Morbidity and mortality in HIV - infected children admitted at Moi Teaching and Referral Hospital in Western Kenya

  • W Nyandiko
  • AW Mwangi
  • E Sang
  • R Vreeman
  • I Marete
  • E Nabakwe
  • C Tenge
  • P Gisore
  • J Songok
  • J Kiplagat
  • E Walumbe
  • S Ayaya


Background: HIV-infected children are at higher risk of opportunistic infections that could result in hospitalisation. The outcomes of hospitalisation are variable and depend on the admission diagnosis, the patients’ immune status and whether or not the patient is on anti-retroviral drugs.
Objective: To describe the characteristics and causes of hospitalisation and mortality for HIV infected children admitted to Moi Teaching and Referral hospital in western Kenya.
Design: a retrospective study of prospectively collected data.
Setting: The paediatric wards of Moi Teaching and Referral Hospital (MTRH). A Kenyan National Referral Hospital.
Subjects: HIV-infected children admitted the paediatric wards.
Interventions: Treatment with combination anti-retroviral therapy (cART), treatment of common opportunistic infections.
Main outcome measures: Demographic and clinical data, including diagnosis, immune status, and treatment with combination anti-retroviral therapy (cART), were extracted from hospital admission records of HIV-infected children registered with the USAIDAcademic Model Providing Access to Healthcare (AMPATH) partnership. We conducted descriptive statistical analyses and used chi-square and fisher’s exact tests to assess for associations between categorical variables and each of the independent variables.
Results: Between December 2006 and May 2009, 396 HIV-infected children were admitted to MTRH. Median age at admission was 2.0 years (range 0-15); 236 (59%) were male; 36 (15%) of available 236 orphan status entries were orphaned; 198 (73%) were in CDC categories B and C and 61 (16%) of available 386 had been on ART. Among 108 patients with documented immunologic status, the mean CD4 cell percentage was 16% (SD 10.8). Among the 396 children, 104 (15%) were diagnosed with pneumonia, 92 (14%) with gastroenteritis, 36 (9%) with tuberculosis and 37 (9%) with malaria. Deaths occurred in 28(7%) of the patients. The median duration of hospitalisation was seven days (range 1- 516) for discharged patients and six days (range 0-72) for those who died. A significantly higher percentage of the children who were not previously enrolled in AMPATH died, signifying 14 (15%) mortality among this population of admitted patients, p = 0.0017. Of those who died, (17%) were diagnosed with pneumonia and 22 (79%) of them were not on cART.
Conclusion: The common diagnoses at hospitalisation included pneumonia, gastroenteritis, malaria and tuberculosis. Higher mortality occurred among those diagnosed with pneumonia and those not previously enrolled in the HIV care programme. Aggressive treatment and prevention of the most prevalent diseases and early enrollment into HIV care are recommended for HIV-infected children. A follow-up study to investigate the pathological causes of death in this population is recommended.

Journal Identifiers

eISSN: 0012-835X