Risk-of-bias assessment of the randomized controlled trials*
| Study | Adequate sequence generation? | Concealment of allocation? | Blinding of particpants, personnel, outcome assessors? | Incomplete health data addressed? | Free of selective reporting? | Free of other bias? | Overall risk of bias |
|---|---|---|---|---|---|---|---|
| Dormandy et al.7 (PROactive study) | Unclear | Yes | Yes | Yes | Unclear | Unclear (industry sponsored) | Unclear |
| Kahn et al.16 (ADOPT) | Unclear | Yes | Unclear | No | Yes | No (very high attrition rate; industry sponsored) | High |
| Home et al.17 (RECORD) | Yes | Yes | Unclear | Yes | Yes | No (differential withdrawls; industry sponsored) | High |
| Sanofi-Aventis trial21 | Unclear | Unclear | No | Yes | Yes | Unclear (industry sponsored) | High |
Note: ADOPT = A Diabetes Outcome Progression Trial, PROactive trial = PROspective pioglitAzone Clinical Trial in macroVascular Events, RECORD = Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycemia in Diabetes.
↵* Risk of bias was assessed with use of the Cochrane risk-of-bias tool.12