ISSN: 2078-5909 Volume 2 Number 2 August 2015

The East and Central Africa ...... MEDICAL JOURNAL

Published by Kenyatta Universitii y, SchoolEast of and EaMedicines tCentral and Ce nAfricatral A fMedicalrica Med Journalical Jour n2015;al 2014; 2: 1(2): i EDITORIAL BOARD

Editor -in-Chief Dr. Ng’ethe Muhoho BSc, MSc, PhD

Editors:

1. Prof. N. Gikonyo TCM BSc, MSc, PhD, PDF

2. Prof. B.M. Okello-Agina MBChB, MMed (Obs/Gyn), IVF/ET, PhD 3. Dr. S. Gwer MD, PhD 4. Dr. C.N. Wamae BSc, MPH, PhD

Editorial Board 1. Prof. E. Kabiru 2. Dr. T. Were Kabianga University Kenya 3. Dr. A.K. Wanyoro Kenyatta University Kenya 4. Prof. J. Waundo Kenyatta University Kenya 5. Dr. A. Ikol Kenyatta university Kenya 6. Prof. O.J. Olobo Makerere university 7. Prof. J. Wangombe Nairobi University Kenya 8. Dr. J.O. Ayugi University of Botwana 9. Prof. V. Mwapasa University of Malawi 10. Prof. E.E. Kaaya Muhimbili University 11. Prof. Peter Mason University of Zimbabwe

Editorial Manager Mr. David Ng’ethe

Communications to:

The Editor-in-Chief The East and Central Africa Medical Journal The School of Medicine Kenyatta University P.O. Box 43844 - 00100 Nairobi, Kenya Fax: + 254 20 8711575 | Tel: +254 020 2453591 Email: [email protected] Website: www.ku.ac.ke Published by Kenyatta University

i East and Central Africa Medical Journal 2014; 1(1): i East and Central Africa Medical Journal 2014; 1(2): i ii i East and Central Africa Medical Journal 2015; 2: ii Content Editorial Antimicrobial resistance in human health Ogutu JO ...... 59

Original Research Antibiotic susceptibility profiles of bacteria isolates obtained from patients presenting with diarrhoea in Machakos Distric Hospital, Kenya Juma BW, Kariuki S, Bulimo WD, Mutugi WM, Wurapa EK, Waiyaki PG ...... 60

Antimicrobial resistance patterns among E. coli isolates from children presenting with diarrhoea at a cosmopolitan hospital in Kenya Kangethe SK, Kiiru J, Kabiru EW, Kariuki S ...... 64

Pattern of morbidity and mortality of childhood illnesses at the children emergency room of Teaching Hospital, Aba, Okoronkwo NC, Chappjumbo AU ...... 70

Determinants of early infant diagnosis and treatment of HIV among exposed infants in informal settlements in Nairobi, Kenya Makau GM, Okwara FN, Oyore JP ...... 74

Eclampsia at N’Djamena South District Hospital, Chad: epidemiological aspects and foeto-maternal prognosis Gabkika BM, Djongali S, Buambo G, Fankep DC, Tchoubou BM ...... 80

Efficacy of total lymphocyte count as a surrogate for CD4 cell counts in HIV-infected adults in the era of higher treatment cutoffs and less funding: a cross-sectional study Mwenda V, Njuguna J, Musa M ...... 84

Knowledge, practices and perception on trachoma and its influence on health seeking behaviour of the pastoralist patients in Kajiado Central Division, Kenya Munguti PN, Ng’ang’a Z , Muttunga J ...... 89

Identification of enterotoxigenicStaphylococcus aureus strains from meat and dairy products by multiplex PCR and reverse passive latex agglutination test in Nairobi, Kenya Mathenge JM, Okemo PO, Ng’ang’a PM, Mbaria JM, Gicheru MM ...... 97

Case Reports Traumatic sex with vulval haematoma formation: case report and review of literature Ngatia JW ...... 104

Cryptococcal meningitis in a none-HIV infected five month old infant with rickets: Case report Okwara FN, Makewa S, Karo E ...... 106

Journal Policy and Guidelines ...... 109

iii East and Central Africa Medical Journal 2015; 2: iii Editorial

Antimicrobial resistance in human health Contribution of bacteria in human disease are leading the pack in coming up with as well as certain aspects of antimicrobial new and workable strategies to tackle the resistance that could have major challenge [3-6]. implications in the control of microbial According to WHO, the microbiology diseases is well covered in this issue of laboratory is a key partner in the East and Central African Medical Journal. fight against antimicrobial resistance. In less than a century after the landmark Establishment and support of the discovery of penicillin the world is today microbiology laboratory has been again faced with many types of microbes identified as a fundamental priority that have developed resistance to various in guiding and assessing intervention antimicrobial agents and with a potential efforts in the fight against antimicrobial to diminish or even reverse the many resistance. However, there are still glaring outstanding inventions from antibiotic gaps in research development and little research. Continued evolution, emergence awareness of the potential negative socio- and dissemination of drug-resistant strains economic and financial impacts of drug such as methicillin or vancomycin-resistant resistance, more so in low-income. Staphylococcus aureus, vancomycin- To combat this menace of drug resistant Enterococci, extended-spectrum resistance, well-coordinated efforts beta-lactamase producing bacteria, from governments, multi-national carbapenem-resistant Enterobacteriaceae organizations, civil society and all stake- multidrug or extensively drug-resistant holders must be geared towards infection Mycobacterium tuberculosis and Neisseria prevention and control, antibiotic gonorrhoea and so on is a worrying trend. resistance surveillance, research on According to World Health Organization mechanisms of evolution and emergence [1], if the current antimicrobial resistance of new resistance traits, discovery of new problem is not promptly and effectively effective antimicrobials including those dealt with the situation could soon slide of plant origin, invention of new vaccines into the reality of the “post-antibiotic and novel diagnostic tools and proper era”. Antibiotic resistance mainly stems antibiotic stewardship. from misuse or overuse of antibiotics, for example in livestock growth Dr. J O Ogutu, PhD, MSc, BSc, Medical promoters, from the environment as in Microbiologist, Department of Pathology, the scavenging free-roaming indigenous School of Medicine, Kenyatta University, chicken and in hospitals. Food handlers PO Box 43844 – 00100, Nairobi, Kenya are also an important conduit in the spread Email: [email protected] of antimicrobial resistant strains like Staphylococcus aureus. References Antimicrobial resistance menace is global and widespread although the extent 1. http://www.who.int/drugresistance/ of the problem varies from country to amr_global_action_plan/en/; country and region to region. Increasing 2. (http://www.weforum.org/reports/ cases of resistant organisms continues to global-risks-report-2013-eighth- be exported from one country or region edition ), there is still little to another due to international travel 3. http://www.who.int/drugresistance/ and trans-border transport. The World amr_global_action_plan/en/; Economic Forum identified antibiotic 4. https://www.gov.uk/government/ resistance as a global risk beyond the uploads/system/uploads/attachment_ capacity of any organization or nation to data/file/385733/UK_AMR_annual_ manage or mitigate alone. A number of report.pdf; international and national bodies such as 5. http://www.whitehouse.gov/sites/ WHO, the European Commission, the default/files/docs/carb_national_ UK government and the White House strategy.pdf

59 East and Central Africa Medical Journal 2015; 2: 59 Original Research

Antibiotic susceptibility profiles of bacteria isolates obtained from patients presenting with diarrhoea in Machakos Distric Hospital, Kenya

Juma BW1,5 , Kariuki S1,3 , Bulimo WD2,4, Mutugi WM1, Wurapa EK2, Waiyaki PG1,3 concern due to the compromised 1 Abstract Jomo Kenyatta University of efficacy of antimicrobial agents used in Agriculture and Technology, Introduction: Bacterial diarrhoeal the treatment of infectious diseases [1]. Nairobi, Kenya. diseases are the cause of morbidity The emergence and dissemination of 2 US Army Medical Research and mortality in Africa, but data on antimicrobial resistance in bacteria has Unit Kenya antimicrobial susceptibility patterns of been well documented as a serious problem 3Centre for Microbiology enteric bacterial aetiologies are limited. worldwide [2]. Selective pressure favoring Research, Kenya Medical This study was carried out in Machakos antimicrobial-resistant phenotypes is Research Institute, Nairobi, District Hospital from September 2010 applied whenever antimicrobials are Kenya to March 2013 on bacteria isolated from used, including treating disease in clinical 4Department of patients with diarrhoea where 300 patients medicine and preventing disease and Biochemistry, University of were studied. promoting growth in animal husbandry Nairobi, Kenya Methodology: Bacterial isolates [1]. 5Centers for Disease Control were identified biochemically and Reports have suggested the use of and Prevention – Kenya E. coli pathotypes detected through tetracycline, sulfa drugs, cephalosporins, Multiplex Polymerase Chain Reaction. and penicillins to be a major factor in Corresponding author: Susceptibility testing was done following the emergence and dissemination of Dr BW Juma, PO Box both disc diffusion and minimum antimicrobial-resistant E. coli [3-5]. inhibitory concentration methods. 54840-00200, Nairobi, However, a relative paucity of information Results: Bacterial isolates detected Kenya exists regarding antimicrobial resistance included: Enteroaggregative E. coli in E. coli from nosocomial and non- Email: xwl2@cdc. (EaggEC 13.7%), Enterotoxigenic E. hospital sources, especially those from gov;[email protected]; coli (ETEC 11%), Enteroinvasive E. animal sources. In this study, antimicrobial [email protected] coli (EIEC 8.3%), Enteropathogenic E. susceptibility profiles were determined coli (EPEC 4.3%), Shigellae (22.3%), for Shigellae, Salmonella spp, Klesiella Salmonella spp (0.3%), Proteus spp spp and E. coli isolates of ETEC, EHEC, (0.6%) and Klebsiella spp (0.3%). The EPEC, EAEC and EIEC. Strains of E. coli isolates were resistant to more than four pathotypes may cause urinary tract and antibiotics chloramphenical (28%), enteric infections in humans and have been cotrimoxazole (78%), co-amoxilav (70%) implicated in infections [1]. The isolates erythromycin (98%), ciprofloxacin (4%), used in the present study, were collected cefotoxime (18%) and tetracycline (56%). from participants attending Machakos Conclusions: The detection of high District Hospital in Kenya. None of the resistance found to commonly used isolates were from food, water or animals. antibiotics should serve as a warning call for close surveillance, identification and Materials and Methods understanding of the epidemiology of Study patients: Three hundred study the resistance with a view to setting up participants comprising of children as preventive strategies that can minimize or young as four months old and adults stop the emerging and spread of resistance from Machakos County seeking medical to the antibiotic arsenal currently in use. attention during the period 2010 to 2013. For the first time in more than ten years, The purpose and nature of the study was gentamicin has shown to be susceptible and can be utilized for management of explained to the participants. Informed diarrhoeal illness resulting from bacterial consent was obtained from the mature infection. participants while parents/guardians gave signed consent for their children to Key words: Antimicrobial resistance, participate in the study. Research ethical E. coli pathotypes, Shigellae approval was granted by the Ethical Introduction Review Committee of Kenya Medical Research Institute (KEMRI SSC No. 989). Bacterial antibiotic resistance is an Sample collection: Stool samples were emerging and serious public health collected in a sterile stool container from

60 East and Central Africa Medical Journal 2015; 2: 60-63 all patients who met the inclusion criteria and consented to participate in the study. For those patients who could not void stool at the time of their visit to the hospital, a rectal swab was obtained by qualified personnel and placed in Carry Blair transport medium for shipment to the laboratory for culture and identification. Detection of bacteria: Stool samples were plated on different freshly prepared and quality- controlled media (MacConkey, Xylose Lysine Desoxycholate (XLD) and sub cultured on nutrient agar following growth after 24 hours. Following 24 hours growth, different colonies from each media were selected as follows: Lanes L: 100 BP ladder, Lane 1; Pooled positive control, Two to three colonies (black, medium and small size Lane 2: ST1 and ST2, Lane 3: LT, pink or clear) from XLD for suspected Salmonella spp Lane 4: CNF1, Lane5: eaeA, Lane6: negative control, and Shigella spp; 5-6 colonies (5-lactose and one non- lactose fermenters) from lactose MacConkey for Lane 7: ST2 and ipaH, Lane8: EAEC, Lane 9: ST2, Lane 10: ST1. different E. coli pathotypes including ETEC; both yellow and green colonies from TCBS for Figure 1: MPCR amplification of different E. coli Vibrio cholerae and Vibrio parahemolyticus respectively, pathotypes toxins and one clear colony from Sorbital MacConkey to detect Drug susceptibility patterns E. coli 0157:H7. Isolates confirmed as E. coli using TSI

(A/A, -H2S, ± gas), Simmon citrate (green), motile, and indole positive were confirmed for different pathotypes including ETEC by use of mPCR. Multiplex Chain Reaction amplification of toxin genes for E. coli pathotypes: DNA extraction from E. coli strains was done by boiling the whole cell lysate. The complete %Resistance multiplex PCR assay required 5 master mixes. For every toxin a master mix was made up of 15.2μl distilled water, 2.5μl of 10X PCR buffer, 1.5μl , 0.5μl dNTPs ( 25mM Drug each nucleotide), primers, 0.3μl taq polymerase and 4 μl Note: The figure shows increased susceptibility to template DNA. Sterile distilled water and positive control gentamicin which has been resistant for more than ten strains of E. coli pathotype toxins were used as negative years. and positive controls, respectively. For amplification, the PCR program was set at 950C for 1min followed by 720C Figure 2: Drug susceptibility profiles on ETEC, other for 1 min for 5 cycles; 950C for 1 min followed by 620C for E. coli pathotypes and other enteric bacterial isolates 1min and then 720C for 1 min for 20 cycles; and 720C for from Machakos District Hospital 5 min for final elongation of the amplified DNA product. All E. coli pathotypes had raised MICs to The PCR products were then separated by agarose gel different antibiotics ranging from cotrimoxazole to electrophoresis for 1hour. The DNA was visualized by fluoroquinolones. However, there was reduced MICs to exposing the gel to ultra violet light and photographed on gentamicin. alpha imager gel documentation machine. Antibiotic susceptibility testing: The antibiotic Table 1: Minimum Inhibitory Concentration for 250 susceptibility profiles were performed according to E. coli isolates the Kirby-Bauer disk diffusion method. The following Drug MIC 50 MIC 90 Ranges antibiotics were tested: ampicilin, ciprofloxacin, STX 36 38 0.002 - 32 norfloxacin, nalidixic acid, amikacin, tetracycline, AM 16 16 0.16 - 256 cefuroxime, cotrimoxazole, chloramphenicol and GEN 4 6 0.16 - 256 CHL 24 64 0.16 - 256 gentamycin. We used the standard strain of E. coli (ATCC CIP 0.08 0.16 0.002 - 4 25922) for media quality and disk potency. In addition CXM 16 16 0.16 - 256 we performed Minimum Inhibitory Concentration for the CTX 32 32 0.16 - 32 same range of antibiotics. CAZ 32 32 0.16 - 256 CRO 16 16 0.16 - 256 Results TET 8 24 1.5 - 256

Out of the 300 subjects recruited during the study period, Key: AMC= Augumentin, W= Trimethoprim it was observed that different pathotypes existed (Figure SXT= Sulfamethoxazole, CN= Canamycin, CXM= 1) and the isolates resisted more than four antibiotics Cefuroxime, CIP= Ciprofloxacin, AMP= Ampicilin, C= (Figure 2) among them fluoroquinolones, cephalosporins Chloramphenicol, NA= Nalidixic acid, TE= Tetracyclin, and amino glycosides (Figure 2, Tables 1, 2 and 3). CAZ= Cefetazidine.

61 East and Central Africa Medical Journal 2015; 2: 60-63 Shigellae isolates showed the highest raised and chloramphenicol was observed. The results are minimum inhibitory concentrations compaired to ETEC similar to those from other studies [2,6-8]. However, the and other E. coli pathotypes especially on cephalosporins current study differs from other studies in that the isolates (P<0.05, an indication of limited options in the treatment were sensitive to gentamycin while in the other four to reduce the shading of resistant strains from people studies the organisms were shown to be resistant. The infected with different Shigellae strains. current study also found 4% resistance to ciprofloxacin whereas Sang et al. [7] and Kariuki et al. [2] reported Table 2: Minimum Inhibitory Concentration ranges for zero resistance. The data also indicate that tetracycline, shigellae isolates ampicillin, augmentin, trimethoprim/sulfamethoxazole Drug MIC 50 MIC 90 Ranges and chloramphenicol are least likely to be effective in Contrimoxazole >2/38 >4/38 0.002 - 32 the treatment of infections due to diarrhoea in Machakos Amikacin >16 >16 0.16 - 256 County. Similar observations have been reported from different regions in other parts of this country [2,6,7]. Gentamicin 4 8 0.16 - 256 Similarly, high proportions of multidrug-resistant E. coli, Chloramphenicol >16 64 0.16 - 256 Shigella, Salmonella Klebsiella isolates in children and Ciprofloxacin 0.16 0.4 0.002 - 4 adults have also been reported from Lagos and Ogun states in Nigeria [1,9], Accra region of [10] and Cefuroxime >16 >16 0.16 - 256 Lima in Peru [11]. In contrast to the studies in Ghana, Cefotoxime > 8 32 0.16 - 32 the Kenyan isolates were not only resistant to the drugs Ceftazidime 8 32 0.16 - 256 mentioned but also to quinolones and third generation cephalosporins. Tetracycline 32 64 1.5 - 256 Susceptibility testing of E. coli isolates revealed high resistance to the locally used antibiotics comparable The table shows raised minimum inhibition concentration to those shown by the closely related Shigellae. This to different antibiotics by 67 isolates of Shigella spp. underlines the usefulness of E. coli as a surveillance tool for infections due to gram negative bacteria. A good Table 3: Minimum Inhibitory Concentration ranges for percentage of E. coli isolates 507/1450 (34.9%) from 52 Salmonellae isolates patients belonged to diarrhoeagenic strains and possessed Drug MIC 50 MIC 90 Ranges virulent genes. More than half of the E. coli isolates in Contrimoxazole >2/38 >4/38 0.002 - 32 this study presented as normal stool E. coli since bacteria Amikacin >16 >16 0.16 - 256 isolated from both children and adults as nonpathogenic Gentamicin 4 8 0.16 - 256 E. coli were also resistant to the same antibiotics. Chloramphenicol >16 64 0.16 - 256 Moreover, both diarrhoeagenic and commensal Ciprofloxacin 0.16 0.4 0.002 - 4 resistant E. coli may constitute a potential reservoir for resistance genes that can be transmitted horizontally to Cefuroxime >16 >16 0.16 - 256 other bacteria. The high proportions of resistant bacteria Cefotoxime > 8 32 0.16 - 32 particularly those resistant to tetracycline, which is Ceftazidime 8 32 0.16 - 256 generally not used in children less than 5 years indicate Tetracycline 32 64 1.5 - 256 the acquisition of resistant bacteria by the children rather than resistance induced through antimicrobial treatment. The table shows raised minimum inhibition concentration The findings from the current study agree with those to different antibiotics by Salmonella spp. from a population-based study, which indicate that

All the isolates showed increased minimum children acquire resistant E. coli isolates from household inhibition concentrations at both MIC50 and MIC90 contacts [12, 14]. Resistance to β-lactam antibiotics or indicating that 50% or 90% of the microorganisms could chloramphenicol was observed more frequently among be inhibited respectively. This indicates a possibility of isolates obtained from infants below 5 years when limited choice in antimicrobial agents for management of compared with older children that were between 6-10 and bacterial diarrhoeal diseases especially in the curtailing 11-16 years. The association was statistically significant of shedding of enteric pathogens. (X2 = 3.38; P<0.05) and this group may be representative for the overall harboring of resistant enteric bacteria by Discussion children in the County. These data suggest that infants From the current study, multiple antimicrobial resistance may acquire the commensal enteric flora from their patterns of the Shigella spp, E. coli and Salmonella parents who have likely been more exposed to antibiotics spp to the drugs most frequently used in Kenya that is, than their older children. Notably, a high prevalence of amoxicillin / ampicillin, trimethoprim/sulfamethoxazole resistant E. coli has been shown for adults from other

62 East and Central Africa Medical Journal 2015; 2: 60-63 parts of Africa [12-16]. With increasing age, the children 4. Niyogi S. Increasing antimicrobial resistance—an may lose some of the resistant strains acquired originally. emerging problem in the treatment of shigellosis. This view was supported by findings from older children Clin Microbiol Infec. 2007; 13: 1141-1143. between the ages of 11-16 years for whom E. coli, Shigella 5. Galland JC, Hyatt DR, Crupper SS and Acheson spp and Salmonella spp showed reduced resistance. DW. Prevalence, antibiotic susceptibility, and The data also indicate that the commonly used diversity of Escherichia coli O157: H7 isolates from antibiotics can no longer be considered as first- line a longitudinal study of beef cattle feedlots. Applied treatment options for infections due to E. coli, Shigella Environ Microbiol. 2001; 67: 1619-1627. and Salmonella in people with diarrhoeal illness in 6. Kariuki S, Revathi G, Kariuki N, Muyodi J, Mwituria J, et Machakos County. In addition, 18/300 (6%) isolates al. Increasing prevalence of multidrug-resistant non- produced extended-spectrum β-lactamases. Similar typhoidal salmonellae, Kenya, 1994–2003. Intern J findings have alsoee b n reported in by Antimicrobial Agents. 2005; 25: 38-43. Gangoue [16], though their figures were 2% higher than 7. Sang W, Oundo J, Mwituria J and Waiyaki P. those reported in the current study. Multidrug-resistant enteroaggregative escherichia Conclusion and recommendations coli associated with persistent diarrhoea in Kenyan children. J Diarrh Dis Res. 1997; 15: 190-192. Due to the high prevalence of resistance to most of the 8. Sang W, Saidi S, Yamamoto H, Ezaki T, Iida T, et al. antibiotics used in Machakos County, their continued use Haemorrhagic colitis due to Escherichia coli O157: may not be appropriate for the treatment of infections H7 in Kenya. J Trop Pediatr. 1996; 42: 118. caused by E. coli, Shigellae or Salmonella. Other 9. Opintan J and Newman MJ. Distribution of effective drugs should be identified and adopted for use serogroups and serotypes of multiple drug resistant on a regular basis. The data underlines the importance Shigella isolates. Ghana Med J. 2007; 41: 8. of regular antimicrobial surveillance in district hospital 10. Mills-Robertson F, Addy ME, Mensah P and Crupper settings such as Machakos. Consequently, the required SS. Molecular characterization of antibiotic resistance laboratory infrastructureand protocols for surveillance of in clinical Salmonella typhi isolated in Ghana. FEMS resistance must be established, monitored, evaluated and Microbiol Letters. 2002; 215: 249-253. sustained. Since over-the-counter sale of commonly used 11. Rivera F, Ochoa T, Maves R, Bernal M, Medina A, antibiotics without prescription is one of the contributing et al. Genotypic and phenotypic characterization factors for the spread of resistance, the practice must be of enterotoxigenic Escherichia coli strains isolated curtailed if the antibiotic arsenal available to physicians from Peruvian children. J Clin Microbiol. 2010; 48: is to be effective. 3198-3203. Acknowledgements 12. Harris AM, Chowdhury F, Begum YA, Khan AI, Faruque AS, et al. Shifting prevalence of major To the Centre for Microbiology Research, Kenya Medical diarrheal pathogens in patients seeking hospital care Research Institute staff for their cooperation during the during floods in 1998, 2004, and 2007 in Dhaka, study, the MoH Machakos District Hospital for allowing Bangladesh. The Amer J Trop Med Hyg. 2008; 79: us to carry out the study in the hospital and Director, 708. Kenya Medical Research Institute for allowing 13. Diniz-Santos DR, Santana JS, Barretto JR, us to publish this work. Andrade MGM and Silva LR Epidemiological and Conflict of interest: No conflict of interest is declared. microbiological aspects of acute bacterial diarrhea References in children from Salvador, Bahia, Brazil. Brazilian J Infect Dis. 2005; 9: 77-83. 1. Cohen ML. Epidemiology of drug resistance: 14. Bopp CA, Brenner F, Fields P, Wells J and Strockbine N. implications for a post—antimicrobial era. Science. Escherichia, Shigella, and Salmonella. Manual Clin 1992; 257: 1050-1055. Microbiol. 2003; 8: 654-671. 2. Kariuki S, Revathi G, Gakuya F, Yamo V, Muyodi J, 15. Newman M and Seidu A. Carriage of antimicrobial et al. Lack of clonal relationship between non-typhi resistant Escherichia coli in adult intestinal flora. Salmonella strain types from humans and those West Afr J Med. 2001; 21: 48-50. isolated from animals living in close contact. FEMS 16. Gangoué-Piéboji J, Bedenic B, Koulla-Shiro S, Immunol Med Microbiol. 2002; 33: 165-171. Randegger C, Adiogo D, et al. Extended-spectrum-β- 3. Sang WK, Oundo V and Schnabel D. Prevalence and lactamase-producing Enterobacteriaceae in Yaounde, antibiotic resistance of bacterial pathogens isolated Cameroon. J Clin Microbiol. 2005; 43: 3273-3277. from childhood diarrhoea in four provinces of Kenya. J Infec Develop Countries. 2012; 6 (7):572-578.

