All-cause mortality after five years on antiretroviral therapy (ART) – 1000 simulations of 1000 patients in cohorts with or without routine viral load monitoring.

ART, antiretroviral therapy; VL, routine viral load monitoring. A (baseline scenario): identical virologic failure rates in both monitoring strategies, switch to second-line ART immediately after confirmed failure. B (delayed switching): identical virologic failure rates in both monitoring strategies, switch to second-line ART after a realistic delay (see Table 1 for parameters). C (higher virologic failure rates with CD4 monitoring): rate of virologic failure set to be 2 or 3 times higher with CD4 monitoring by adjusting the scale parameter of the Weibull distribution (Table 1), switch to second-line ART immediately after confirmed failure. *Uncorrected mortality: mortality based on observed mortality from data. **Corrected mortality: mortality based on observed mortality, observed LTFU and estimated mortality among patients lost [22]. ***Ratios of uncorrected mortality, comparing VL with CD4 monitoring.