Hazard ratios were corrected for regression dilution bias with the use of a non-parametric method. Cox proportional-hazards regression versions were used to estimate hazard ratios for clinical end factors as a function of APOC3 genotypes. Versions were adjusted for multiple factors as explained above; for vascular disease end points, the models were adjusted for triglyceride levels additionally. The median follow-up period was 34 years. The observed hazard ratio for vascular risk associated with a 1-unit increment in log-transformed triglyceride levels was used to calculate theoretically predicted risks of ischemic vascular disease and ischemic cardiovascular disease corresponding to the magnitude of the changes in nonfasting triglyceride levels associated with APOC3 genotype.Patients receiving adherence counseling had been 29 % less inclined to experience poor adherence compared to those who received no guidance. Furthermore, those getting intensive early adherence counseling were 59 % less inclined to experience viral failing. However, there is no significant difference in mortality or significant differences in CD4 counts at 18 months follow-up between those that received counseling and the ones who did not. There were no significant variations in adherence also, time to viral failure, mortality, or CD4 counts in individuals who received alarm devices compared to those who did not. The authors conclude: ‘As antiretroviral treatment treatment centers expand to meet a growing demand for HIV care and attention in sub-Saharan Africa, adherence counseling ought to be implemented to diminish the development of treatment failure and spread of resistant HIV.’..