اإلرضاع، أو ُم َو َاص َلة معاجلة ُوصفت من قبل أثناء احلمل. وإذا كان ّثمة حاجة لوصف املعاجلات ملدة طويلة أثناء اإلرضاع الطبيعي من الثدي، فإن من املطلوب وضع أسلوب اسرتاتيجي لذلك.

RÉSUMÉ Le lait maternel, en plus de ses propriétés nutritives et immunologiques indéniables, est un élément important pour le développement du nourrisson. Des mises en garde parfois exagérées empêchent la mère et l’enfant de profiter à la fois des bienfaits de l’allaitement et d’une médication appropriée. Cette attitude est justifiée par l’absence totale d’investigations sur l’excrétion des médicaments dans le lait maternel lors des essais cliniques. Le risque réel a été évalué par plusieurs classifications. Une estimation du passage des médicaments dans le lait et une estimation de l’exposition du nourrisson au médicament sont possibles, constituant ainsi un indice très fiable pour mieux juger le risque encouru par le nourrisson. Généralement, nous sommes confrontés à deux situations différentes : une nouvelle prescription durant l’allaitement ou le maintien d’un traitement déjà prescrit lors de la grossesse. Si le traitement sera prescrit pour une longue période pendant l’allaitement, la nécessité d’une démarche stratégique s’impose.

Medicines and breastfeeding: assessing the risk of medicines to infants

ABSTRACT Milk, in addition to its undeniable nutritional and immunological benefits, is an important element for the development of the infant. Warnings, sometimes exaggerated, may prevent the mother and child from enjoying the benefits of both breastfeeding and appropriate medication. This approach is justified by the total absence of investigations on the excretion of drugs in breast milk during clinical trials. The actual risk was evaluated by several classifications; an estimate of the passage of drugs into milk and estimation of infant exposure to the drug are possible and can provide a reliable indicator to better judge the risk to the infant. Generally we are faced with two different situations: a new treatment during lactation or maintenance treatment already prescribed during pregnancy. If treatment needs to be prescribed for a long period of breastfeeding, a strategic approach is required.

1Laboratoire de Pharmacologie et de Toxicologie, Faculté de Médecine et de Pharmacie, Université Sidi Mohammed Ben Abdellah, Fès (Maroc) (Correspondance à adresser à Y. Khabbal : [email protected] ou [email protected]). 2Laboratoire de Pharmacologie et de Toxicologie, Laboratoire de Recherche PCIM (Pneumo-cardio-immunopathologie et Métabolisme), Faculté de Médecine et de Pharmacie, Université Cadi Ayyad, Marrakech (Maroc). 3Laboratoire de Pharmacologie et de Toxicologie, Faculté de Médecine et de Pharmacie, Université Mohammed V, Rabat (Maroc). Reçu : 21/02/12; accepté : 23/02/12

186

Book 19-2.indb 186 2/21/2013 11:39:28 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction l’exposition au médicament pendant les avantages de l'utilisation chez l’allaitement. Citons à titre d’exemple les mères qui allaitent peuvent Soucieux d’éviter les effets indésirables la classification de Briggs, une être acceptables malgré le risque chez le nourrisson, le prescripteur est classification pratique et rapide : cet pour l'enfant (par exemple, si le contraint dans plusieurs situations à auteur classe le risque lié à l’exposition médicament est nécessaire dans une ne pas utiliser chez la femme allaitante au médicament pendant l’allaitement situation potentiellement mortelle certaines molécules ou à indiquer de A à X et décrit la pharmacocinétique ou dans le cas d’une maladie grave le sevrage précoce du nourrisson, de la molécule dans le compartiment pour laquelle des médicaments plus empêchant ainsi la mère et l’enfant lacté [3]. Cette même graduation est sûrs ne peuvent pas être utilisés ou de profiter à la fois des bienfaits de adoptée par Delaloye. sont inefficaces). l’allaitement et de la médication Le manuel de référence reste celui • L5 : contre-indiqué car des études appropriée. de Thomas Hale qui propose une basées sur l'expérience humaine ont Cette attitude est justifiée par classification claire et pratique [4]. Le démontré chez les mères qui allaitent l’absence totale d’investigations sur risque lié à l’exposition au médicament qu'il existe un risque documenté l’excrétion des médicaments dans le lait pendant l’allaitement est classé de L1 important pour le nourrisson, ou il maternel lors des essais cliniques. à L5 : s'agit d'un médicament qui présente un risque élevé de dommages Dans ce travail, nous exposons • L1 : (médicament) le plus sûr, importants chez le nourrisson. la démarche classique pour évaluer - car la molécule a été largement Le risque de l'utilisation du s’il existe bien sûr - le risque lié à utilisée par de nombreuses médicament chez les femmes qui une exposition médicamenteuse femmes allaitantes sans que l’on ait allaitent l’emporte sur tout avantage du nourrisson pendant l’allaitement observé une augmentation d’effets possible de l'allaitement maternel. maternel. indésirables chez le nourrisson. Les Le médicament est contre-indiqué études contrôlées chez les femmes chez les femmes qui allaitent un allaitantes ne démontrent ni risque nourrisson. Classification du risque ni possibilité d’effets néfastes pour le médicamenteux nourrisson, ou la biodisponibilité de la molécule prise par voie orale est En plus de ses propriétés nutritives négligeable. Facteurs modulant et immunologiques indéniables, • L2 : sûr, car les données sont le passage des le lait maternel est un élément limitées chez la femme allaitante. médicaments dans le important pour le développement Cependant, il n’a pas été mis en compartiment lacté du nourrisson [1]. Des mises en évidence d’augmentation des effets garde parfois exagérées empêchent indésirables chez le nourrisson, et/ Le lait et le plasma sont deux la mère et l’enfant de profiter à la fois ou la preuve d'un risque avéré, suite compartiments biologiques des bienfaits de l’allaitement et d’une à l'utilisation de ce médicament indépendants mais il y a assez médication appropriée. Mis à part chez une femme qui allaite, est souvent une corrélation parfaite entre l’hypersensibilité qui survient à des faible. la concentration des médicaments concentrations sub-thérapeutiques, les • L3 : moyennement sûr. Il n'y a pas dans les deux compartiments [4] : effets pharmacologiques à surveiller d'études contrôlées chez les femmes si le taux plasmatique augmente, le pendant l’allaitement (léthargie, qui allaitent mais le risque d'effets taux lacté augmente aussi, et si au somnolence, insomnie, hyperactivité) indésirables chez le nourrisson est contraire le taux plasmatique baisse, surviennent lorsque la concentration possible, ou des études contrôlées le médicament repasse dans le minimale efficace est atteinte. Certains montrent un faible risque d’effets compartiment plasmatique et le taux symptômes, comme la diminution des indésirables non menaçants. La lacté baisse logiquement. Plusieurs réflexes de succion, la perte de poids molécule ne doit être administrée à facteurs influent sur la concentration du nouveau-né, permettront aussi, la femme allaitante que si le bénéfice des médicaments dans le lait mais de manière indirecte, de mettre justifie le risque potentiel chez maternel. en évidence un effet pharmacolo l’enfant. gique [2]. • L4 : potentiellement dangereux. Il La composition du lait Plusieurs classifications sont y a des preuves positives de risque La composition du lait joue un rôle proposées pour évaluer le risque lié à pour le nourrisson allaité, mais prépondérant pour expliquer la

187

Book 19-2.indb 187 2/21/2013 11:39:28 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

valeur des ratios lait/plasma des contribution de la lactoferrine et de la central qui ont généralement une concentrations de médicament. caséine est faible [2]. structure physico-chimique qui les rend La composition du lait maternel très liposolubles [4]. Le degré d’ionisation du en ions hydrogènes, en lipides et en médicament La liaison aux protéines protéines varie principalement en plasmatiques fonction du stade de maturation du La différence de pH entre le lait [5]. En tout début de lactation, plasma et le lait maternel a un effet La liaison des médicaments aux l’épithélium alvéolaire est beaucoup sur la distribution puisque seule la protéines du plasma et du lait influe plus perméable car les connections molécule non ionisée peut traverser sur la distribution dans le lait et sur la intercellulaires ne sont pas jointives ; la une membrane biologique. vitesse avec laquelle ces médicaments concentration lactée des médicaments Le pH du lait est acide (6,6 à passeront au travers de l’épithélium peut donc être plus élevée, mais étant 7,3) comparativement à celui du alvéolaire. Seuls les médicaments non donné le faible volume de colostrum plasma (7,4). Le lait maternel favorise liés diffusent à travers les membranes. secrété et consommé, la dose totale donc l’accumulation des bases faibles Les médicaments fortement liés aux reçue par l’enfant reste probablement alors que la concentration d’acides protéines plasmatiques (> 90 %) ont un faible. La composition du lait évolue faibles est généralement plus basse dans faible passage lacté (anti-inflammatoires au cours du temps. Le contenu moyen le lait que dans le plasma. Les substances non stéroïdiens, paroxétine, propranolol, en lipides du lait augmente durant fortement acides de pKa égal à 2,7 sont warfarine). Cette fixation aux protéines la maturation du lait, passant de plus ionisées dans le plasma et pénètrent plasmatiques dépend beaucoup des 3 % dans le colostrum à 5,4 % dans donc très peu dans le lait, alors que les caractéristiques acido-basiques du le lait mature. Cette concentration bases fortes s’accumulent dans le lait médicament (Tableau 2). maternel. En pratique, comme déjà en graisses varie parallèlement à la Le poids de la molécule vidange du sein et est à l’origine de mentionné, le pH du lait est légèrement variations importantes liées en grande plus acide que celui du plasma. Les Quand le poids moléculaire (PM) partie aux capacités de stockage médicaments qui sont des bases est très faible (< 200 daltons) mammaire. Ces variations touchent faibles (barbituriques, β-bloquants) comme l’éthanol, il y a un passage essentiellement les substances peuvent se trouver piégés dans le par diffusion directe par l’espace liposolubles [5]. compartiment lacté [5] (Tableau 1). intercellulaire ; par contre, les substances dont le PM est supérieur à 800- La concentration moyenne en La liposolubilité protéines est à son plus haut niveau 1000 daltons passent plus difficilement dans le colostrum (18-23 g/L). Elle Les médicaments les plus liposolubles dans le lait ; pour les substances dont le ne varie pas de manière significative diffusent plus facilement dans les PM est très élevé (25 000 à 200 000), le du début à la fin de la tétée. Les lipides. Donc les variations de la passage lacté est pratiquement nul [5]. concentration des lipides du lait influent protéines du petit lait comptent La demi-vie pour 70 % à 80 % des protéines du sur les quantités de médicaments dans lait et incluent l’albumine (0,4 g/L), le lait maternel ; l’augmentation de la Le risque de passage d’une molécule l’alpha-lactalbumine (2,7-3,3 g/L), concentration de lipides dans le lait dans le lait maternel est proportionnel la lactoferrine (1,4 à 1,9 g/L), les maternel déplacera le ratio à des valeurs à la demi-vie. Les substances ayant immunoglobulines A - IgA (1-3 g/L) supérieures ; ce type de substances est une demi-vie courte de l’ordre de et les lysoenzymes (0,4 g/L). Par représenté surtout par les médicaments 1-3 heures ont une élimination rapide. ailleurs, la caséine, un mélange actifs au niveau du système nerveux Si le médicament est pris juste après la hétérogène de protéines, compte pour le reste. L’α -glycoprotéine 1 Tableau 1 PH de certains médicaments acide n’est pas présente en quantité Médicaments alcalins pH > 7 Médicaments acides pH < 7 significative. Aciclovir (Zovirax ®) Amiodarone (Cordarone®) La concentration en protéines Furosémide (Lasilix®) Dobutamine (Dobutrex®) totales du lait est plus basse que Ganciclovir (Cymevene®) Dopamine celle du plasma ; ainsi la liaison Phénytoïne (Phenhydan®) Doxorubicine (Adriblastine®) protéique aux protéines du lait est plus faible que la liaison au plasma et Sulfamides (Bactrim®) Midazolam (Dormicum®) se fait principalement à l’albumine ; la Thiopental (Pentothal®)

