Context: The burden of tuberculosis (TB) disease in children, particularly in HIV-infected children, is poorly described because of a lack of effective diagnostic tests and the emphasis of public health programs on transmissible TB.
Objectives: The objectives of this study were to describe the observed incidence of and risk factors for TB diagnosis among HIV-infected children enrolled in a large network of HIV clinics in western Kenya.
Design: Retrospective observational study.
Setting: The USAID-Academic Model Providing Access to Healthcare (AMPATH) Partnership is Kenya's largest HIV/AIDS care system. Since 2001, the program has enrolled over 70,000 HIV-infected patients in 18 clinics throughout Western Kenya.
Patients: This analysis included all HIV-infected children aged 0 to 13 years attending an AMPATH clinic.
Main outcome measure: The primary outcome was a diagnosis of any TB, defined either by a recorded diagnosis or by the initiation of anti-TB treatment. Diagnosis of TB is based on a modified Kenneth Jones scoring system and is consistent with WHO case definitions.
Results: There were 6535 HIV-infected children aged 0 to 13 years, eligible for analysis, 50.1% were female. Of these, 234 (3.6%) were diagnosed with TB at enrollment. There were subsequently 765 new TB diagnoses in 4368.0 child-years of follow-up for an incidence rate of 17.5 diagnoses (16.3-18.8) per 100 child-years. The majority of these occurred in the first 6 months after enrollment (IR: 106.8 per 100 CY, 98.4-115.8). In multivariable analysis, being severely immune-suppressed at enrollment (Adjusted Hazard Ratio [AHR]: 4.44, 95% CI: 3.62-5.44), having ever attended school AHR: 2.65, 95% CI: 2.15-3.25), being an orphan (AHR: 1.57, 95% CI: 1.28-1.92), being severely low weight-for-height at enrollment (AHR: 1.46, 95% CI: 1.32-1.62), and attending an urban clinic (AHR: 1.39, 95% CI: 1.16-1.67) were all independent risk factors for having an incident TB diagnosis. Children receiving combination antiretroviral treatment were dramatically less likely to be diagnosed with incident TB (AHR: 0.15, 95% CI: 0.12-0.20).
Conclusions: These data suggest a high rate of TB diagnosis among HIV-infected children, with severe immune suppression, school attendance, orphan status, very low weight-for-height, and attending an urban clinic being key risk factors. The use of combination antiretroviral treatment reduced the probability of an HIV-infected child being diagnosed with incident TB by 85%.