Confound it: latent lessons from the Mwanza trial of STD treatment to reduce HIV transmission

In 1995, an international team reported that improved syndromic management of sexually transmitted disease (STD) in Mwanza, Tanzania, had reduced HIV incidence by 38% in intervention compared to control communities. However, the team has not addressed confound: project interventions might have reduced HIV transmission during health care through provision of syringes and benzathine (replacing short acting) penicillin and through interactions with a coeval safe injection initiative. Mwanza's success in lowering HIV incidence is a puzzle, since it was achieved with only minor reductions in observed STD prevalence. Despite incomplete analyses, reports from Mwanza have encouraged expansion of STD treatment. However, should success be attributed to injection safety rather than to decreased STD prevalence-an hypothesis that fits published data-expanded STD treatment without attention to injection safety could, ironically, increase rather than decrease HIV incidence. To control for confound, additional data and analyses from the Mwanza study are warranted.