Table 3.
Risk of virologic failure or death over 2 years in the trial and in persons with HIV diagnosed in the United States during 2008–2014.
| ACTG A5095 trial |
Target population |
||||||
|---|---|---|---|---|---|---|---|
| 2-year risk (%) | 2-year RD (%) (95% CI) | HR (95% CI) | 2-year risk (%) | 2-year RD (%) (95% CI) | HR (95% CI) | ||
| Conventional ITT effecta | Three-drug armb | 24.5 | 0 (REF) | 1 (REF) | 24.4 | 0 (REF) | 1 (REF) |
| Four-drug armc | 23.1 | −1.4 (−8.0, 5.1) | 0.91 (0.68, 1.24) | 24.0 | −0.4 (−6.2, 5.1) | 0.97 (0.70, 1.34) | |
| Dropout-weighted ITT effect | Three-drug arm | 25.3 | 0 (REF) | 1 (REF) | 25.1 | 0 (REF) | 1 (REF) |
| Four-drug arm | 23.2 | −2.1 (−8.8, 4.6) | 0.89 (0.66, 1.20) | 24.2 | −0.9 (−6.9, 5.3) | 0.95 (0.68, 1.32) | |
| Per-protocol effectd | Three-drug arm | 24.6 | 0 (REF) | 1 (REF) | 24.4 | 0 (REF) | 1 (REF) |
| Four-drug arm | 23.0 | −1.7 (−8.2, 5.9) | 0.90 (0.67, 1.23) | 23.7 | −0.7 (−6.7, 5.5) | 0.96 (0.69, 1.34) | |
RD, risk difference; CI, confidence interval.
a
Assuming independent censoring.
b
Three-drug standard-of-care-regimen included zidovudine, lamivudine, and efavirenz.
c
Four-drug regimen including zidovudine, lamivudine, efavirenz, and abacavir.
d
Effect accounted for informative dropout, protocol deviation in the trial, and additionally accounted for sampling bias in the target population.