Table A2. Stratified analyses by blinding and intention to treat analysis.
| Outcome | Trials (patients) | RR* (95% CI) | P- value | P- value for interaction test |
|---|---|---|---|---|
| Blinding | ||||
| Proven or probable IFI | 0.52 | |||
| Blinded | 4 (2284) | 0.62 (0.42, 0.91) | 0.01 | |
| Not blinded | 14 (2518) | 0.71 (0.52, 0.98) | 0.27 | |
| Invasive aspergillosis | 0.60 | |||
| Blinded | 4 (2284) | 0.47 (0.23, 0.98) | 0.04 | |
| Not blinded | 11 (22196) | 0.59 (0.37, 0.95) | 0.03 | |
| IFI-related mortality | 0.88 | |||
| Blinded | 3 (1686) | 0.68 (0.19, 2.40) | 0.54 | |
| Not blinded | 12 (2586) | 0.75 (0.49, 1.15) | 0.19 | |
| IA-related mortality | 0.60 | |||
| Blinded | 2 (1084) | 0.84 (0.15, 4.91) | 0.85 | |
| Not blinded | 7 (1530) | 0.46 (0.12, 1.77) | 0.26 | |
| Overall mortality | 0.23 | |||
| Blinded | 4 (2284) | 0.92 (0.76, 1.11) | 0.36 | |
| Not blinded | 12 (2586) | 1.07 (0.90, 1.28) | 0.44 | |
| ITT analysis | ||||
| Proven or probable IFI | 0.83 | |||
| ITT | 4 (1445) | 0.74 (0.46, 1.17) | 0.2 | |
| No ITT | 14 (3357) | 0.69 (0.44, 1.07) | 0.1 | |
| Invasive aspergillosis | 0.55 | |||
| ITT | 5 (1800) | 0.47 (0.27, 0.80) | 0.006 | |
| No ITT | 10 (2703) | 0.58 (0.37, 0.91) | 0.02 | |
| IFI-related mortality | 0.60 | |||
| ITT | 5 (1280) | 0.60 (0.27, 1.33) | 0.21 | |
| No ITT | 10 (2992) | 0.77 (0.49, 1.22) | 0.27 | |
| IA-related mortality | NA | |||
| ITT | 1 (59) | Not estimable | NA | |
| No ITT | 8 (2555) | 0.62 (0.23, 1.71) | 0.36 | |
| Overall mortality | NA | |||
| ITT | 0 (0) | Not estimable | NA | |
| No ITT | 16 (4870) | 1.00 (0.88, 1.13) | 0.96 | |
Abbreviations: CI=confidence interval; IA=invasive aspergillosis; IFI=invasive fungal infection; ITT=intention-to-treat; NA=not applicable; RR=relative risk. *RR<1 represents an advantage of mould-active coverage compared with fluconazole using a random-effects model.