Background:  Cotrimoxazole prophylaxis prolongs survival and prevents opportunistic infections, malaria, and diarrhea in HIV-infected persons. Many countries recommended that individuals on ART discontinue cotrimoxazole when CD4 counts rise above 200 cells/mm³. However, this practice has never been evaluated in Sub-Saharan Africa where background incidence of infectious diseases is high.

Methods:  The study was conducted among patients in the Home-Based AIDS Care (HBAC) program in Eastern Uganda. Patients initiated ART if they had a CD4 cell count ≤250 cells/mm³ or WHO stage III or IV disease, excluding pulmonary TB. In the program’s fourth year, patients were randomly assigned, by household, to continue or discontinue cotrimoxazole. For patients who consented and were assigned to discontinue prophylaxis, it was required that their 2 most recent, consecutively drawn quarterly CD4 counts exceeded 200 cells/mm³. They were followed for episodes of malaria (fever with accompanying parasites on thick-smear) and diarrhea (self-report of >3 loose or watery stools/24 hours). Planned follow-up period was at least 2 years. Clinical outcomes were compared by arm using Fisher’s Exact test and rate ratios were calculated.

Results:  At randomization for this sub-study, eligible patients (n = 836) had been receiving ART for a mean of 3.7 years, with a median CD4 count of 489 cells/mm³. 94% had a viral load <400 copies/mL. They were followed for 116 days after randomization, at which time the study was discontinued at first review by the Data Safety Monitoring Board. Among those continuing (n = 452) vs discontinuing (n = 384) cotrimoxazole, 0.4% (2 episodes in 2 patients) vs 12.3% (55 episodes in 47 patients) had at least 1 episode of malaria (p <0.0001) and 14% (79 episodes in 64 patients) vs 25% (120 episodes in 95 patients) had at least 1 episode of diarrhea (p <0.0001). Compared to those remaining on cotrimoxazole, patients who discontinued had a relative risk of malaria of 28 (95%CI 6 to 105) and of diarrhea of 1.8 (95%CI 1.3 to 2.3). There were 4 deaths (3 on cotrimoxazole and 1 off); no deaths were related to malaria, diarrhea, or opportunistic infections.

Conclusions:  HIV patients on ART with CD4 counts over 200 cells/mm³ who live in a malaria-endemic area of Sub-Saharan Africa and who discontinue cotrimoxazole prophylaxis have an increased incidence of malaria and diarrhea compared to those who continue prophylaxis.