Table 4 Reporting of important other analytic considerations
| Â | Number of articles n (%) |
|---|---|
Analysis of studies with hierarchical clustering (N = 37) | |
 No clustering consideration | 28 (76) |
 Variance corrected method | 4 (11) |
 Frailty models | 3 (8.1) |
 Multilevel regression | 2 (5.4) |
Censoring description | |
 Yes | 50 (66) |
 No | 26 (34) |
Test of effect modification/interaction (N = 4) | |
 Likelihood ratio test | 2 (50) |
 Chi-square test of homogeneity | 1 (25) |
 Not reported | 1 (25) |
Handling of missing exposure data | |
 Single imputation | 2 (2.6) |
 Multiple imputation | 3 (4.0) |
 A missing category included | 1 (1.3) |
 Not reported | 70 (92) |