Table 3.
Knowledge of systematic reviews among individuals in charge of European biomedical graduate programs
| Survey items | Yes n (%) | No n (%) | I’m not sure n (%) | I don’t know n (%) |
|---|---|---|---|---|
| It is sufficient to search one database to produce a systematic review (correct answer: no) | 6 (8.6%) | 57 (80.3%) | 5 (7.0%) | 3 (4.2%) |
| Systematic reviews must be produced by one author only (correct answer: no) | 4 (5.6%) | 57 (80.3%) | 6 (8.5%) | 4 (5.6%) |
| Systematic reviews must contain meta-analyses (correct answer: no) | 31 (43.7%) | 22 (31.0%) | 15 (21.1%) | 3 (4.2%) |
| Systematic reviews must have duplicate screening and data extraction (correct answer: yes) | 30 (42.3%) | 3 (4.2%) | 25 (35.2%) | 13 (18.3%) |
| A list of both included and excluded studies must be provided (correct answer: yes) | 48 (67.6%) | 5 (7.0%) | 13 (18.3%) | 5 (7.0%) |
| Quality of included studies must be assessed (correct answer: yes) | 60 (84.5%) | 4 (5.6%) | 4 (5.6%) | 3 (4.2%) |
| In the case of meta-analyses, a heterogeneity test must be done to ensure the results of studies can be combined (correct answer: yes) | 46 (64.8%) | 3 (4.2%) | 13 (18.3%) | 9 (12.7%) |
| Results of meta-analyses must be presented as a funnel plot (correct answer: no) | 26 (36.6%) | 7 (9.9%) | 25 (35.2%) | 13 (18.3%) |
| Results of publication bias analysis must be presented as a forest plot (correct answer: no) | 23 (32.4%) | 8 (11.3%) | 27 (38.0%) | 13 (18.3%) |