63 East and Central Africa Medical Journal 2015; 2: 60-63 Original Research

Antimicrobial resistance patterns among E. coli isolates from children presenting with diarrhoea at a cosmopolitan hospital in Kenya

Kangethe SK1,2, Kiiru J3, Kabiru EW4 Kariuki S3

Abstract diarrhoea were more likely to exhibit resistance to co-trimoxazole than those 1 School of Health Sciences Background: Diarrhoea is a serious with chronic diarrhoea (p<0.003 CI=1.41- Mount Kenya University, infection that kills at least 2 million 5.113, OR=2.68). In addition, isolates P.O.Box 342-01000 Thika, children globally. Children under the age from acute diarrhoea were also more Kenya likely to exhibit combined resistance to 2 of 5 years are particularly predisposed to Department of Pathology, diarrhoea especially if the surrounding SXT/FEP/NA or CIP/ and to at least one School of Medicine, environment is contaminated or polluted aminoglycoside than those from chronic Kenyatta University, with human sewerage. Due to frequent diarrhoea (p<0.024, CI=0.025-0.66, P. O. Box 43844 – 00100, episodes of diarrhoea, children may OR=0.13). Isolates from children who Nairobi, Kenya had used antibiotics 3 months prior to 3 unjustifiably be subjected to antibiotics Centre for Microbiology and this could in turn lead to emergence of sampling were more likely to be resistant Research, Kenya Medical multidrug resistant strains. In this study, to multiple antimicrobials compared to Research Institute, P.O Box we isolated and determined antimicrobial those from children who had not (OR: 0.9; 19464-00202, Nairobi, susceptibility profiles of 384 E. coli 95% CI=0.016-1.22; p

Corresponding author: presenting with diarrhoea at Thika Level resistant to major classes of antimicrobials Mr. Stanley K. Kangethe, 5 Hospital in Kiambu County. such as β-lactams, aminoglycosides and School of Health Sciences Methods: E. coli strains were isolated fluoroquinolones. The study therefore Mount Kenya University, from stool specimen or anal swabs using indicates a need to conduct culture and P.O.Box 342-01000, Thika, conventional techniques and antibiotic susceptibility testing with the view of

Kenya. Email: skangethe@ susceptibilities determined using the determining aetiologic agents of diarrhoea mku.ac.ke disc diffusion technique. Questionnaires and their effectiveness in situations where were used to collect data on history of antimicrobial use is justified. hospitalization, antibiotic use history, access to clean drinking water and toilets. Key words: Antimicrobial resistance, Results: Resistance prevalence was Hygiene, Diarrhoea, Antimicrobials, high for ampicillin, sulphamethoxazole- E. coli

trimethoprim combinations chloramphenicol (23%), amikacin (10%) and kanamycin Introduction (21%). Streptomycin (S) was the Diarrhoeal diseases are leading causes least effective aminoglycoside while of morbidity and mortality worldwide, gentamicin (CN) and amikacin (AK) especially in the developing countries. were effective against at least 70% of all There are an estimated 2 million deaths isolates. Close to 32% of all isolates were per year among children globally, resistant to amoxicillin-clavulanic acid majority of whom are in developing combinations (AMC) while 63% were countries [1]. Antibiotic treatment is resistant to ampicillin. At least 52% of recommended for use in amoebic and the isolates were resistant to more than Shigella dysentery, but not for watery three classes of antimicrobials tested. diarrhoea associated with E. coli Resistance was lowest for ciprofloxacin infections [2]. This recommendation is (4%), nalidixic acid (5%), cefepime (5%), frequently flouted in most developing ceftriaxone (7%) and ceftazidime (7%). countries because many health centers are A total of 47 (12%) of the 384 isolates not equipped to diagnose the aetiologic exhibited the ESBL phenotype. Isolates agent associated with diarrhoea. In such recovered from children with acute countries therefore, treatment of diarrhoea

64 East and Central Africa Medical Journal 2015; 2: 64-69 is generally syndromic and includes use of antibiotics Acute diarrhoea was defined as an abnormally frequent besides nutritional rehabilitation and replacement of fluid discharge of semisolid or fluid faecal matter from the and electrolytes in the patient [3]. Use of antimicrobials bowel, lasting less than 14 days. Chronic diarrhoea was in the absence of Culture and Susceptibility Testing defined as diarrhoea lasting for more than two weeks. (C&ST) exposes patients to two main dangers: - first, Specimens were only collected from children whose there is a high chance of inappropriate antimicrobials parents or guardians gave a written consent. In addition, use, a phenomenon that promotes emergence and specimens were only obtained from patients who had not spread of resistant strains. Secondly, treatment of E. coli taken antibiotics at least one week before the onset and infections can result in adverse reactions especially if during the period of diarrhoea. Using this criterion, we Enterohemorrhagic E. coli (EHEC) strains are implicated recruited 384 children. in the illness. The lysis of EHEC strains in blood after Specimen collection and processing: Stool specimen were collected from patients in sterile plastic containers antimicrobial treatment can release toxins that trigger a and processed immediately. For children who could not potentially fatal immune response [4]. readily produce stool specimen, faecal material was Young children, especially those under the age collected from around the anal opening using sterile moist of 5 years are particularly vulnerable to diarrhoea and cotton swabs. The specimens were plated on MacConkey therefore represent a group frequently subjected to agar and incubated at 37°C for up to 24 hours. A antimicrobial treatment. It is thus not surprising that maximum of 3 colonies were randomly selected per plate studies conducted in many developing countries indicate for each patient and purified further before confirmation that pathogens recovered from stool from children are of species identity using standard biochemical methods as resistant to multiple classes of antimicrobials. Since previously described [5]. The biochemical tests included children who suffer frequent diarrhoea episodes may gram staining, oxidase test, citrate utilization and growth have other underlying conditions including malnutrition, characteristic on Lysine Iron Agar (LIA), Triple Sugar Iron they are likely to suffer frequent bouts of diarrhoea agar (TSI), and motility test on indole ornithine media (Oxoid and they are particularly susceptible to invasive E. coli LTD, Basingstoke, Hampshire, England). infections. Resistance to antimicrobials among intestinal Antimicrobial susceptibility testing: Antimicrobial E. coli is particularly important because, strains from the susceptibility testing was done on Mueller-Hinton agar using gut may find their way into other sites such as blood and the disc diffusion method according to the recommendation of urethra where they cause septicemia and urinary tract Clinical and Laboratory Standards Institute [6]. The following infections. Such infections may therefore necessitate the antimicrobials were used: - ampicillin (10 μg), ceftazidime (30 use of antimicrobials. It is therefore prudent to determine μg), and cefepime ((30 μg). Sulphamethoxazole-trimethoprim resistance profiles of intestinal E. coli regardless of combination is the ration of 30:5.2 μg, amoxicillin- whether they are implicated in diarrhoea or not. The aim clavulanic acid (in the ratios of 20:10 μg) respectively, of this study was to determine resistance profiles among ceftriaxone (30 μg), tetracylines (30 μg), chloramphenicol E. coli strains recovered from children with diarrhoea (30 μg), nalidixic acid (30 μg), ciprofloxacin (5 μg), under the age of 5 years seeking treatment at a large kanamycin (30 μg), streptomycin (30 μg), amikacin hospital in a cosmopolitan setting. We also investigated (30 μg) and gentamicin (10 μg). All the antibiotic discs if there is a relationship between access to clean drinking were obtained from Oxoid (Basingstoke, Hampshire, water, toilets, family composition and carriage of England). In order to identify potential β-lactamase- multidrug resistant (MDR) strains. producers, one plate was inoculated with only β-lactam antibiotics while the second plate contained all other Materials and Methods antimicrobials. The β-lactam antibiotics were placed Study population: This cross-section study obtained adjacent to the amoxicillin/clavulanic (AMC) disc at stool specimen from children younger than five years of inter-disc distances (centre to centre) of 20 mm. A clear age presenting with diarrhoea at a Thika Level 5 Hospital extension of the edge of a cephalosporin disc zone in Kiambu County. towards the AMC (also described as an appearance of The hospital has a bed capacity of 245 in the general a ghost zone) was interpreted as positive for Extended wards and 51 in the paediatric ward but patients may Spectrum Beta Lactamase (ESBL) production. share beds. The outpatient department treats an average Statistical analysis: For the purpose of analysis, of 4200 children per month. Stool or rectal swabs both intermediate and resistant results for antibiotic specimens were taken from diarrhoeic children with susceptibility testing were grouped together as “resistant”. laboratory request. Information was obtained for each Differences in proportion of isolates bearing different patient regarding age, sex, onset of diarrhoea, history of elements was analyzed using the Chi (χ2) tests while treatment and hospitalization, access to clean drinking Fisher’s exact test were used for smaller sample sizes. water, composition of the family and availability and The Odds Rations (OR) and the 95% Confidence Intervals type of toilet facilities at the home of the child. Improved (CIs) accompanying the χ2 tests were determined using toilets were defined as in door flush toilets connected to the approximation of Woolf. The null hypothesis was a sealed septic tank or to the municipal sewerage system. rejected for values of p ≥ 0.05. Statistical analysis was For consistency, diarrhoea was defined as defecation of performed using Statgraphics plus version 5 (StatPoint liquid or semi-liquid stool three or more times a day. Technologies, INC, Warrenton, VA, USA).

65 East and Central Africa Medical Journal 2015; 2: 64-69 Results Key: Amoxicillin-Clavulanic Acid (AMC), Ceftazidime Resistance to antimicrobials: Table 1 shows (CAZ) Sulphamethoxazole-Trimethoprim (SXT), Am- antimicrobial-susceptibility profiles of 384 isolates picillin (AMP), Tetracylines (TET), Ceftriaxone (CRO), analyzed in this study. Resistance to ampicillin, Gentamicin (CN), Chloramphenicol (C), Nalidixic Acid sulfonamides, tetracylines, streptomycin and amoxicillin- (NA), Ciprofloxacin (CIP), Kanamycin (K), Streptomycin clavulanic acid (AMC) was noted in over 30% of (S), Amikacin (AK), and Cefepime (FEP). all isolates (Figure 1). Resistance was also high for chloramphenicol (23%), amikacin (10%) and kanamycin Resistance to β-lactams in relation to other antimicrobials: (21%). At least 52% of the isolates were resistant to This study also revealed that 32% of all isolates were more than three different classes of antimicrobials resistant to amoxicillin-clavulanic acid combinations tested and were thus classified as Multi-Drug Resistant (AMC) while 63% were resistant to ampicillin. Resistance (MDR). Resistance was lowest for ciprofloxacin (4%), to other β-lactam antibiotics was much lower and only 47 nalidixic acid (5%), cefepime (5%), ceftriaxone (7%) and (12%) of the 384 isolates exhibited the ESBL phenotype. ceftazidime (7%). This study also showed that 5% of isolates were resistant to cefepime (FEP). Majority of isolates resistant to β-lactams Resistance (%) to various antimicrobails among isolates analysed were also resistant to sulphamethoxazole and trimethoprim 63 (SXT) and to at least one more class of antimicrobial 56 53 especially tetracycline (TET) and chloramphenicol (C).

37 Streptomycin (S) was the least effective aminoglycoside 32 while gentamicin (CN) and amikacin (AK) were effective 23 21 against at least 70% of all isolates. In addition, all isolates 8 10 7 7 5 4 5 resistant to either nalidixic acid (NA) or ciprofloxacin (CIP) were also resistant to at least a 3rd generation AMC CAZ SXT AMP TE CRO CN C NA CIP K S AK FEB cephalosporin. Such multi-drug resistant isolates were from both children who had taken antibiotics than those Figure 1: Percentage resistance among the 384 E. coli who had not, and resident in both urban and rural settings strains isolates (Table 1).

Table 1: Number (%) of strains from different clinical backgrounds exhibiting combined resistance to multiple classes of antimicrobials History of hospitalization last Antibiotic use Types of diarrhoea 6 months last 3 months Yes. Last 6 No. Last 3 Acute Chronic Yes No months months Number of isolates 338 (%) 46 (%) 158 (%) 226 (%) 174 (%) 210 (%) resistant to SXT 59* 35 91* 31 75* 40 4th generation 5 0 9* 1 7 2 β-lactams (FEB) NA/ CIP 4 7 20* 7 18* 7 ≥1 aminoglycosides 66* 29 87* 44 75* 50 SXT/FEP/ NA/ CIP/ ≥1 1* 7 3 <1 3 <1 aminoglycoside ESBL phenotype 14* 0 18* 8 21* 5 Key: (*) indicates X2 values for dichotomous variables (e.g. chronic vs. acute diarrhoea) that are statisticall significant (p<0.05)

66 East and Central Africa Medical Journal 2015; 2: 64-69 Resistance pattern for acute or chronic diarrhoea, were not previously hospitalized (Table 2). Similarly, and exposure to antibiotics: This study did not seek isolates from children who had used antibiotics in the to establish if indeed the isolates analyzed were the last 3 months before sampling were more likely to be aetiologic agents of diarrhoea. However, our data resistant to multiple antimicrobials compared to those indicates that isolates obtained from acute diarrhoea from children who did not. Our observations indicate that stools were significantly resistant to a broader spectrum majority of the children presenting with acute diarrhoea of antimicrobials than those obtained from children with received β-lactam antibiotics including AMC while chronic diarrhoea. Isolates recovered from children with others received co-trimoxazole syrup. acute diarrhoea were more likely to exhibit resistance Access to a toilet is important in reducing to co-trimoxazole than those with chronic diarrhoea environmental contamination with intestinal tract (p<0.003 CI=1.41- 5.113, OR=2.68 (Table 2) but no pathogens that cause diarrhoea and other illnesses. However, this study shows that resistance patterns of significant difference was recorded in the resistance isolates from children whose family use a pit latrine were to fourth generation cephalosporins (e.g. cefepime) almost similar to those of isolates from children whose or to nalidixic acid and ciprofloxacin among isolates family use improved latrines. This study also shows from urban and rural patients. Strains obtained from that there were no significant differences in resistance children with acute diarrhoea were more likely to exhibit patterns of isolates obtained from children from rural resistance to at least one or more aminoglycosides than or urban settings. Similarly, the resistance patterns of those from children with chronic diarrhoea. Similarly all isolates from children living in single rooms were not the 47 ESBL isolates identified were from stool specimens significantly different from those living in multiple-room obtained from children with acute diarrhoea and so were houses. Practice of treating or boiling water for drinking the isolates from acute diarrhoea. Combined resistance to was higher in the rural areas (66%) than in the urban SXT/FEP/NA or CIP/ and to at least one aminoglycoside areas (13%). However, over 66% of families indicated than those from chronic diarrhoea (p<0.024, CI=0.025- that they treated drinking water by boiling or chlorination 0.66, OR=0.13). while only 13% from urban areas did the same. This Data on isolates from children who were study also shows that the number of family members hospitalized one time or another in the last 6 months were was not related to carriage of MDR strains or episodes likely to be more resistant than those from children who of diarrhoea.

Table 2: Number (%) of strains from different socio-demographic backgrounds exhibiting combined resistance

Resistance Toilet type Family members Type of house Social setting

Single Multiple Pit latrine Improved ≤ 4 > 4 Rural Urban room room

Number of isolates 238 146 234 150 171 213 167 217 SXT 57* 30 55 57 46* 64 59 54 4th generation β-lactam 5 3 3* 7 5 4 (FEB) 13 NA or CIP 10 5* 23 13 11 10 13

One aminoglycosides 68 51 70* 48 53 68 61 62

SXT/FEP/ NA or CIP/ ≥ 2* 1 1 2 1 2 1 2 aminoglycoside ESBL phenotype 12 12 5* 23 16 15 11 13

Key: The symbol (*) statistically significant(p<0.05)

67 East and Central Africa Medical Journal 2015; 2: 64-69 Discussion burden and those from houses where hygiene is poor [13, 14]. Same studies also indicate that young children From the findings in this study a high proportion of were likely to acquire MDR strains from their siblings, isolates from children with diarrhoea are significantly especially from those who attend school [13, 14]. resistant to older generation of antimicrobials such Another study found that children from crowded houses as sulfonamides, tetracylines, chloramphenicol, have lower odds for MDR [15]. One possible explanation trimethoprim and streptomycin but these resistances to our observations is that families with fewer members were lower than those reported from other developing (majority of these being children) are more likely to seek countries [7,8]. Other studies have also reported high treatment and may afford to purchase antibiotics for resistances among strains from children and it is their children than those with many children. If this is therefore apparent that young children, especially those true, it is plausible to assume that children from families with diarrhoea may serve as reservoirs of MDR strains with few members may be significantly more exposed [9, 10]. Although antibiotics are not recommended for to antimicrobials, hence the probabilities of their misuse treatment of E. coli diarrhoea, empiric antimicrobials are high. There was no significant difference in the level interventions are frequently applied in the management of of antimicrobial resistance between the urban and diarrhoea in most developing countries. A previous study the rural children, unlike the findings of a recent study by Brooks et al [11] also conducted locally, showed that that urban health practitioners prescribed fewer, more over 70% of patients with diarrhoea are likely to receive effective antibiotics, and were more likely to prescribe an antimicrobial to which the aetiologic agent is not oral rehydration therapy for bloody diarrhoea [16] susceptible. Such practices are likely to contribute to the hence deduced probabilities of acquiring antimicrobial emergence of MDR strains such as those observed in this resistance. The difference between the two surveys study. Identification of strains with combined resistance could be in the study setting. All the patients in our to β-lactams, fluoroquinolones and aminoglycosides, study came from the same hospital. The catchment area especially in children without a history of antibiotic use for Thika level 5 Hospital extends to the adjoining rural or hospitalization is worrying. Such strain may seriously communities where urban lifestyle may influence the limit the treatment options especially if infections lifestyle in the adjoining rural communities. On the other become invasive. We did not record any evidence that hand, the second study was conducted in two different any of the children had been treated using nalidixic acid settings, one in the urban area hospital and the other one or ciprofloxacin yet 4% of the isolates were resistant in the rural hospital setting. Our observations may be to these antimicrobials. Co-resistance to multiple interpreted to mean that antibiotic use patterns and other antimicrobials suggests that use of a single class of factors that govern emergence and spread of resistance antimicrobials against these isolates could co-select for are similar in rural and urban settings. resistance to other unrelated antimicrobials as previously A study conducted in Kenya by cruz et al [17] reported in a related study [12]. identified several factors that contribute to disease Contrary to related studies, the current study transmission in relation to home and toilet hygiene suggests a positive correlation between use of antibiotics including, not boiling drinking water, not washing hands and carriage of MDR strains [13,14]. It was also observed after using a toilet, not using soap to wash hands after that strains from acute diarrhoea are more resistant than using the toilet, use of dirty and inappropriate toilets those from chronic cases. It would be expected that and lack of toilets,. This study showed that majority of children who suffer acute episodes of diarrhoea are more rural families are likely to treat drinking water using likely to use antibiotics than those with mild diarrhoea chlorination method or boiling than those in urban that is not seen as potentially life-threatening. It is also areas. This is partially because most people living in possible that mothers are more likely to seek hospital urban areas had access to piped municipal water. This treatment for children with acute diarrhoea than those water is assumed to be potable and palatable and many with non-severe chronic diarrhoea. urban dwellers do not bother chlorinating or boiling it. A past study by Seidman et al [15] showed that However, the water pipes burst frequently and sewerage susceptibilities profiles of common gut pathogens from and other contaminants may reach the consumers thereby children in rural and urban settings are generally similar exposing them to potential pathogens [17]. It was also and that in both settings, antibiotics are prescribed for noted that majority of urban families had access to a flush diarrhoea cases. Similarly, the current study shows toilet while most of the rural families use pit latrines that that there were no significant differences between the are deemed less safe for containment of human sewerage resistances of isolates from rural children and those due to seepage especially during floods that disperse from urban settings or between those living in single and human waste exposing people to diarrhoea [17]. multiple-roomed houses. This study also shows that the number of family members in a house was not related Conclusions to carriage of MDR strains or episodes of diarrhoea. Contrary to our findings, other studies reported high This study provided data that indirectly link carriage of episodes of diarrhoea among children living in large MDR strains to various socio-demographic factors and families, those attended by care-givers of low education antibiotic use patterns. While the study does not suggest level, those from parents with high responsibility

68 East and Central Africa Medical Journal 2015; 2: 64-69 possible interventions for reducing diarrhoea and for 8. Bartoloni A, Pallecchi L, Benedetti M, Fernandez C, reducing carriage of MDR strains, the data provided here Vallejos Y, et al. Multidrug-resistant commensal can be vital in formulation of antimicrobial use policies Escherichia coli in children, Peru and Bolivia. Emerg for children and for identifying several issues that can Infect Dis. 2006; 12: 907-913. be addressed with a view to reducing diarrhoea cases in 9. Putnam SD, Riddle MS, Wierzba TF, Pittner BT, young children. Some of these factors include: proper use Elyazeed RA, et al. Antimicrobial susceptibility of antimicrobials, access to proper and clean toilets and trends among Escherichia coli and Shigella spp. provision of clean water for domestic use and potable isolated from rural Egyptian paediatric populations water for drinking. with diarrhoea between 1995 and 2000. Clin Microbiol Infect. 2004; 10:804-810. Acknowledgements 10. Bartoloni A, Cutts F, Leoni S, Austin CC, Mantella A, Guglielmetti P, et al. Patterns of antimicrobial use We would like to thank the management of Mount Kenya and antimicrobial resistance among healthy children University for financing the study, doctors and nurses in in Bolivia. Trop Med Int Health. 1998; 3:116-123. the paediatric wards and outpatient clinics for assisting 11. Brooks JT, Shapiro RL, Kumar L, Wells JG, Phillips- in identifying the study patients and specimen collection. Howard PA, et al. Epidemiology of sporadic bloody We also thank mothers/guardians and their children for diarrhoea in rural Western Kenya. Am J Trop Med agreeing to participate in this study. We are grateful to Hyg. 2003; 68: 671-677. the laboratory technical staff for their invaluable task in 12. Kubin R. Safety and efficacy of ciprofloxacin in pae- specimen collection and assistance in susceptibility tests. diatric patients-review. Infection. 1993; 21: 413-421. The authors declare no conflict of interest. 13. Lietzau S, Raum E, von BH, Marre R and Brenner H. References Household contacts were key factor for children’s colonization with resistant Escherichia coli in 1. Ochoa TJ, Ruiz J, Molina M, Del Valle LJ, Vargas M, community setting. J Clin Epidemiol. 2007; 60:1149- et al. High frequency of antimicrobial drug resistance 1155. of diarrhoeagenic Escherichia coli in infants in Peru. 14. Fornasini M, Reves RR, Murray BE, Morrow AL and Am J Trop Med Hyg. 2009; 81:296-301. Pickering LK. Trimethoprim-resistant Escherichia 2. Handbook IMCI: Integrated Management of coli in households of children attending day care Childhood Illness: WHO and UNICEF; 2006:25. centers. J Infect Dis. 1992; 166: 326-330. 3. Guarino A and De Marco G. Persistent diarrhoea, in 15. Seidman JC, Anitha KP, Kanungo R, Bourgeois AL pediatric gastrointestinal disease. BC Decker: USA; and Coles CL. Risk factors for antibiotic-resistant 2004:180-193. E. coli in children in a rural area. Epidemiol Infect. 4. Kimmitt PT, Harwood CR and Barer MR. Toxin 2009; 137: 879-888. gene expression by shiga toxin-producing Eschericia 16. Beatty ME, Ochieng JB, Chege W, Kumar L, Okoth G, coli: the role of antibiotics and the bacterial SOS etal. Sporadic paediatric diarrhoeal illness in urban response. Emerg Infect Dis. 2000; 6:458-465. and rural sites in Nyanza Province, Kenya. East Afr 5. Behiry IK, Abada EA, Ahmed EA and Labeeb RS. Med J. 2009; 86: 387-398. Enteropathogenic Escherichia coli associated with 17. Cruz MC, Cacciabue DG, Gil JF, Gamboni O, Vicente diarrhoea in children in Cairo, . Scien World J. MS, et al. The impact of point source pollution on 2011; 11:2613-2619. shallow groundwater used for human consumption 6. Clinical Laboratory Standards Institute (CLSI): in a threshold country. J Environ Monit. 2012; 14: Performance standards for antimicrobial 2338-2349. susceptibility testing; 15th informational supplement 18. Kinuthia GK, Gicheru MM, Ngure PK and Kabiru (M100-S15). Wayne, PA; 2007. EW. Lifestyles and practices that enhance malaria and 7. Zaidi MB, Zamora E, Diaz P, Tollefson L, Fedorka- typhoid fever in Njoro District, Kenya. J Community Cray PJ and Headrick ML. Risk factors for fecal Health. 2012; 37:224-233. quinolone-resistant Escherichia coli in Mexican children. Antimicrob Agents Chemother. 2003; 47:1999-2001.