188

Book 19-2.indb 188 2/21/2013 11:39:28 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Tableau 2 Caractéristiques influençant la fixation des médicaments aux protéines Caractéristique Type 1 Type 2 Nature du médicament Acide faible Base faible/substance non ionisable Protéine fixatrice Albumine Albumine - AAG Affinité Forte Faible Nombre de sites de fixation Petit Grand Possibilité de saturation Oui Non Possibilité d’interaction Possible Improbable

AAG : α1-glycoprotéine acide.

tétée, le taux plasmatique au moment retrouvées dans le lait par rapport à en lui-même mais peut être utilisé, en de la tétée suivante aura probablement celles retrouvées dans le plasma. l’absence de mesure de la concentration beaucoup diminué (conseil que nous Les dosages doivent être faits dans d’un médicament dans le lait, pour donnons fréquemment pour les le plasma et le lait en même temps et calculer la concentration qui correspond femmes allaitantes ayant un traitement idéalement après avoir atteint l’équilibre. à la concentration plasmatique de chronique). Nous sommes souvent Les ratios sont obtenus en faisant le médicament chez la mère. contraints à évaluer la quantité qui reste rapport des concentrations analysées, Un médicament qui est quatre fois dans l’organisme, surtout dans les cas où soit en prélevant un échantillon de plus concentré qu’un autre dans le lait la femme allaitante a pris un traitement lait au moment où la concentration n’est pas nécessairement quatre fois pour une durée déterminée. plasmatique est maximale, c’est- à-dire plus actif chez le nouveau-né. Tout Généralement, on considère qu’au au T max, soit en comparant les aires dépend du devenir du médicament bout de quatre demi-vies, environ 10 % sous la courbe des concentrations dans chez l’enfant. Lorsque la dose ingérée de la substance reste dans l’organisme le lait et dans le plasma en fonction du par l’enfant est évaluée, cette dose peut et qu’il faut pratiquement sept demi- temps. Ces façons de faire comportent être comparée avec celle que l’on donne vies pour qu’elle soit complètement chacune des avantages et des habituellement à l’enfant en mg par kg éliminée (Tableau 3). inconvénients. Elles donnent cependant dose pédiatrique ou au pourcentage de une approximation fort acceptable pour la dose maternelle afin de déterminer si la comparaison des médicaments entre la dose reçue par le nouveau-né via le lait Estimation du passage eux ; la méthode mathématique de peut produire un effet pharmacologique. des médicaments simulation par ordinateur et la méthode Généralement, on considère dans le lait in vitro permettent également d’évaluer qu’une dose inférieur à 10 % de celle les ratios lait/plasma (L/P) [2]. prise par la mère ne comporte pas de L’évaluation directe in vivo du transfert Le ratio L/P est utilisé comme un risque pour le nourrisson lorsqu’on ne des médicaments du plasma vers le lait indice de l’importance du transfert connaît pas la cinétique du médicament après atteinte de l’équilibre est le moyen d’un médicament dans le lait maternel. chez le nouveau-né ; lorsque les ratios le plus classique d’établir les ratios Le ratio L/P le plus soigneusement rapportés sont différents d’un auteur à des concentrations de médicament mesuré n’a pas de signification clinique l’autre, le risque peut être évalué en se basant sur la cinétique maternelle pour étudier le cas. Tableau 3 Temps nécessaire pour qu’une molécule soit complètement éliminée de l’organisme Nombre de demi-vie (t1/2) Taux éliminé (% dose administrée) 1 50,0 Estimation de 2 75,0 l’exposition du 3 87,5 nourrisson au 4 90,0 médicament 5 97,0 7 99,0 La quantité de médicament ingérée par le

t1/2 : temps de demi-vie (temps nécessaire pour que la concentration plasmatique d’une substance diminue de nourrisson peut être calculée si on connaît moitié). la concentration du médicament dans le

189

Book 19-2.indb 189 2/21/2013 11:39:28 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

plasma maternel (Cmat), le volume de 3 à 4 kg si ce médicament a une longue des effets indésirables surviennent chez lait ingéré (VLI) par unité de poids et demi-vie chez le nouveau-né. un enfant allaité en post-partum, ils sont par jour en mL/kg/jour et le rapport des Un nouveau-né âgé d’une semaine plus probablement la conséquence de concentrations entre le lait et le plasma de consomme au plus 40 mL aux 4 heures, l’exposition au médicament in utero. la mère (ratio L/P). La concentration du alors qu’à six semaines il consomme Les médicaments les plus concernés sont médicament dans le plasma maternel qui 150 mL aux 6 heures et qu’à 16 semaines généralement les psychotropes : dans est indiquée dans les études pharmaco il ingère de 150 mL à 225 mL ; 4 à 6 fois une telle situation, une surveillance sera cinétiques peut être utilisée en pratique et par jour, la quantité de substance ingérée programmée avant la naissance. multipliée par le ratio L/P pour donner la par séance d’allaitement, négligeable au Dans l’autre cas, le médecin est concentration par millilitre de lait. début, pourra avoir un effet sur l’enfant contraint à prescrire un traitement On calcule ensuite directement la plus tard au cours de l’allaitement [8]. pendant l’allaitement. Si le traitement dose ingérée par l’enfant en multipliant Souvent, dans la pratique est destiné à traiter une pathologie la concentration dans le lait par le quotidienne, lorsqu’une conduite à tenir aiguë, une suspension temporaire volume de lait ingéré par kg : Cmat × explicite figure dans la monographie pourra être envisagée avec deux simples ratio L/P × VLI = dose kg/jour. d’un produit « allaitement possible » précautions : maintenir la lactation en Ce mode de calcul de l’exposition ou « allaitement contre-indiqué », tirant et en jetant le lait et reprendre le a été proposé par Wilson et al. [6]. elle doit être suivie dans la majorité traitement en considérant les 7 demi- Il donne la quantité de médicament des cas ; mais lorsque la conduite à vies nécessaires à une élimination ingérée par le nourrisson sur une période tenir dans la monographie est moins totale du traitement. Par ailleurs, si le de 24 heures. Il est certes utile mais claire « allaitement déconseillé » traitement sera prescrit pendant une les paramètres pharmacocinétiques ou « simples données cinétiques » longue période au cours de l’allaitement, doivent absolument être évalués en ou « rubrique non renseignée », la une démarche stratégique s’impose et tenant compte de l’immaturité du décision d’allaiter ou de poursuivre un consiste pratiquement à : nourrisson [7]. allaitement maternel sous traitement doit être le plus souvent prise au cas • évaluer de façon individuelle le par cas, en accord avec la mère, après rapport bénéfice/risque en tenant Appréhension pratique l’avoir informée des risques éventuels. compte des données cliniques comme Il faut alors tenir compte de l’activité l’âge de l’enfant, son immaturité La concentration maximale dans le lait pharmacologique du médicament et de hépatique, ses capacités d’épuration, et des données pharmacologiques, correspond aux pics de concentration son profil cinétique, du profil des effets surtout la classe pharmacologique dans le plasma maternel. Il est donc indésirables du médicament, de l’âge du du médicament et son potentiel à préférable de ne pas allaiter lorsque nourrisson, du niveau d’allaitement, de engendrer des effets indésirables ; la concentration du médicament la possibilité de la surveillance et de suivi dans le plasma maternel est à son du nourrisson et de la compréhension • connaître les médicaments ayant une maximum. De plus, la quantité de de la mère. affinité lactée et pouvant donner des médicament liposoluble transmise à Généralement, nous sommes concentrations élevées dans le lait ; l’enfant varie en fonction de la durée confrontés à deux situations • être vigilant avec les médicaments de l’allaitement. différentes : une nouvelle prescription qui peuvent entraîner une sédation Pendant les dix à quinze premières durant l’allaitement ou le maintien ou une dépression respiratoire ou un minutes de tétée, le lait contient moins d’un traitement déjà prescrit lors de syndrome de sevrage ; de lipides de sorte que dans les séances la grossesse. En fait, si la mère souffre • penser aux médicaments qui peuvent plus longues, l’enfant reçoit moins de d’une affection chronique pour entraîner une diminution de la substance liposoluble au début qu’en laquelle le traitement est pris au long production lactée ; fin de tétée. Enfin, la quantité de lait cours, et généralement antérieure à la • promouvoir les médicaments absorbée par le nourrisson influe sur la grossesse, dans cette situation le risque utilisables chez l’enfant ; quantité de substance ingérée. Même paraît moindre pendant la période de si la concentration du médicament l’allaitement relativement à la période • promouvoir les médicaments ayant par mL de lait est faible, la dose de la grossesse car les concentrations qui des données sur leur concentration accumulée dans 600 à 1000 mL de lait atteignent le fœtus sont supérieures à lactée ; par 24 heures peut entraîner un effet celles qui atteignent le nourrisson par le • prendre le médicament à la fin de la pharmacologique chez un enfant de biais du transfert lacté. Par contre, quand tétée.

