From: Analysis and reporting of stratified cluster randomized trials—a systematic survey
Outcome | n = 185 |
---|---|
Type of primary outcome; n (%) | |
Continuous | 64 (35) |
Binary | 88 (48) |
Count | 28 (15) |
Time to event | 3 (2) |
Other | 2 (1) |
Unit of analysis; n (%) | |
Cluster-level | 28 (15) |
Individual-level | 143 (77) |
Not clear | 14 (8) |
Primary approach of analysis; n (%) | |
Intention-to-treat | 96 (52) |
Per-protocol | 9 (5) |
Not available | 80 (43) |
Primary method/reported effect estimate adjusted for clustering or stratification; n (%) | |
Clustering and stratification | 71 (38) |
Clustering only | 92 (50) |
Stratification only | 2 (1) |
None | 20 (11) |
Type of adjustment for stratification; n (%) [n = 73] | |
As a covariate | 66 (90) |
Stratum-specific estimate and then combine | 5 (7) |
Stratum-specific estimate | 2 (3) |
Imputed missing data; n (%) | |
No | 150 (81) |
Yes | 35 (19) |
Performed sensitivity analysis; n (%) | |
No | 127 (69) |
Yes | 58 (31) |
Intervention effect significant; n (%) [2-arm trials only; n = 140] | |
No | 76 (54) |
Yes | 64 (46) |
Significance of intervention effect among those adjusted for both clustering and stratification; n (%) [n = 58] | |
No | 32 (55) |
Yes | 26 (45) |
Significance of intervention effect among those not adjusted for both clustering and stratification; n (%) [n = 82] | |
No | 44 (54) |
Yes | 38 (46) |