(a) Imputation performed |
Yes | 56 (33) |
Worst-case scenario | 19 (34) |
Multiple imputation | 11 (20) |
Last observation carried forward | 8 (14) |
Complete case analysis | 6 (11) |
Best-case scenario | 2 (4) |
Last observation carried forward and worst-case scenario | 2 (4) |
Best-case/worst-case scenario | 3 (5) |
Mean imputation | 1 (2) |
Complete case analysis, multiple imputation using propensity scores and multiple imputation using regression modelling | 1 (2) |
Other and worst-case scenario | 1 (2) |
Other | 1 (2) |
No | 12 (7) |
Not reported | 99 (59) |
Unclear | 1 (1) |
Including NEJM protocols (N=61) |
Yes | 22 (36) |
No | 7 (11) |
Not reported | 31 (51) |
Unclear | 1 (2) |
(b) Sensitivity analyses performed |
Yes | 64 (38) |
Patient population | 13 (20) |
Competing risks | 2 (3) |
Statistical modelling | 2 (3) |
Adjusted for baseline variables | 1 (2) |
Excluded protocol violations | 1 (2) |
On-treatment | 1 (2) |
Patient population/other | 1 (2) |
Unclear | 2 (3) |
Other | 15 (23) |
Missing data | 27 (42) |
Best-case/worst-case scenario | 5 |
Complete case analysis | 3 |
Imputation of missing values | 3 |
Multiple imputation | 3 |
Worst-case scenario | 3 |
Baseline observation carried forward | 1 |
Baseline observation carried forward and complete case analysis | 1 |
Complete case analysis, multiple imputation using propensity scores and multiple imputation using regression modelling | 1 |
Complete case analysis and missing not at random | 1 |
Complete case analysis and best-case scenario | 1 |
Different methods | 1 |
Last observation carried forward | 1 |
Modelling | 1 |
Observed failure | 1 |
Worst-case scenario and last observation carried forward | 1 |
No | 103 (61) |
Unclear | 1 (1) |
Including NEJM protocols |
Yes | 38 (62) |
No | 23 (38) |