190

Book 19-2.indb 190 2/21/2013 11:39:28 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Conclusion d’investigations pharmacocinétiques actifs et se méfier de l’automédication. sur l’excrétion dans le lait que l’existence Il convient de toujours privilégier, dans Les paramètres pharmacocinétiques d’observations cliniques. Certaines une classe médicamenteuse donnée, varient pour chaque médicament et les règles simples peuvent aider les le médicament passant le moins dans études en cas d’allaitement maternel professionnels de santé confrontés à le lait, sans métabolite actif, à demi-vie manquent pour un nombre important la prescription de médicaments chez courte ou ne s’accumulant pas dans de spécialités, en particulier quand leur la femme allaitant son enfant, comme le compartiment lacté. Enfin, la voie commercialisation est récente. Ainsi, la proscrire les médicaments non d’administration occasionnant le plus mention « ne pas utiliser chez la femme indispensables, éviter les spécialités faible passage systémique doit être allaitante » signifie plutôt l’absence contenant une association de principes préférée.

Références

1. Department of Health and Human Service Office on Women’s 5. Gremmo-Féger G, Dobrzynski M, Collet M. Allaitement ma- Health. Benefits of breastfeeding. Nutrition in Clinical Care, ternel et médicaments [Breastfeeding and drugs]. Journal de 2003, 6:125–131. Gynécologie, Obstétrique et Biologie de la Reproduction, 2003, 2. Leblanc PP et al. Traité de biopharmacie et pharmacocinétique. 32(5):466–475. 3e édition [Treatise on biopharmaceutics and pharmacokinetics, 6. Wilson J. Drugs in breast milk. New York, ADIS Press, 1981. 3rd ed.]. Montréal et Paris, Presses de l’Université de Montréal 7. Saulnier JL et Comité de Rédaction du Centre National et Éditions Vigot, 1997:307–310. d’Information sur le Médicament Hospitalier (CNIMH). Évalu- 3. Briggs GG, Freeman RK, Yaffe SJ. Drugs in pregnancy and lacta- ation thérapeutique - Médicaments et allaitement [Therapeu- tion. Baltimore, Williams and Wilkins, 1999:1595. tic evaluation – drugs and breastfeeding]. Dossier du CNIMH, 4. Hale TW. Medications and mothers’ milk. Amarillo, Pharmasoft 1996, XVII(5-6). Publishing, 2002:812.

191

Book 19-2.indb 191 2/21/2013 11:39:28 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Revue Diagnostic et évaluation de l’hépatite virale C chez l’hémodialysé A. Bahadi,1 O. Maoujoud,1 Y. Zejjari,1 A. Alayoud,1 K. Hassani,1 D. Elkabbaj 1 et M. Benyahia 1

مراجعة تشخيص وتقييم فريوس التهاب الكبد يس بني مرىض الديال الدموي عبدالعايل هبادي، عمر موجود، يارس زجياري، أمحد العيود، كوتر حساين، دريس الكباج، حممد بن حييى 3 " " اخلالصـة: ُي َع ُّد التهاب الكبد يس من املشكالت الصحية الواسعة االنتشار يف شتى أرجاء العامل. ِّوتقدر منظمة الصحة العاملية أن %من سكان العامل مصابون بالعدوى بفريوس التهاب الكبد "يس ." ولدى مرىض ِّالديال، قد يصل معدل العدوى يف بعض البلدان إىل 80 ، %ويف املغرب يصل إىل %32 وفق السجل املغريب ِّللديال. َّوتتسم السرية الطبيعية للعدوى بالتهاب الكبد "يس" لدى مرىض ِّالديال الدموي املزمن بتفاقم صامت. وهناك طرق خمتلفة لتشخيص التهاب الكبد "يس" وحساسيتها تتزايد باطراد، وهناك اختبارات غري مبارشة ألضداد الفريوس "يس" اللتهاب الكبد، ويمكن تكرارها، ولكن تشيع النتائج السلبية الكاذبة؛ وهناك طرق غري مبارشة الختبار الرنا الفريويس، وهي طرق أكثر حساسية ولكنها ال تكون متوافرة ًدائام، وتكون أغىل ًثمنامن غريها. وقد تم تصنيف استعامالت هذه االختبارات من خالل التوصيات اجلديدة التي أصدرهتا اجلمعيات العلمية؛ إال أن تقييم أمراض الكبد اليزال موضع جدل، والتزال خزعة الكبد ِّمتثل املعيار الذهبي. وتستعرض هذه الورقة أساليب تشخيص وتقييم مرىض التهاب الكبد "يس" اخلاضعني ِّللديال الدموي.

RÉSUMÉ L’hépatite virale C constitue un problème de santé à l’échelle mondiale. L’Organisation mondiale de la Santé estime que 3 % de la population mondiale est infectée par le virus de l’hépatite C (VHC). Chez les hémodialysés, la prévalence atteint 80 % dans certains pays. Au Maroc, la prévalence est de 32 % selon le registre marocain de dialyse. L’histoire naturelle de l’infection par le virus de l’hépatite C chez l’hémodialysé chronique est caractérisée par une évolution silencieuse. Les moyens de diagnostic du VHC sont divers et de plus en plus sensibles : les tests indirects à la recherche des anticorps anti-VHC qui sont reproductibles mais les cas de faux négatifs sont très fréquents, et les tests directs recherchant l’ARN viral qui sont de plus en plus sensibles mais ne sont pas toujours disponibles et ont un coût élevé. L’utilisation de ces tests a été codifiée grâce aux nouvelles recommandations des sociétés savantes. Cependant, l’évaluation de l’atteinte hépatique demeure controversée et la ponction-biopsie hépatique reste la méthode de référence. Le présent article passe en revue les approches pour le diagnostic et l’évaluation de l’hépatite C chez l’hémodialysé.

Diagnosis and evaluation of hepatitis C virus among haemodialysis patients

ABSTRACT Hepatitis C is a health problem worldwide. The World Health Organization estimates that 3% of world’s population is infected with hepatitis C virus (HCV). In haemodialysis patients, the prevalence reaches 80% in some countries. In Morocco, HCV prevalence is 32% according to the Moroccan register of dialysis. The natural history of hepatitis C infection in chronic haemodialysis patients is characterized by a silent evolution. There are different methods to diagnose HCV and they are becoming increasingly sensitive. There are indirect tests for antibodies to HCV: these are reproducible but false negatives are common and there is direct testing of viral RNA: this is more sensitive but not always available and is more expensive. The use of these tests has been categorized through new recommendations from learned societies. However, the evaluation of liver disease is still controversial and liver biopsy remains the gold standard. This paper reviews the approaches for diagnosing and evaluating hepatitis C in haemodialysis patients.

1Service de Néphrologie, Dialyse et Transplantation rénale, Hôpital militaire d’instruction Mohammed V, Rabat (Maroc) (Correspondance à adresser à A. Bahadi : [email protected]). Reçu : 28/02/12; accepté : 19/06/12

192

Book 19-2.indb 192 2/21/2013 11:39:29 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction en dialyse est de 32 % et l’incidence recherche est négative et une infection de 9,4 % par an, d’où l’intérêt de cette évolutive quand on détecte l’acide Depuis que Freinstone et al. [1] ont question. ribonucléique (ARN) viral (Figure 2). mis en évidence, en 1975, l’existence Dans ce cadre, plusieurs tests ont été d’un nouvel agent inconnu causant développés avec un seuil de détection des hépatites non A non B, baptisées Diagnostic dans la qui ne cesse de baisser, en particulier NANB, plusieurs études ont permis population générale la PCR (Polymerase Chain Reaction) de faire certaines avancées [2-5]. Mais en temps réel avec un seuil qui atteint l’avancée majeure fut la caractérisation Le test de diagnostic le plus utilisé 12 UI/mL (Tableau 1). Ces tests directs du génome du virus par Choo et al. en est la recherche des anticorps dirigés sont sensibles mais ne sont pas toujours 1989 [6]. L’agent causal des NANB, contre les protéines virales par méthode disponibles et posent le problème de ainsi identifié, prit alors le nom de virus indirecte, notamment par ELISA. Ce coût élevé. Il fallait donc trouver d’autres de l’hépatite C (VHC). D’un point test utilise une plaque imprégnée de marqueurs fiables et reproductibles, de vue médical et épidémiologique, protéines du virus (représenté dans la surtout dans les laboratoires où on ne cette découverte permit de mettre au figure 1 par des triangles). Le sérum du dispose pas de techniques de biologie point un test ELISA pour détecter la patient est ajouté à cette plaque. Si ce moléculaire. Pour cela, Tanaka et al. présence du VHC chez les patients et sérum contient un anticorps anti-VHC, ont évalué un nouveau test biologique surtout de tester les produits sanguins, ce dernier va se fixer aux protéines qui permet de détecter l’antigène core et donc d’éliminer la première cause virales. Ce complexe est détecté ensuite du virus et ont trouvé une corrélation de transmission du virus dans les pays par un anticorps spécifique et cette linéaire entre les concentrations de développés [7]. fixation va entraîner un changement de l’antigène core et de l’ARN pour les L’infection par le virus de coloration du marqueur (Figure 1). génotypes 1, 2 et 3 [14]. Cette recherche l’hépatite C (VHC) est considérée La sensibilité de l’ELISA est proche a été ensuite combinée avec la recherche comme un problème majeur de de 100 % chez les immunocompétents d’anticorps pour développer le test santé publique à l’échelle mondiale. mais peut être faussement négative ou ELISA 4e génération. L’Organisation mondiale de la Santé positive chez les immunodéprimés et au À travers ces moyens de diagnostic estime qu’environ 3 % de la population cours des maladies auto-immunes [13]. direct et indirect, on a pu identifier, générale est infectée par ce virus, avec Cependant, un anti-VHC positif ne avec presque certitude, l’histoire de 130 à 170 millions de porteurs renseigne pas sur le caractère aigu ou naturelle du VHC dans la population chroniques [8]. En dialyse, le problème chronique de l’infection et encore peut générale (Figure 3). Malgré ces est beaucoup plus ample, avec une être faussement positif. Pour cela, il moyens de plus en plus sensibles, prévalence qui peut atteindre 80 % est indispensable de réaliser un test de Lerat et al. ont mis en évidence une et une incidence de plus de 9 % par confirmation par la recherche d’acide nouvelle entité biologique qui échappe an [9-12]. Au Maroc, en estime que nucléique. Ce test direct pourra faire la aux moyens de diagnostic direct et la prévalence de l’hépatite virale C part entre une infection guérie quand la indirect : l’hépatite C occulte [15].

» Test immuno-enzymatique ELISA de troisième génération : » –– Sensibilité : 97 %-100 % –– Spécificité : 99 % » Faux négatifs : » –– Etats d'immunosuppression –– IRC hémodialysés –– Co-infection VIH –– Transplantation » Faux positifs : rares » –– Affections auto-immunes Source : NIH Management of Hepatitis C Consensus Conference Statement. June 2002.

