| Keywords |
• General medical • Statistical • Cochrane • Risk of bias |
| Problem(s) |
• Meta-analyses and forest plots presented without considering risk of bias / quality |
| Article Type | Empirical |
| Article Subtype | Cross-sectional survey/Methodological systematic review |
| First Author Country | United States |
| Aim | To assess the number, style and content of forest and funnel plots presented in a previously studied cross-sectional sample of 300 systematic non-Cochrane systematic reviews and Cochrane reviews. |
| Level of Investigation | Descriptive |
| Summary of Findings | All Cochrane reviews had forest plots, but 10% contained no data and 80% had 3 or fewer studies; 10% had no studies). Plots with 0 or 1 studies serve little purpose other than to convey that more research is needed. Conversely 84% of non-Cochrane reviews did not contain even one forest plot. Only 3% of both Cochrane and non-Cochrane reviews presented funnel plots. All Cochrane review plots and 44% of non-Cochrane review plots presented studies in either alphabetical order or some other order that had little potential for illuminating the meaning of the data. |
| Number of systematic reviews included | 300 |
| Number of eligible systematic reviews assessed | 300 |
| Treatment impacted | No |
| Treatment impacted description | |
| Interpretation impacted | Not Applicable |
| Interpretation impacted description |
