Table 1 Reporting approaches used in the description of the trial results
Author, year | Diagram of SWT rollout | CONSORT-style diagram | Assessment of balance | Reported k/ICC | Group summaries between crossovers | Secular trend reported |
|---|---|---|---|---|---|---|
Bacchieri, 2010 [29] | No | No | No | No | No | Yes, graph |
Bashour, 2013 [22] | No | Yes | Yes: individual-level covariates at enrolment between conditions | Yes | No | Yes, fixed effects from model and a graph of the Likert responses by group |
Durovni, 2013 [12] | No | Yes | No | Yes | No | No |
Fuller, 2012 [23] | Yes | Yes | No | No | No | Yes, graphed modelled effect by condition |
Gruber 2013 [21] | Yes | Yes | Yes: individual and household-level covariates at baseline by condition, weighted for time in condition | No | No | Yes, graphed by condition |
Horner, 2012 [27] | Yes | No | Yes: cluster-level at baseline by group | No | Yes | Yes, model parameters, sub-groups, and graphs by group |
Mhurchu, 2013 [24] | Yes | Yes | Yes: all schools described, with group indicated, but not summarised by group | No | Yes | Yes, using step-group summaries |
Roy, 2013 [25] | Yes | Yes | Yes: cluster-level covariates and individual-level covariates at enrolment by condition | No | No | No |
Schultz, 2014 [28] | Yes | No | Yes: individual-level at enrolment by condition | No | No | No |
Stern, 2014 [26] | Yes | Yes | Yes: cluster-level covariates at baseline and individual-level covariates at enrolment by condition | No | No | No |