Figure 1 Détection des anticorps anti-VHC par ELISA [13] (AC : anticorps ; VHC : virus de l’hépatite C ; IRC : insuffisance rénale chronique ; VIH : virus de l’immunodéficience humaine)

193

Book 19-2.indb 193 2/21/2013 11:39:29 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

anti-VHC

Symptômes +/-

ARN-VHC Titre

ALT

Normal

0 1 2 3 4 5 6 1 2 3 4 Mois Années

Figure 2 Cinétique des marqueurs biologiques au cours de l’hépatite virale C (VHC : virus de l’hépatite C ; ARN : acide ribonucléique ; ALT : alanine aminotransférase)

Les auteurs ont recherché l’ARN viral le taux des transaminases est bas au 70 % de la limite supérieure normale dans le tissu hépatique et dans les cours de l’insuffisancerénale chro- du laboratoire est fortement prédictif cellules mononucléées périphériques nique (IRC) [16]. Cette diminution a d’hépatite virale C avec une sensibilité chez 100 patients ayant une élévation été attribuée à la carence en vitamine de 67 % et une spécificité de 75 %. inexpliquée des transaminases et chez B6 et à la présence de toxines urém- Comme dans la population générale, lesquels la recherche d’anticorps et de iques dans le sang qui pourrait altérer le diagnostic de l’hépatite virale C par les l’ARN viral dans le sérum était négative : la détection des transaminases [17-19]. anticorps anti-VHC est confronté à de 57 % avaient l’ARN viral au niveau de Cependant, on a remarqué que malgré faux positifs et de faux négatifs, avec un leur tissu hépatique. cette baisse, le taux des transaminases taux de 4 % et 9 % respectivement [21]. est élevé chez les patients IRC porteurs On peut donc conclure que la PCR est le d’une hépatite C en comparaison avec moyen idéal pour détecter précocement les anti-VHC négatifs même dans la une hépatite C. Cependant, la recherche Diagnostic chez fourchette normale du laboratoire. d’ARN viral n’est pas toujours l'hémodialysé Gouveia et al. ont comparé le taux disponible dans les laboratoires, chronique d’ALAT (alanine aminotransférase) notamment en périphérie, et a un impact chez 202 hémodialysés dont 15 anti- économique considérable. Fabrizi et al. L’histoire naturelle de l’hépatite virale VHC positifs [20]. Les auteurs ont con- ont voulu savoir si l’antigène du virus de C chez l’hémodialysé est, par contre, staté que le rapport des ALAT sur la l’hépatite C peut améliorer la situation. caractérisée par une évolution à bas limite supérieure de la normale est de Ils ont constaté que dans la population bruit. En effet, et comme l’ont dé- 0,7 chez les patients infectés par le virus. des hémodialysés aussi, il y avait une montré Fabrizi et al. depuis 10 ans, Ils ont conclu qu’un taux dépassant corrélation linéaire entre l’antigène core

Tableau 1 Techniques de détection de l’ARN du virus de l’hépatite C et seuil de détectabilité Technique Seuil de détectabilité RT-PCR Amplicor® HCV v. 2.0 (Roche) 50 UI/mL Version semi-automatisée Cobas® Amplicor® HCV v. 2.0 (Roche) 50 UI/mL TMA Versant HCV RNA (Bayer) 10 UI/mL PCR en temps réel (Cobas® TaqMan®) 15 UI/mL PCR en temps réel (Abbott) 12-30 UI/mL

ARN : acide ribonucléique. RT-PCR : reverse transcription-polymerase chain reaction ; HCV : hepatitis C virus ; TMA : transcription-mediated amplification ; RNA : ribonucleic acid.

194

Book 19-2.indb 194 2/21/2013 11:39:29 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Infection VHC

Chronicité Guérison 60 % à 85 % 15 % à 40 %

Cirrhose Stable 10 % à 15 % 85 % à 90 %

Progression lente 75 % Complications 25 % CHC (2 % - 4 %)

Figure 3 Histoire naturelle de l’hépatite virale C (VHC : virus de l’hépatite C ; CHC : carcinome hépatocellulaire)

et l’ARN viral [22]. Ce travail confirme précoce de quatre patients (six mois temps réel. Alors que faire pour dépister la fiabilité et la forte sensibilité du test avant). le VHC en hémodialyse ? ELISA de quatrième génération chez Malgré tous ces moyens, de plus en les hémodialysés. Sur le plan pratique, plus pertinents, l’hépatite virale C est l’apport de ce test a été étudié récem- peut être sous-estimée en hémodialyse. Recommandations ment en Inde chez 250 hémodialysés Barril et al. ont recherché l’infection de dépistage chroniques(HDC) [23]. Dans ce par le virus de l’hépatite C occulte chez travail, on a remarqué que 13 patients 109 patients ayant une élévation des Un grand nombre d’experts en la négatifs pour l’anticorps contre le virus enzymes hépatiques inexpliquée [24]. matière se sont réunis dans le cadre de l’hépatite C (Ac anti-VHC) ont été L’ARN viral a été recherché dans les de la Fondation KDIGO (Kidney détectés par l’antigène core malgré la cellules mononucléées sanguines Disease : Improving Global Outcomes) faible charge virale. Les auteurs ont suivi périphériques. Cette recherche était pour établir des recommandations ces patients pendant six mois et ont positive chez 45 patients et 26 patients concernant la prévention, le diagnostic, constaté que ce test a permis le diagnostic seulement étaient positifs par PCR en l’évaluation et le traitement de l’hépatite

Minimum costs using recombinant immunoblot assay (RIBA®) 3.0 only Minimum costs using nucleic acid test (NAT), followed by 7 RIBA on NAT-negatives Maximum costs using RIBA only 6 Maximum costs using NAT, followed by RIBA on NAT-negatives 5

Incremental cost increase (US$) cost increase Incremental 0 0 5 10 15 20 25 30

Anti-HCV prevalence (%)

Figure 4 Estimation du coût du dépistage du VHC en fonction de la prévalence selon les Centers for Disease Control and Prevention (CDC) [26]

195

Book 19-2.indb 195 2/21/2013 11:39:30 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

virale C [25]. Selon ces d’hémodialyse où la prévalence du VHC doivent être soumis à un test recommandations, les patients en VHC est basse, la recherche du virus moléculaire (recommandation hémodialyse chronique (maladie devrait être initiée par un test immuno- forte). Un deuxième test moléculaire rénale chronique stade 5D) enzymatique (suivi en cas de positivité est suggéré 2 à 12 semaines après un doivent être testés pour le VHC par un test moléculaire à la recherche premier test négatif (recommandation à l’initiation de l’hémodialyse ou de l’ARN du VHC) (recommandation faible). Un algorithme a été lors du transfert d’une autre unité modérée). Dans les unités d’hémo- également proposé par ces experts, d’hémodialyse (recommandation dialyse où la prévalence du VHC est dans lequel on a repris toutes ces élevée, un test moléculaire doit être forte). Reste le moyen de diagnostic recommandations (Figure 5). ou de dépistage : faut-il rechercher envisagé d’emblée (recommandation l’anticorps ou l’ARN viral ? Dans les modérée). études publiées, la sensibilité des Pour le suivi, il est conseillé de méthodes immuno-enzymatiques varie retester tous les 6 à 12 mois par test Évaluation de entre 53 et 100 % et la spécificité entre immuno-enzymatique les patients en l’hépatite virale C 85 et 99 %. Si on considère la prévalence hémodialyse qui sont négatifs pour le chez l’hémodialysé au centre de dialyse, on constate que VHC (recommandation modérée) le nombre de faux négatifs augmente et un test moléculaire pour le VHC Après diagnostic positif de l’hépatite avec la prévalence de l’hépatite C dans le doit être réalisé chez les patients virale C chez un patient en hémodialyse, centre. De plus, selon une étude réalisée hémodialysés qui ont une élévation l’évaluation d’un hémodialysé porteur par les Centers for Disease Control and inexpliquée des transaminases d’une hépatite virale consiste en ce qui Prevention (CDC) (Figure 4), le coût du plasmatiques (recommandation suit [27] : dépistage augmente en cas de recherche forte). Cependant, si un nouveau cas d’acide nucléique, en particulier d’infection VHC dans une unité • Recherche de co-infection par si la prévalence est basse [26]. En d’hémodialyse est suspecté d’être le VHB (virus de l’hépatite B), le s’appuyant sur ces deux considérations, nosocomial, tous les patients qui VIH (virus de l’immunodéficience on recommande que dans les unités pourraient avoir été exposés au humaine)

CKD Stage 5 HD

Admission to HD facility Transfer from other HD facility Testing every 6-12 months

Low-prevalence setting High-prevalence setting HCV test

(-) EIA NAT (+)

(-) (+)

Normal Abnormal If HCV ALT/AST (+) outbreak, repeat NAT in 2-12 weeks

(-) Consider antiviral treatment

Figure 5 Algorithme proposé par KDIGO et résumant les moyens de dépistage de l’hépatite virale C en hémodialyse (http:// www.kdigo.org/guidelines/hepc/guide1.html#alg1) (CKD: chronic kidney disease; HD: haemodialysis; HCV: hepatitis C virus ; EIA: enzyme immunoassay; NAT: nucleic acid testing; ALT: alanine aminotransferase; AST: aspartate aminotransferase)

196

Book 19-2.indb 196 2/21/2013 11:39:30 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

