Table 3.
Treatment effect estimates for the effect of azithromycin on PEx estimated across study design scenarios and with varying regression methods.
| Treatment Effects | |||||
|---|---|---|---|---|---|
| Study Design Scenarios | OPTIMIZE AZM (n=77) vs. |
Cox Regression HR (95%CI) p-value |
Poisson Regression HR (95% CI) p-value |
||
| Naive1 | Robust2 | Naive1 | Robust2 | ||
| OPTIMIZE Only | N=86 OPTIMIZE Placebo | 0.55 (0.35, 0.87) p=0.01 |
N/A | 0.55 (0.34, 0.86) p=0.01 |
N/A |
| Pooling OPTIMIZE with EPIC | N=86 OPTIMIZE Placebo with N=304 EPIC | 0.83 (0.56, 1.20) p=0.33 |
0.60 (0.37, 0.95) p=0.03 |
0.85 (0.57, 1.22) p=0.40 |
0.60 (0.46, 0.77) p<0.001 |
| Augmenting OPTIMIZE with EPIC 3 | N=43 OPTIMIZE Placebo with N=304 EPIC | 0.85 (0.58, 1.26) p=0.43 |
0.62 (0.39, 1.00) p=0.054 |
0.87 (0.59, 1.29) p=0.51 |
0.63 (0.48, 0.82) p<0.001 |
| Replacing OPTIMIZE with EPIC | N=0 OPTIMIZE Placebo with N=304 EPIC | 0.89 (0.60, 1.33) p=0.58 |
0.63 (0.37, 1.08) p=0.09 |
0.92 (0.61, 1.34) p=0.69 |
0.63 (0.50, 0.80) p<0.001 |
AZM= Azithromycin; HR= Hazard Ratio
1
Naive approach does not account for baseline differences between EPIC and OPTIMIZE study populations.
2
Robust approach accounts for baseline differences between EPIC and OPTIMIZE study populations using inverse probability weighting (IPW).
3
Results presented are based on the average of 20 repetitions of sampling 50% of OPTIMIZE control participants.