69 East and Central Africa Medical Journal 2015; 2: 64-69 Original Research

Pattern of morbidity and mortality of childhood illnesses at the children emergency room of Abia State University Teaching Hospital, Aba, Nigeria

Okoronkwo NC, Chappjumbo AU Abstract Introduction Department of Paediatrics, Introduction: In order to improve the Abia State University Knowledge of pattern of morbidity and quality of prompt child care and to assess Teaching Hospital Aba, mortality in a health institution is a the efforts towards the realization of Nigeria valuable source of information which Millennium Development Goal (MDG)4, aids policy making and intervention a regular appraisal of the morbidity and Corresponding author: strategies. Carrying out such studies in mortality pattern at our emergency units Dr Nneka C Okoronkwo, is important. Findings of this study can our emergency units will not only assess Department of Paediatrics, inform policy decisions on resource our preparedness for prompt quality care, Abia State University and manpower allocations as well as but also appraises our efforts towards Teaching Hospital Aba, P.O. preventive measures in the bid to reduce achieving the MDG-4 of reducing under Box 21953, Aba, Abia State, morbidity and mortality from childhood 5 mortality rate. Nigeria. Email:nnekaceo@ diseases. Infections and communicable rocketmail.com Objective: To describe the morbidity and diseases have remained top causes of mortality pattern of childhood illnesses as childhood morbidity and mortality in seen at the children emergency room of Africa [1,2]. Pattern of high morbidity Abia State University Teaching Hospital, and mortality in under 5 children in Aba, Nigeria. Nigeria warrants regular review in Methods: This was a descriptive different health facilities [3,4]. Previous prospective study done over a 5 month studies were done mainly on children period. The ages, gender, duration of illness, diagnosis and disease outcome of admitted in the major paediatric wards all the children aged 29 days to 14 years of hospitals [5,6]. With little information that presented to the children emergency from the children emergence room where room were consecutively documented. numerous paediatric emergency cases first Data was organized using SPSS Version report, this study aims at determining the 20 and Excel 2010. pattern of morbidity, trend and outcome of Results: There were 198 patients, 55.1% childhood illnesses treated at the Children of whom were males and 44.9% females. Emergency Room of Abia State University Complicated malaria (50%), sepsis Teaching Hospital, Aba, Nigeria, over a (16.2%) and diarrhoea (6.1%) were the period of 5 months. major causes of admission. Mortality rate within the study period was 9.6%, Materials and Methods discharges 20.7% while 69.7% were transferred to the children’s ward. Of the Patients setting: This was a prospective children who died, 94.8% of the deaths hospital based study conducted over 5 occurred among the 5 years or less. Severe months, from July to November 2008, malaria (40%) and septicaemia (25%) at the Children Emergency Room of the were the leading causes of death. Abia State University Teaching Hospital Conclusion: Under-5 mortality still (ABSUTH), Aba, Nigeria. ABSUTH remains high in our environment with serves as a general/referral centre for malaria infection being the major cause of patients resident in Aba metropolis, and morbidity and mortality. adjoining cities and communities. The

Key words: Pattern, Morbidity,Mortality, Children Emergency Room (CHER) is a Children Emergency Room 12-bed unit, which caters for paediatric

70 East and Central Africa Medical Journal 2015; 2: 70-73 emergencies (excluding trauma cases). Approximately Table 2: Distribution of various illnesses in children 700 children register at the facility for treatment each year admitted in the Children Emergency Room at ABSUTH [7]. It is covered 24 hours daily by a team consisting (n=198) of a paediatric registrar and 2 house officers, supervised Diagnosis No. of (%) by a senior registrar in the morning shifts; and overseen patients by 2 consultants. Patients stay in the CHER between 24- Severe malaria 97 49.0 48 hours or less before being transferred to the major Septicaemia 33 16.7 ward, or discharged. Diarrhoea 12 6.1 All children aged 29 days to 14 years whose parents/ guardians gave signed consent were included in the Uncomplicated malaria 12 6.1 study. Ethical clearance was obtained from the Ethics Bronchopneumonia 11 5.6 Committee of ABSUTH, Aba. A study performa was Meningitis 6 3.0 used to record age, sex, and duration of illness before Acute severe asthma 6 3.0 presentation, presenting complaints, diagnosis and P.E.M* 6 3.0 outcome. Diagnoses were based on clinical assessment and other necessary investigations. Individual cases were AGN** 3 1.5 managed according to standard paediatric management Miscellaneous 12 6.1 protocols. Outcome was classified as ‘Died at the Total 198 100 facility (CHER), ‘transferred to the paediatric ward’ or ‘discharged’. PEM* = protein energy malnutrition, AGN** =Acute Data was analyzed using SPSS Version 20 and Excel glomerulonephritis.

2010. Frequency tables were generated for all major Based on the outcome of illness, 69.7% of the patients variables of interest. Percentages, ratios and means were were transferred to the children’s ward, 20.7% were also calculated. discharged, while 9.6% died (Table 3). Among those that died, 63.2% were males and 36.8% females, infants were Results 73.7%, between 1 and 5 years 21.1% and only one death occurred among those above 5 years (Table 4). A total of 198 children, presenting at the child emergency room, between July and November 2008 were studied. Table 3: Outcome of the admitted patients Distribution of the patients by age and gender indicated Outcome No. of patients (%) that 43.9% aged less than one year, 42.9% between one Transferred to 138 69.7 year and five years, while 13.2% were those more than wards five years of age, median age one year. Distribution by Discharged 41 20.7 gender was 55.1% males and 44.9% females (Table 1). Died 19 9.6 Table 1: Distribution of patients by age and gender Total 198 100 (n=198) Table 4: Disease outcome in patients by age and Age group Gender Total gender (years) Male(%) Female (%) Ages(years) Dead Alive Total < 1 24.2 19.7 43.9 <1 14(73.7%) 73(40.8%) 87(43.9%) 1-5 23.7 19.2 42.9 1-5 4(21.1%) 81(45.3%) 85(42.9%) >5 1(5.3%) 25(14.0% 26(13.2%) > 5 7.1 6.1 13.2 Sex distribution Total 55.1 44.9 100 Female 7(36.8%) 82(45.8%) 89(44.9%) Children were admitted for various types of illnesses Male 12(63.2%) 97(54.2%) 109(55.1%) in which severe malaria (49.0%) recorded the highest number of patients followed by sepsis (16.7%), diarrhoea Severe malaria, septicaemia, meningitis, diarrhoea and PEM were the main death causing diseases in that (6.1%) and uncomplicated malaria (6.1%) respectively sequence (Table 5). (Table 2).

71 East and Central Africa Medical Journal 2015; 2: 70-73 Table 5: The pattern of diseases causing death in the Protein energy malnutrition comprised 3% of the children presenting to the CHER total cases seen in this study. Malnutrition alone or in Diagnosis No. of No. of Mortality combination with other diseases contributed to 4% of the cases deaths (%) total morbidity in a study done at Enugu [9] which is in a nearby state in South-eastern Nigeria whereas morbidity Severe malaria 97 8 42.1 rate of 2.1% was recorded in Benin [19] in mid-western Septicaemia 33 5 26.4 Nigeria. The low prevalence of PEM at our CHER could Diarrhoea 12 2 10.5 be explained by the fact that most malnutrition cases Meningitis 6 2 10.5 were seen at the children outpatient department and later PEM 6 2 10.5 admitted directly into the main paediatric wards. We recorded an overall mortality rate of 9.6%, which Total 154 19 100 was comparable to the findings at Enugu, Ibadan and Discussion Benin, all in southern Nigeria [9, 19, 20]. A recent study repeated at the same centre in Benin showed a much Malaria and septicaemia were the major causes of lower mortality rate of 4.4% as against the earlier Benin morbidity and mortality and that male infants were more findings of 10% [8]. This highlights the need of regular affected, a finding that was comparable to previous related follow up studies, even in the same centre in order to studies [8, 9]. This could be explained by the observations assess our score cards in the fight against child mortality. that many families in Nigeria and certain parts of Africa More than 80% of such deaths occurred among the and Asia have a tendency and attitude to seek health care under-fives which re-emphasizes the need to work for the earlier and more frequently for their male children, than survival of this age group. for the female children [10, 11], besides the male infants being more susceptible to diseases than females due to Conclusion difference in immune response. Similarly, morbidity rate Mortality for the children admitted at the Children (86.8%) and mortality rates (9.6%) were higher in under- Emergency Room at the study hospital was high (9.6%) fives which compares with other similar studies where where severe malaria took a high toll (42.1%). The study more than 80% of the morbidities and mortalities were did not examine the capacity of the health facility to reported among the under-fives [5, 6, 8, 9, 12]. handle such childhood illnesses but the large number of High malaria morbidity and mortality in our study, children presenting with severe malaria was an indicator like in the previous reports [5,6,8,9,12-17] pointed that there was a significant level of delay for the care at a dire need to improve on the disease prevention takers to seek medical attention. Poor health seeking and intervention strategies and the related health behaviours are generally associated with poor health seeking behaviour. Occurrence of severe malaria (49%) awareness of the caregivers, which could also apply compared to the uncomplicated malaria (6.1%) was in this study. There is therefore a need to put in place an indicator that there were delays in medical attention programmes which will address the challenges of MDGs seeking behaviour, which, in turn, could be associated 4 & 5 of reducing under-five mortality. with low level of awareness about disease prevention and interventions. Septicaemia morbidity and mortality were References relatively high and this could also be associated with delayed health seeking behaviour. Bacteria related infection especially those causing 1. Ojukwu JU, Ogbu CN and Nnebe-Agumadu UH. diarrhoea and meningitis are preventable diseases that Post-neonatal medical admissions into the paediatric are associated with poor environmental sanitation. ward of Ebonyi State University Teaching Hospital, This study did not examine the social background Abakiliki: The initial experience and outcome. Niger of the patients but poor sanitation and low levels of J Paed. 2004: 31; 79-86. health awareness are associated with poor social- 2. Menge I, Essamai F, Van Renken D and Anabwani G. economic status [21]. Poor environmental sanitation and Pediatric morbidity and mortality at Eldoret Hospital, surroundings could influence the onset of pulmonary Kenya. East Afr Med J. 1995; 72: 165-169. diseases hence a reasonable number of children suffered 3. Petit PL and Van Ginneken JK. Analysis of hospital from bronchopneumonia in our study. records in four African countries, 1975-1990, with More recent studies reported low mortality rate and emphasis on infectious diseases. J Trop Med Hyg. morbidity which is an indicator that improved health 1995; 98: 217-227. delivery systems and improved health awareness in the 4. Adetunji J. Trends in under-5 mortality rates and the communities could go a long way towards achieving HIV/AIDS epidemic. Bull WHO. 2000; 78: 1206. the MDG-4. Such can only be achieved with total 5. Chapp-Jumbo AUN. Pattern of morbidity requiring collaboration from all stakeholders like Integrated admission into the children’s ward of Abia State Management of Childhood Illnesses (IMCI) and National University Teaching Hospital, Aba. JOMIP. 2003; Program on Immunization (NPI). 4:23-25.

72 East and Central Africa Medical Journal 2015; 2: 70-73 6. Ejike O and Ohanenye OC. Pattern and outcome of 14. Adeyokunnu AA, Taiwo O and Antra AU. Childhood paediatric admissions in the children’s ward of Abia mortality among 22, 255 consecutive admissions in State University Teaching Hospital, Aba; over a two- the University College Hospital, Ibadan. Niger J. year period. Programme of events/Book of Abstracts Paed. 1980; 7: 7-15. th of the 38 Annual General & Scientific Meeting of 15. Elusiyan JBE, Obiajunwa PO, Adejuigbe EA, the Paediatric Association of Nigeria held at Nnewi, Olowu WA, Adeodu OO, et al. Pattern of morbidity rd th Nigeria. 23 -27 January 2007. and mortality among children hospitalized at the 7. Children Emergency Room Annual Report 2003- Obafemi Awolowo University Teaching Hospital, Ile 2007. ABSUTH Records Department. Ife. Niger J Paed. 2009; 36:22-28. 8. Abhulimhen-Iyoha BI and Okolo AA. Morbidity of childhood illnesses at the emergency paediatric unit 16. Hodges M and Williams RA. Registered infants and of the University of Benin Teaching Hospital, Benin under-five deaths in Freetown, from City. Niger J Paed. 2012; 39(2): 71-74. 1987 to 1991 and comparison with 1969- 1979. West 9. Chukwu BF, Chinawa JM, Ikefuna AN and Emodi Afr J Med. 1998; 92: 958. IJ. Pattern and outcome of paediatric medical 17. Muhe L, Byass P, Freiji L, Sandstorm A and Wall S. admissions at the University of Nigeria Teaching A one-year community study of under-fives in rural Hospital (UNTH), Ituku-Ozalla, Enugu: a five year : Patterns of morbidity and public health risk retrospective review (2007-2011). Niger J Paed. factors. Public Health. 1995; 109: 99-109. 2013; 40(4): 354-359. 18. Nathoo KJ, Bannerman CH and Pirie DJ. Pattern of 10. Obi CC. Gender differences in educational oppor- admissions to the paediatric medical wards (1995- tunities: The case of girl-child education in Nigeria. 1996) at Harare Hospital, Zimbabwe. Cent Afr J Afr Econs Bus Rev. 2009; 7:1-12. Med. 1999; 45: 258-260. 11. Kam-lun EH and Edmund ASN. Gender disparity Ayoola OO, Orimadegun AE, Akinsola AK and in paediatric hospital admissions. Ann Acad Med 19. Singapore. 2006; 35: 882-888. Osinusi K. A five year review of childhood mortality 12. Ibeziako SN and Ibekwe RC. Pattern and outcome at the University College Hospital, Ibadan. West Afr of admissions in the Children’s Emergency Room of J Med. 2005; 24: 175-179. the University of Nigeria Teaching Hospital, Enugu. 20. Diakparome MA and Obi JO. The pattern of Niger J Paed. 2002; 29: 103-107. paediatric emergencies in the University of Benin 13. Fagbule D and Joiner KT. Pattern of childhood Teaching Hospital. Niger J Paed. 1980; 7: 43-45. mortality at the Teaching 21. World Health Organization. The World Health Hospital. Niger J Paed. 1987; 14: 1-5. Report: 2003: Shaping the future. WHO, Geneva

73 East and Central Africa Medical Journal 2015; 2: 70-73 Original Research

Determinants of early infant diagnosis and treatment of HIV among exposed infants in informal settlements in Nairobi, Kenya

Makau GM1, Okwara FN1, Oyore JP2 Abstract facility (p = 0.000, OR 0.171, CI 0.065- 0.451), receiving of psychosocial support (P = 0.000, Background: Paediatric HIV infection is OR 0.173, CI 0.075- 0.398), high maternal 1Department of a growing health challenge, especially in knowledge on PMTCT (p = 0.001), mothers Paediatrics and Child sub-Saharan Africa. Most paediatric HIV on ARV therapy (P= 0.017, OR 0.284, Health, Kenyatta infections are perinatally transmitted. Early CI 0.101- 0.795) and mother on ARV University, Nairobi, prophylaxis (p = 0.020, OR 0.306173, Kenya Infant Diagnosis (EID) and immediate CI 0.113- 0.830). Factors associated with 2Department of initiation of treatment minimizes deaths. early initiation of treatment were delivery Community Health, Though EID services are widely available in public health (P=0.005) and receiving Kenyatta University, in Kenya, uptake remains low, especially of psychosocial support (P=0.000). Nairobi, Kenya in disadvantaged communities. Objectives: To evaluate determinants of Conclusion: Knowledge of on PMTCT Corresponding author: EID and early treatment initiation among and EID was low, and this lead to low EID Dr F.N Okwara, P.O. HIV exposed children from informal uptake. We recommend integration of Box 19533-00202, settlements in Nairobi, Kenya. PMTCT and paediatric HIV care services Nairobi, Kenya. Email: Methodology: A descriptive cross to MNCH settings. florenceoringe@gmail. sectional study was done; where HIV com infected mother-infant pairs attending Key words: HIV, Infant, Diagnosis, Treatment health care facilities were recruited. Consent was sought. Non-biological Introduction caretakers were excluded. Structured questionnaires were administered to obtain Paediatric HIV infection is a growing information on socio-demographics, health challenge worldwide, with 2 million knowledge and uptake of EID, Prevention children infected by 2010 [1]. Sub-Saharan of Mother to Child Transmission (PMTCT) Africa is home to 68% of all people living and Antiretroviral (ARV) therapy. with HIV. The region accounts for more Results: A total of 238 mother-infant than 70% of new HIV infections globally, pairs were interviewed. Majority, (69.2%) 90% of who are children [2]. Most, were aged below 30 years, 75% had below secondary level of education, (90%) of paediatric HIV infections are 67.6% were married, and 71.4% were of perinatally transmitted. Mother to Child poor social backgrounds. Most (77.4%) Transmission (MTCT) rates in developing had HIV diagnosis made in the preceding countries were reported to stand at 25-40% year, 68.5% of them during pregnancy. [3]. Kenya has experienced one of the Knowledge on importance of EID was world’s harshest epidemics. By 2010, the poor, and only 53.8% had knowledge of National HIV prevalence stood at 6.3%, PMTCT interventions. Only 38.7% had with higher prevalence reported among received ARV prophylaxis in pregnancy, females (8%) [4]. Nairobi County, which while 37.4% were on ARV therapy. is largely urban and the Carpital City, had Majority, (63.5%) had delivered in a the second highest HIV prevalence in the health facility, but only 56.7% had EID at 6 country at 8.2% after Lake Victoria region weeks. While 19.7% of infants tested HIV [5]. Current PMTCT strategies focus on positive, only 10.6% of infected infants initiation of ARV therapy for infected were started on treatment immediately. pregnant women, and subsequent The main determinants of EID at 6 weeks prophylaxis for infant’s. However, were maternal delivery at a public health less than 50% of eligible pregnant

74 East and Central Africa Medical Journal 2015; 2: 74-79 women receive either intervention [5]. Low Discussions (FGDs) with HIV infected women and Key utilization of Ante-Natal Care (ANC) services (47%) and Informant Interviews (KII) with staff at the health centers low rate of hospital deliveries which stands at 44% were provided secondary data. Ethical approval was obtained reported to contribute to this status [6]. from Kenyatta University Ethics Review Committee, the Previous studies indicated that infected children Kenya National Council of Science and Technology, and deteriorate rapidly due to immature immune systems the health facility in-charges. Signed consent was sought [7] and that if untreated, one third of infected children from the mothers after successful explanation of purpose die by the end of the first year, and 50% by the end of of the study. the second year [8,9]. It was further observed that Early Data collected was coded, entered and analyzed Infant Diagnosis (EID) enables infected children to be using Statistical software; SPSS version 19.1. Household identified prior to development of clinical disease and that characteristics are presented as frequencies, percentages, early treatment reduces infant mortality by 76% and HIV means and ratios. Cross tabulations were done and progression by 75% [10]. In a South African study, associations were established by chi-square (X2). 80% of infected well children at 6 weeks progressed Inferential statistics (P-value, confidence intervals,2 X ) to become eligible to ARV therapy by 6-12 months were used to establish factors that predicted EID uptake. of age [11]. Statistical significance level was fixed at P<0.005. Kenyan guidelines recommend EID at 6 weeks Logistic regression was used to test the relationship of and immediate initiation of treatment within 2-4 weeks variances of EID uptake and treatment. of diagnosis to minimize deaths [12]. Notwithstanding, Maternal characteristics: There were a total of 238 by 2010, only 28% of HIV Exposed Infants (HEI) mother – infant pairs enrolled; age range was 18 – 45 were tested within 2 months, and only 23% received years of whom the majority (69.2%) were aged below cotrimoxazole prophylaxis [13]. Efforts were put in place 30 years (Table 1). Most (76.0%) had primary level of to provide wider access of EID services and treatment education or below, 67.6% were married, and 71.4% were for HEI, including those in resource poor settings. These from poor social backgrounds, with monthly household services became availed for free in all public facilities income of less than US $ 22.2. Regarding the timing of [12]. By 2011, 64% exposed infants had EID, and of the HIV diagnosis, 37.4% were diagnosed more than 2 170,000 Kenyan children eligible for treatment, only years prior to the study, 39.9% 1-2 years and 27% were 21% were started on treatment [2] which is still much diagnosed less than one year earlier. Based on the term of below the required figure according to WHO guidelines. pregnancy, HIV diagnosis was made during pregnancy in The objective of this study, therefore was to determine 68.5% of the mothers while 19.7% were diagnosed during the factors that influence uptake of these services in a sickness episode. Regarding their PMCTC, knowledge informal settings in Nairobi, Kenya. level, only 5.9% recorded high levels whereas 53.8% recorded low awareness levels. Only 77.3% respondents Materials and Methods had heard about EID, 55.5% of who heard had about it during pregnancy from a health worker. None reported to This was a cross sectional descriptive study conducted have heard of it from mass media. On importance of EID, at two informal settlements in Nairobi, Kenya. The majority (68.1%) reported that it allowed them to know two health facilities selected offered EID and HIV the child’s HIV status, while 22.7% did not understand treatment services for free. Subject enrollment started its importance. On the treatments recommended for HIV from January 2013 to June 2013. Random sampling of Exposed Infants (HEI), 70.4% of the women knew of HIV infected women attending health care facilities at ARV therapy and only 48.9% knew about cotrimoxazole Child Welfare Clinics (CWC), Comprehensive Care prophylaxis. However, 18.9% were not aware of any Clinics (CCC) or sick child services was conducted. medications required by HEI. Findings on their PMTCT Structured interview schedules were used to get socio- care information indicated that 38.7% received ARV demographics information, HIV diagnosis and treatment, prophylaxis, 37.4% were on ARV therapy and 23.9% PMTCT and EID knowledge and practice. Non-biologic were not on any intervention. Most (63.5%) had delivered caretakers were excluded in the study. Focus Group in a health facility while 26.5% delivered at home.

