School of Public Health

Anjuli D. Wagner

Uptake, acceptability, and cost-effectiveness of a targeted pediatric HIV testing strategy


Prevention of mother to child transmission of HIV (PMTCT) programs have scaled up and resulted in fewer new infant infections globally. However, children born prior to the scale up of PMTCT and early infant diagnosis (EID) systems, as well as those who are missed by the complex systems, remain undiagnosed. In the absence of treatment, pediatric HIV infection has high morbidity and mortality; prompt testing and treatment are essential. Universal testing strategies are relatively inefficient; provider-initiated testing and counseling identifies many positive children but tends to identify those children who are already symptomatic and least likely to survive even with appropriate treatment. Targeted strategies are needed to increase the efficiency of pediatric HIV case detection.


In a HIV care clinic in Nairobi, HIV-infected adults were systematically assessed to determine whether they had children of unknown HIV status ages 0-12. Eligible caregivers were referred for enrollment in the study and elected to test their children in either a home- or clinic-based setting. Uptake and yield of testing were compared before and after the systematic assessment and active referral intervention (Chapter 1). Health care workers from PMTCT, Voluntary Counseling and Testing (VCT), and HIV care clinics completed focus group discussions and a subset of caregivers completed in-depth interviews about barriers and facilitators to pediatric HIV testing. Themes were identified and compared with barriers and facilitators identified in the enrollment survey (Chapter 2). Cost-effectiveness of passive referral, active referral, and active referral plus home-based testing was estimated (Chapter 3).


Chapter 1‚Äč: A substantial proportion of adults in care had children of unknown status. Following implementation of active referral, the rate of pediatric HIV testing increased 3.8-fold (p<0.001). However, 86% of caregivers did not test their children during the intervention. HIV prevalence among 108 tested children was 7.4% and median age was 8 years (IQR: 2-11); one child was symptomatic at the time of testing. Chapter 2: Interview and focus group data revealed barriers and facilitators that applied to three periods of the pediatric HIV testing process: concerns about the decision-making process to test, the test visit itself, and coping during the post-test period. The greatest concerns that emerged from both qualitative and quantitative sources were inaccurate information about the likelihood of child infection and prognosis, challenges with permission and disclosure with partners and children, experiences with discouraging provider attitudes, lack of clear pediatric testing guidelines around consent/assent and disclosure, insufficient staff and inappropriate space for children, concerns about privacy, and perceived costs of testing and care. Chapter 3: Uptake of testing and average costs were highest in the active referral plus home-based testing model, followed by the active referral model, with passive referral being the least costly and least effective. The incremental cost effectiveness ratio (ICER) comparing active referral plus home-based testing to active referral was $58 per child tested and $789 per HIV-infected child identified from the Ministry of Health perspective. The ICER comparing active referral to passive referral was $57 per child tested and $768 per HIV-infected child identified. The ICER comparing active to passive referral was robust to variations in cost and effectiveness in sensitivity analyses; the ICER comparing active referral plus home-based testing to active referral was not.


Referring HIV-infected parents in care to have their children tested revealed many untested children and significantly increased the rate of pediatric testing; prevalence of HIV was high. Despite increases in pediatric testing, most adults did not complete testing of their children. Interventions are needed to address client, provider, and clinic-level barriers to pediatric HIV testing. Assisted disclosure services for caregivers, partners, and children; small financial incentives to compensate for lost wages and transport costs; peer support groups and counseling before and after testing; in-service provider training on pediatric HIV testing; and weekend and school holiday clinic dates merit evaluation to address barriers to pediatric testing. Costs for case detection by active referral plus home-based testing and active referral alone were comparable to costs per person tested under adult and EID models; further research is needed to determine whether these models are cost-effective based on disability-adjusted life years averted.