• Détermination de la charge virale par la population générale. Dans une été correctement classés par rapport PCR-RT étude randomisée, publiée il y a 2 ans, au score Métavir [35]. Et on remarque • Identification du génotype et qui incluait plus de 300 patients aussi que la valeur prédictive positive dont 78 hémodialysés, Pawa et al. ont est augmentée quand le score est plus • Évaluation de la maladie hépatique démontré que la PBH était sans risque important (score > 0,6). D’autres –– Plaquettes, taux de chez les hémodialysés [32]. tests biologiques ont été étudiés mais prothrombine (TP), bilirubine, De plus, la PBH transjugulaire reste aucune corrélation n’a été retrouvée albumine, échographie un moyen alternatif en cas de risque avec le degré de fibrose hépatique chez –– Ponction-biopsie hépatique/ hémorragique accru surtout chez les l’hémodialysé [36]. FibroTest® et FibroScan®. patients obèses, en présence d’ascite L’élastométrie impulsionnelle Concernant la charge virale, ou en cas de thrombopénie. Dans ce ultrasonore représente une nouvelle des précautions sont nécessaires ; cadre, Ahmad et al. ont comparé les étape dans le développement de tests notamment, les prélèvements doivent complications de la PBH entre deux diagnostiques non invasifs de la fibrose être réalisés avant le branchement groupes : 46 patients ayant bénéficié hépatique. Avec un résultat instantané, car l’héparine inhibe la transcriptase d’une PBH par voie transjugulaire une technique indolore non invasive, inverse (reverse transcriptase- RT) et et 32 patients d’une PBH par voie renouvelable et un apprentissage pourrait contribuer à de faux négatifs ou à percutanée [33]. Il n’y avait aucune aisé, l’élastométrie hépatique est réduire la charge virale. La centrifugation différence statistique entre les deux véritablement le « stéthoscope doit être réalisée en moins de 4 heures groupes. Les résultats sont surprenants hépatologique », avec déjà de après prélèvement et la conservation puisqu’aucune complication nombreuses applications. après, de préférence, à moins hémorragique n’a été recensée dans le La valeur de l’élastométrie de 20 °C [26]. groupe de PBH par voie transjugulaire. impulsionnelle n’a été évaluée au cours Le problème majeur est le suivant : La différence entre les deux groupes est de la maladie rénale qu’une seule fois comment évaluer la fibrose ? La significative. par Alric et al. [37]. En effet, les auteurs ponction-biopsie hépatique (PBH) est Le problème majeur de la PBH ont démontré chez 38 transplantés que la méthode de référence pour évaluer reste sa reproductibilité. Calès et al. le FibroScan® a permis la classification la sévérité de l’atteinte hépatique par le ont démontré la grande variabilité du correcte dans 90 % des cas en cas de VHC [28]. Elle permet l’évaluation de la diagnostic histologique en pratique fibrose minime (< F2). Cependant, sévérité de l’inflammation, la recherche courante [34]. En effet, dans cette le diagnostic est incorrect dans de maladie associée et surtout la série de 205 hépatites chroniques C, le environ 40 % des patients avec fibrose mesure de la fibrose (stade). Pour cela, diagnostic initial posé par le pathologiste sévère (F3/F4). on s’est basé depuis plusieurs années généraliste ou de première ligne a été Dans un diagnostique, le FibroScan® sur le score de Ishak et surtout le score comparé à celui d’une relecture par un pourrait être complémentaire des tests Métavir [29,30] qui évaluent la fibrose expert du groupe Métavir. Si l’accord biologiques, chacun ayant ses propres en quatre stades et l’activité en trois n’est pas mauvais aux stades extrêmes, limites. Pour cela, Castéra et al. ont stades. il devient très médiocre pour les stades étudié la sensibilité et la spécificité Dans la population des intermédiaires, en particulier pour le de l’association de ces deux moyens hémodialysés, cet examen invasif stade 2, puisque, lorsque l’expert dit [38]. Les auteurs ont constaté que le peut aboutir à de nombreuses stade 2, le pathologiste de première stade de fibrose a été correctement complications, surtout hémorragiques. ligne peut dire 0, 1, 2, 3, ou 4 ! Ceci établi par cette association dans 88 % Ozdogan et al. ont recensé les explique que la concordance globale en et 95 % des cas de fibrose débutante complications de la PBH chez 150 pratique courante est médiocre, donc et avancée respectivement. Il est donc patients, dont 74 patients atteints insuffisante. trop tôt pour préciser la place de ces d’IRC [31]. Les auteurs ont été D’autres moyens biologiques tests non invasifs par rapport à la biopsie confrontés à sept complications pour évaluer la fibrose ont prouvé leur hépatique et jusqu’à ce jour, il n’y a pas hémorragiques dans le groupe IRC efficacité dans la population générale de recommandation claire concernant alors qu’aucun saignement n’a été et le FibroTest® en est le chef de file. ce sujet. Selon l’AASLD (American objectivé chez le groupe témoin. Mais Varaut et al. ont utilisé ce test et ont Association for the Study of Liver Diseases), c’est un risque potentiel puisque des retrouvé que pour 50 hémodialysés la ponction-biopsie hépatique chez un études récentes ont démontré que le chroniques et 60 transplantés, 88 % et hémodialysé doit être discutée au cas risque est moins important que dans 77 % des patients respectivement ont par cas [39].

197

Book 19-2.indb 197 2/21/2013 11:39:30 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Conclusion diagnostiques ont bénéficié des progrès Concernant l’évaluation de la fibrose de la biologie moléculaire, cependant hépatique, il est trop tôt pour préciser la L’hépatite virale C constitue un beaucoup de questions restent sujets à place des tests non invasifs par rapport à la souci majeur en dialyse. Les moyens controverse, notamment l’hépatite occulte. biopsie hépatique chez l’hémodialysé.

Références

1. Freinstone SM et al. Transfusion associated hepatitis not due of viral hepatitis. American Journal of Kidney Diseases, 2001, to viral hepatitis A or B. New England Journal of Medicine, 1975, 38(5):1009–1015. 292:767–770. 17. Perico N et al. Infection and Chronic Renal Diseases. Clinical 2. Alter HJ et al. Transmissible agent in non-A, non-B hepatitis. Journal of the American Society of Nephrology; CJASN, 2009, Lancet, 1(8062):459–463. 4:207–220. 3. André P et al. Characterization of low- and very-low-density 18. Tang S, Lai KN. Chronic viral hepatitis in hemodialysis patients. hepatitis C virus RNA-containing particles. Journal of virology, Hemodialysis International. International Symposium on Home 2002, 76(14):6919–6928. Hemodialysis, 2005, 9:169–179. 4. Bartenschlager R et al. Nonstructural protein 3 of the hepatitis 19. Furusyo N et al. Liver damage in hemodialysis patients with C virus encodes a serine-type proteinase required for cleavage hepatitis C virus viremia: A prospective 10-year study. Digestive at the NS3/4 and NS4/5 junctions. Journal of virology, 1993, Diseases and Sciences, 2000, 45:2221–2228. 67(7):3835–3844. 20. Cavalcanti Gouveia E et al. Identificacao de ponto de corte 5. Behrens SE, Tomei L, De Francesco R. Identification and prop- no nivel serico da alanina aminotransferase para rastrea- erties of the RNA-dependent RNA polymerase of hepatitis C mento da hepatite C em pacientes com insuficiencia renal virus. The EMBO Journal, 1996, 15(1):12–22. cronica em hemodialise [Identification of the cutoff value 6. Choo QL et al. Isolation of a cDNA clone derived from a for serum alanine aminotransferase in hepatitis C screen- blood-borne non-A, non-B viral hepatitis genome. Science, ing of patients with chronic renal failure on hemodialysis]. 1989, 244(4902):359–362. Revista da Sociedade Brasileira de Medicina Tropical, 2004, 7. Kuo G et al. An assay for circulating antibodies to a major 37:18–21. etiologic virus of human non-A, non-B hepatitis. Science, 1989, 21. Hanuka N et al. Hepatitis C virus infection in renal failure 244(4902):362–364. patients in the absence of anti-hepatitis C virus antibodies. 8. Organisation mondiale de la Santé. Document WHA62/22. Journal of Viral Hepatitis, 2002, 9(2):141–145. Hépatite virale [Viral hepatitis]. Soixante-Deuxième Assem- 22. Fabrizi F et al. Novel assay using total hepatitis C Virus (HCV) blée mondiale de la Santé, Genève, 2009 (http://apps.who. core antigen quantification for diagnosis of HCV infection int/gb/ebwha/pdf_files/A62/A62_22-fr.pdf, consulté le 29 in dialysis patients. Journal of Clinical Microbiology, 2005, octobre 2012) 43(1):414–420. 9. Fabrizi F, Poordad FF, Martin P. Infection and the patient with 23. Medhi S et al. Diagnostic utility of hepatitis C virus core an- end-stage renal disease. Hepatology (Baltimore, Md.), 2002, tigen in hemodialysis patients. Clinical Biochemistry, 2008, 36(1):3–10. 41:447–452. 10. Magredial (Maroc Greffe et Dialyse) Registre de l’Insuffisance 24. Barril I et al. Occult hepatitis C virus infection among hemo- rénale chronique terminale [Morocco Dialysis and Transplant dialysis patients. Journal of the American Society of Nephrology, « Magredial » Registry]. Rabat, Ministère de la Santé, Direction 2008, 19:2288–2292. des Hôpitaux et des Soins ambulatoires (Communication 25. KDIGO clinical practice guidelines for the prevention. diagno- orale des résultats de 2005), 2009. sis, evaluation, and treatment of hepatitis C in chronic kidney 11. Sekkat S et al. Prévalence des anticorps anti-VHC et incidence disease. Kidney International. Supplement, 2008, (109):S1–S99. de séroconversion dans cinq centres d’hémodialyse au Maroc 26. Recommendations for preventing transmission of infections [Prevalence of anti-HCV antibodies and seroconversion inci- among chronic hemodialysis patients. MMWR Recommenda- dence in five haemodialysis units in Morocco].Néphrologie & tions and Reports, 2001, 50(RR–5):1–43. Thérapeutique, 2008, 4(2):105–110. 27. Perico N et al. Infection and Chronic Renal Diseases. Clinical 12. Fissell RB et al. Patterns of hepatitis C prevalence and sero-� Journal of the American Society of Nephrology; CJASN, 2009, conversion in hemodialysis units from three continents: The 4:207–220. DOPPS. Kidney International, 2004, 65:2335–2342. 28. Saleh HA, Abu-Rashed AH. Liver biopsy remains the gold stand- 13. NIH Consensus Statement on Management of Hepatitis C: ard for evaluation of chronic hepatitis and fibrosis. Journal of 2002. NIH Consensus and State-of-the-Science Statements, Gastrointestinal and Liver Disease, 2007, 16:425–426. 2002, 19(3):1–46 (http://consensus.nih.gov/2002/2002Hepa titisC2002116PDF.pdf, accessed 7 September 2012). 29. Ishak K et al. Histological grading and staging of chronic hepatitis. Journal of Hepatology, 1995, 22:696–699. 14. Tanaka E et al. Evaluation of a new enzyme immunoassay for hepatitis C virus (HCV) core antigen with clinical sensitivity 30. Bedossa P, Poynard T. An algorithm for the grading of activity in approximating that of genomic amplification of HCV RNA. chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology (Baltimore, Md.), 2000, 32:388–393. Hepatology (Baltimore, Md.), 1996, 24:289–293. 15. Lerat et al. Occult hepatitis C virus infection in patients in 31. Ozdogan M et al. H. Telatar percutaneous liver biopsy compli- whom the etiology of persistently abnormal results of liver- cations in patients with chronic renal failure. Nephron, 1996, function tests is unknown. Journal of Infectious Diseases, 2004, 74:442–443. 189(1):3–6; 7–14. 32. Pawa S et al. percutaneous liver biopsy is safe in chronic hepa- 16. Fabrizi F et al. Decreased serum aminotransferase activity in titis C Patients with end-stage renal disease. Clinical Gastroen- patients with chronic renal failure: Impact on the detection terology and Hepatology, 2007, 5(11):1316–1320.