75 East and Central Africa Medical Journal 2015; 2: 74-79 Table 1: Maternal demographic and heath related Infant characteristics: Out of the 238 study infants, 43.3% characteristics (n=238) were aged below 9 months and the median age was 6 Characteristic ( % ) weeks, and male: female ratio was 1:1. Majority, (96.0%) Age(years) were on exclusive breastfeeding. On postnatal PMTCT interventions received, 65.5% were on ARV prophylaxis, <20 15 while 47.8% were on cotrimoxazole. However, only 21-30 54.2 56.7% of the infants were examined for HIV by dried 23-40 28.6 blood spot at 6 weeks. Diagnosis by PCR HIV DNA tested <40 2.1 positive for 19.7% of the infants. Treatment data indicated Level of education that only 10.6% of the infected infants were started on Non-formal 4.6 treatment immediately. Primary 71.4 Determinants of EID in HIV exposed infants: Socio- Secondary 18.9 demographic factors that were significantly associated Tertiary 5.0 with EID were mother’s age (p = 0.024) and household Marital status income (p = 0.000). Mother’s knowledge of EID (p = 0.000), PMTCT intervention (p = 0.000), and public Married 67.6 facility delivery (p = 0.000) were other statistically Single 12.6 significant determinants. The factors that determined EID Separated/divorced 17.6 is presented in Table 2. Widowed 2.1 Monthly household income(Ksh.) Table 2: Maternal factors that determined early infant <2000 71.4 diagnosis at 6 weeks 2001-5000 16.4 EID at 6 weeks 5001-10,000 9.2 Characteristic No. (%) No. (%) Chi-square >10,000 2.9 Age (years) Maternal PMTCT knowledge <20 22 16.3 14 13.6 X2=9.423 Low 53.8 21-30 82 60.7 47 45.6 df = 3 23-40 28 20.7 40 38.8 p = 0.024 Average 40.3 <40 3 2.2 2 1.9 High 5.9 Level of education Maternal knowledge on importance of EID Non-formal 5 3.7 6 5.8 X2=2.327 To know child’s status 68.1 Primary 96 71.1 74 71.8 df = 3 identiy the HIV infected 8.4 Secondary 25 18.4 20 19.4 p = 0.507 No importance 22.7 Tertiary 9 6.7 3 2.9 Prevent HIV infection in child 0.8 Marital status Knowledge of treatment of HEI Married 13 9.6 17 16.5 X2=5.819 Single 97 71.1 64 62.1 df = 3 ARV 70.4 Seperated/Di- 24 17.1 18 17.5 p = 0.121 Cotrimoxazole 48.9 vorced 1 0.7 4 3.9 Other 5.0 Widowed Time since last HIV diagnosis(Months) Monthly household income(Ksh) 2 <6 5.9 <2000 81 60.0 89 86.4 X =21.589 2001-5000 30 22.2 9 8.7 df = 3 6-12 16.8 5001-10,000 18 13.3 4 3.9 p = 0.000 12-24 39.9 >10,000 6 4.4 1 1.0 >24 37.4 Maternal PMTCT When HIV diagnosis was made Knowledge During pregnancy 19.7 Low 46 34.1 82 79.6 X2=52.981 Average 76 56.3 20 19.4 Sickness/Unwell 19.7 df = 2 High 13 9.6 1 1 p = 0.000 VCT 9.6 Home based DTC 0.8 Maternal PMTCT intervention during PMTCT intervenntion received pregnancy ARV prophylaxis 38.7 ARV prophylaxis 58 56.0 34 26.2 X2=29.478 ARV therapy 37.4 ART 62 45.9 27 40.8 df = 2 None 23.9 No intervention 15 48.1 42 p = 0.000 Place of delivery Place of delivery 2 Home 26.5 Public health 65 48.1 21 20.4 X =33.793 facility 18 13.3 45 43.7 df = 2 Public health facility 36.1 at home 52 38.5 37 35.9 p = 0.000 Private health facility 37.4 Private facility

76 East and Central Africa Medical Journal 2015; 2: 74-79 Infant characteristics associated with EID uptake were and CCC settings in Kenya [15] and concluded that age of infant (p = 0.000), immunization status (p = 0.000), HEI were more adherent to follow-up if they attended ARV prophylaxis therapy (p = 0.000), and cotrimoxazole MNCH rather than CCCs. Other reports observed that prophylaxis therapy (p=0.000). Service delivery lack of integration of EID and HIV treatment services determinants of adherence to EID and treatment of HEI in MNCH setting was a great hindrance to adherence to were belonging to a support group (p=0.000). On logistic EID [16]. Young maternal age was associated with poor regression to determine variables that had a relationship EID uptake [17]. with EID, the factors that showed statistical significance The site of study in the disadvantaged community area were delivery at a health facility (p = 0.000, OR 0.171, of Nairobi City was reported to have a large population CI 0.065 - 0.451), receiving of psychosocial support of people with low levels of formal education or none (p = 0.000, OR 0.173, CI 0.075 - 0.398), high maternal at all [18]. However, low levels of formal education knowledge on PMTCT (p = 0.001), and mothers on ARV or no formal education did not significantly influence therapy (p = 0.017, OR 0.284, CI 0.101- 0.795) and the uptake of infant early diagnosis (P= 0.507) in this mothers on ARV prophylaxis (p = 0.020, OR 0.306173, study. At least 17.8% of respondents were separated or CI 0.113- 0.830). divorced, and these tended to have lower uptake. This Determinants for initiation of ARV treatment: Factors site had much higher separation/divorce rates compared associated with initiation of treatment were mainly to the national separation/divorce rate among women health facility factors. Those who had delivered in public of 6% [6]. Mother’s monthly household income was health facility were more likely to start treatment (p found to be significantly associated with EID uptake. =0.005). Membership to a support group was associated Improvement of household economic stability through with early treatment initiation (p=0.000). From the income generating activities could improve EID uptake FGDs and KII, the factors that constrained adherence by eliminating competing priority interests. to EID included accessibility of the health facilities, Majority, 93.4% of respondents had attended at long waiting times especially at the pharmacy, lack of least one ANC clinic visit and 97% were tested during training of staff on EID and ARV therapy, weak linkages pregnancy but only 38.5% of women accomplished the and communications between children attending MNCH number of visits recommended by WHO [4-6], in order to CCCs for prophylaxis or ART or cotrimozaxole to ensure improved PMTCT knowledge necessary to prophylaxis for HEI. On the other hand, reasons given for improve the EID [12]. Our data is more or less similar to failure to initiate ARV therapy was mainly lack of time the national estimates where pregnant women attendance to take child to the CCC (54.8%), lack of information of a single ANC was 92%, and that of HIV counseling from health care providers (26.0%), and poor counseling and testing was 92% [6]. Regarding the timing of offered by staff (11.9%). Denial of mothers to accept diagnosis of mother’s HIV status, 39.9% had been their own and subsequently infant’s HIV status was also diagnosed in the last 12-24 months, while 5.9% were a common barrier to uptake. diagnosed in the preceding 6 months. This meant that they were diagnosed after the index pregnancy, indicating Discussion missed opportunities to test pregnant women in perinatal In Kenya 170,000 children were eligible to treatment by period. Majority, 53.8% of respondents demonstrated 2011, but only 21% accessed it [2]. Kenyan guidelines low knowledge of PMTCT which could be the cause of recommend EID at 6 weeks and immediate initiation none adherence to EID. Only 76.1% of women received of treatment within 2-4 weeks of diagnosis to minimize a PMTCT intervention, probably due to delayed ANC infant mortality [12]. The 56.7% of adherence observed attendance and low PMTCT knowledge. Some 77.3% of in the current study, which is higher than the national the respondents had heard about EID, 55.5% of whom had average of 35% [14], is yet too low to make a significant heard about it during pregnancy, from a health worker, impact in the reduction of HIV infant mortality rate. which was below the national average of 82% [6]. As regards PMTCT treatments instituted, 65.5% of the The method of delivery of PMTCT could influence the infants were on ARV prophylaxis, while 47.8% were assimilation of the pertinent knowledge to the mothers. on cotrimoxazole. Majority, 93.4% of respondents had Despite having undergone PMTCT intervention, other attended at least one ANC clinic visit and 97% were tested studies observed that most of the mothers had not heard during pregnancy and the rates were more or less similar about EID and only 18.9% respondents had knowledge to the national figures of ANC attendance of 92%, and that of care of HEI [7]. Good maternal knowledge of EID, of HIV counseling and testing of 92% [6]. The improved and initiation of PMTCT interventions in pregnancy EID uptake was attributed to successful integration were both predictors of early EID uptake in our study. of EID to routine Maternal Neonatal and Child Health High delivery at public health facilities were reported (MNCH) activities. Similar findings were reported in an in this study at 73.5% compared to national estimates earlier report that compared service delivery in MNCH of 44% [6]. This was attributed to Output Based Aid

77 East and Central Africa Medical Journal 2015; 2: 74-79 (OBA), reproductive health vouchers that were available (Kariobangi and Babadogo Health Centers) for their time in this community through the provincial administration. and invaluable input during data collection. Delivery in public health facility was associated with References improved EID uptake. This could have been related to better staff familiarization with ARV guidelines. 1. UN joint program on HIV/AIDS (UNAIDS), (2011) Moreover, being a member of support groups promoted - Global plan towards elimination of new HIV adherence to EID, observations was also reported in an infections among children by the year 2015 and earlier report [19]. Two main themes on role of support keeping their mothers alive, Geneva. groups emerged, mainly, the socio-emotional support 2. WHO /UNAIDS/ UNICEF (2011). Global HIV/ received from other members and they also served to AIDS Response: Epidemic update and health sector provide health education information for mothers of HEI. progress towards universal access 2011, Geneva. Infants coming for the 6 week immunizations program 3. UN joint program on HIV/AIDS (UNAIDS (2011) - Global Report on the global AIDS Epidemic: 2010, were more likely to adhere to EID [15,20]. Majority, Geneva. 43.3% of infants were aged between 6-9 months, the age 4. National AIDS Control Council (NACC), (2010): group likely to be attending CWC. In the data on ARV, United Nations General Assembly. Special session on 66.4% of the children were initiated to ARV prophylaxis HIV and AIDS (UNGASS) Country Report- Kenya. at birth, which compares to the national figure of 63% 5. National AIDS and STI control program (NASCOP), [14]. Various constraints expressed by the respondents (2011). Guidelines for antiretroviral therapy in as negative factors influencing EID uptake included Kenya, 4th edition, Nairobi. inaccessibility of health services due to long distances, 6. National AIDS and STI control program (NASCOP), lack of transport fare and long waiting time at the clinic. (2011). National Health indicators for Kenya, This was in line with the previous report that ineffective Nairobi. linkages of MNCH and CCC clinics factored negatively 7. Kenya Demographic Health Survey (KDHS), on EID uptake [21] in which high attrition rates from (2008/9). PMTCT care, high MTCT rates, and late infant diagnosis 8. World Health Organization. Antiretroviral therapy were noted in cases of separate programs for maternal for HIV infections in infants and children: Towards and infant HIV prevention and care services. However, universal access: Recommendations for a Public our study focused on mother–infant pairs attending health Approach, 2010 Revision, Geneva. health care services and may not be a reflection of those 9. Marie-Louise N, Coovadia H, Cortina-Borsa who kept away from these services or declined testing. M, Rollinset N. et al. Mortality of infected and Conclusions uninfected infants born to HIV – infected mothers in Africa: A pooled analysis. The Lancet. 2004; 364 The level of knowledge on PMTCT and in particular (9441):1236 -1243. EID among the study population was low. The main 10. Pendergast A, Tudor- Williams G, Jeana P, Burchett S determinants of EID were maternal age, average and Goulder P. International perspectives, progress, household income, maternal knowledge and practices and future challenges of paediatric HIV infection. on PMTCT and EID. The study identified poor referral Lancet. 2007; 370: 68- 80. systems and lack of integration of PMTCT and 11. Violari AM, Mark FC, Gibb DM, Babiker AG, paediatric CCC as the main service delivery barriers. Steyn J, et al. Early antiretoviral therapy and We recommend integration of PMTCT and paediatric mortality among HIV- infected infants. New Eng J HIV care and treatment services in the MNCH setting as Med. 2008; 359 (21): 2233- 2244. well as establishment of effective linkages and referral 12. World Health Organization. Antiretroviral therapy mechanisms between PMTCT and treatment services. for treating pregnant women and preventing HIV Competing interests infections in infants, Recommendations for a public health approach, 2010 version, Geneva. (WHO/ The authors declare that they have no competing interests. NASCOP, 2011). Acknowledgements 13. Ministry of Health (2010): Prevention of mother to child transmission and paediatric HIV care and We are grateful to the Medical Officers in-charges of the treatment, Joint IATT technical review Mission two health facilities in which the study was conducted Report, Nairobi.

78 East and Central Africa Medical Journal 2015; 2: 74-79 14. National AIDS and STI control program (NASCOP), 18. African Population and Health Research Center (2012). Sentinel surveillance for HIV for HIV and (2008), Nairobi Urban Demographic Surveillance syphilis among pregnant women, Nairobi. System (NUHDSS). 15. Ongech JO, Heather JH, Kose J, Audo M et al. 19. CDC/UNICEF report on paediatric HIV (2008). Provision of services and care for HIV- exposed 20. Rollins N, Mzolo S, Moodley T, Esterhuizen T, and infants: a comparison of maternal and child health van Rooyen H. Universal HIV testing of infants at clinic and HIV comprehensive care clinic models. immunization clinics: an acceptable and feasible J Acq Immune Def Synd. 2012; 61(1): 83- 89. approach for early infant diagnosis in high HIV 16. Cherutich P, Inwani I, Nduati R and Mbori-Ngacha D. prevalence settings. AIDS. 2009; 23(14): 1851- 1857. Optimizing paediatric HIV care in Kenya: challenges 21. Braun M, Kabue MM, McCollum ED, Ahmed S, in early infant diagnosis. Bull World Health org. Kim M, et al. Inadequate co-ordination of maternal 2008; 86 (2): 155-160. PMID 18297171. and infant HIV services detrimentally affects early 17. Amin SH, Sakwa EM, Nabwera HM, Taegtmeyer Infant diagnosis outcomes in Lilongwe, Malawi. J Acq MM, Kimutai RM, et al. Dynamics and constraints Immune Def Syndrome. 2011; 56(5):122-128. of early infant diagnosis of HIV in rural Kenya. AIDs Behav. 2012; 16(1): 5-12.

79 East and Central Africa Medical Journal 2015; 2: 74-79 Original Research

Eclampsia at N’Djamena South District Hospital, Chad: epidemiological aspects and foeto-maternal prognosis

Gabkika BM1, Djongali S2, Buambo G1, Fankep DC2, Tchoubou BM1 Abstract Introduction 1 N’Djamena South District According to 2004 report the maternal’ Hospital,Chad Background: Eclampsia is the second death rate in Chad was estimated to be 2 N’Djamena Mother and cause of maternal death in Chad. It 1089 for every 100,000 living births [1] Child N’Djamena constitutes one of the most difficult and it is the most elevated in the sub- Chad. obstetric emergencies at N’Djamena south Saharan Africa region. Eclampsia was district hospital. Corresponding Author: reported to be the second highest cause Objective: The aim of this work was to Dr. Bray G Madoue, of maternal deaths [2]. Pre eclampsia’ determine the eclampsia’ epidemiologic N’Djamena South complications were also reported to District Hospital, Chad. aspects, and the foeto-maternal prognosis. constitute the most difficult obstetric Email:[email protected] Methods: This was a retrospective emergencies at N’Djamena south district descriptive and analytic survey for one hospital. Such eclampsia related deaths st st year (January, 1 2014 to December, 31 could reduce with improvement of 2014) conducted at N’Djamena south prenatal care services and programs. district hospital. The study sample General lack or insufficient prenatal consisted of two groups: Survey group care services coupled by poor health composed of eclampsia patients; Control facilities and poor treatment make the group in which three women recorded prognosis poor [4]. Death related to after every eclampsia positive case were eclampsia occurs in 0.1 to 10% of the included systematically. Chi-square cases in Chad [5,6]. However, exact 2 (X ) test (p<0.05) was used to compare epidemiological situation of eclampsia variables. in the country is scanty or none existent. Results: The incidence of eclampsia was The aim of this study therefore was to 02.55%. The age of the patients varied determine the epidemiology of foeto- between 15 and 43 years, average 28.1 maternal eclampsia and their prognosis at years. The age group between 15-19 years N’Djamena south district hospital. were more represented 46%. Forty-eight per cent of the survey group have never Materials and Methods done prenatal consultation. The majority of patients (60%) presented eclampsia This was a retrospective descriptive and during pregnancy. Fifty eight per cent of analytic study covering the period of newborns had no abnormality but 10% had one year, January to December 2014. intrauterine death. The mother prognosis The population consisted of pregnant was marked by four maternal deaths (8%). women who gave birth or were admitted Conclusion: The incidence of eclampsia is post-partum in N’Djamena south district increasing in Chad in relation to previous hospital during the period of the survey. reports. Foeto-maternal prognoses are The study sample consisted of two marked by high mortality rate, which calls groups: (i) The study group composed of for increased awareness education and eclampsia patients and (ii) The control improved prenatal care programs. group in which three eclampsia negative pregnant women or newly delivered Key words: Eclampsia Epidemiology, women were recorded after every Foeto-maternal prognosis eclampsia positive case.

80 East and Central Africa Medical Journal 2015; 2: 80-83 Data analysis was done by Epi info 6.0 French. Chi- The majority of patients (60%) were diagnosed with square (X2) test (p<0.05) was used to compare variables. eclampsia during pregnancy. Term of pregnancy and health status: Fifty eight (58%) Results carried pregnancy to full term (superior to 37 weeks) Prevalence of eclampsia: Overall 1957 child birth cases while this was 34 weeks and 36 weeks 6 days for 16%. were recorded during the study period in which 50 (2.5%) Most of the women (78%) were conscious while 10% were eclampsia positive (Table 1). were in a coma. Table 1: Age distribution of eclampsia positive patients Method of delivery: About half (52%) of the women had and none positive controls Caesarean section. Foetal prognosis: In the data on foetal prognosis, 64% Age Eclampsia (+) Eclampsia(-) P-value had a score of Apgar ≥8 at 1st minute while 4% of the group No. (%) No. (%) foetuses scored Apgar ≤ 3 at 1st minute (Table 4). 15-19 23(46) 20(13.3) 0.647 20-24 12(24) 57(38) Table 4: Foetal prognosis 25-29 6(12) 35(23.3) Fœtal prognosis Eclampsia (+) Eclampsi (-) P value 30-34 4(8) 29(19.3) No. (%) No. (%) ≥35 5(10) 9(6) 0.285 Normal 29 (58) 114 (76) Total 50(100) 150(100) Small for term 8 (16) 13 (8.7) 0.275 Prematurity 4 (8) 8 (5.3) Age distribution of the patients varied between 15 Intrauterine 5 (10) 4 (2.7) 0.738 and 43 years, mean 28.1 years in which more women death were recorded in the age group 15 to 19 years. Precocious neo- 1 (2) 5 (3.3) 0.102 Mode of admission and source: More than half of the natal death patients (n=27/50 i.e. 54%) lived in rural zone. Majority Neonatal 3 (6) 6 (4) (76%) of the patients were reffered from peripheral health asphyxia facilities. Data on antenatal survey indicated that 48% Total 50 (100) 150 (100) of the eclampsia positive women did not have antenatal consultations and only 16% had 4 or more antennal Majority (58%) of newborns did not show any consultations (Table 2) as required by WHO. abnormality but 10% of fetus had died intrauterine as compared to 76% without abnormalities and only 2% Table 2: Surveillance of pregnancy and the parity fetal mortality rate in the control group. Maternal prognosis: Eight (8%) in the group of eclampsia Prenatal or Eclampsia No eclampsia (-) P value positive died (three presented a cerebral vascular accident visits (No.) (%) (No.) (%) another one a renal deficiency). In the control group we recorded one death (0.7%, p=0.179). None 24 (48) 23 (15.3) 0.884 1-3 18 (36) 30 (20) 0.083 Discussion ≥4 8 (16) 97 (64.7) Total 50 (100) 150 (100) Eclampsia incidence of 2.5 was recorded in the current Parity Eclampsia No eclampsia P-value study out of 1957 births that took place within a period of one year in N’djamena south district hospital, Chad. The (No.) (%) (No.) (%) rate was higher than those reported in other studies on Uniparas 21 (42) 36 (24) 0.465 eclampsia in Africa [6-9]. Such high incidence could be Pauciparas 16 (32) 64 (42.7) 0.000 associated with the high number of patients visiting the Multiparas 13(26) 50 (33.3) study hospital and its proximity to the rural communities whose record of antenatal visits was poor [6]. Total 50 (100) 150 (100) Adolescence constitutes an important risk factor for onset of eclampsia [10]. Adolescence age group (15- Table 3: Period of contracting eclampsia 19 years) recorded significantly higher incidence of eclampsia than the control counterparts, 46% vs. 13.3% Contracting period No. (%) (p=0.647). This was also higher than that recorded in the Pregnancy 30 60 older age groups, an observation that was also reported in earlier studies [6-11]. Per partum (during delivery) 17 34 Data on parity in relation to incidence of eclampsia Post-partum 3 06 indicated significantly higher incidence (42%) in Total 50 100 primiparous against 24% in the control group (p=0.465).

81 East and Central Africa Medical Journal 2015; 2: 80-83 This was contrary to earlier report that multiparas Acknowledgements constituted more than half of the number of patients with eclampsia [9]. Other studies also in Africa [7, 12, Conflict of interest: All authors have declared that there 13] reported the primiparous as the main high risk group was no conflict of interest. constituting more than half of the eclampsia patients. Funding: No financial assistance or grants were solicited Prenatal consultations play an important role in or obtained during the course of preparing this article. Consent: For this survey we got the permission of the the prevention of eclampsia such that the higher rate director of N’Djamena south district hospital. of prenatal consultations the lower the incidence of eclampsia [4]. We recorded significantly higher number References

(48%) of the eclampsia patients in those who did not 1. Institut National des Statistiques, des Etudes Eco- have any prenatal consultation against 15.3% in the nomiques et Démographiques. (INSEED). Enquête control group (p=0.884). However, this was a reverse of démographique et de santé au Tchad II. Ministère de earlier report in which eclampsia was reported in 48.75% l’économie, du plan et de la coopération.2004 : 414p of patients who had been followed during pregnancy [6]. 2. Memadji M, Dzbernis L, Welfens C, and al. Résultats Such discrepancy could point to other factors which were préliminaires de cause de décès maternels enregistrés à la not considered in either studies including the study set up maternité de Moundou de 2001à 2009.SAGO 2010. or the delivery of services in the relevant facilities. 3. Merger R, Levy L, Melchior R. Césarienne. Précis The findings in this study recorded 8% mortality d’obstétrique Ed Masson, Paris 1995; 597: 415-443. rate from the eclampsia mothers and 0.7% for controls 4. Atade J, Adisso S. L’éclampsie à la maternité du (p=0.179). The rate was higher than that reported in CHDU de Parakou Bénin. Incidence et létalité. various studies conducted elsewhere [8, 13-15] but lower Fondation Genevoise pour la Formation et la than 17.9% reported in Dakar [7]. The high mortality Recherche Médicales. Août 2006. http://www.gfmer.ch/ rate could be attributed to delays to seek medical care Membres_GFMER/pdf/Eclampsie_Adisso_2006.pdf. and to the traditional believes that associate eclampsia 5. Bah AO, Diallo MH, Diallo AAS and al. Hypertension symptoms with witchcraft. artérielle et grossesse : Aspects épidémiologiques et Other clinical signs observed included hypotrophy facteurs de risques. Med. Afr. Noire. 2000; 47(10): in 16% of the new-borns from eclampsia mothers com- 422-425. 6. Ministère de la santé publique. Département des pared to 8.7% in the control mothers (p=0.275) a finding statistiques de santé et de l’information. Annuaire des that was also observed in an earlier study [12]. statistiques de santé au Tchad. 2005; 19: 201-203. For foetal prognosis, significantly high number 7. Cissé. C-T, Faye Dieme M.-E, Ngabo D and al. (10%) of intrauterine death was recorded in the test Moreau. Indications thérapeutiques et pronostic group but only 2.7% in the control group (p=0.738)The de l›éclampsie au CHU de Dakar. J Gynécologie rate of stillbirth was close to that reported in an earlier Obstétrique et Biologie de la Reproduction mai. study [14]. 2003 ; 32(3): 239-245. Significant negative correlation in neonatal death 8. Akpadza K, Baeta S, Kotor KT and al. L’éclampsie rate was recorded in foetuses from the eclampsia cases à la Clinique de Gynécologie Obstétrique du CHU (2%) compared to 3.3% in the control group, (p=0.102). Lomé- Tokoin. Médecine d’Afrique Noire. 1996; The discrepancy could be due to poor services, poor 43(3):166- 169. health facilities and poor access to health facilities from 9. Harioly Nirina MOJ, Rasolonjatovo TY, Andrianirina the rural areas. M, Randriambololona DMA and al. Profil épidémiologique des pré-éclampsies et des Conclusion éclampsies admises à la réanimation des adultes de The prevalence of eclampsia was reported in this study la maternité de Befelatanana. Revue d’Anesthésie- is higher than previously reported. Prognosis for foetus Réanimation et de Médecine d’Urgence. 2009 (July- August); 1(3): 22-24. and the mother was poor and associated with high 10. Beaufils. M, Haddad. B, Bavoux. F. Hypertension mortality rate. Incidence of eclampsia and fœto-maternal artérielle pendant la grossesse: aspects complications reduced with improvement of prenatal physiopathologiques et pronostic à long terme 5-036- surveillance consultations. It is therefore important to A-10. introduce early the prenatal education for awareness of 11. Lansac J, Magnin G, Senthilhes L. Hypertension et eclampsia related symptoms like severe headaches, fuzzy grossesse. Traité d’obstétrique-6e Ed Masson, paris vision, and generalized oedema. 2013; 197-209.

82 East and Central Africa Medical Journal 2015; 2: 80-83 12. Samaké BM, Mangané M, Togola M. Les as- 14. Lokossou.A, Avode.DG, Komongui DG and al. pects épidemio-clinique et thérapeutiques de Prise en charge des manifestations neurologiques de l’éclampsie en réanimation au C.H.U. Gabriel la pré- éclampsie sévère et de l’éclampsie par le sulfate de magnésium à Cotonou. Afr J Neurol Sci. 2006; 25(1): Touré. RAMUR Tome 17, n°4-2012 (spécial 78-82. congrès) : Page 10. 15. Villar J, Say L, Gülmezoglu AM, and al. Eclampsia 13. Lankoande J, Toure B, Ouedraogo A, and al. and pre- eclampsia: a worldwide health problem for Les éclampsies à la Maternité du CHU Yalgado 2000. In: preeclampsia. Edited by Hilary Critchley, Ouedraogo de Ouagadougou. Médecine d’Afrique Allan MacLean, Lucilla Poston and James Walker. Noire. 1998; 45(6):359- 402. RCOG Press, London 2003:189-207.

