Table 4 Challenges in estimating antimalarial drug efficacy and possible alternatives
Challenges | Current approach | Alternative approach | Software |
|---|---|---|---|
Competing risk event | Censored on the day of event [3] | Cumulative incidence function | cmprsk package R [30] (Additional file 1: Section B) |
Interval censoring | Ignored | Interval censored survival estimates | survival package R [57] (Additional file 1: Section C) |
K–M for multicentre studies | No specific recommendation | Use of meta-analysis approach [40] | MetaSurv package R [41] (Additional file 1: Section D) |
Comparing K–M estimates | Current comparison based on whole survival curve (log-rank test) | Comparison at a fixed point in time based on complementary log–log transformation [42] | R script available as additional file (Additional file 1: Section E) |
Demonstrating non-inferiority | Based on the cured proportion | Based on the difference of two K–M estimates after complementary log–log transformation (or on hazards ratio scale) | R script available as additional file (Additional file 1: Sections F, G, H) |