The 4-year stratification

was selected a priori based on the fact that the epidemic in IDUs began in 1998. The analysis was repeated using the year 1997 as an alternative cut-off in order to analyse the possible effect of cART. The interval 1998–2001 was defined as the first stage of the sub-epidemic among IDUs, whereas the interval 1985–1989 was defined as the first stage of the sub-epidemic among MSM and heterosexual cases. The other periods were defined as later stages of the epidemic. SPSS 15.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Chi-square tests, t-tests and Mann–Whitney tests were used to test for differences between the groups. Multivariate logistic regression analysis using backward and forward selection procedures served to identify factors independently associated with late HIV diagnosis and delayed entry to care. Variables

found to be significant BKM120 (P<0.2) in univariate analysis were included in the models. Out of all 934 cases, the reported transmission risk was IDU in 26%, heterosexual transmission in 31% and MSM in 42%. The transmission risk was other or unknown for 11 cases. The characteristics of the study population divided into 4-year calendar periods of HIV diagnosis are shown in Table 1. The study population Belnacasan represents 77% of IDU, 66% of MSM and 42% of heterosexually transmitted HIV cases reported to the NIDR surveillance system nationwide between 1985 and 2005 (n=1597). The annual number of newly diagnosed HIV cases in the study population follows the same trends as those for the whole country (Fig. Fludarabine mw 1). Out

of 934 patients, 62% had their CD4 cell count measured on the day of the first clinic visit or within 90 days after the visit. Thirty-eight per cent had their CD4 cell count measured prior to the clinic visit. The median CD4 count was 419 cells/μL. Of all cases, 21% presented with a CD4 count ≤200 cells/μL, 6% with CD4 <50 cells/μL and 9% presented with an AIDS-defining illness. Altogether, 23% were classified as diagnosed late (CD4 <200 cells/μL, or AIDS within 3 months of HIV diagnosis). Within the first year after HIV diagnosis, 11% had been diagnosed with AIDS. Late diagnosed cases by calendar year of diagnosis and HIV transmission risk are presented in Fig. 1. The proportions of individuals diagnosed late, and predictors of late diagnosis in the multivariate model are presented in Table 2. In the multivariate analyses, individuals diagnosed late were more often older, non-Finnish and less likely to have been tested previously. Compared with female IDUs, male IDUs, male heterosexuals and MSM were at risk of a late diagnosis. Cases diagnosed late were more often diagnosed in primary health care, secondary health care or at an unknown site compared with STD clinics, and in more recent calendar periods. Late diagnosis was rare before 1990 and between 1998 and 2001.