Anaemia was defined as a haemoglobin level ≤12 or ≤14 mg/dL for women and men, respectively [17]. Patients could develop anaemia or, for those with anaemia, worsening anaemia was defined as a haemoglobin level ≤8 mg/dL. For the liver function tests, 40 IU/L was taken as the ULN (for Pexidartinib research buy both ALT and AST) [18]. Patients were followed until they experienced an event or to the date of their last measurement for each clinical or laboratory marker in EuroSIDA. It should be noted that not all patients in all groups had information on these markers available for all analyses; therefore, the number of patients included in each analysis

differed according to the availability of data. Patients with the event at baseline were excluded from analyses. Any factor that was significant at the 10% level in univariate analyses Staurosporine in vivo (P<0.1) was included in multivariate analyses. In multivariate analyses, statistical significance was attained

if P<0.05. All analyses were performed using sas 9.1 (SAS Institute, Cary, NC, USA). A total of 6634 patients started a nevirapine- (1600; 24%), efavirenz- (3109; 47%) or lopinavir- (1925; 29%) based cART regimen after 1 January 2000. A total of 1750 patients (26%) were excluded from the analysis because they had no CD4 cell count or viral load measurement prior to starting treatment: 410 (26%) on nevirapine, 888 (29%) on efavirenz, and 452 (23%) on lopinavir. A total of 1039 patients (21%) were excluded because of previous exposure to any of the three also drugs: 339 on nevirapine (28%), 297 on efavirenz (13%) and 403 on lopinavir (27%). Nine hundred and fifty-nine

patients (25%) did not achieve suppression, had stopped treatment within the first 3 months or did not have sufficient follow-up and were therefore excluded: 248 (29%) on nevirapine, 459 (24%) on efavirenz, and 252 (24%) on lopinavir. Thus, a total of 2886 patients were included in the analysis; 603 of these patients (21%) were on a nevirapine-based cART regimen, 1465 (51%) on an efavirenz-based cART regimen, and 818 (28%) on a lopinavir-based cART regimen. Patients excluded from the analysis had similar characteristics to those included, but were more likely to have previous cART exposure (64%vs. 57%, respectively; P<0.0001) and to have a prior AIDS diagnosis (32%vs. 26%, respectively; P<0.0001). Table 1 compares the characteristics of the patients in each group at the time of starting their new regimen. A lower proportion of patients starting nevirapine were treatment naïve: 28%, compared with 38% of patients starting efavirenz and 38% of patients starting lopinavir. Patients on nevirapine had a higher median CD4 count [359 cells/μL; interquartile range (IQR) 230–583 cells/μL] and a lower median viral load (2.70 log10 copies/mL; IQR 1.70–4.56 log10 copies/mL) compared with those on efavirenz [median CD4 count 323 cells/μL (IQR 190–535 cells/μL) and median viral load 3.59 log10 copies/mL (IQR 1.70–4.