83 East and Central Africa Medical Journal 2015; 2: 80-83 Original Research

Efficacy of total lymphocyte count as a surrogate for CD4 cell counts in HIV-infected adults in the era of higher treatment cutoffs and less funding: a cross-sectional study

Mwenda V1, 2, Njuguna J1, Musa M1 Abstract 1Department of Medicine, Introduction Consolata Hospital Background: Total Lymphocyte Count Nkubu, Meru County, (TLC) has been previously found to be a HIV/AIDS continues to be a major Kenya 2Comprehensive Care suitable surrogate for CD4 counts in the disease burden particularly in resource Center, Consolata management of HIV in resource limited limited settings, and more specifically, in Hospital Nkubu, Meru settings. With the new treatment cutoff of sub-Saharan Africa. In the last few years, County, Kenya 500 cells/mm3, its performance needs to two major events that will have a lasting be further evaluated. impact in sustaining the gains in control of Corresponding author: Objective: To determine the efficacy of the pandemic have taken place. First, the Dr V Mwenda, P.O. Box 205-60200, Meru, Kenya. TLC as a surrogate for CD4 counts in a US government decided that it had reached Email: valmwenda@ resource-limited African setting. a turning point in its emergency response gmail.com Methods: A retrospective, cross-sectional to HIV/AIDS and announced a gradual study was carried out using the medical reduction of its funding to many of the database at the comprehensive care clinic countries benefitting from the President’s for HIV patients in Consolata Hospital Emergency Plan for Aids Relief Nkubu, Meru county, Kenya. (PEPFAR) with final aim of transferring Results: Of the 234 patients included the care of patients back to clinics run in the study, 72.2% were females while by the respective governments [1]. 27.8% were males. 69.2% had a CD4 of Second, in June 2013, the WHO 500 cells/mm3 or less. The mean age was released new guidelines for the 35.8 years (34.0-37.6). The median CD4 management of HIV which raised the count was 427 cells/mm3 (Interquartile minimum cut-off for initiation of Highly range-IQR, 261-676) while that of TLC Active Antiretroviral Therapy (HAART) was 1900 cells/mm3 ( IQR, 1400-2500). from 350 to 500 cells per mm3 [2]. This Correlations between CD4 count and is an unfortunate paradox as more patients TLC (r=0.582, p<0.0001) and CD4 will be enrolled for HAART when the count and age (r=-0.344, p<0.0001) were donor funds supporting the initiative significant while that between CD4 count in Africa continue to dry up. Most of and haemoglobin level (0.046, p=0.484) the benefitting countries may lack both was not. TLC cutoff of 2000 cells/ financial and human resource capabilities mm3 was found to best predict CD4 500 to effectively manage the transition, as cells/mm3 or less with a sensitivity of it has already become apparent in South 78.1%, specificity of 35.9% and Positive Africa [3]. In Kenya, the lowest level of Predictive Value (PPV) of 66.1%. the healthcare system that can effectively Conclusion: TLC still retains some handle these patients is the health centers, usefulness in detecting CD4 counts of less owing to their countrywide distribution than 500 cells/mm3, though not as strongly [4]. However this level of the Kenyan as it performs with lower cutoffs. healthcare system may not have both the personnel and equipments needed in Key words: CD4, Total lymphocyte count, managing HIV patients in the complexity Surrogate, Resource-limited settings of provision of HAART and both screening

84 East and Central Africa Medical Journal 2015; 2: 84-88 and treating opportunistic infections [5]. This includes Data analysis: Categorical variables were described physicians and equipments for checking CD4 counts using percentages while continuous variables were and viral load determinations. Therefore there is need described using mean, median and interquartile ranges for concerted efforts from all stakeholders to develop (IQR). Statistical Package for Social Sciences (SPSS cost-effective, sustainable, accessible and feasible health version 17, Chicago, IL, USA) was used to analyze the systems to support these patients to avoid discontinuation data. Both linear and logistic regressions were used to of care. Total Lymphocyte Count (TLC) has long been determine whether TLC was a predictor of CD4 count in considered as a suitable surrogate marker for CD4 counts the study population. Pearson correlation coefficient was when the latter is not available, especially in resource determined for age, haemoglobin level and TLC against limited settings [6-8]. However, the same has not been CD4 count. demonstrated in the new WHO cutoffs and with the urgency of transition from donor funded to government Results funded care programmes. Our study was aimed at evaluating the TLC cut- A total of 234 patients were included, among which off that would best approximate a CD4 level of 500/ 72.2% were females while 65(27.8%) were males, mm3, as well as checking the relationship between age, mean age was 35.8 years (34.0-37.6)3. Overall 69.2% haemoglobin level and CD4 counts. Unlike CD4 cell of the patients demonstrated CD4 counts of 500 cells/ count determinations, TLC is easier to perform, requires mm3 and less. The median CD4 count was 427 cells/ cheaper equipment and expertise and is part of a total 3 blood count screen that can provide additional valuable mm (IQR; 261-676), median TLC was 1900 cells/mm information. However it is less precise in determining (IQR; 1400-2500), while the median haemoglobin level the actual degree of immune suppression in HIV infected was 13.1 g/dl (IQR; 12.1-14.3), (Table 1). Correlation patients. coefficient r between CD4 count and TLC (r=0.582, p<0.0001), (Figure 1) as well as CD4 count and age Materials and Methods (r=-0.344, p<0.0001) was found to be significant while the correlation between CD4 count and haemoglobin Study setting and participants: The study was carried level (r=0.046, p=0.484) was not statistically significant using the electronic database of patients enrolled for care (Table 2). at the comprehensive care center of Consolata Hospital Nkubu, a faith-based, regional referral hospital in Meru Table 1: Median and interquartile ranges for CD4 count, County, Kenya. For a medical record to be included, the TLC and Hb patient had to be above 18 years of age, enrolled into Variable Median IQR care between 2005 and 2014, and both a CD4 cell count CD4 count(cells/ 427 261,676 and a full blood count had been performed at least once mm3) and entered into the database on the same visit. In cases TLC(cells/mm3) 1900 1400,2500 of multiple laboratory records, the latest one was used Hb (g/dl) 13.1 12.1,14.3 for the study. The criteria of exclusion/inclusion was: Records of patients below 18 years of age, those with Table 2: Correlation coefficients(r) between the TLC, missing demographic data of age and sex, evidence of Age, Hb) and CD4 count opportunistic infection in the preceding three months, as well as entries of CD4 cell counts and full blood Variable r P-value counts performed and entered on different dates. Ethical TLC 0.582 <0.001 approval for the study was granted by the institutional Age -0.344 <0.001 Ethics and Research Committee (ERC/MIN 4/2015). No Hb 0.046 0.484 consent from the patient was necessary since there was no interaction between patients and the research team.

85 East and Central Africa Medical Journal 2015; 2: 84-88 HIV patients [7, 9-12], though not with higher HAART

initiation cutoffs of 500 cells/mm3. This is important since TLC as part of full blood count tests is widely available at an affordable cost in many African settings. It also requires relatively less expertise in performance and interpretation as compared to CD4 count or viral load determinations especially on a large scale. Although this was also the case in our study, the sensitivity, specificity and predictive value were lower than in other studies, which had used lower CD4 cut-offs previously [7,8]. A TLC cutoff of 1500 cells/mm3 would have a considerably high sensitivity (91.4%) in detecting patients with CD4 counts below 500 cells/mm3, but specificity and positive predictive value are quite low (40.2% and Figure 1: Scatter plot depicting the positive correlation 39.5% respectively). Raising the TLC threshold to 2000 between TLC and CD4 count. cells/mm3, the sensitivity falls to 78.1%, meaning ability to detect 8 in 10 patients in need of HAART, with an Table 3 is the sensitivity, specificity, Positive Predictive improved positive predictive value of 66.1%, implying Value (PPV) and Negative Predictive Value (NPV) that in every 10 patients with TLC levels below 2000/ for TLC cutoffs of 1500, 2000 and 2500 cells/mm3 in mm3, at least 6 will have a CD4 count of less than 500 predicting CD4 count of less than 500 cells/mm3. cells/mm3. Further raising the TLC threshold lowers the For a TLC cutoff of 1500 cells/mm3, the sensitivity sensitivity more without any meaningful improvement in and NPV are high (91.4% and 75.8% respectively), the other parameters. To balance between the detection of while the specificity and PPV are both relatively low at patients qualifying for HAART and the burden of over- approximately 40%. When the cutoff is raised to 2000 classifying people as requiring treatment, which includes cells/mm3, the sensitivity falls to 78% while the PPV side effects and increase cost to running of programmes, rises to 66%. For detection of a CD4 level of less than TLC cutoff of 2000 cells/mm3 seems to be the most 200 cells/mm3, a cutoff of 1000 cells/mm3 has a high suitable to predict CD4 counts of below 500 cells/mm3. NPV (95.5%) while that of 1500 has specificity, PPV and The correlation coefficient between TLC and CD4 NPV of 91.5, 61.1 and 75.8 % respectively (Table 4). count was strongly positive (r=0.582, p<0.001); similar results reported in other studies [2,7,13]. However in Table 3: Sensitivity, specificity, PPV and NPV for our study, the correlation between haemoglobin level various TLC cutoffs to detect CD4 count of 500 cells/ and CD4 count was weak (r=0.046, p=0.484), contrary mm3 or less to the findings elsewhere [14]. Such could probably be TLC cutoff Sensitivity Specificity PPV NPV due to the fact that our study included patients already (cells/mm3) (%) (%) (%) (%) on care whether on HAART or not, with a high chance 1500 91.4 40.2 39.5 75.8 of continuing nutritional and medical care as opposed to 2000 78.1 35.9 66.1 58.3 new patients with coexisting burden of other infectious 2500 67.9 26.0 77.2 18.1 diseases and poor nutrition. The correlation between age and CD4 count in our study was significant (r=-0.344, Table 4: Sensitivity, specificity, PPV and NPV for p<0.001), contrary to findings elsewhere [7]. various TLC cutoffs to detect CD4 counts of 200 cells/ For CD4 less than 200 cells/mm3, which still is 3 mm or less an important consideration in opportunistic infection TLC cutoff Sensitivity Specificity PPV NPV prophylaxis, a TLC cutoff of 1500 cells/mm3 displayed a (cells/mm3) (%) (%) (%) (%) sensitivity of 31.4%, specificity of 91.5%, PPV of 61.1% 1000 52.6 12.1 27.8 95.5 and NPV of 75.8%. TLC cutoff of 1400 cells/mm3, was 1500 31.4 91.5 61.1 75.8 also reported elsewhere but with a higher sensitivity of 73% [9]. Other studies conducted in East Africa indicated Discussion better performance with higher cutoffs for detecting CD4 Several studies conducted elsewhere reported that counts below 200/mm3. A TLC level of 2100 cells/mm3 TLC is a suitable surrogate marker for CD4 counts in was shown to best predict CD4 count of <200 cells/mm3

86 East and Central Africa Medical Journal 2015; 2: 84-88 (sensitivity 83%, specificity 77%, PPV 92%, NPV 57%) Conflict of interest: The authors declare no conflict of as opposed to that of 1200 cells/mm3 reported in a study interests. from Uganda [15]. Sensitivity, specificity, PPV, NPV References improved to 81%, 90%, 90% and 80% respectively with the raising of the TLC cutoff to detect CD4 count 1. Ingrid TK, Ingrid VB and Alexi AW. PEPFAR in <200 cells/mm3 from 1200 to 1900 cells/mm3 in a study transition: implications for HIV care in South Africa. conducted in Kenya [10]. However, our study did not N Engl J Med. DOI: 10.1056/NEJMp1310982. test the performance of these cutoffs above 1500 cells/ 2. Consolidated guidelines on the use of antiretroviral 3 mm . The higher cutoffs required in African settings as drugs for treating and preventing HIV infection: compared to WHO suggested cutoffs was suggested recommendations for a public health approach. to be possibly due to background burden of infectious 2013 revision. Geneva, World Health Organization, disease, because WHO recommendations were largely 2013.(http://www.who.int/hiv/pub/guidelines/ driven by findings from the western countries [16]. artadultguidelines.pdf, accessed April 2015. Some studies observed TLC to be an imprecise 3. Kavanagh M. The politics and epidemiology of surrogate marker for predicting CD4 counts, but advised transition: PEPFAR and Aids in South Africa. J Acq that it can still be used in resource limited settings Immun Def Synd. 2014; 65: 3.1 with no cheaper or feasible alternative, like Kenya 4. Kenya‘s Health Policy Framework: The Second [17]. This is even of utmost urgency as Governments National Health Sector Strategic Plan (NHSSP II), in sub-Saharan Africa prepare to take full control of 2005-2010. HIV treatment programmes in their respective countries 5. Reversing the trends: Norms and standards for health [1], remembering that even with massive support from service delivery. Health sector reform secretariat, donors like PEPFAR, more than half of these countries Kenya Ministry of Health. www.hsrs.health.go.ke, still reported ART coverage of less than 50% in 2010 assessed, June 2014. [18]. Since our study included patients on HAART and 6. WHO approved draft treatment guidelines for HIV- those not on HAART, it provides useful preliminary infected people in resource limited settings. In: The information to not only initiate HAART but also to Hopkins HIV report. The John Hopkins University monitor patients already on treatment. However, an Aids Service 2002; 14:1-4. important weakness of the study was the fact that it was 7. Karanth S, Rau N, Gupta A, Kamth A, Shanbhoque V cross-sectional in nature therefore could not give more and Pruthvi B. Utility of total lymphocyte count as a information on the behaviour of TLC as CD4 counts surrogate for absolute CD4 count in the adult Indian change in the course of treatment. HIV population: A prospective study. Avicenna J Conclusion and recommendations Med. 2014; 4:1-4. 8. Mwamburi D, Mayurika G, Fauntleroy J, Gorbach S The transition from donor funded to government driven and Wanke C. Predicting CD4 count using total HIV treatment programmes in sub-Saharan Africa has lymphocyte count: A sustainable toll for clinical already started. With the state of health care systems decisions during HAART use. Am J Trop Med Hyg. in most of these countries, it is highly likely that the 2005; 73(1): 58-62. transferred patients will be devolved to level three or 9. Kumarasamy N, Mahajan A, Flanigan T, Hemalatha health centers for continuity of care. This level of the R, Mayer K, Carpenter C, Thyagarajan S and health system is characterized by considerable country- Solomon S. TLC is a useful tool for the timing of wide distribution but with meager financial, equipment opportunistic infection prophylaxis in India and and human resources. It is imperative that the relatively other resource constrained settings. J Acq Immun Def junior clinical staff in these centers is supported with Synd. 2002; 31: 378-383. cheaper and easy to use tool in readiness for this huge 10. Gitura B, Joshi M, Lule G and Anzala O. Total number of patients from the donor clinics and also lymphocyte counts as surrogate marker for CD4+ T occasioned by the revising upwards of CD4 cutoffs for cells count in initiating HAART at Kenyatta National initiating HAART. Our study indicated that a TLC cutoff Hospital, Nairobi. East Afr Med J. 2007; 84: 466- 3 of 2000 cells/mm is efficacious in this respect and the 472. same needs to be corroborated in further, larger studies 11. The President’s Emergency Plan for Aids Relief, with the new higher WHO treatment cutoffs of CD4 (PEPFAR), Next Generation Indicators Reference 3 counts below 500 cells/mm . Guide. (http://www.pepfar.gov).

87 East and Central Africa Medical Journal 2015; 2: 84-88 12. Flanigan T, Mahajan A and Kumarasamy N. TLC 15. Kamya M and Semitala F. TLC of 1200 is not a as a surrogate for CD4 count to initiate and monitor sensitive predictor of CD4 lymphocyte count among HAART in resource limited countries. 9th Conference patients with HIV disease in Kampala. Uganda. Afr on Retroviruses and Opportunistic Infections, Seattle, Health Sci. 2004; 2: 941-101. February 2002. 16. Githinji N, Obimbo E, Nderitu M, Wamalwa D and 13. Akanmu A, Akinsete I, Eshofonie A, Davies A and Ngacha D. Utility of total lymphocyte count as a Okanny C. Absolute lymphocyte count as surrogate surrogate marker for CD4 counts in HIV-1 infected for CD4+ cell count in monitoring response to children in Kenya. BMC Infect Dis. 2011; 11:259. antiretorviral therapy. Niger Postgrad Med J. 2001; 17. Daka D and Loha E. Relationship between total 8:105-111. lymphocyte count (TLC) and CD4 count among 14. Spacek L, Griswold M, Quinn T and Moore R. people living with HIV, Southern Ethiopia: A Total lymphocyte count and hemoglobin combined retrospective evaluation. AIDS Res Ther. 2008; 5:26 in an algorithm to initiate the use of highly active 18. Antiretroviral Therapy Coverage in sub-Saharan antiretroviral therapy in resource limited settings. African countries, 2010. WHO/UNAIDS/UNICEF AIDS. 2003; 17:1311-1317. Global HIV Response, 2011.

88 East and Central Africa Medical Journal 2015; 2: 84-88 Original Research

Knowledge, practices and perception on trachoma and its influence on health seeking behaviour of the pastoralist patients in Kajiado Central Division, Kenya

Munguti PN1, Ng’ang’a Z1 , Muttunga J2 1Jomo Kenyatta University Abstract and poor practices are important risk of Agriculture and factors that sustain the infection in Kajiado Technology, Department Background: Trachoma is an infection of Central Division in Kajiado County. There of Public Health, P.O. Box the eyes caused by bacterium Chlamydia is a dire need to enhance and strengthen 62000 – 00200, Nairobi, trachomatis and it is a major cause of the existing programmes on prevention Kenya blindness in the world especially in and control of the infection. 2Kenya Medical Research developing countries. It is endemic in over Institute, P.O. Box 54840- 30 districts in Kenya including Kajiado. Introduction 00200, Nairobi, Kenya This study aims at assessing the influence of knowledge, practices and community Trachoma is an infection of the eyes caused Corresponding author: perception on health seeking behavior of by bacterium Chlamydia trachomatis Peninah N. Munguti, Jomo patients and caretakers of children. that results in blindness after repeated Kenyatta University of Objectives: To determine the knowledge, infections. Currently the disease is a Agriculture and Technology, disease preventive practices, perception public health concern especially in Kenya Department of Public on trachoma and the factors that influence where it is endemic in over 30 districts Health, P.O. Box 62000 health seeking behaviour of the pastoralist located in dry areas with scarcity of water, – 00200 Nairobi Kenya. patients in Kajiado Central Division Kajiado District included [1]. The disease Email:peninah_munguti@ Methods: This was a descriptive cross- is characterized by swelling of the eyelids, yahoo.co.uk. sectional study conducted in Kajiado sensitivity to light and eventual scarring Central Division in 2011 among of the conjunctivae and cornea of the 230 children aged 0-17 years old eyes and spreads easily from an infected with infectious trachoma. Structured person to un-infected person and disease questionnaire interview administered is categorized mainly as active/infectious, to the caretakers of the minors (parent/ affecting minors and blinding trachoma in guardian), focused group discussion and adults [1]. key informant interview guide were used A previous study reported trachoma to obtain the survey information. prevalence of 17.4% for infectious Results: Data for a total of 230 caretakers trachoma and 3.3% for blinding trachoma with children aged 0 –17 years with in Kajiado county in 2009 [2] which was infectious/active trachoma was analyzed. high compared to WHO manageable Health care seeking behaviour by the levels of 10% and 1%, respectively [1]. caretakers was significantly associated The aim of the study was to determine with knowledge on linkage between the knowledge, practices and perception trachoma and animals (AOR=2.80; 95% of trachoma and its influence on health- CI: 1.07 – 7.33; p=0.036), spending 2 hours seeking behaviour of the patients in or more to the water source (AOR=3.85; Kajiado central division. The findings 95% CI: 1.85 – 7.69; p<0.001), and provide valuable information on the having a clean compound (AOR=4.25; existing gaps, to projects targeting 95% CI: 1.32 – 13.68; p=0.015); A large prevention and elimination of the disease. number of caretakers sought health care within a period of two months or more Materials and Methods from the onset of symptoms. Seeking health care within one month was The study was a cross sectional descriptive significantly associated with; being single survey among trachoma patients with or widowed (AOR=3.62; 95% CI: 1.58 active/infectious in minors aged 1 to – 8.26; p=0.002), not being trained on 17 years in Kajiado central division in trachoma (AOR=50.00; 95% CI: 7.14 – Kajiado central district. The area was 333.33; p<0.001). selected purposively based on the high Conclusion: Poor social economic factors, prevalence of trachoma published in a lack of education and poor knowledge on previous study [2]. Trachoma patients trachoma, poor environmental sanitation were randomly selected from 7 of the 18

89 East and Central Africa Medical Journal 2015; 2: 89-96 infected clusters registered in the division under AMREF analyzed for this study. The ratio of infected minors was trachoma control programme, which included; Enkaroni, 48.7% males and 51.3% females. Social demographic Ortarikati, Emuktani, Kiroyian, Oltanai, Paranai and characteristics of the caretakers of the infected minors Oletepesi [2]. (parents and guardians) were examined, 75.1% were Both qualitative and quantitative data collection married and 22.2% were single. The predominant methods were applied to determine the levels of occupation of the caretakers was housewife (41.6%), and knowledge, practices and perception on trachoma and its some form of business (41.6%). Teaching profession was influence on the health seeking behaviour of the caretakers recorded in 45.8% of the husbands of the care takers while (parents or guardians) of the infected children. The study 33.5% did some form of business. A large number of the population comprised of caretakers of children aged 1 to caretakers (73.1%) did not have any formal education, 17 years with active/infectious trachoma. A structured 24.2% reported to have primary level of education. Over questionnaire, FGD and key informant interview guide half of them (55.2%) earned not more US$28 per month. written in English and translated to Maasai language were used to collect data Table 1: Socio demographic characteristics of trachoma A written informed consent was obtained from the infected children (n=230) caretakers to the minors and adult patients with trachoma Characteristics Variables (%) before administering the tools. AMREF trachoma control health monitors from the area were recruited as research Marital status Single 22.2 assistants and trained on how to collect data prior to the Married 75.1 start of the study. Widowed 2.6 The respondents were asked about their demographics Occupation Housewife 41.6 including; age, gender, education background, marital status, family monthly income, occupation, knowledge, Business 41.6 perception and practices on trachoma and the influence Farmer 7.5 this has on their health seeking behaviour. Teacher 8.0 Data collected in the questionnaires was entered in MS CHW * 1.3 Access and cleaned before transferring to SPSS version 12.0 (statistical Package for Social Sciences). Six FGDs Education level No education 73.1 and ten key informant interviews were done, transcribed Primary 24.2 and categorized thematically based on emerging themes Secondary 1.8 to triangulate the information. Tertiary 0.9 Observational check list method was used to record environmental cleanliness around the compound of Monthly income (US$) <12 14.8 the interviewed households, latrine coverage and use, 12>28 40.4 distance to water sources, cleanliness of the children’s ≥28 -56 19.6 faces and size of the containers used for collecting and >56 20.9 storing water at household level. Approval to carry out the study was sought from Kenya Don’t know 4.3 Medical Research Institute (KEMRI) Scientific Steering *= Community health workers Committee and National Ethical Review Committee. Further approval to carry out the study was sought from Knowledge on trachoma: About two thirds of the the AMREF trachoma control unit in Kajiado, from the respondent caretakers of the infected minors (65.7%) Medical Officer of Health (MOH) Kajiado and from the reported to have heard about trachoma and indicated the district ophthalmologist’s in charge of the study area. main signs to be: red eyes (34.8%), watery eyes (26.5%), Children with infectious trachoma whose care takers and poor eye sight (20.5%). Most reported causes of consented to participate were interviewed trachoma among children were contact with flies (33.5%) and dirty faces (23.0%). Majority (70.0%) reported to Results have a local name for trachoma and 43% reported to have Socio demographic characteristics of the caretakers of a local treatment for trachoma. Children aged less than the infected minors: Data for caretakers of 230 minors 10 years were reported to be the most affected by 81.7% aged between 0-17 years with infectious trachoma was of the respondents (Table 2).