198

Book 19-2.indb 198 2/21/2013 11:39:30 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

33. Ahmad A et al. Transjugular liver biopsy in patients with end- infected with hepatitis C virus. Transplantation proceedings, stage renal disease. Journal of Vascular and Interventional Radi- 2004, 36(1):50–52. ology, 2004, 15:257–260. 37. Alric L et al. Comparison of liver stiffness, fibrotest and liver 34. Calès P et al. Reproductibilité des examens usuels pour la fibrose biopsy for the assessment of liver fibrosis in renal transplant hépatique en pratique clinique : PBH, tests sanguins et Fibroscan. patient with chronic viral hepatitis. Transplant International, [Reproductibility of routine examinations for liver fibrosis 2009, 22(5):568–573. in clinical practice: liver biopsy, blood tests and Fibroscan]. 38. Castéra L et al. Prospective comparison of transient elastog- Communication orale présentée aux Journées francophones raphy, fibrotest, APRI, and Liver biopsy for the assessment de pathologie digestive (JFPD) (Résumé), Paris, 2008 (http:// of fibrosis in chronic hepatitis C. Gastroenterology, 2005, www.biols.fr/uploads/rte/File/JFPD2008VARIOB9.pdf, 128:343–350. consulté le 29 octobre 2012). 39. Ghany MG et al. Diagnosis, management, and treatment of 35. Varaut A et al. Diagnostic accuracy of the fibrotest in hemo- hepatitis C: an update. Hepatology (Baltimore, Md.), 2009, dialysis and renal transplant patients with chronic hepatitis C 49:1335–1373. virus. Transplantation, 2005, 80(11):1550–1555. 36. Boyacioğlu S et al. Investigation of possible clinical and laboratory predictors of liver fibrosis in hemodialysis patients

199

Book 19-2.indb 199 2/21/2013 11:39:31 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

Report Evaluation of the national tuberculosis surveillance system in Afghanistan K.M.I. Saeed,1,3 R. Bano 2 and R.J. Asghar 3

تقييم النظام الوطني ُّلرتصد السل يف أفغانستان خواجة مري إسالم سعيد، راشدة سيد بانو، رانا جواد أصغر

اخلالصـة:يوجد يف أفغانستان نظامان ُّلرتصد السل؛ الربنامج الوطني ملكافحة السل، ونظام املعلومات لإلدارة الصحية. وقد أجرى الباحثون ًتقييام 2010 – هلذين النظامني ُّللرتصد يف الفرتة كانون الثاين/يناير شباط/فرباير ُّللتعرف عىل َم َواطن القوة ومكامن الضعف ولصياغة التوصيات. وتم تقييم مكونات كل نظام باستخدام الدالئل اإلرشادية ملراكز مكافحة األمراض والوقاية منها يف الواليات املتحدة األمريكية، ووجد الباحثون أن منافع ومرونة الربنامج الوطني ملكافحة السل جيدة، وأن استقراره ُوح ْ ن سمتثيله وجودة املعطيات فيه معتدلة، أما البساطة واملقبولية ومالءمة التوقيت فيه فكانت سيئة؛ إذ يتجاوز ُّالتأخر يف إصدار التقارير ثالثة أشهر، وكانت القيمة التنبؤية اإلجيابية 11 %واحلساسية 70 .% كام وجد الباحثون أن نظام املعلومات لإلدارة الصحية بسيط ومقبول ومستقر، ُت َع ُّ دتقاريره يف الوقت املناسب، وقد كان إعداد التقارير واإلمداد باملعلومات االرجتاعية ً جيداألن هذا النظام كان َحيظى ٍبدعم حكومي جيد. وكانت املرونة وجودة املعطيات والتمثيل معتدلة، وكانت القيمة التنبؤية اإلجيابية 10 %واحلساسية %68. ومل يكشف ٌّأي من النظامني أية فاشية. واتضح للباحثني أن النظامني ِّيمثالن ازدواجية جلهود ُّالرتصد وأهنام ال يغطيان القطاع اخلاص.

ABSTRACT Afghanistan has 2 tuberculosis surveillance systems, the National Tuberculosis Control Programme (NTP) and the Health Management Information System (HMIS). An evaluation of these surveillance systems in January/ February 2010 was done to identify their strengths and weaknesses and to formulate recommendations. Attributes of the programmes were evaluated using US Centers for Disease Control and Prevention guidelines. Usefulness and flexibility of the NTP system were good; stability, representativeness and data quality were average. Simplicity, acceptability and timeliness were poor. Reporting delays regularly exceeded 3 months. Positive predictive value and sensitivity were 11% and 70% respectively. The HMIS system was simple, acceptable and stable, with timely reporting. Reporting and feedback were good, as this system has strong government support. Flexibility, data quality and representativeness were average. Positive predictive value and sensitivity were 10% and 68% respectively. No outbreaks were detected by either system. The NTP and HMIS surveillance systems are duplicative and neither covers the private sector.

Évaluation du système national de surveillance de la tuberculose en Afghanistan

RÉSUMÉ L'Afghanistan dispose de deux systèmes de surveillance de la tuberculose, le Programme national de lutte antituberculeuse et le système d'information pour la gestion de la santé. Une évaluation de ces systèmes de surveillance en janvier/février 2010 a été menée afin d'identifier leurs forces et leurs faiblesses et de formuler des recommandations. Les caractéristiques des programmes ont été évaluées à l'aide des recommandations des Centers for Disease Control and Prevention américains. L'utilité et la souplesse du programme national de lutte antituberculeuse étaient satisfaisantes ; la représentativité et la qualité des données étaient moyennes. La simplicité, l'acceptabilité et la ponctualité étaient médiocres. Le retard de transmission des notifications dépassait régulièrement trois mois. La valeur prédictive positive et la sensibilité étaient respectivement de 11 % et 70 %. Le système d'information pour la gestion de la santé était simple, acceptable et stable, et transmettait les notifications en temps voulu. Les notifications et les commentaires étaient satisfaisants, car le système bénéficiait d'un appui soutenu du gouvernement. La souplesse, la qualité des données et la représentativité étaient moyennes. La valeur prédictive positive et la sensibilité étaient respectivement de 10 % et 68 %. Aucune flambée n'a été détectée par ces deux systèmes. Le programme national de lutte antituberculeuse et le système d'information pour la gestion de la santé font double emploi mais aucun ne couvre le secteur privé.

1Afghanistan National Public Health Institute, Ministry of Public Health, Kabul, Afghanistan (Correspondence to K.M.I. Saeed: [email protected]). 2Communicable Disease Surveillance and Response Unit, World Health Organization Country Office, Kabul, Afghanistan. 3Field Epidemiology and Laboratory Training Programme, National Institute of Health, Islamabad, Pakistan. Received: 29/08/11; accepted: 21/02/12

200

Book 19-2.indb 200 2/21/2013 11:39:31 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Introduction describes the national tuberculosis criteria and ranked as poor, average surveillance systems in Afghanistan, or good. The main documents of each Tuberculosis (TB) is a public health identifies strength and weakness of the system (strategic plan, guidelines, problem which is more common in de- systems and provides recommenda- manuals, annual reports and data- veloping countries due to the disease’s tions for improvement. bases) were reviewed and consulta- close links to poverty, housing status and tion with stakeholders was carried out. access to health services. Furthermore, In addition 5 key informants from the disease has been re-emerging in Methods each system were interviewed from developed countries in recent decades. among the directors, managers and Based on World Health Organization The review was undertaken during officers. Furthermore 1 or 2 staff at (WHO) estimates, 9.27 million new January to February 2010. In order to departments of the MoPH (research, cases of TB occurred in 2007 (139 per evaluate the TB surveillance systems at surveillance, preventive health, policy 100 000 population), compared with the national level the guidelines for the and planning and health care service 9.24 million new cases (140 per 100 evaluation of public health surveillance delivery) were interviewed; they were 000 population) in 2006 [1]. systems developed by the United States selected based on their involvement in Centers for Disease Control and Pre- Afghanistan is recovering from the and relevance to the TB surveillance vention (CDC) [5] were used. These aftermath of more than 2 decades of system. guidelines include the following steps: civil war; nevertheless, there is ongoing security instability and economic • Engage the stakeholders in evalua- difficulties in the country. Control of tion. Description of the TB communicable diseases is one of the • Describe the surveillance system to surveillance systems highest priorities for the Ministry of be evaluated. The Afghan NTP was established in Public Health (MoPH) because these • Focus the evaluation design. account for 60%–80% of all curative 1954 with technical and financial sup- outpatient visits and over half of all • Gather credible evidence regarding port from WHO. The NTP is a techni- deaths in Afghanistan [2]. The country the performance of the surveillance cal department of the MoPH that leads remains one of the TB high-burden system. and carries out prevention, detection, countries with a high ratio of females • Justify and state conclusions and diagnosis and treatment of TB patients. to males infected (1.85:1). TB control make recommendations. Its vision is a TB-free country, with services are an integral part of the pack- • Ensure use of evaluation findings and elimination of the disease as a public age of services delivered through the share lessons learned. health problem by 2050 [6]. The major primary health care system at district Using the guidelines, the surveil- objectives of the programme are to and provincial levels. The sustainability lance systems for one disease were reduce the risk of infection, morbid- of activities is unclear, however, given taken into acount, the sampling was ity and mortality due to TB by having the unstable security situation with its purposeful and the head of depart- 100% coverage of DOTS treatment current reliance on donor support. ments were approached to obtain nec- by the end of year 2015 and ensuring TB control is a priority for the coun- essary information. Both the Afghan the cure rate and detection rate of new try, being one of the main 7 compo- NTP and HMIS were evaluated by ap- sputum smear-positive pulmonary TB nents of the Basic Package of Health plying these guidelines. We did a desk cases at over 85% and 70% respectively. Services for Afghanistan [3]. TB care review before interviewing managers There are 13 recording and reporting is provided free of charge in all public in both systems. The managers of other forms which are used by the surveil- health facilities. It is very important to relevant departments were also inter- lance department for data management. strengthen surveillance systems in order viewed. We did not use any statistical The programme claims that data are to find TB patients as early as possible tests of significance; however, selected collected, analysed and disseminated and treat them properly. There are sur- indices were used and calculated as quarterly in review meetings. Originat- veillance 2 systems which, in addition to described in the guidelines [5]. Ten ing from public health facilities, the other activities, are responsible for TB attributes (usefulness, simplicity, flex- data flows to the provinces and then surveillance in Afghanistan: the Health ibility, data quality, predictive value, to regions and finally to the NTP at the Management Information System sensitivity, timeliness, acceptability, national level, where multiple national (HMIS) [4] and the National Tuber- representativeness and stability) were indicators are calculated and reported culosis Control Programme (NTP). evaluated in both systems. Attributes to the MoPH and WHO Regional This paper is an evaluation report which were scored based on pre-defined Office for the Eastern Mediterranean.

