Background:  Due to financial and infrastructure considerations, many providers in resource-limited settings do not use routine viral loads or CD4 cell counts to monitor ART. We evaluated the utility of routine laboratory monitoring in rural Uganda.

Methods:  HIV-infected adults with a CD4 count of <250 cells/mm³ or World Health Organization (WHO) stage 3 or 4 were offered ART and randomized to 1 of 3 monitoring groups:  arm A) clinical monitoring and quarterly CD4 cell counts and viral loads; arm B) clinical monitoring and quarterly CD4 cell counts; or arm C) clinical monitoring alone. Trained lay providers delivered ART to participants’ homes weekly, collected data on illness and mortality, and referred participants to the study clinic for care. Quarterly CD4 cell counts and viral loads were done on all participants; however, results were provided to clinicians per study arm protocol. Cox proportional hazards and Poisson regression models were used to compare incidence of new AIDS-defining events and mortality across arms.

Results:  In all, 1094 ART-naïve participants started ART; median baseline CD4 cell count was 130 cells/μL. Median follow-up was 3 years. From the time of randomization, there were 125 deaths (11%) and 148 new AIDS-defining illnesses; 66 deaths (52%) occurred >90 days after starting ART. Virologic failure defined as 2 consecutive viral loads greater than 500 copies/mL after the first 6 months of ART, was experienced by 61 (6%) of participants. A change to second line drugs occurred in 28 (3%) of participants. In an intention-to-treat analysis from the date of ART initiation, adjusting for age, sex, baseline CD4-cell count, viral load, body mass index, and Center for Epidemiologic Studies Depression Scale (CESD) score, the rate to new AIDS-defining event or death was higher in arm C than arm A (hazard ratio [HR] 1.9, p = 0.002) or B (HR 1.5, p = 0.047). No difference between arms B and A (HR 1.3, p = 0.26) was found; we had 80% power to detect a rate ratio as small as 1.75 at p <0.05. Although overall mortality in Arm C during the 3 years of ART was low (13%), there was a non-significant trend towards higher mortality between arm C and Arms A (HR 1.6, p = 0.07) and B (HR 1.4, p = 0.18). There was no difference in mortality between arms B and A (HR 1.1, p = 0.6).

Conclusions:  The use of routine CD4 cell count monitoring was associated with fewer new AIDS-defining events or mortality compared with clinical monitoring alone. These data support WHO guidelines prioritizing access to CD4 cell count measurements.