90 East and Central Africa Medical Journal 2015; 2: 89-96 Table 2: Knowledge on trachoma among child care Table 3: Practices and perception of the children takers (n=230) caretakers (n=230) Characteristics Variables (%) Characteristics/Links Variables (%) Heard of trachoma 65.7 Trachoma and flies 31.7 Knowledge of signs Red eyes 34.8 Trachoma/animals 16.1 Water source River 13.9 of trachoma Eye rashes 6.1 Borehole 12.6 Watery eyes 26.5 Well 1.3 Poor eye sight 13.9 Pool 25.2 Don’t know 18.7 Stream 47.0 Causes of trachoma Flies 33.5 Time taken to fetch 30 mins 5.7 Dirty face 23 water 1 hour 14.8 Contact with 1.3 2-4 hours 48.7 infected persons ≥5 hours 29.6 Don’t know 42.2 Amount of drinking 5 litres 1.3 water Have local name for 70 10 litres 1.7 trachoma 20 litres 49.6 Have traditional 43 >20 litres 47.4 treatment Frequency of washing Once 67.4 face Twice 17.8 Most affected by Children<10years 81.7 Thrice 3.1 trachoma Teens 13-17 years 3 None 6.5 Adults (over 18 5.6 Others 5.2 years) Child washed 35.2 Everybody 4.8 Have pit latrine 2.2 Older people 2.6 Times face washed Once 62.6 Don’t know 2.1 Twice 17.0 Thrice 8.7 Practices on trachoma: Data on practices and perception None 5.2 of the children caretakers indicated that 31.7% of the caretakers knew the link between flies and trachoma and Others 6.5 only 16% of them knew the link between trachoma and Waste handling Pit 2.6 animals/livestock (Table 3). Water pools and streams were Latrine 7.8 the main source of water for 72.2% of the respondents for Burn 45.2 which 48.7% spend 2-4 hours at the water source and Bush 44.3 97% fetched it in one 20 liter jerrican per day. Over two thirds of the caretakers washed their face once a day and Trained on trachoma 13.5 35.2% reported to be washing or cleaning their children Clean compound 13.0 of whom 62.6% did it only once a day. The number Rituals for trachoma 6.1 with households pit latrine was only 2.2% and the most Traditional believes 10.9 commonly reported ways of domestic waste disposal was by burning (45.2%) and throwing it in the bush (44.3%). Food restrictions 1.3

91 East and Central Africa Medical Journal 2015; 2: 89-96 Comparison of different factors that influenced the was clean (86.7%) compared to those whose compound caretakers’ choice for health care (Table 4) revealed was not clean (65.0%), (OR=3.50; 95% CI: 1.17 – 10.43; significant increase in the proportion of caretakers seeking p=0.018) and significance difference among those who healthcare at heath facilities among those who spent 2 washed or cleaned their children (77.8%) compared to hours or more to the water source (72.7%) compared to those who did not (62.4%), (OR=2.11; 95% CI: 1.13 – those spending one hour and below (48.9%), (OR=2.78; 3.92; p=0.017). 95% CI: 1.43 – 5.26; p=0.002), those whose compound Table 4: Comparison of different risk factors in relation to health care seeking behaviour Health Other 95% CI Variables /Links facility(n=156) (n=74) OR p value n % n % Lower Upper Trachoma and flies Yes 49 67.1 24 32.9 0.95 0.53 1.73 0.876 No 107 68.2 50 31.8 1.00 Trachoma and animals Yes 30 81.1 7 18.9 2.28 0.95 5.46 0.060 No 126 65.3 67 34.7 Time to water source <1 hour 23 48.9 24 51.1 1.00 2 hours or more 133 72.7 50 27.3 2.78 1.43 5.26 0.002 Times they wash face Once 103 64.0 58 36.0 0.54 0.28 1.02 0.056 Twice or more 53 76.8 16 23.2 1.00 Have pit latrine Yes 2 40.0 3 60.0 0.31 0.05 1.88 0.331 No 154 68.4 71 31.6 1.00 Compound clean Yes 26 86.7 4 13.3 3.50 1.17 10.43 0.018 No 130 65.0 70 35.0 1.00 Heard of trachoma Yes 100 66.2 51 33.8 0.81 0.45 1.45 0.472 No 56 70.9 23 29.1 1.00 Marital status Single/widowed 35 74.5 12 25.5 1.84 0.88 3.85 0.101 Married 87 61.3 55 38.7 1.00 Gender Female 83 70.3 35 29.7 1.27 0.73 2.21 0.402 Male 73 65.2 39 34.8 1.00 Level of education No education 110 65.1 59 34.9 0.61 0.31 1.18 0.139 Primary and higher 46 75.4 15 24.6 1.00 Trained on trachoma Yes 19 61.3 12 38.7 0.72 0.33 1.57 0.402 No 137 68.8 62 31.2 1.00 Washed or cleaned Yes 63 77.8 18 22.2 2.11 1.13 3.92 0.017 No 93 62.4 56 37.6 1.00

92 East and Central Africa Medical Journal 2015; 2: 89-96 From the multivariate analysis, significance differences (AOR=3.85; 95% CI: 1.85 – 7.69; p<0.001), and having were observed in: the knowledge on linkage between a clean compound (AOR=4.25; 95% CI: 1.32 – 13.68; trachoma and animals (AOR=2.80; 95% CI: 1.07 – 7.33; p=0.015), were significantly associated with where the p=0.036), spending 2 hours or more to the water source caretaker sought health services.

Table 5: Timing for seeking health care at the health facilities in relation to different characteristics <1 month >1 month 95% CI Variables/Links (n=120) (n=110) OR p value n % n % Lower Upper Link between trachoma and flies Yes 25 34.2 48 65.8 1.00 No 95 60.5 62 39.5 2.94 1.64 5.26 <0.001 Link between trachoma and animals Yes 12 32.4 25 67.6 1.00 No 108 56.0 85 44.0 2.63 1.25 5.56 0.009 How long to water source <1 hour 12 25.5 35 74.5 1.00 2 hours or more 108 59.0 75 41.0 4.17 2.04 8.33 <0.001 Times they wash face Once 82 50.9 79 49.1 0.85 0.48 1.49 0.565 Two or more 38 55.1 31 44.9 1.00 Have pit latrine Yes 1 20.0 4 80.0 0.22 0.02 2.02 No 119 52.9 106 47.1 1.00 Compound clean Yes 16 53.3 14 46.7 1.05 0.49 2.28 0.892 No 104 52.0 96 48.0 1.00 Heard of trachoma Yes 81 53.6 70 46.4 1.19 0.69 2.05 0.538 No 39 49.4 40 50.6 1.00 Marital status Single/widowed 32 68.1 15 31.9 1.91 0.95 3.82 0.067 Married 75 52.8 67 47.2 1.00 Gender Female 60 50.8 58 49.2 0.90 0.53 1.50 0.679 Male 60 53.6 52 46.4 1.00 Level of education No education 85 50.3 84 49.7 0.75 0.42 1.36 0.343 Primary and higher 35 57.4 26 42.6 1.00 Trained on trachoma Yes 3 9.7 28 90.3 1.00 No 117 58.8 82 41.2 12.50 3.85 50.00 <0.001 Children washed or cleaned Yes 70 86.4 11 13.6 12.60 6.13 25.91 <0.001 No 50 33.6 99 66.4 1.00

93 East and Central Africa Medical Journal 2015; 2: 89-96 Bivariate analysis of the factors associated with to wash faces of the children and this could attract a host health care seeking behaviour (Table 5) revealed that of flies on the face thus a rise in transmission potentials. significantly high number of those who were likely to Data on knowledge in the current study gave low seek health care in the health facilities within one month results compared to the qualitative report in Turkana of initial symptoms were those who reported not to know which indicated that majority of the respondents could the link between trachoma and flies (60.5%) compared associate trachoma with some of the known causes to those on the affirmative (34.2%), (OR=2.94; 95% like; dirt, flies, dust, lack of water and latrines [5]. CI: 1.64 – 5.26; p<0.001), among those who did not Knowledge of signs and symptoms of trachoma among know the link between trachoma and animals (56.0%) the respondents in our study was low where 26.5% of compared to those who did (32.4%), (OR=2.63; 95% child caretakers did not know signs and symptoms of CI: 1.25 – 5.56; p=0.009). Those who spend 2 hours or trachoma. However, majority of the respondents (70%) more to reach the water source (59.0%) compared to had a local name (Enkoe) for the infection and about half those who spent 1 hour or less (25.5%), (OR=4.17; 95% of them (46%) of the respondents knew local treatment CI: 2.04 – 8.33; p<0.001), those who washed or cleaned (Oseki and Ortikariti) for trachoma. Identification of the their children (86.4%) compared to those who did not disease by local language could translate into increased (33.6%), (OR=12.60; 95% CI: 6.13 – 25.91; p<0.001 and knowledge of the symptoms [5]. No data was collected to those who did not have education on trachoma (58.8%) compare the distribution of infection based on different compared to those who were trained (9.7%), (OR=12.50; social demographic characteristics of caretakers but 95% CI: 3.85 – 50.00; p<0.001). the fact that there was a recognized local name and treatment for the disease was an indication that trachoma Discussion is a recognized problem in the community. It could be of interest to study the effectiveness of local treatment and The finding in this study demonstrated a wide distribution potentials of promoting it to reduce transmission of the of infectious trachoma in Kajiado Division Central infection. despite strong coordinated campaign by AMREF/ Water availability and safety was reported as important Ministry of Health to prevent and control the infection. factors in the transmission of trachoma [10]. Significantly Significantly low levels of knowledge (P=0.045) high number (47%) among the caretakers of the infected perception about trachoma and practices among infected minors collected domestic water from water pools and persons were important factors in the transmission and streams and about half of them collected less than 20 liters sustaining of the infection in the community. The present a day to be shared among the family members (P <0.001). study did not collect data from none infected persons Among the caretakers (49.6%) fetched water in one 20-liter but there were significant links between trachoma low container per day, 48.7% of whom walked for over 2 socio-economic status (P=0.000), long distance to water hours to fetch water for the family. Previous studies in source, inadequate amounts of water used per family, pastoralist communities in Turkana region Kenya [4] washing of face of the children, health seeking behaviour Gambia and Northern Tanzania reported large families absence of latrines and poor practices (p=0.004). as a major risk factor in the transmission of trachoma [6, This study did not consider the number of minors 7]. Such could be associated with the amount of water infected in each household but data from caretakers of available to each family member each day. Long distance 230 children with infectious trachoma was analyzed. According to the caretakers, children less than 10 years to water source, low, inadequate amounts of water used of age were the most infected. Similar observations were per family to clean, lack of latrines and poor hygiene made in a study conducted in Turkana, which recorded behaviour and practices were major contributing factors low knowledge on signs and symptoms of trachoma and to high transmission rate of trachoma [7]. significantly high cases of trachoma among children less Education levels of the caretakers influenced the than 10 years of age compared to other ages [6]. practices towards transmission of trachoma. The practice Two thirds of the respondents among child caretakers of washing face was significantly higher in those with in our study reported to have heard about trachoma, primary and higher level of formal education (61.2%) but four out of ten of them did not know the causes of than those with no education (p= 0.031) in the current trachoma. Flies as a major cause of trachoma were study. The results of this study concur with studies done mentioned by 33.5% and 23% for dirty faces. Knowledge in Gambia and Tanzania that showed a direct association about contact with flies and dirty faces was significantly between improved act of washing of face among children associated with those with primary level of education and adults with trachoma infection [3]. and beyond (39.3% and 28%) compared to those with Caretakers also tended to seek treatment when signs no education (30.9% and 22%) respectively. Only one and symptoms become critical (27%). The husband third of them knew the link between flies and trachoma or partner of respondents was the first contact person (P<0.001). Flies are important agents in the transmission notified when symptoms were noted (52%), followed by of the infectious trachoma [10]. Flies covered faces were a visit to the health facility (33.5%) (P = 0.019). However, common sights in children under five years (personal 20% of them sought treatment after failure of traditional observation). About 35% of the caretakers reported not methods.

94 East and Central Africa Medical Journal 2015; 2: 89-96 Long distance to health facilities could be a deterrent The practices and believes were similar for those with to seeking health attention early. Majority (81.7%) primary education and those with no education. These of the respondents of the caretakers walked for over 2 findings support similar observations in an earlier study hours to the nearest health care centre for treatment. on the traditional believes in the same community [3]. An earlier study in Mali reported a direct association Health seeking behaviour were significantly affected between proximity to health clinics and reduced levels by knowledge on the link between trachoma and animals, of trachoma [8]. the time it took them to a water source, the compound Traditional believes and culture were reported to cleanliness, the frequency of washing their faces and influence health seeking behaviour of the caretakers some presence of pit latrine (P = 0.002). Seeking health care of whom attributed the cause of infection to a generational within a month was significantly associated with their curse, family inheritance from ancestors and a disease of knowledge on the link between trachoma and animals the poor people who live with livestock. Consequently, link between trachoma and flies, time it took them to 10% of the caretakers sought treatment from a traditional water source, cleanliness of their compound, frequency healer (P = 0.009). Our observation could be compared of washing their faces (P<0.001) and if they were trained with the report in a study in Turkana, Kenya which on trachoma. This concurs with a study done in Tigray, associated trachoma to people with livestock where the Northern Ethiopia in 2004 on prevalence and risk factors administrative regions with high prevalence of active in which it was indicated that poor health seeking and potentially blinding trachoma were the main grazing behaviour and lack of access to eye care services could areas for cattle [6]. More studies to establish the role of be responsible for the high prevalence of trachoma in the domestic animals in the transmission of the infection are study districts [9]. needed in the grazing communities. Prolonged delays in care seeking among respondents The caretakers who were able to link trachoma with with infectious trachoma could affect prevention and animals, those taking 2 or more hours to fetch water, control programmes. Early detection and treatment those with no pit latrine, were more likely to seek health reduces the burden of blindness while treatment of care at a health facility. It is not clear how the observed active trachoma reduces the incidence of the infection trend takes place. However, such an explanation could (9). Our study concurred with the Ugandan study, which be related to the intensity of infection, chronicity or advocated behavioural changes that could reduce the constant re-infection due to high risk factors. Self transmission. Rapid spread of the infection was attributed medication could also play a role in the observed delays to the tendency of many community members to seek to seek medical care especially for the enlightened group treatment from traditional healers or self-medications who could be more aware of some other conventional in response to the initial symptoms instead of going to remedies and would be ready to try them before seeking health centres, only to go to the centers when the former medical attention. Failure of such home remedies could failed [4]. resort to exacerbation of the symptoms. This study did not examine the intensity of infection in relation to Conclusion different characteristics of the respondents. Such studies could shed some light on the causes of delays especially Health seeking behaviour against infectious trachoma in the enlightened group in the community. was poor in all the study villages in Kajiado Central Poor sanitation is important in the transmission of Division as a result of which transmission is high. trachoma [10]. Only 2.2% of the caretakers reported to Numerous environmental and human factors contributed have household latrines, which could be attributed to to the transmission and delayed health seeking behaviour. traditional believes that it is a taboo to mix men and women Initial tendency to consult traditional healers deterred faeces and that there is enough bush around for one to hide early medical care attention a situation that could for execrator disposal hence no energy should be wasted play a significant role in the spread of the infection. digging and building latrines. It was established that the Concerted efforts are required to address the problem of few latrines in existence are as a result of organizations’ transmission by all stakeholders especially those directed efforts on awareness on the importance disposing human to behavioural change for early diagnosis and treatment excreta in pit latrines. Further studies could be conducted of the infection. to examine the rate of faecal contamination in the region and the effects not only in the domestic animals but also Declaration in the wild animals including rats, which are common The authors declare that there is no conflict of interest in pests in the human habitat. this work. Traditional believes could influence transmission of infections especially if there are no proven local remedies. Acknowledgements Over 5% of the caretakers reported having rituals related to trachoma such as smoking of the patients eyes with Our gratitude and indebtedness to all those who made burned cow dung and application of “Oseki” and about this study possible. Special thanks to African Medical 1% reported having food restriction such as eggs for and Research Foundation (AMREF), Kenya trachoma trachoma patient’s especially in expectant women. control project manager for sharing valuable statistics and

95 East and Central Africa Medical Journal 2015; 2: 89-96 information on locations with high burden of trachoma 5. Harding-Esch EM, Edwards T, Mkocha H, and to all the study participants for their willingness to Munoz, B, Holland M, et al. Trachoma prevalence take part in the study and for their invaluable contribution. and associated risk factors in Gambia and Tanzania: Baseline results of a cluster randomised controlled References trial. PLoS Negl Trop Dis. 2010; 11: e861. 6. Karimurio, J. and Rono, H. Baseline trachoma 1. World Health Organization (2004). Report of the prevalence survey in the larger Turkana district, Eighth Meeting of the WHO Alliance for the Global Kenya. AMREF. 2010. Elimination of Trachoma. Geneva: World Health 7. Mahande J, Mazingo D and Kweke J. Association Organization, Geneva. between water related factors and active trachoma in 2. Bailey R, Downes, B, Downes, R. and Mabey, D. Hai District, Northern Tanzania. Infect Dis Poverty. Trachoma and water use; a case control study in a 2012; 1:10. Gambian village. Trop Med Intern Health. 1999; 8. Schemann J, Sacko D, Malvy D, Momo G, Traorel 32:58-59. BO, Coulibaly S and Banou A. Risk factors for 3. Karimurio J, Ilako HS, Gichangi, M. and Kilima P. Trachoma in Mali. Int. J. Epidemiol. 2010; 1: 194- Prevalence of trachoma in Kenya. East Afr Med J. 201. 2004; 83(4): 63-68. 9. Taylor and Francis group, LLC. Ophthalmic 4. Cromwella, E. A. Courtrightb, P. Kinga, J. D. Epidemiology. 2006; 13.173-181. Rotondoa, L. A. Ngondic, J. and Emerson, P.M. The excess burden of trachomatoustrichiasis in women: a systematic review and meta-analysis. Trans Royal Soc Trop Med Hyg. 2009; 1183: 1-8.

96 East and Central Africa Medical Journal 2015; 2: 89-96 Original Research

Identification of enterotoxigenic Staphylococcus aureus strains from meat and dairy products by multiplex PCR and reverse passive latex agglutination test in Nairobi, Kenya

Mathenge JM1, Okemo PO2, Ng’ang’a PM3, Mbaria JM4, Gicheru MM1 1 Kenyatta University, P.O. Abstract Staphylococcal enterotoxin B (seb) was Box 43844-00100, Nairobi, Background: Foods of animal origin, the least occurring gene in the S. aureus Kenya especially meat and dairy products, isolates examined (13.9%). Genes 2Department of are sometimes associated with food occurring in pairs included Sea/See Microbiology, Kenyatta borne diseases. In many countries, (21.2%), Sea/Sed (9.8%), Sed/See (2.1%), University, Nairobi, Kenya Staphylococcus aureus is considered the Sea/Sec (0.7%) and Seb/Sec (0.5%) among 3Division of Vector-Borne second or third most common pathogen others. A relatively low number 4 (2.1%) and Neglected Tropical responsible for outbreaks of food of discrepancies between the results of Diseases, Ministry of poisoning. In Kenya enterotoxigenic multiplex PCR and RPLA were found Health, P.O. Box 19982– staphylococcal food poisoning poses a where by the sed genes were expressed by 00202, Nairobi, Kenya. potential health hazard to the consumers PCR but the corresponding toxins were 4University of Nairobi, of meat and dairy products but little data not detected by RPLA. College of Agriculture and is available about the strains involved in Conclusions: The study clearly indicated Veterinary Medicine, P.O. food poisoning. that meat and milk products marketed Box 29053, Nairobi, Kenya Objective: The primary objective of this in and around Nairobi, Kenya were study were to investigate the occurrence contaminated with enterotoxigenic Corresponding author: of different strains of Staphylococcus S. aureus posing a high risk of food Mr. John M. Mathenge, aureus in meat and milk products and to poisoning to the consumers. Equally, these Kenyatta University, P.O. determine the existence of gene coding, data demonstrated that multiplex PCR and Box 43844-00100, Nairobi, the classical staphylococcal enterotoxins. RPLA are useful methods for detection Kenya. Email: mathenge. Methodology: The survey was a cross- of enterotoxigenic potential of S. aureus. [email protected] sectional descriptive study targeting meat, There is need for strict hygienic and dairy products and meat processing plant. preventive measures to the manufacturer, The study was confined in randomly distributors and consumers of meat and selected meat and milk outlets in the milk products since the contamination of Central Business District of Nairobi S. aureus is greater than other pathogenic (CBDN) and its surroundings. Four bacteria previously reported. hundred and twenty food samples of Keywords: Staphylococcal enterotoxin, animal origin, comprising of meat and PCR, RPLA meat, Dairy products dairy products, were randomly sampled. Genes coding classical staphylococcal Introduction enterotoxins were profiled using multiplex Foods of animal origin, especially milk and Polymerase Chain Reaction (PCR) and dairy products, are associated with food the production by Reversed Passive Latex borne diseases [1,2]. In many countries, S. Agglutination (RPLA). Two hundred and aureus is considered the second or third fifty one samples of raw pork and packed most common pathogen responsible for pork products from a local processing outbreaks of food poisoning [3]. Strains factory were similarly tested. of S. aureus can produce one or more Results: Data from a total 671 samples staphylococcal enterotoxins (SEs), which were examined and 37.4% of them were causes food poisoning in humans [4]. contaminated with S. aureus in which Due to a combination of toxin mediated the contamination rate was higher in virulence, invasiveness and antibiotic meat products than in the milk products. resistance S. aureus infections are difficult Enterotoxin production was detected in to control [5]. 74.5% of the isolated strains of S. aureus. To date, 21 staphylococcal enterotoxins Staphylococcal enterotoxin genes (ses) (SE) or enterotoxin-like proteins (SEls) were detected in 77.3% of the total have been identified [6]. On the basis isolates. The most frequent gene was sea of antigenic characters Staphylococcal (61.8%) followed by see (33.1%), sed enterotoxins, on the basis of antigenic (17.5%) and sec (15.9%) respectively. characters are classified into five main

97 East and Central Africa Medical Journal 2015; 2: 97-103 serological types characterized by the initials SEA, 20 uncooked hams, and 20 hot dogs, 172 raw pork) SEB, SEC , SED and SEE (enterotoxins “classic”) [7,6]. were randomly collected from a meat processing plant However, in recent years, the existences of “new types” during processing stage were included in the study for of SE (SElG, SElH, SEI, SElJ, SElK, SElL, SElM, SElN, comparison. All the samples were placed in sterile plastic SElO, SElP, SElQ, SER, SES, SET, SElU and SElV) bags and transported to the laboratory in cool boxes at have been reported [6]. The toxic shock staphylococcal 4oC. Analysis was done immediately the samples arrived toxin, (TSST-1) initially designated as SEF, lacks emetic in the laboratory or within 4 hours of arrival. activity [8]. The common feature of SEs is high stability Isolation and detection of Staphylococcus aureus: For and resistance towards most proteolytic enzymes, such each sample, 10 grams of the product was weighed and as pepsin or trypsin, allowing protection of their activity transferred into a sterile plastic bag and 90ml of 0.1% in gastrointestinal tract [9]. The SE also retains their sterile peptone water was added. Each sample was biological activities even after pasteurization [1]. homogenized for 2min in a stomacher blender (Elekta Enterotoxigenic strains of S. aureus are estimated Ltd., Japan 400). Homogenates of 0.1mls of the sample to account for about 25% of all isolated strains [10]. Data were spread plated on Baird-Parker agar with egg from various studies of food matrices gives different and York Tellurites Emulsion (Himedia M 043). For the sometimes higher values depending on the SE under dairy products (fresh milk collected from super market consideration [11]. The predominance of Staphylococcal dispensers, yoghurt and packets of pasteurized milk), enterotoxin A (SEA) is well documented in different samples of approximately 10 ml were collected using countries [12]. The SEA in low concentration of 0.6 ng/ sterile sample collection bottles and 0.1ml aseptically ml is the enterotoxin most commonly reported in foods, spread on Baird-Parker agar. After incubation at 37oC, and is also considered as the main cause of SFP, probably suspected S. aureus colonies with convex, black, shiny due to its extraordinarily high resistance to proteolytic appearance with narrow white margin surrounded by enzymes, followed by SEB, SEC and SED [10]. The clear zone were regarded as S. aureus. Five typical fifth classical enterotoxin, SEE, has been infrequently colonies with similar morphologies were isolated and reported in foods and food producing animals, and its cultured separately in blood Agar plates. These colonies involvement in Staphylococcal Food Poisoning (SFP) were confirmed asS aureus by conducting Gram staining, outbreaks has only been demonstrated in rare occasions coagulase test, catalase test, DNase and anaerobic [10,12]. Staphylococcal enterotoxins can be detected by utilization of glucose and mannitol [15]. All the isolated immunoassay Enzyme Linked Immunosorbent Assay bacteria were kept frozen at -70oC until the time of use. (ELISA), and latex agglutination but availability of these DNA extraction from bacteria: Total DNA was extracted methods are limited to commercial tests for classical SEs from 5 ml of S. aureus culture grown at 35± 2ºC for [13]. The DNA-based approach (PCR assays) is thought 24 hours in Brain Heart Infusion (Oxoid) broth. DNA to be an essential tool for investigating SE genes. Studies purification DNeasy blood and tissue kit (QIAGEN conducted elsewhere established the strains harbouring Group, Beckman Instrument, and Icl, USA) and lisozyme classical genes to be predominant [14]. Although several 10mg. mL-1 (Sigma Aldrich) were used for DNA isolation studies on staphylococcal enterotoxin genes (ses) of both as per manufacturer’s instructions. Enterotoxigenic S. classical and the new have been reported in other parts of aureus strains ATCC 13565 (sea), ATCC 14458 (seb), the world currently, there is limited information regarding ATCC 19095 (sec), ATCC 23235 (sed) and ATCC 27664 this in Kenya. The aims of this study was to investigate (see) were used as positive controls and Staphylococcus the occurrence of enterotoxigenic S. aureus isolates from xylosus ATCC 29971 as negative control. meat and dairy products in Nairobi, Kenya, and to detect Multiplex PCR conditions: Primer mixes were prepared the genes encoding the classical SEs by multiplex PCR. according to the master mixes of components [16] with slight modifications to the given instructions. The Materials and Methods volume of this mix was adjusted to 50μl with sterile water. Evaporation of the reaction mixture was prevented Meat and milk products samples collection: A total of by addition of 100μl of sterile mineral oil. 671 samples of meat and dairy products were used in Amplification conditions: Amplification was carried this study. Out of these, 420 samples comprising of meat out in a PCR 96 well Thermocycler (GeneAmp 9700 products (n=280) were purchased randomly from various ABI) with the following thermal cycling profile: Initial butcheries and supermarkets (80 beef chunks, 40 pork, denaturation at 94°C for 5 min was followed by 35 60 fish, 40 sausages, 60 poultry carcasses). Likewise, cycles of amplification (denaturation at 94°C for 2 min, dairy products (n=140), which includes 40 raw milk, annealing at 57°C for 2 min, and extension at 72°C from supermarket dispensers, 60 packets of yoghurt for 1 min), ending with a final extension at 72°C for 7 and 40 samples of pasteurized milk were purchased mins. The amplified PCR products were separated by from, retail shops and supermarkets in Nairobi City and electrophoresis (Major Science Mini Electrophoresis its surroundings were sampled between the June 2012 System CO., Ltd Taiwan) at 80 V for 70 min in 2% and May 2013. A total of 251 samples (79 finished agarose gel (Loba Chemie V201504, India) and stained pork products of 19 cooked salami, 20 fresh sausages, with ethidium bromide. Gels were visualized in a UV

98 East and Central Africa Medical Journal 2015; 2: 97-103 transilluminator (Utra-Lum Electronic Paramount Ca (Sony DCHX 400). Primer pairs used in this study and USA) and images digitalized with a digital camera corresponding multiplex PCR are as described in Table 1.

