Table 3 Comparison of reporting with PRISMA 2009 reporting items
Suggestions in PRISMA 2009a | Total (n = 152) | Cochrane (n = 75) | Non-Cochrane (n = 77) |
|---|---|---|---|
Provide data extraction form | 10 (7%) | 7 (9%) | 3 (4%) |
Report use of a dedicated data extraction form | 77 (51%) | 49 (65%) | 28 (36%) |
Report whether the data extraction form was developed a priori | 49 (32%) | 33 (44%) | 16 (21%) |
Report whether the data extraction form was piloted | 24 (16%) | 14 (19%) | 10 (13%) |
Report whether a training exercise was undertaken | 0 (0%) | 0 (0%) | 0 (0%) |
Report the data collection process (in duplicate, independently or not) | 133 (88%) | 75 (100%) | 58 (75%) |
Report how disagreements were resolved | 107 (70%) | 70 (93%) | 37 (48%) |
Report who extracted the data | 116 (76%) | 69 (92%) | 47 (61%) |
Report the process used to obtain data from investigatorsb | 103 (68%) | 75 (100%) | 28 (36%) |
Report all variables for which data were collectedc | 58 (38%) | 36 (48%) | 22 (29%) |
Report methods used to deal with multiple reports of the same study | 47 (31%) | 33 (44%) | 14 (18%) |