12/100 pyr; 95% CI 22.73–78.05/100 pyr) BMN-673 and 350–499 cells/μL (41.12/100 pyr; 95% CI 30.44–55.55/100 pyr). Similarly, the incidence of bacterial infections (including pneumonia) also varied by whether an individual was on ART, and the time on ART. Among HIV seroconverters not on ART, the overall incidence was 8.61/100 pyr (95% CI 6.95–10.67/100 pyr) from 1990 to 2008. In comparison, the incidence was significantly higher among individuals on ART for <12 months (17.66/100 pyr; 95% CI 9.84–31.68/100 pyr; adjusted HR 2.05; 95% CI 1.13–3.71), and slightly lower among those on ART for 12 months or longer (8.11/100 pyr; 95% CI 4.24–15.48/100 pyr; adjusted HR 0.94; 95% CI 0.49–1.81). The incidence

rate of any WHO stage-defining disease among HIV seroconverters was 14.39/100 pyr (95% CI 11.14–18.58/100 pyr) in 1990–1998, and increased to 45.97/100 pyr (95% CI 37.71–56.04/100 pyr) in 1999–2003. The

incidence then declined to 36.42/100 pyr (95% CI 27.14–48.86/100 pyr) in 2004–2005 (the period soon after ART introduction) and 28.29/100 pyr in 2006–2008 (the later period after ART introduction). The reduction in incidence after the introduction of ART persisted after adjustment for confounders, with a significant reduction in incidence in the later period after ART introduction (adjusted HR 0.59; 95% CI 0.39–0.89, compared with 1990–1998; Table 4). Further adjustment for an individual’s ART status attenuated selleck screening library the reduction in incidence rates by calendar time (adjusted

HR 0.72; 95% CI 0.46–1.13, comparing 2006–2008 with 1990–1998), and the decreased incidence among those individuals on ART for longer than 12 months persisted (adjusted HR 0.35; 95% CI 0.20–0.61, compared with those not on ART). In this population of HIV seroconverters, the incidence of any WHO stage-defining disease was substantially lower following ART introduction (2004–2008) compared with the period before ART (1990–2003), and was lowest in the later period (2006–2008). Our analyses suggest Chloroambucil that some of the decline in incidence rates could be attributable to being on ART rather than population-level availability. The overall decline in morbidity following the introduction of ART is similar to that seen in developed countries [15–17]. The incidence rate of having any WHO stage-defining disease in seroconverters increased over time before ART introduction (1990–2003). This is probably attributable to increasing immunosuppression with duration of HIV infection. Previous studies in industrialized settings corroborate this decline in incidence rates after the introduction of ART, although those studies involved prevalent HIV-infected patients and adjusted for CD4 cell count at the time of recruitment [15–17], rather than duration of HIV infection, as we have done in this study. Although incidence rates declined after the introduction of ART, the rate observed during the latest period analysed was twice that in the earliest period analysed.