Table 1: Nucleotide sequences, gene locations, and anticipated sizes of PCR products Gene Primer Oligonucleotide sequence (5-3) Location within gene Size of amplified product (bp) ggttatcaatgtgcgggtgg 349–368 Sea sea- sea-2 102 cggcacttttttctcttcgg 431–450 seb-1 gtatggtggtgtaactgagc 666-685 Seb 164 seb-2 ccaaatagtgacgagttagg 810-829 sec-1 agatgaagtagttgatgtgtatgg 432-455 Sec 451 sec-2 cacacttttagaatcaaccg 863-882 sed-1 ccaataataggagaaaataaaag 492-514 Sed 451 sed-2 attggtattttttttcgttc 750-769 see-1 aggttttttcacaggtcatcc 237-257 See 290 see-2 cttttttttcttcggtcaatc 425-445 fema-1 aaaaaagcacataacaagcg 1444-1463 femA 132 fema-2 gataaagaagaaaccagcag 1556-1575

Enterotoxin production detection by RPLA: Bacteria tested Chi-square (X2) tests. The threshold for statistical were centrifuged and supernatant assayed using significance was set at p<0.05. Staphylococcal enterotoxin Reversed Passive Latex Agglutination (SET-RPLA) kit (Oxoid, England) Results according to manufacturer’s instructions. The analysis was designed to detect four staphylococcal enterotoxins Isolation of S. aureus: Of the 671 samples examined, (SE’s), one each for enterotoxins A (SEA), B (SEB), C 37.4% were contaminated with S. aureus. Overall, 36.2% (SEC) and D (SED). Micro titer plates were sealed with and 39.4% of the samples collected from, the food outlets a plate sealer and shaken to mix the contents of the wells. and processing factory were contaminated with S. aureus, Plates were incubated immediately at room temperature respectively. The proportions of contamination of animal on a vibration-free surface and the agglutination reactions products from the two sources were not significantly read after 22–24 hours according to the manufacturer’s different (p=0.400). Analysis of S. aureus contamination instructions. Reference strains of S. aureus were used for by type of product revealed that significantly more verification of SEs production using RPLA. contamination was observed in meat and meat products Data analysis: Statistical package for social sciences (40.7%) than in the dairy products (25.0%) with p=0.001 (SPSS program version 11) was used to analyze the (Table 2). data. Analysis of variance of different variables were

Table 2: Distribution of S. aureus contamination Characteristic Total number of S. aureus p-value samples -ve (n=420) +ve (n=251) Source Seller 420 268(63.8%) 152(36.2%) Processor 251 152(60.6%) 99(39.4%) 0.400 Type of animal product Meat products 531 315(59.3%) 216(40.7%) Dairy products 140 95(67.9%) 35(25.0%) 0.001

99 East and Central Africa Medical Journal 2015; 2: 97-103 Detection of enterotoxigenic genes: Using multiplex Table 3: Gene coding for S. aureus enterotoxins PCR, the reaction with each individual primer pair Characteristic Frequency (%) resulted in the amplification of single products when Genes coding for toxins (n=251) DNA from each reference strain used as a template Absent 58 23.1 (Figure 1). Reference strains known for the production Presence of one or more genes 193 76.9 of se were used for the verification of this multiplex PCR No. of genes coding for toxins (n=193) (ATCC 25923-FEMA, ATCC 3565-SEA, ATCC 14458- 1 65 33.7 SEB, ATCC 19095-SEC, ATCC 23235-SED, ATCC 2 96 49.7 27664-SEE). The sizes of the products obtained from 3 31 16.1 control strains in multiplex PCR designs corresponded 4 1 0.01 to the predicted sizes. Reproducibility was observed in Specifications of the genes for enterotoxins all tested strains. Although there were variations with (n=357) Sea 155 61.8 band intensity, their presence and sizes were the same. Seb 35 13.9 For providing an assurance against false negative results, Sec 40 15.9 Staphylococcus aureus ATCC 25923 known to have Sed 44 17.5 femA gene was included as an internal positive control. See 83 33.1 In order to determine the specificity of the primers, and detection of the genes with multiplex PCR femA, and a The specifications of the combinations of genes encoding mixer of each of sea, seb, sec, sed and see were amplified for various exotoxins are listed in Table 4. Some isolates together. had only one gene encoding for enterotoxins, which included Sea (27.5%), Seb (2.6%), Sed (0.5%) and N 1 2 3 4 5 6 N See (3.1%). Genes occurring in pairs included Sea/ See (21.2%), Sea/Sed (9.8%), Sed/See (2.1%), Sea/Sec (0.7%) and Seb/Sec while only (0.5%) Sea/Sec/Sed/See coded for enterotoxin. Table 4: Specifications of the combinations of genes 451 bp encoding for enterotoxin bp 500 Sec Frequency 290 bp Genes (%) See (n=193) 278 bp Sed 164 bp Sea 53 27.5 Seb 132 bp Seb 5 2.6 102 bp Fem A sea Sed 1 0.5 bp100 See 6 3.1

Sea/Seb 8 4.1 Sea/Sec 9 4.7 Sea/Sed 19 9.8 Figure: 1 Agarose gel electrophoresis patterns of Sea/See 41 21.2 multiplex PCR amplification products for Staphylococcal Seb/Sec 1 0.5 enterotoxin genes in standards reference strains. Lanes Seb/Sed 11 5.7 N, DNA molecular size marker (100-bp ladder; Thermo Seb/See 1 0.5 Scientific Fermentas™ lanes 1 to 6, PCR amplicons. Sec/See 2 1.0 Lanes: 1, sea plus fem A; 2, seb plus femA; 3, sec plus Sed/See 4 2.1 femA; 4, see plus femA; 5, sed plus femA; 6, sea, seb, sec, Sea/Seb/Sec 2 1.0 sed, see, and femA simultaneously; N, negative control. Sea/Seb/See 4 2.1 Sea/Sec/Sed 1 0.5 Table 3 shows the distribution of genes coding for Sea/Sec/See 16 8.3 classical enterotoxins in S. aureus using the primers for Sea/Sed/See 1 0.5 enterotoxin A to E. Out of the 251 S. aureus isolates that Seb/Sec/Sed 1 0.5 were analysed, 76.9% were found to have one or more Seb/Sec/See 2 1.0 genes encoding for enterotoxins. The most frequent gene Sec/Sed/See 4 2.1 was sea (61.8%) followed by see (33.1%), sed (17.5%) Sea/Sec/Sed/ 1 0.5 and sec (15.9%) respectively. Staphylococcal enterotoxin See B (seb) was the least occurring gene in the S. aureus isolates examined (13.9%).

100 East and Central Africa Medical Journal 2015; 2: 97-103 Determination of SE using RPLA: Of the 251 isolates study. A previous study in Italy, found a higher frequency 187(74.5%) tested positive for the production of one or value of 67% of the S. aureus strains isolated from milk more SEs (RPLA (SEA to SED) as determined by the and dairy products positive for the toxin genes [27]. results of SET-RPLA (KITOXOID; Thermal scientific, Results in our study indicated that of the 270 S. UK) Staphylococcal enterotoxins most frequently aureus isolates, gene coding for sea (61.8%) were the detected from the 187 enterotoxigenic strains was SEA most frequent, followed by see (33.1%), sed (17.5%), 90 (48.1%), followed by a combination of SEA and and sec (15.9%) respectively. The least occurring gene in SEC 22 (11.8%). Sixteen strains (8.6%) were found to the S. aureus isolates examined was seb (13.9%). Similar synthesize a combination of SEA and SEB while fifteen results were reported in studies conducted elsewhere strains (8.0%) synthesized SEA and SED. [5,12,25]. It was clear that 33.5% and 16.5% strains of Additionally, nine strains (4.8%) synthesized SEB and S. aureus had one and three genes coding for enterotoxins, SED. When comparing the results of multiplex PCR for respectively. One strain (0.5%) encoding a combination isolated strains harboring genes of classical SEs (sea-see) of four enterotoxins (Sea/ Sec/ Sed /See) was realized. and SET-RPLA (SEA-SED) for enterotoxin expression, Recent studies have reported that most isolates harboured the results showed 100% correlation particularly more than one toxin gene and 15 different toxinotypes enterotoxin SEA, SEB, and SEC. The results indicated were recorded with “sec” being the most frequent gene a minimal discrepancy 4 (2.1%) were PCR detected alone (28.6%), “sea, sed, ser, selj” (20%), “seg, sei” and encoding genes for sed but enterotoxin production was “seh” (8.6%), which reverse of our study [28]. not shown by the RPLA by the same strains. As observed in other studies, the enterotoxin type Discussion (SE) of 16 enterotoxigenic staphylococci isolates were found to produce SEA (9), SEA+SEB (3), SEA+SED (2), In this study, the overall occurrence of S. aureus in SEC+SED (1), and SEA+SEB+SEC (1) [29]. Of the 27 the analysed samples was 36.2% for food outlets and S. aureus isolated from 420 chicken nugget samples from 39.4% in the processing factory. The proportions the results showed that the most commonly detected of contamination of the samples of animal products genes were sea (25%), sea + seg (8.33%), sec (12.50%), from the two sources were not significantly different sea + sed (12.50%), and sea + sec + sej (12.5%) but (p=0.400). Significantly higher level of contamination no see gene was detected [30]. The main cause of was observed in meat and meat products (40.7%) as differences in the frequency of genes encoding strains in compared to dairy products (25.0%) (p=0.001). Such this study as well as other studies might be the origin of high contamination levels were similarly reported in bacteria isolation, which could vary in animals, humans, other studies conducted in Turkey and USA [17-19], in foods or environment. This study overlooked new SEs milk in Zimbabwe (82%) and in Malaysia (60%) [20,21]. that are being discovered which could have increased The findings of the current study differs from the results the percentages of potentially enterotoxigenic S. aureus. obtained in Egypt [22] in which the contamination rate The role of the newly described SEs in staphylococcal of the raw milk was higher than that of the raw meat food poisoning is still not clear in most parts of the world (58% vs. 18%). Lower rates than those of the current probably because of the limitation of the diagnostic kits study were reported in Norway [2]. The considerable for their detection. Further studies are needed to confirm differences between their results and this study may be the occurrence of these SEs and to assess the role they due to differences in geographic as well as study setting. play in SFP. The high contamination rate of S. aureus in meat and The results obtained by multiplex PCR for harboring dairy products indicates contamination which could be genes of classical SEs (sea-see) and SET-RPLA (SEA- introduced by food handlers who may be habouring the SED) for toxin type were compared in this study. The bacteria because 50% of the human population carry S. frequency of detecting SEA, SEB, and SEC enterotoxin aureus as commensals. Other contamination sources are was consistent with their corresponding genes. The soil, water, dust and air during processing which may results showed a minimal discrepancy 4 (2.1%) for vary in different environmental condition. The midline, PCR detected sed genes but the RPLA did not detect the neck and hind portion of the carcass during slaughter enterotoxin by these strains. Staphylococcal enterotoxins were reported to be areas heavily contaminated with are similar in structural and biological properties but differ microorganisms [23, 24]. in amounts produced where SEB and SEC are expressed In the current study Staphylococcus aureus isolates in greater quantities than SEA and SED [31]. Other were screened by multiplex PCR in order to detect the studies conducted elsewhere indicated a small number of classical enterotoxin genes (sea, seb, sec, sed and see). discrepancies (<5%) between the results using PCR and The results demonstrated that 76.9% coded for one or the RPLA assays [2, 32]. Markedly higher discrepancies more SE genes with 22 genotypes. A frequency close to of between 15 to 32% have been reported elsewhere [27, this study was observed in Switzerland [25] where 172 33] between the results of multiplex PCR and RPLA samples of goat and sheep milk, 65.2% of 296 strains assays, particularly on account of SEA. This discrepancy were positive for the of genes that encode enterotoxins. could be explained by the production of enterotoxin in Similar findings were reported in Japan, 67.8% genes a quantity that was below the limit of detection of the coding for one or more enterotoxins, were recorded [26] RPLA test or its non-expression. but such were lower than those observed in the current

101 East and Central Africa Medical Journal 2015; 2: 97-103 Conclusions 3. Atanassova V, Meindl A and Ring C. Prevalence of Staphylococcus aureus and staphylococcal This study clearly indicated that meat and milk enterotoxins in raw pork and uncooked smoked products marketed in and around Nairobi, Kenya were ham–a comparison of classical culturing detection contaminated with S. aureus, posing a high risk of food and RFLP-PCR. Int J Food Microbiol. 2001; 68: poisoning. S. aureus could be present in these products 105-113. as a result of the animal suffering from disease condition 4. Al-Jumaily EF, Saeed NM, Hussain H and Khanaka and sheding the organism or due to unhygienic conditions HH. Detection of enterotoxin types produce by during production, processing, storage on handling of coagulase positive Staphylococcus species isolated these products. More than half of all detected isolates from mastitis in dairy cows in Sulaimaniyah region. displayed the presence of Staphylococcal enterotoxin Appl Sci Report. 2014; 2(1): 19-26. genes (ses) with sea being the most frequent gene. 5. Qiu J, Feng H, Lu J, Xiang H, Wang D, et al. The results demonstrated a consistency of detecting Eugenol reduces the expression of virulence-related enterotoxin production and their corresponding genes exoproteins in Staphylococcus aureus. Appl Environ with a relatively low number of discrepancies between Microbiol. 2010; 76(17): 5846–5851. the results of multiplex PCR and RPLA. It is important 6. Schelin J, Wallin-Carlquist N, Thorup Cohn M, to note that the PCR is only able to demonstrate the Lindqvist R and Barker, GC. The formation existence of enteroxoxin genes from the isolates but does of Staphylococcus aureus enterotoxin in food not prove that the production of SEs proteins occurs. environments and advances in risk assessment. Recommendations Virulence. 2011; 2: 580-592. 7. Loncarevic S, Jørgensen HJ, Løvseth A, Mathisen T, More detailed investigations on the prevalence of and Rørvik LM Diversity of Staphylococcus aureus newly discovered staphylococcal enterotoxin gene are enterotoxin types within single samples of raw milk needed in this country since contamination of meat and and raw milk products. J Appl Microbiol. 2005; 98: dairy products with new enterotoxigenic strains of this 344-350. bacterium are increasingly being reported in many other 8. Oliveira L, Rodrigues AC, Hulland C, and Ruegg parts of the world. The study suggests the need for strict PL. Enterotoxin production, enterotoxin gene hygienic and preventive measures to the manufacturer, distribution, and genetic diversity of Staphylococcus distributors and consumers of meat and milk products aureus recovered from milk of cows with subclinical since the contamination is greater than other bacteria mastitis. Am J Vet Res. 2011; 72: 1361-1368. pathogens previously reported. 9. Clarisse T, Michèle S, Olivier T, Valérie E, Vincent M, et al. Detection and quantification of staphylococcal Acknowledgements enterotoxin A in foods with specific and sensitive polyclonal antibodies. Food Control. 2013; 32: 255- This study was supported by a grant provided from 261. Government of Kenya, Ministry of Education, Science 10. Argudín MÁ, Mendoza MC and Rodicio MR. Food and Technology through National Commission for poisoning and Staphylococcus aureus enterotoxins. Sciences, Technology and Innovation grant (NACOSTI). Toxins. 2010; 2: 1751-1773. rd NCST/5/003/3 CALL PhD/170. 11. Le Loir Y, Baron F. and Gautier M. Staphylococcus aureus and food poisoning. Genet. Mol. Res. 2003; Conflict of interest statement 31: 63–76. 12. Vernozy-Rozand C, Mazuy-Cruchaudet C , Bavai C, No conflict of interest exists between the authors or and Richard Y. Comparison of three immunological institutions with other third parties. methods for detecting staphylococcal enterotoxins

References from food. Lett Appl Microbiol. 2004; 39: 490-494. 13. Omoe K, Ishikawa M, Shimoda Y, Hu DL, Ueda S,and 1. Asao T, Kumeda Y, Kawai T, Shibata T, Oda H, et al. Shinagawa K Detection of Seg, Seh and Sei genes in An extensive outbreak of staphylococcal food poi- Staphylococcus aureus isolates and determination of soning due to low-fat milk in Japan: estimation of enterotoxin productivities of Staphylococcus aureus enterotoxin A in the incriminated milk and powdered isolates harboring Seg, Seh, or Sei genes. J Clin skim milk. Epidemi Infec. 2003; 130: 33–40. Microbiol. 2002; 40: 857-862. 2. Jorgensen HJ, Mork T, Hogasen HR, et al, 14. Rosec JP and Gigaud O. Staphylococcal enterotoxin Enterotoxigenic Staphylococcus aureus in bulk milk genes of classical and new types detected by PCR in in Norway. J Appl Microbiol. 2005; 99: 158–167. France. Int J Food Microbiol. 2002; 77 (1/2): 61-70.

102 East and Central Africa Medical Journal 2015; 2: 97-103 15. Bennett RW and Lancette GA (2001). Staphylococcus 26. Katsuda K, Hata E, Kobayashi H, et al. Molecular aureus, Chapter 12. In: FDA bacteriological analytical typing of Staphylococcus aureus isolated from th manual, 8 ed., Rev. A. AOAC International, bovine mastitic milk on the basis of toxin genes Gaithersburg, MD. 16. Mehrotra M, Wang G, and Johnson WM. Multiplex and coagulase gene polymorphisms. Vet Microbiol. PCR for detection of genes for Staphylococcus 2005; 105: 301–305. aureus enterotoxins, exfoliative toxins, toxic shock 27. Morandi S, Brasca M, Lodi R, Cremonesi P and syndrome toxin 1, and methicillin resistance. J Clin Castiglioni B. Detection of classical enterotoxins and Microbiol. 2000; 38: 1032-1035. identification of enterotoxin genes in Staphylococcus 17. Aydin A, Sudagidan M, and Muratoglu K. Prevalence of staphylococcal enterotoxins, toxin genes and aureus from milk and dairy products. Vet Microbiol. genetic-relatedness of foodborne Staphylococcus 2007; 124: 66–72. aureus strains isolated in the Marmara Region of 28. Carfora V, Caprioli A, Marri N, Sagrafoli D and Turkey. Int J Food Microbiol. 2011; 148: 99-106. Boselli C. Enterotoxin genes, enterotoxin production, 18. Kiymet G, Mehmet BM, Aysel G, et al. Occurrence and methicillin resistance in Staphylococcus aureus and characterization of Staphylococcus aureus isolated from meat and dairy products consumed in isolated from milk and dairy products in Central Turkey J. Food Safety. 2010; 30: 196–212. Italy. Int Dairy J. 2015; 42: 12-15. 19. Lubna SA, Harrington W and Mohamed KF. 29. Gücükoğlu AA, Kevenk TO, Uyanik T, Ozgur, Staphylococcus aureus is more prevalent in retail beef Cadirci O, et al,. Detection of enterotoxigenic livers than in pork and other beef cuts. Pathogens. Staphylococcus aureus in raw milk and dairy 2015; 4: 4182-4198. 20. Gran HM, Wetlesen A, Mutukumira AN, Rukure G. products by multiplex PCR. J Food Sci. 2012; and Andnarvhus JA. Occurrence of pathogenic 77(11): M620-623. bacteria in raw milk cultured pasteurised milk and 30. Madahi H, Rostami, F, Rahimi E and Dehkordi FS. naturally soured milk produced at small-scale dairies Prevalence of enterotoxigenic Staphylococcus aureus in Zimbabwe. Food Control. 2003; 14: 539–544. isolated from chicken nugget in Iran. J Microbiol. 21. Chye FY, Abdullah A. and Ayob MK. Bacteriological quality and safety of raw milk in Malaysia. Food 2014; 7(8): 10237. Microbiol. 2004; 21: 535–541. 31. Bergdoll MS, Staphylococcus aureus. In: Foodborne 22. Nashwa OK, Fatma IE, Hanan AF and Barakat Bacterial Pathogens (Doyle, M.P., ed.). Marcel AMA. Epidemiological and genetic studies of Dekker, Inc., New York, NY, USA, 1989 pp. 463- enterotoxigenic Staphylococcus aureus isolated from 523. goats and humans. Am. J. Inf. Dis. Microbiol. 2015; 3: 32-37. 32. Pinto B, Chenoll E and Aznar R. Identification 23. Arbuthnott JP. Staphylococcal toxins in human and typing of food-borne S. aureus by PCR-based disease. J Appl Bacteriology (Suppl). 1990;69: techniques. Systematic Appl Microbiol. 2005; 28: 101S–107S. 340–352. 24. Gill CO, McGinnis JC and Badoni M. Assessment of 33. Zouharova M and Rysanek D. Multiplex PCR and the hygienic characteristics of a beef carcass dressing process. J Food Protection. 1995; 59: 136-140. RPLA Identification of Staphylococcus aureus 25. Scherrer D, Corti S, Muehlherr JE, et al. Phenotypic enterotoxigenic strains from bulk tank milk. Zoonoses and genotypic characteristics of Staphylococcus Public Health. 2008; 55(6):313–319. aureus isolates from raw bulk-tank milk samples of goats and sheep. Vet. Microbiol. 2004; 101: 101–107.

103 East and Central Africa Medical Journal 2015; 2: 97-103 Case Report

Traumatic sex with vulval haematoma formation: case report and review of literature

Ngatia JW

Consultant Obstetrician/ Abstract later, she noticed a painful vulval swelling. The swelling progressively increased in Gynecologist, Lecturer The rich vascular supply to the vulva size for about 2 hours and subsequently, Kenyatta University places it at risk for bleeding from trauma. she noticed per vaginal bleeding. She P.O. Box 43844-00100, Vulval haematomas are the most common was changing approximately 1 pad every Nairobi, Kenya. Email: sequelae. In adult women, the labia 2 hours. She had been managed for [email protected] majora are comprised of large fat pads, pulmonary tuberculosis in 2013. which act to protect the vulva against On examination, she was in pain and injury. In contrast, children lack well- was walking with a waddling gait. She developed fat pads in this area and often was not pale and the blood pressure was engage in play activities predisposing normal. She had no lymphadenopathy. them to vulval trauma; thus, they are She had very tender left vulval swelling more likely to sustain vulval injuries approximately 7x4cm (Figure 1). There than adults eg “Straddle injuries” [1-4]. was a bleeding site on the surface of the The case presented was of traumatic sex swelling. The impression was vulval with vulval haemorhage and haematoma haematoma. Her haemogram, urea and formation. It was successfully managed electrolytes, urinalysis, random blood by surgical evacuation, ligation of the sugar were normal. HIV test, VDRL and bleeding sites, and use of antibiotics and pregnancy tests were negative. analgesics. She was prepared for operation. Under epidural anaesthesia, vulvo- Key words: Coitus, Vulval haematoma, vaginal toilet was done and she was Lacerations catheterized. An incision was made on the mucocutaneous junction, haematoma was drained and bleeding sites singularly Introduction ligated using absorbable monocryl 2/0 suture. Obliteration of the cavity Any female with a complaint of vulvo- was done using the same suture. She vaginal pain, bleeding or swelling should was put on Betapyn, Augmentin and undergo a careful examination to look for Clindamycin (Dalacin C). Sitz bath was vulval or vaginal trauma or laceration. done twice daily. She was discharged Patients may not be forthcoming with on 1st post-operative day. Subsequent details of the events that caused the follow-up revealed complete recovery. trauma. Therefore, identifying those at risk is a crucial step in management. The possibility of sexual abuse or assault must always be considered. A case of a young lady who had “quick” coitus on her way home with subsequent vulval haematoma is presented.