201

Book 19-2.indb 201 2/21/2013 11:39:31 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

WHO and other MoPH partners National TB Control Programmme

Regional level Regional level Regional level Regional level

Provinvial levels - (34 provinces)

Data from facility levels totally 1013 till 2010

Figure 1 Flow of data at the National TB Control Programme (NTP) in Afghanistan (MoPH = Ministry of Public Health)

The flow of information is presented illustrates the flow of information. The registries (logbooks) are the source of in Figure 1. The system is supported by system is supported by USAID and the data. HMIS collects few variables WHO, the United States Agency for World Bank. whereas NTP uses consistent, standard International Development (USAID), forms to collect more variables regard- the Japan International Cooperation ing morbidity and mortality of TB Agency, the Global Fund to Fight AIDS, Attributes of the systems including age, sex and geography. The Tuberculosis and Malaria and some HMIS database is linked and fully inte- other donors. A summary of the evaluation of the 10 grated with other databases at MoPH, The Afghan health information sys- attributes from the CDC guidelines are while the NTP database is still in the tem was upgraded to HMIS in 2003 presented in Table 1. pilot stage. Prior to sending to the national level, data are edited, analysed and [4] and focuses on standard report- Usefulness ing forms containing the minimum shared with stakeholders at provincial information that needs to be collected, HMIS data are useful for planning and levels and submitted electronically to analysed and reported routinely. The monitoring but are less useful for de- the MoPH quarterly. The HMIS is sim- system is computerized, which allows tecting TB outbreaks, whereas proper ple, quick and easy to enter and analyse data on priority health problems at the analysis of NTP data will detect and the data. The Microsoft Access database local levels to be aggregated on a data- allow a response to outbreaks. TB data is familiar to staff and facilitated by a base at the provincial and national levels collected by HMIS are poorly linked to series of simple drop-down menus. and rapid distribution of analysis copies action, while NTP data are used by both However, data collection and flow of to all levels. The main aims of the system national and international stakeholders. information is more complex. Taking are better management of health servic- Based on research department state- into account the very high turnover es and resources, facilitating supervision ments at MoPH they have not used TB of staff the officers are well-trained at and assisting in planning, monitoring surveillance data for the development all levels. All implementers at primary and evaluation. In HMIS data are col- and conduct of studies; however, NTP health care facilities, including govern- lected about suspected TB cases and the surveillance data are used by the TB ment, national and international NGOs number of slides examined. Data gener- research department and WHO for are involved in the reporting of data and ated at the facility level are reported to various studies. timely quarterly feedback is provided to provincial levels monthly for analysis them from the national level. and feedback and afterwards the reports Simplicity Using multiple forms for data col- go to the national level quarterly for Standard case definitions are well- lection makes the NTP system more feedback and dissemination. Figure 2 utilized in both systems and facility complex both in its structure and mode

202

Book 19-2.indb 202 2/21/2013 11:39:31 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

Other MoPH partners

HMIS central database National hospital Quarterly

Provincial levels - (34 provinces) Provincial public hospitals Monthly Community level - BHCs, CHCs and district hospitals

Figure 2 Flow of data at the Health Management Information System (HMIS) in Afghanistan (MoPH = Ministry of Public Health, BHC = basic health centre, CHC = comprehensive health centre)

of operation. Data are sent as hard Being supported by the government for missing and illegal values (out of copies to all levels, with delays at each and having staff in the government a predetermined range) at different level. Vital TB indicators such as case salary system, the HMIS is sustainable, levels. Data management, including notification rate, case detection rate, unlike a project based approach in data collection, entry, editing, analysis TB conversion rate and treatment suc- which the funding is withdrawn when and feedback, is good due to timely cess rate are calculated using available the project is finished. It also conforms feedback and training. data in NTP. With proper feedback with planned new software systems in The NTP has expanded rapidly mechanisms the main vehicle for com- which common files will link to other but there are still blank and illegal val- munication is discussing the key issues databases such as training, expendi- ues in the forms despite checks on the using data in the quarterly review meet- ture, pharmacy, human resources and quality of data at provincial, regional ings at provincial, regional and national others. Variations in funding have had and national levels in quarterly re- levels. Staff are well-trained regarding little effect on the functioning of the view meetings. It covers more com- TB surveillance but there is still a need system. prehensive data compared with the to conduct training at the grass-roots In contrast, the NTP system is flexi- HMIS. However, there are difficulties level in health facilities for effective ble and able to adapt and accommodate in data management at NTP at all TB case management. Outbreak de- changes when there is need for addi- levels, including reporting, analysis tection and investigation throughout tional information or reporting in TB and feedback. the country is not recorded either by forms. However, the system is consider- HMIS or NTP. Some other issues ably donor-dependent and cessation of Acceptability include the complex data collection funding and support is a concern for its HMIS benefits from acceptability (many forms and ways of reporting), sustainability. For example, there was an among all MoPH, NGOs and interna- flow of information (delays) and labo- interruption in funding for laboratory tional agencies. Data are collected from ratory algorithms (different categories, services and no reporting was done in registries and there is no need to inter- not just positive or negative). There are late 2008. view and collect the data for routine also issues with the NTP conducting cases. A high rate of data reporting is crosschecks of the national reference Data quality observed at all facilities and almost laboratory data and random samples While the HMIS has progressed since no refusals in participation are seen. of slides from regions which are using it was set up in 2003 in terms of quality Complete and timely data are available. different indicators. of data, some blank cells and illegal However, HMIS is less acceptable by values are still observed in the forms; TB-related stakeholders due to prob- Flexibility for instance, it is known that some lems with incomplete and duplicate HMIS seems to be less flexible for facilities have reported that they are data. accommodating changes when there performing sputum smear examina- The NTP on the other hand is a need for additional information or tions when in reality they do not have is not fully acceptable at all levels of modes of operation, however, remov- microscopes available. By and large health workers due to the multitude of ing questions (variables) is easier than the HMIS is collecting complete and forms and high workload at primary adding new questions to the forms. valid data from registries by checking level health facilities. The system even

203

Book 19-2.indb 203 2/21/2013 11:39:32 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale Justification National Programme TB Control Availability ofAvailability information comprehensive used for and data are feedback at planning at national and international levels. Poor lower levels. due to complicated, multipleSystem is complex forms. Staff is a challenge. No integration withtraining HMIS. Still using paper forms. Some easily. changes and can accommodate System is flexible ongoing.changes are Rapid expansion. copies with indicators. Hard System covers all necessary with rechecking data improper values. Poor blank and missing management. used but health lack forms are workers WHO recommended and sometimes Delayed reporting about procedures. awareness shows refusals. no reporting Estimated sensitivity (2008 data) 73% is consistent with global targets Estimated PPV (2008 data) 11%. Majority of public health facilities but data missing covered are information Age and sex is available with private sector. from data for multiple TB indicators. but and feedback is supposed to be done quarterly Report delays with timereview identifiable. easily lag are Quarterly meeting and feedback. is mechanism for reporting is available for long-term. Development ofDonor support and feedback Data reporting database is in progress. electronic copies. Staffis done by hard on data management is training needed. Good Poor Good Average Poor Good Average Average Poor Average Evaluation Justification

Health Management Information System Standard case definitions with guidelines. Poor case definitionsStandard with guidelines. to action. poorly linked Data are coordination with NTP. definitions Case followed and consistent forms are are collected. Staff used but limited TB data are trained. are to understand. System is simple and easy of storage Proper remove variables. to add or Difficult data with link to other databases. System is not able to changes easily accommodate properly. and checked registries from taken Data are and data management. Demographic Appropriate Low socioeconomic status information insufficient. are values. unknown, blank and missing High level of of awareness with stakeholders respect and and their No refusal to procedures ownership. All public health rate. almost 95% reporting facilities participate. Estimated sensitivity (2008 data) 68% is close to global target. Estimated PPV (2008 data) 14%. health facilities but data missing Public covered are with areas little System favours rural private sector. from and secondary hospitals. to urban settings attention of recording and behavioural demographic Poor variables. Monthly collection and level and reporting at provincial feedback is timely and there is no time (quarterly) lag. Inconsistency in feedback and not annual reports. Data management without System is fully failure. functional with sustainable technical and financial electricity Database is stable despite regular support. power interruptions. Poor Good Average Good Good Good Average Average Good Good Evaluation a Comparison ofComparison of attributes the Health Management Information (TB) (NTP) in Afghanistan for monitoring tuberculosis System (HMIS) and National Programme TB Control Attribute Usefulness Simplicity Flexibility Data quality Acceptability Sensitivity predictive Positive value Representativeness Timeliness Stability Using classificationPrevention of and Control Centers for Disease [5]. the Table 1 Table data a HealthWHO = World Organization; PPV = positive predictive value.

204

Book 19-2.indb 204 2/21/2013 11:39:32 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

