Nine studies (more than 20,239 participants, range 211 to 14,097 where reported) were included in the review: one randomised trial, four prospective cohort studies and four retrospective cohort studies. Six studies were deemed moderate quality and three studies were deemed high quality. There was no evidence of publication bias for the outcome of death. Six studies had a follow-up of less than 13 months, one study followed up participants for two years and one study followed up for almost five years.
Cotrimoxazole reduced the incidence of death (OR 0.42, 95% CI 0.29 to 0.61, Ι²=93%; seven studies). Survival was increased by continuing cotrimoxazole at antiretroviral therapy initiation or starting cotrimoxazole at the same time as antiretroviral therapy or initiating cotrimoxazole when the patient was stable on antiretroviral therapy. Sensitivity analysis indicated that differing length of follow-up and HIV disease stage did not explain the heterogeneity of results.
Other results for secondary outcomes and adverse events were presented in the review.