Case Report

On 13th September 2014, a 19 year old para 0+0 single lady presented at St Francis Community Hospital at Nairobi with vulval pain, swelling and bleeding for 8 hours. On her way home, she had passed through her boyfriend’s house and Figure 1: Left vulval haematoma and had coitus for a few minutes. One hour associated blood clots in the vulva

104 East and Central Africa Medical Journal 2015; 2: 104-105 Discussion analgesics. Sitz bath helped to reduce the discomfort and accelerated the healing process. The highly vascular nature of the vulva places it at risk for bleeding from trauma. Haematoma of the vulva is a Conclusion and Recommendation possible sequelae [5]. In adult women, the labia majora are comprised of large fat pads, which act to protect the A case of vulval haematoma due to traumatic sex was vulva against injury. Among the adult women, obstetric presented as one of the injuries that may be a consequence injuries are the commonest causes of vulval haematomas. of violent coitus (consensual and non-consensual). In In contrast, children lack well-developed fat pads in this view of rich vascular supply to the perineum, patients area and often engage in play activities predisposing them should be encouraged to seek medical attention when to vulvar trauma; thus, they are more likely to sustain they sustain vulval injuries. It is recommended that there vulval injuries than adults eg Straddle injuries [3,4]. should be modest approach to sexual intercourse with In most cases direct vulvo-vaginal trauma is caused sufficient foreplay to encourage lubrication of the female by direct blunt trauma to an area containing a rich vascular external genitalia. This may not be accomplished when net-work. Such injuries usually result from coitus. either of the parties is in extreme hurry. Suggested predisposing factors that may result in such injuries include virginity, insertion of foreign bodies, self Acknowledgement mutilation, disproportion of male and female genitalia, atrophic vagina in postmenopausal women, congenital To the administration of St Francis Community Hospital. abnormalities stenosis and scarring of the vagina because Written consent was obtained from the patient. of previous surgery, or pelvic radiation therapy. Other factors include rough and violent thrusting of the penis References during intercourse (consensual and non-consensual). This traumatic sex may result from hastily performed 1. Ian SCJ and Alan O. Non-obstetric vaginal trauma. coitus [1,2,3,6]. This case is presented to highlight one of Open J Obstet Gynecol. 2013; (3): 21-23.(http:// the injuries that may be a consequence of such. www.script.org/journal/ojog) Non-obstetric injuries of the female genital tract 2. Singhal V, Bhargava S, Saxena N. et al. Case Report constitute up to 0.8% of all gynaecologic admissions. – Traumatic vulval injury : A rare case. Intern J Approximately 40% of such admissions are due to Gynae Plastic Surg. 2014; 4 (1): 30-32. coital injuries [2,7]. Non-obstetric vulva/vaginal trauma 3. Odawa FXO. Post-coital vulval haematoma: Case can span a continuum of severity from minor trauma report and review of literature J Obstet Gynaecol resulting from normal sexual intercourse to major East Central Afr. 2012; 24(2): 55-56. vaginal lacerations. The true incidences of such injuries 4. Propst AM and Thorp JM Jr. Traumatic vulvar are difficult to ascertain, especially because the nature of hematomas: conservative versus surgical vulvo-vaginal injury usually remains undisclosed (8). management. South Med J. 1998; 91(2):144-146. Diagnosis is usually not a problem when there is 5. Dash S, Verghese J and Nizami DJ. Severe proper co-relation with the history, but sometimes, the haematoma of the vulva: A report of two cases and a vulval swelling could be mistaken for a Bartholin’s clinical review. Kathmandu University Med J. 2006; abscess [5]. In our case, the diagnosis was presumed 4(2): 228-231. from the clinical picture and confirmed at operation. 6. Jeyalakshmi P. Case report: Postcoital bleeding - Patients with large haematomas require operative an unusual cause for abdominal apoplexy. J Obstet management. Incision is made on the inner aspect of the Gynecol India. 2009; 59(4): 360-361. vulva. Evacuation of the clots and separate ligation of the 7. Rabinerson D, Fradin Z, Zeidman A. et al. Vulvar individual bleeding points are done. Then the dead space hematoma after cunnilingus in a teenager with is obliterated with absorbable suture. Small haematomas essential thrombocythemia: a case report. J Reprod with no active bleeding are managed conservatively with Med. 2007; 52(5):458-459. ice packs, bed rest and analgesics [1,3-5). In our case, 8. Marvin MS, Mersedeh K and Pedram I. Non obstetric ligation of the bleeders and obliteration of the cavity lacerations of the vagina. J Amer Osteopath Ass. was done. In view of the possible infection which could 2006; 106(5): 271-273. have occurred, the patient was put on antibiotics and

105 East and Central Africa Medical Journal 2015; 2: 104-105 Case Report

Crypotococcal meningitis in a none-HIV infected five month old infant with rickets: Case report

Okwara FN1, Makewa S2, Karo E3 Abstract crytococcus neoformans var gatti is mainly found in the tropics, where it causes 1 Department of Paediatrics Crypotococcal is an invasive fungal endemic disease [2]. Cryptococcus gattii and Child Health, Kenyatta disease, now endemic in the tropics. It is causes infections in immunocompetent University, P. O. Box largely transmitted through inhalation, but people but C. neoformans v. grubii, and v. 19533, 00202, Nairobi, can be transmitted locally through skin and neoformans usually only cause clinically Kenya eyes. Mostly, it causes disease in immune evident infections especially in persons compromised individuals, especially with compromised immune system [3]. 2Department of Paediatrics older children and adults, where it causes Seroprevalence studies show limited and Child Health, Kenyatta disseminated disease. exposure in infants, but exposures occur Baby JG aged 5 months presented with National Hospital, Nairobi, in less than 5% in children over 5 years a prodrome of respiratory symptoms. Kenya and 60% in adults [2-6]. The disease His anterior fontanel was wide and occurs in 5-10% of HIV infected adults, bulging, and had poor muscle tone. 3Department of mainly when CD 4 counts fall below A week later, he developed convulsions, 3 Paediatrics,Melchizadeck 200mm [7]. It also occurs in other forms and a depressed sensorium. Haemogram of immune suppression such as those with Hospital, Nairobi, Kenya showed a leucocytosis. Bone metabolism malignancy, or on immune suppressants. showed serum low phosphate and high Corresponding author: Paediatric cases are evenly distributed alkaline phosphatase. Cerebrospinal among the immunocompromised and Dr. Florence N. Okwara, fluid biochemistry was unremarkable, Department of Paediatrics the immunocompetent individuals [4- but microscopy was positive for Indian 6]. The disease is transmitted mainly and Child Health, ink stain and crypotococcal antigen. through inhalation, but can also be locally Kenyatta University, P. O. HIV PCR test was negative. Clinical transmitted through the skin and eyes [3]. Box 19533, 00202, Nairobi. improvement was observed on institution Pulmonary disease is the commonest site, Email: florenceoringe@ of antimeningitic therapy, and intravenous but dissemination occurs in the immune gmail.com Kenya fluconazole, vitamin D3 and calcium compromised hosts to involve the brain, supplementation, but another spike was the meninges, skin, eyes, and the skeletal noted on day 7 of therapy. The findings system. Lung disease presents with fever, of crypotococcal meningitis in HIV sero- cough, chest pain, and constitutional negative infant is very rare. Immune symptoms. Neurologic disease causes a reconstitution syndrome may occur during sub-acute or chronic meningitis presenting treatment. with headaches, malaise, convulsions or High index of suspicion for altered mental status [3]. Sepsis syndrome cryptococcosis is needed in high risk may occur in the immunocompromised. children with sub-acute presentations of meningitis, and a relatively normal CSF Case Presentation cell counts and biochemistry. Routine fungal screening of CSF for all suspected JG was a 5 month old male child, born children is justified. and raised in Nairobi, Kenya The mother reported complains of irritability, poor Key words: Crytococcosis, Leucocytosis, feeding, fever, nasal congestion, and Sero-negative, Antimeningitic cough for 7 days prior to admission. The child was initially seen in an outpatient Introduction clinic as case of respiratory tract infection Crytococcosis is an invasive fungal and put on normal saline nose drops, disease caused by a monomorphic amoxicillin, and paracetamol, but his encapsulated yeast. There are various condition remained poor. On the morning variants, but crytococcus neoformans var prior to admission, the child developed neoformans is the commonest worldwide convulsions, and was brought back to [1]. It is commonly found in soil, in avian hospital. This was the first episode of droppings, on fruits and vegetables. fits. The child was born SVD at term, in

106 East and Central Africa Medical Journal 2015; 2: 106-108 hospital, had a birth weight of 3.5kg. Perinatal history lumbar puncture was done on day 28 of treatment and was unremarkable. The mother was a housewife, and was negative for all stains including indian ink. CSF exclusively breastfed the child. The child’s immunization -CRAG was also negative. The child was discharged was up to date. His growth chart revealed weight faltering on oral fluconazole for up to 6 weeks, as well as oral in the previous two months. Findings on admission calcium, iron and vitamin D supplements. 0 revealed a sick looking child, was febrile- 39 C, Discussion weighed 5.8 kg, had some dehydration and mild pallor. On neurologic examination he was drowsy, hypotonic, Indian test result and the clinical manifestations in the and poorly responsive to pain. The anterior fontanelle child were indicative of cryptococcal infection. None was wide and bulging. The head circumference was the less, the finding of cryptoccoccal meningitis in 41cm. The neck was equivocal. The child had widened HIV seronegative infant is very rare. However, this was documented in some studies previously [3,4]. wrists, mild frontal boozing, and poor trunk support Cryptoccocosis is rare, and especially in infants, but it but no rickety rosary. In the respiratory system, he had does occur in 1% of HIV positive children. High prevalence tachypnoe of 56 breaths /min, had no chest retraction, but was documented in children with immunocompromising had bilateral transmitted sounds. He was mouth breathing conditions such as those with acute leukaemia, primary with hypertrophied inferior turbinates, but had a normal immunodeficiency, children on immune suppressive throat exam. His abdomen was full, but not tender. The therapy, immunosuppression from rheumatic disorders, cardiovascular system was unremarkable. malnutrition, and celiac disease [4,6]. Pulmonary disease The initial impression was meningitis in a child is commonest in immune competent individuals, and with rickets. A lumbar puncture done immediately was tends to be assymptomatic. Meningitis is the commonest not under pressure. Cerebrospinal (CSF) fluid analysis presentation of disseminated disease [3]. This tends to revealed protein of 30g/dl, glucose of 160mg/dl, be caused by Cryptococcus neoformans var gratti, a leucocytes 0-3phf, Gram stain and Zeehil Nelsen stains species which is found more commonly in the tropics were negative, but the Indian ink stain and crypotococcal [3]. The coincidental finding of rickets in this child may have contributed to some immune suppression, which antigen (CRAG) test were positive. Blood slide for malaria then predisposed the child to the disseminated disease. In was negative, on the haemogram done; leucocytosis 3 retrospect, the initial respiratory symptoms in this child of 900/mm , with a relative neutrophillia of 78%. may have been the pulmonary phase before progression Haemoglobin level was 9.7g/dl, microcytic hypochromic to the meningitis. However this was not considered at picture, Widal test -negative, a random blood sugar was that point. Detection of cryptococcal antigen (capsular 6.7mmol/l, blood urea nitrogen and electrolytes were material) by culture of CSF, sputum and urine provided normal, Liver Function Tests (LFTs) were normal except definitive diagnosis [9]. Blood cultures may be positive an elevated alkaline phosphatase of 502mmol/l, serum in heavy infections. Indian ink of the CSF is a traditional calcium was low at 1.15mmol/l, serum phosphorus was microscopic method of diagnosis [10], although the 1.85mmol/l, urinalysis was normal. Blood, CSF, and urine sensitivity was poor in early infection, and may miss cultures were taken. X-ray of the wrists showed features 15-20% of patients with culture-positive cryptococcal of rickets. He was started on antimeningitic therapy with meningitis [11]. Cryptococcal antigen from CSF is the intravenous ceftriaxone, amikacin, and fluconazole. He best test for diagnosis of cryptococcal meningitis in terms of sensitivity [12]. However, species isolation was was also put on phenobarbitone, paracetamol, calcium also not done in this case. gluconate, vitamine D3 injection, betamethasone/ Routine screening for HIV infected patients with neomycin nose drops and cetrizine. meningitis is advised [13]. Serial lumbar punctures may The fever dropped within 72 hours, but on day 7 have been useful for monitoring [12,13]. Standard therapy of treatment, the child started spiking fever again. The remains a combination of intravenous amphotericin B culture reports back were all negative. A chest X-ray was and either fluconanazole or flucytosine for 2 weeks, then ordered but the result was normal. Repeat white blood followed with oral flucozole for 8-10 weeks [9,15,16]. cell counts were normal, C-Reactive protein was 46mg/ The Immune Reconstitution Inflammatory dl and the Erythrocyte Sedimentation Rate (ESR) was Syndrome (IRIS) was described in immunocompetent 26mm/hr. A CT scan of the head revealed a normal brain hosts with meningitis caused by C. gattii and C. grubii. parenchyma and para-nasal sinuses. Human Immune Several weeks or even months into appropriate treatment, Deficiency Virus (HIV) RNA PCR test was negative. there was sudden onset deterioration with worsening Intravenous ampotericin B was added to his treatment for meningitis symptoms and progression or development of new neurological symptoms. IRIS was however much 14 days. Hydrocortisone injection was added to prevent more common in immune-compromised hosts (25% inflammation caused by immune reconstitution reactions. versus 8%). In severe IRIS cases, treatment with systemic Serial monitoring of electrolytes was done during corticosteroids was utilized. treatment, and remained normal. The fevers gradually Mortality rate from cryptococcal meningitis is settled and the child clinically improved. A repeat about 15% - 30% in developed countries, but may rise

107 East and Central Africa Medical Journal 2015; 2: 106-108 up to 70% in low income countries [3]. Mortality rates 7. Park BJ, Wannemuehler KA, Marston BJ, Govender N, are higher in HIV infected patients, who have relapse Pappas PG and Chiller TM. Estimation of the current rates of more than 50% but was reported to be unusual global burden of cryptococcal meningitis among in adequately treated in non-HIV infected patients. persons living with HIV/AIDS. AIDS. 2009; 23 (4): However, neurologic sequelae such as hydrocephalus, 525–30. doi:10.1097/QAD.0b013e328322ffac. deafness, cranial nerve palsies, visual defects, seizures, 8. Dromer F, Mathoulin S, Dupont B, et al. Epidemiology ataxia are common. of cryptococcosis in France: A 9 year survey. Clin Conclusion Infect Dis. 1996; 23:82. 9. Cryptoccoccal meningitis in HIV sero-negative infant Rhein, J. and Boulware DR. Prognosis and was diagosed in a 5 month infant whose initial impression management of cryptococcal meningitis in patients was like meningitis in a child with rickets. Clinicians with HIV infection. Neurobehavioral HIV Med. must have a high index of suspicion for cryptococcossis, 2012; 4: 45. doi:10.2147/NBHIV.S24748. especially when patients have sub-acute or chronic 10. Zerpa, R, Huicho L and Guillén A. Modified India presentations of meningitis, relatively normal CSF ink preparation for Cryptococcus neoformans in cell biochemistry, and poor response to initial course cerebrospinal fluid specimens. J Clin Microbiol. of antibiotics. In view of high prevalence of high risk 1996; 34 (9): 2290–1. PMID 8862601. children including the malnourished, preterms and HIV 11. Boulware DR, Rolfes MA, Rajasingham R, von infected children in our settings, routine screening of Hohenberg M, Qin Z. et al. Multisite validation CSF for all suspected children is justified. of cryptococcal antigen lateral flow assay and quantification by laser thermal contrast. Emerg Conflict of interest: Authors declare that we have no Infect Dis. 2014; 20 (1): 45–53. doi:10.3201/ conflict of interest. eid2001.130906. PMC 3884728. PMID 24378231. 12. Antinori S, Radice A, Galimberti L, Magni C, Fasan Acknowledgement M and Parravicini C. The role of cryptococcal antigen To Jamaa Mission Hospital administration, where the assay in diagnosis and monitoring of cryptococcal case was managed. The study received no funding. meningitis. J Clin Microbiol. 2005; 43 (11): 5828–9. doi:10.1128/JCM.43.11.5828-5829.2005. References PMC 1287839. PMID 16272534. 13. Rajasingham R, Meya DB and Boulware DR. 1. Robert M. Kliegman, Bonita M.D. Stanton, Joseph Integrating cryptococcal antigen screening and pre- St. Geme, Nina Schor, and Richard E. Behrman. emptive treatment into routine HIV care. J Acq Nelson textbook of Paediatrics, edition 19 pp 2069, Immune Def Syndromes. 2012; 59 (5): e85–91. year 2011, Elsever, ISBN: 978-1-4377-0755-7. 14. World Health Organization. Rapid advice: Diagnosis, 2. “Meningitis: cryptococcal: Overview”. Medical prevention and management of cryptococcal disease Reference: University of Maryland Medical Center. in HIV-infected adults, adolescents, and children. 2011. September 2010. 15. Practice Guidelines for the Management of 3. Centers for Disease Control and Prevention (CDC), Cryptococcal Disease. Infectious Disease Society National Center for Emerging and Zoonotic of America 2010. http://www.idsociety.org/ Infectious Diseases http://www.cdc.gov/fungal Organism/#CryptococcalDisease 4. Speed BR and Kaldor J. Rarity of cryptococcal 16. Rajasingham R, Rolfes MA, Birkenkamp KE, Meya infection in children. Pediatr Infect Dis J. 1997; DB and Boulware DR. “Cryptococcal meningitis 16:536. treatment strategies in resource-limited settings: A 5. Gonzalez CE, Shetty D, Lewis L, et al. Crypotococcus cost-effectiveness analysis. In: Farrar, Jeremy. PLoS in Human immunodeficiency virus-infected children. Med. 2012; 9 (9): e1001316. doi:10.1371/journal. Paediatr Infect Dis J. 1996; 15:796. pmed.1001316. PMC 3463510. PMID 23055838. 6. Ndiaye M, Diagne NR, Seck LB, Sow AD, Sène MS, Retrieved 26 September 2012. et al. Cryptococcal meningitis in children: description of 3 cases. Med Trop (Mars). 2011; 71(2):176-178.

108 East and Central Africa Medical Journal 2015; 2: 106-108 JOURNAL POLICY AND GUIDELINES Vision Statement Original research articles should be between two and three The East and Central Africa Medical Journal is a thousand and not more that 30 references. The number general medical peer-reviewed journal providing a of tables and other illustrations should not exceed eight. communication forum of research results in clinical The message carried in the manuscript should be clear medicine, health sciences in public health, biomedical and relevant to the promotion and development of health research, pharmaceutical sciences and related disciplines sciences and practice of medicine especially in Africa. and a continuing medical education to those involved Reviews which should be contributed by authorities and in medical research and related professions through analysis of the subject reviewed and balanced. Reviews review articles and updates necessary for development, should not exceed 4000 words and 40 references. improvement and promotion of health in Africa. Abstract: A structured abstract of not more than three Mission Statement hundred words must be included. For original articles, the structured abstract should have the following headings A commitment to publishing scholarly articles in the background, objective (s), methodology, results, and promotion and development of medical research and conclusion. For reviews, the headings should be; achieving new horizons in medicine, drug development objective(s) data source, study selection, data extraction, and delivery and general improvement in community data synthesis and conclusions. health. AUTHORS GUIDELINES Editorials: These will be commissioned. Unsolicited communications of up to one thousand words can be Manuscript: Original research papers in clinical submitted and will be subject to peer review. case reports, health and related biomedical research Case reports: Reasons for their presentation must be authoritative reviews, diagnosis exercises, short reports, clear, especially their relevance to clinical practices. letters to the editor and book reviews written in English They must be authentic, adequately and appropriately will be considered for publication. illustrated and the identity of any individual concealed according to ethical requirements. They should not be The E&C.A.M.J. peer review process: All manuscripts more than two thousand words with a prose summary of submitted to E&C.A.M.J will be peer reviewed. Upon about one hundred words. reception, the articles will be reviewed in-house by the editors. The manuscript may be rejected at this stage due Letters to the editor: Letters to the editor, especially to several reasons including lack of originality and lack those responding to recent articles in the journal are encouraged but should be less than six hundred words. review will be sent to at least two external peer reviewers selected from database of medical experts and scientists. Peer reviewers, after reading the manuscript will have four Authors of original articles, reviews etc. should declare possible options: accept; accept after minor revision; ask author(s) to make major revisions and resubmit; or reject. judgment, where applicable. Once all the reviews are obtained, the Editor assigned to the manuscript will make a preliminary decision on Ethical Guidelines: the manuscript, and will then forward the manuscript, the Authors are responsible for the views, opinions and reviewer feedback and his/her preliminary decision to the authenticity of their materials published in this journal. E&C.A.M.J aims Local and international guidelines on research ethics months after manuscript submission. When reporting experiments on human subjects, authors General Instructions should indicate whether the procedures followed were in accordance with the ethical standards of the responsible Manuscripts: All manuscripts submitted for committee on human experimentation (institutional and consideration for publication must be submitted national) and with the Helsinki Declaration of 1975, as exclusively to E&C.A.M.J. For original articles and revised in 2008. Evidence of ethical approval from the clinical cases, a written permission must be sought and relevant ethics committee will be required. obtained from the in-charge of the hospital or centre When reporting experiments on animals, authors should where the work was done or the case managed. indicate whether ethical approval was received from All manuscripts must be typed in double line spacing on the relevant animal ethics committee, and whether numbered pages. All authors should provide their full institutional and/or national guidelines for the care and ‘names, address, and e-mail for ease of correspondence. use of laboratory animals were followed. All materials submitted for publication will be edited. East and Central Africa Medical Journal 2014; 1(1): 43-44 43 East and Central Africa Medical Journal 2014; 1(2):(2): 101-102101-102 101109 East and Central Africa Medical Journal 2015; 2: 109-110 Copyright References: Only references which are clearly related to the All authors will be expected to transfer all copyright work of the authors should be quoted. References should ownership of manuscripts to the publishers of E&C.AM.J be cited by consecutive numbers in the text. The numbers in accordance with international copyright regulations should appear in [ ] parenthesis and should appear before and the Copyright Act, Chapter 30 Laws of Kenya or any punctuation. Abbreviate journal names in the style used by the National Library of Medicine. References other relevant country authority or agent. Once accepted should be from peer –reviewed publications that are for publication, the corresponding author, on behalf of all generally available to the reader. Abstracts, proceedings, authors, must submit a scanned copy of the completed works in progress, theses and manuscripts submitted but and signed “Transfer of Copyright Form”. not accepted for publication are not acceptable to cite as references. References are listed numerically in the order Format and style of initial appearance in the manuscript. All authors must Original articles should conform to the AIMRADAR be listed. Do not use “et al”. form (i.e. Abstract, Introduction, Materials and Methods, Results and Discussion). The cover page should contain: Privacy Statement: The names and e-mail address the title of the paper, authors names (s), degrees, entered in this journal site will be used exclusively for designation, name of the institution where the work the stated purposes of this journal and will not be made was carried out, present address, e-mail address and the available for any other purpose or to any other party. contact of the author to whom all correspondence should Disclaimer: Although the E&CAMJ accepts articles and Word or RTF advertisements for products and services related to the improvement of public health and advancement of medical research, all materials published in the journal Tables: should be typed on separate pages, serially represent the opinion of the authors responsible for numbered in Arabic numerals and in parenthesis in the text. Tables must be designed for printing in one-column E&CAMJ. width. If possible this should never require more than full Publication of advertisement of a product in the E&CAMJ page width. does not constitute endorsement of the quality or value of the product or service described therein or any of the Illustrations: Graphs, drawings and photographs representations or claims made by the advertiser. (include patients’ permission to use their photographs or All advertising accepted for publication in the journal blinders to protect their identity) should be numbered and must be presented with highest professional standards entered in the text. and must be in consistent with the goals of the journal.

Acknowledgement

The journal is partially funded by the PRIME-K project, grant number IR24TW008889 of National Institute of Health, U.S.A

East and Central Africa Medical Journal 2014; 1(2): 101-102 110 East and Central Africa Medical Journal 2014; 1(2): 101-102 10244 EaEastst a nandd C eCentralntral A frAfricaica M eMedicaldical Jou Journalrnal 2014; 2015; 1(1): 2: 43-44 109-110

1. 2. 3. http://kupccconference.ku.ac.ke 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Nutrition in Patient Centered Care

East and Central Africa Medical Journal 2014; 1(2): i 112104 East and Central Africa Medical Journal 2015; 2: