Understanding the association between caregiver sex and HIV infection among orphans and vulnerable children in Tanzania: learning from the USAID Kizazi Kipya project.

Background: Tanzania has met only 50.1% of the 90% target for diagnosing HIV in children. The country's pediatric case finding strategy uses global best practices of index testing, provider-initiated counselling and testing, and targeted community testing of at-risk populations to find about 50,000 children living with HIV (CLHIV) who are undiagnosed. However, context-specific strategies are necessary to find the hidden children to meet the full 90% target. This study assesses whether sex of the caregiver is associated with HIV status of orphans and vulnerable children (OVC) as a valuable strategy for enhanced pediatric case findings.

Methods: Data originate from the community-based, United States Agency for International Development (USAID)-funded Kizazi Kipya Project, which works towards increasing OVC's and their caregivers' uptake of HIV/AIDS and other health and social services in Tanzania. Included in this study are 39,578 OVC ages 0-19 years who the project enrolled during January through March 2017 in 18 regions of Tanzania and who voluntarily reported their HIV status. Data analysis involved multi-level logistic regression, with OVC HIV status as the outcome of interest and caregiver's sex as the main independent variable.

Results: Three-quarters (74.3%) of the OVC included in the study had female caregivers, and their overall HIV prevalence was 7.1%. The prevalence was significantly higher (p < 0.001) among OVC with male caregivers (7.8%) than among OVC with female caregivers (6.8%), and indeed, multivariate analysis showed that OVC with male caregivers were significantly 40% more likely to be HIV-positive than those with female caregivers (OR = 1.40, 95% CI 1.08-1.83). This effect was the strongest among 0-4 year-olds (OR = 4.02, 95% CI 1.61-10.03), declined to 1.72 among 5-9 year-olds (OR = 1.72, 95% CI 1.02-2.93), and lost significance for children over age 9 years. Other significant factors included OVC age and nutritional status; caregiver HIV status and marital status; household health insurance status, and family size; and rural versus urban residence.

Conclusions: OVC in Tanzania with male caregivers have a 40% higher likelihood of being HIV-positive than those with female caregivers. HIV risk assessment activities should target OVC with male caregivers, as well as OVC who have malnutrition, HIV-positive caregivers, or caregivers who do not disclose their HIV status to community volunteers. Further, younger HIV-positive OVC are more likely to live in rural areas, while older HIV-positive OVC are more likely to live in urban areas. These factors should be integrated in HIV risk assessment algorithms to enhance HIV testing yields and pediatric case-finding in the OVC population in Tanzania.