Table 2.
Primary outcomes
| Azithromycin plus usual care | Usual care alone | Estimated treatment effect (95% Bayseian credible interval) | Probability of meaningful effect | Probability of superiority | ||
|---|---|---|---|---|---|---|
| Primary outcomes (primary analysis population) | ||||||
| First reported recovery | 402/500 (80%) | 631/823 (77%) | .. | .. | .. | |
| Time to first reported recovery (days) | 7 (3 to 17) | 8 (2 to 23) | 1·08 (0·95 to 1·23)* | 0·23* | 0·89* | |
| Hospitalisation or death at 28 days | 16/500 (3%) | 28/823 (3%) | 0·3% (−1·7 to 2·2)† | 0·042† | 0·64† | |
| Primary outcomes (SARS-CoV-2-positive analysis population) | ||||||
| First reported recovery | 136/186 (73%) | 163/236 (69%) | .. | .. | .. | |
| Time to first reported recovery (days) | 9 (4 to not reached) | 13 (5 to not reached) | 1·12 (0·91–1·38)* | 0·47* | 0·86* | |
| Hospitalisation or death at 28 days | 11/186 (6%) | 17/236 (7%) | 1·6% (−3·1 to 6·2)† | 0·43† | 0·76† | |
Data are n/N (%) or median (IQR). HR=hazard ratio.
*
Estimated HR derived from a Bayesian piecewise exponential model adjusted for age and comorbidity at baseline, with 95% Bayesian credible interval. HR >1 favours azithromycin.
†
Estimated absolute benefit in percentage of hospitalisation or death derived from a Bayesian logistic regression model adjusted for age and comorbidity at baseline, with 95% Bayesian credible interval. A positive value favours azithromycin.