Mortality rates in a cohort of infants attending immunization clinics in Uganda (2017-2019)

Uganda reported a significant reduction in the mortality rate of children under 5 years of age, from 146/1,000 live births in 2000-42/1,000 live births in 2021. With the rollout of Option B+, the vertical transmission rate of HIV decreased from 13.0% (2012) to 6.0% (2019). However, its impact on the mortality rate among children is not well documented. We determined the mortality rate and associated risk factors among infants exposed and not exposed to HIV attending immunization clinics in Uganda.

We conducted an observational prospective cohort study of mother-infant pairs (MIPs) with infants exposed or unexposed to HIV. We enrolled infants aged 4-12 weeks. The inclusion criteria were biological mothers attending health facilities that provide routine immunization for children and/or postnatal care visits who were able to provide signed written informed consent; mothers or infants who were not severely ill; and those who consented to have their infants tested for HIV antibodies at baseline and follow-up visits every 3 months until the children were aged 18 months. Child-HIV infection and death were censored events. Children lost to follow-up or withdrawn from the study were censored from analyses at the last documented study visit. The outcome of interest was child mortality, and the independent variables were mother's age; infant HIV exposure status; infant sex; family socioeconomic status; marital status; education level; malaria during pregnancy; birth attendee; mother's ART initiation; mode of transport to health facilities; breastfeeding pattern; 4 or more ANC visits; and mother's baseline viral load nonsuppression and place of delivery. We used Kaplan-Meier survival curves to estimate cumulative mortality probability and the Wilcoxon log-rank test to compare differences in cumulative survival functions. We used multivariate Weibull proportional hazards and Weibull accelerated failure time (AFT) regression models with 95% confidence intervals (CIs) to identify factors associated with child death.

Among the 16,718 MIPs identified, 11,519 (68.9%) mothers consented to study follow-up. At the 18-month follow-up, 0.7% (79/11,519) of the infants had died, 40.5% (32/79) of whom were exposed to HIV. The overall child mortality rate per 1,000 person-years was 5.0 (95% CI: 4.0--6.2) and was significantly greater among the infants exposed to HIV (14.2; 95% CI: 10.0--20.0) than among the infants not exposed to HIV (3.5; 95% CI: 2.6--4.6). In the adjusted model, the mortality risk factors were HIV exposure status (aHR5.6 95% CI: 3.5--9.4), maternal age < 25 years (aHR1.8; 95% CI: 1.1--2.9), living without a partner (aHR1.8; 95% CI: 1.1--2.9), and delivery at home (aHR2.2; 95% CI: 1.3--4.0).

Single young mothers living with HIV delivering at home increased the risk of child mortality. Identifying mothers with risk factors early for support could reduce the risk of child mortality.