lacks acceptability within the TB We can calculate sensitivity at the level For each TB suspect a sputum sample programme among those working of all TB cases or at the level of sputum is collected and 2 to 3 AFB slides are at grass-roots levels, such as heads of smear-positive cases. examined both for diagnosis and follow health facilities. Additionally private Total TB suspected cases (preva- up. The following 2008 data from the sector acceptability is another issue. lence): 40 000 NTP were identified: WHO, however, is more supportive of Total AFB slides examined: 268 614 NTP and uses the data for its regional Total TB suspected cases detected : and global reports. A weak point is 28 300 Total AFB slides positive: 30 447 that quarterly data sent to the NTP for Sensitivity (first level) PPV (30 447/268 614) ×100 = 11% analysis are delayed and incomplete. (28 300/40 000) × 100 = 71% Representativeness On the other hand, quality control Using data from smear-positive cases practices are good as laboratory slides shows almost the same sensitivity: HMIS covers public facilities includ- are collected from facilities randomly Total TB sputum smear-positive cas- ing government and NGOs; however, for crosschecking. es estimated: 17 885 15% of the country is not covered by public health facilities and a consid- Sensitivity of the surveillance Total TB sputum smear-positive cas- erable proportion of the population system es detected: 13 136 are utilizing the private sector, are We defined sensitivity as the ability of Sensitivity (second level) not seeking medical care and/or are the surveillance system to truly detect (13 136/17 885) × 100 = 73% seeking care through other means. In cases of TB in the country (sensitiv- Predictive value positive of addition the system is not collecting ity = cases detected/cases existing × the surveillance system sufficient information about the demo- 100). No survey has been conducted graphic and socioeconomic status of We defined predictive value positive to establish the exact burden of TB in the population. It therefore cannot be (PPV) as the proportion of acid-fast ba- Afghanistan and the system is using said that the system is representative cilli (AFB)-positive slides out all slides of the whole population. Nevertheless, WHO estimates for case detection and examined by laboratory facilities (PPV HMIS covers fewer variables and is other indicators. = true positives/all positives × 100). HMIS data trying to cover the private sector and HMIS data military health facilities as well in efforts Based on HMIS case definitions for TB After a review of data and interviews towards improving the representative- the patient is recorded as suspect and he with HMIS staff at the central level in ness of the data. or she is sent for sputum examination, 2008 data at HMIS we calculated the NTP has expanded its DOTS cov- but the HMIS is not able to trace the PPV in 2 ways. First, it was based on the erage exponentially and is now collect- contacts. The following is a calculation proportion of AFB slides that were posi- ing data from 1013 facilities, which are of the system sensitivity using data from tive out of all slides examined for AFB: almost 90% of Basic Package of Health the HMIS database for 2008: Total AFB slides examined: 283 831 Services facilities. The system is col- Total TB suspected cases reported: lecting sufficient information regarding Total AFB slides positive: 28 951 146 981 age and sex of cases along with their Total sputum smear-positive cases PPV (28 951/282 831) × 100 = 10% residential areas. There are 2 sources of estimated: 17 885 Secondly, we can calculate the propor- information: basic management units tion of TB cases starting treatment (i.e. (facilities) and laboratory registries. Total TB sputum smear-positive cas- true positives) out of those identified as es detected: 12 229 Similar to the HMIS, the NTP misses suspected TB: data from private sector and military Sensitivity (12 229/17 885) × 100 Total TB cases suspected: 146 981 health facilities. = 68% Total TB cases started treatment : NTP data Timeliness 21 487 The NTP system is collecting epide- Timeliness was quantified as delays miological and laboratory data from all PPV (21 487/146 981) × 100 = 15% more than 3 months. At the community facilities covered by DOTS. The contact NTP data level the amount of time it takes for cli- tracing is another problem which is not There are possibilities of laboratory ex- ents in the community to seek medical done by the system and reduces its amination for AFB at facilities but that care is very difficult to determine. In sensitivity. We calculated sensitivity just is not possible to do culture except at HMIS data are taken from health facili- at the routine reporting level in 2008. the NTP reference laboratory in Kabul. ties and sent to the provinces monthly.

205

Book 19-2.indb 205 2/21/2013 11:39:32 AM EMHJ • Vol. 19 No. 2 • 2013 Eastern Mediterranean Health Journal La Revue de Santé de la Méditerranée orientale

At the provincial level data are ana- Due to delays in reporting of data where many cases come and are lysed and shared with all stakeholders timely information is not available managed. and after cleaning data are forwarded within the NTP. Data are collected at Both systems have approximately as soft copies to the national level health facilities and laboratory facilities equal sensitivity and PPVs. The low at the end of each quarter without in hard copies for each quarter and prevalence of smear-positive cases in delays. More than 95% of facilities are are entered into a Microsoft Excel as the population has affected the PPV, sending their reports regularly. Col- well as a new Access database. These lowering it to 10%. Both systems are lection and examination of sputum databases are very simple for data able to show the seasonal and secu- is also timely at facilities. Feedback is entry and analysis and few resources lar trend of TB cases, which generate provided after the quarterly data col- are needed for maintenance of the hypotheses and trigger research ques- lection from central HMIS to those database. Corruption of databases due tions. facilities in which problems are identi- to viruses is always a concern. The Despite a review of the literature fied in analysis. system needs to be made electronic at we could not find any similar evalu- NTP is collecting information all levels particularly at provincial and ations for comparison, but some quarterly using laboratory and facility regional levels. unpublished reports of public surveil- quarterly formats. It is also reporting lance evaluations from the Field Epi- treatment outcome and conversion demiology and Laboratory Training Discussion of sputum-smear positive cases. How- Programme in Pakistan provide simi- ever, the quarterly reports are not lar results regarding TB surveillance HMIS is a very good system which sent on time from facilities to the systems. provides timely information through provincial level or from the province The following points are recom- to the regions or from the regions to daily recording and monthly and quarterly reporting and feedback. mended based on findings of this the national level. Delay in reporting evaluation. is therefore a concern. Furthermore, NTP is collecting comprehensive in- feedback is not provided timely and formation but with delays in reporting • Attention should be given to in- properly. All data are sent as hard cop- and feedback. So timeliness, as a key tegrate both systems in order to ies from the lower level to the central surveillance system metric, is better avoid duplication and to strengthen NTP, which is not acceptable in the in the HMIS than the TB surveillance mechanisms for coordinating with current challenging conditions of Af- system under the NTP. Unfortunately related departments to use HMIS ghanistan. Timeliness therefore is a the information is only occasionally data for action. concern for the NTP. linked to action in HMIS while an- • Taking the lesson learnt from ac- nual data is mostly used for planning at ceptability and sustainability of Stability the national and international level at HMIS the surveillance system at The HMIS is able to collect, manage NTP. The HMIS system has a high de- NTP should search for mecha- and provide data properly without gree of acceptability among stakehold- nisms for improving its acceptability failure from all its facilities and its fully ers because it has been developed with among health workers. There is a operational with its monthly report- the involvement and participation of need to reconsider the variables in ing at the provincial levels and quar- various HMIS taskforces. The NTP both the HMIS and NTP surveil- terly reporting at the central level. All surveillance system is less acceptable lance systems. stakeholders are entitled to receive the due to its complexity and multitude • The system should assess the reli- HMIS analysis folder for their own of forms. ability of data sources or triangulate analysis of data including TB. The HMIS is collecting fewer vari- the data with other data sources. Microsoft Access database is simple ables regarding TB compared with Refresher training of health workers and user-friendly, is well developed the NTP. There is poor coordination at the lower levels on guidelines and and few resources including human or sharing of information between case definitions of TB will increase and non-human are required for its the 2 surveillance systems regarding their skills and improve the quality maintenance. It takes very little time to TB and this causes duplication of of data reported. enter, edit, transfer, analyse, store and effort, wastage of resources and in- • Better adherence to guidelines is retrieve the data. Problems in extrac- complete information. Both systems needed in both of the systems in tion of information do occur, however, are collecting data from all public terms of collecting 3 samples to due to occasional breakdown in the health facilities and both are missing be sent to laboratories for confir- electricity power supply. data from the private sector facilities mation. NTP should collect infor-

206

Book 19-2.indb 206 2/21/2013 11:39:32 AM املجلة الصحية لرشق املتوسط املجلد التاسع عرش العدد الثاين

mation on contacts and enhance • Establishment and protection of data- Acknowledgements contact tracing. bases from power outrages and com- • Both systems should try to strengthen puter viruses will improve the stability This study was supported by Field public–private partnerships and incor- of the systems. Double entry of data Epidemiology and Laboratory Pro- porate private sector data into their will enhance the data quality and pre- gramme, National Institute of Health, systems in order to increase the utiliza- vent missing and erroneous data. More Islamabad, Pakistan. In addition the tion and representativeness of findings. efforts are needed to ensure the stabil- National TB Control Programme ity and sustainability of the systems. • The concept of timeliness and prompt and Health Management Information reporting should be applied equally • The general coordination, integra- System at Ministry of Public Health to both systems and for dissemination, decentralization and expansion were supportive in the evaluation. tion of reports as well. The end users of activities need to be strengthened I would like to thank all who were should receive the reports in hard and in order to avoid duplication and pro- involved in this study in one way or soft copies as soon as possible. vide information for action. another.

References

1. WHO report 2009. Global tuberculosis control: epidemiology, 5. Updated guidelines for evaluating public health surveillance strategy, financing. Geneva, World Health Organization, 2009 systems: recommendations from the guidelines working (WHO/HTM/TB/2009.411). group. Morbidity and Mortality Weekly Report, 2001, 50 No. 2. Health Management Information System. Data analysis. Kabul, RR-3. Afghanistan, Ministry of Public Health, 2009. 6. Annual report and statistics of tuberculosis in Afghanistan—2008. 3. Basic Package of Health Services (BPHS). Kabul, Afghanistan, Afghanistan. Kabul, Afghanistan, National TB Control Program, Ministry of Public Health, 2006. Ministry of Public Health, 2008. 4. National Health Management Information System procedures manual. Kabul, Afghanistan, Ministry of Public Health, 2006.

207

Book 19-2.indb 207 2/21/2013 11:39:32 AM

Members of the WHO Regional Committee for the Eastern Mediterranean املجلة الصحية لرشق املتوسط Afghanistan . Bahrain . Djibouti . Egypt . Islamic Republic of Iran . Iraq . Jordan . Kuwait . Lebanon هى املجلة الرسمية التى تصدر عن املكتب اإلقليمى لرشق املتوسط بمنظمة الصحة العاملية. وهى منرب لتقديم Libya . Morocco . Oman . Pakistan . Palestine . Qatar . Saudi Arabia . Somalia . South Sudan السياسات واملبادرات اجلديدة ىف اخلدمات الصحية والرتويج هلا، ولتبادل اآلراء واملفاهيم واملعطيات الوبائية Sudan . Syrian Arab Republic . Tunisia . United Arab Emirates . Yemen ونتائج األبحاث وغري ذلك من املعلومات، وخاصة ما يتعلق منها بإقليم رشق املتوسط. وهى موجهة إىل كل أعضاء املهن الصحية، والكليات الطبية وسائر املعاهد التعليمية، وكذا املنظامت غري احلكومية املعنية، واملراكز املتعاونة مع منظمة الصحة العاملية واألفراد املهتمني بالصحة ىف اإلقليم وخارجه.

EMHJ is a trilingual, peer reviewed, open access journal and the full contents are freely available at its website: http://www/emro.who.int/emhj.htm

Editor-in-chief ©World Health Organization 2013 EMHJ All rights reserved WHO Regional Office for the Eastern Mediterranean P.O. Box 7608 Disclaimer The designations employed and the presentation of the material in this publication do not imply the expression of any opinion Nasr City, Cairo 11371 whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of Egypt its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border Tel: (+202) 2276 5000 lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products Fax: (+202) 2670 2492/(+202) 2670 2494 does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar Email: [email protected] nature that are not mentioned. All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either express or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. The named authors alone are responsible for the views expressed in this publication.

Cover 19-2.indd 2 2/25/2013 10:28:24 AM Contents

Acknowledgement of the work of Dr Mohammad Haytham Khayat...... 103 Editorial

Child mortality in the Eastern Mediterranean Region: challenges and opportunities...... 104 Research articles

Médicaments et allaitement maternel : évaluation du risque médicamenteux chez le nourrisson...... 186 Diagnostic et évaluation de l’hépatite virale C chez l’hémodialysé...... 192 Report

Evaluation of national tuberculosis surveillance system in Afghanistan...... 200

املجلد التاسع عرش / عدد Volume 19 / No. 2 2 شباط / فرباير February / Février 2013

Cover 19-2.indd 1 2/20/2013 2:31:21 PM