001). The estimates for calendar year were unaffected by the choice of lagging window (6–12, 12–24 or 24–36

months) for the introduction of new drugs and classes. Similarly, the introduction of an additional variable coding for long delays of >6 months between viral load determinations did not alter the findings. INK-128 This study of a large national observational cohort demonstrated a continuous improvement of virological and immunological effectiveness of ART over recent years. Between 2000 and 2008, the proportion of participants with three consecutive viral load values <50 copies/mL increased from 37 to 64% and the proportion with CD4 counts >500 cells/μL rose from 40 to >50%. In our study we were able to adjust for adherence, treatment interruptions, stable partnership and active hepatitis virus coinfections without

appreciable effects on the time trends, but the improvements Bleomycin chemical structure could only partially be attributed to the numerous predictors tested, including the use of new drugs. Of note, we did not find a relevant dilution effect through new participants entering our open clinical cohort over time. Assigning the most unfavourable outcome to individuals who were lost to follow-up or died did attenuate but not offset the time trends. Because, by definition, the number of individuals lost to follow-up increases, a favourable time trend for virological effectiveness is artificially reduced. Further, in a resource-rich country with universal health care, most individuals will continue to receive adequate care and ART outside the cohort. Our findings are consistent with the results from a collaboration of five HIV clinics analysing time trends of virological success during the early years of combination ART from 1996 to 2002 [11]. The authors attributed some of the observed improvements to better starting regimens, and concluded that additional factors, such as increasing clinical experience, may have played an important role.

Clearly, the experience of care providers continues to improve, and greater physician experience is related to better survival [12], earlier adoption of new treatments [13] and increased adherence 4-Aminobutyrate aminotransferase to treatment [14]. In addition, societal factors such as further reductions of HIV-related stigma and improvement in knowledge of patients may also have played a role [15]. In addition to the superior virological outcome, we found that there was an improvement in immunological status over time, especially after 2004. Contrary to our expectations, time trends for the proportion of individuals with CD4 lymphocyte counts >500 cells/μL did not differ between the open and closed cohorts despite the constant influx of new patients with median CD4 counts of 360 cells/μL in 2001 and 420 cells/μL in 2007 (data not shown). This supports observations from the analyses of the virological endpoint suggesting a negligible bias of time trend analyses by cohort design.