Table 2.
Quality review of studies: Score sheet. SIGN grade was estimated using the general methodology of Harbour and Miller, assigning a grade based on the total score of external validity, internal validity (bias and confounding) as follows: 0-8 (-); 9-12 (+); 13-16 (++)
| Downs and Black, Carter, Carter | ||||||||
|---|---|---|---|---|---|---|---|---|
| Study Quality Assessed | Reporting | External Validity | Internal Validity (Bias) | Internal Validity (Confounding) | Power | Score (of 29) | Harbour and Miller SIGN | Author Comments Reason for Upgrade or Downgrade |
| Anitua et al18 | 9 | 1 | 3 | 3 | 0 | 16 | − | |
| Driver et al19 | 10 | 2 | 6 | 1 | 1 | 20 | − | Downgrade - Efficacy analysis: dropouts high; many treatment violations |
| Friese et al20 | * | |||||||
| Saldalamacchia et al21 | * | |||||||
| Carter et al22 | 8 | 3 | 5 | 3 | 1 | 20 | + | |
| Mazzucco et al24 | 10 | 3 | 5 | 2 | 2 | 22 | + | |
| Margolis et al23 | 8 | 3 | 3 | 4 | 0 | 18 | + | |
| Almdahl et al25 | 11 | 3 | 7 | 5 | 1 | 27 | ++ | |
| Buchwald et al26 | 9 | 0 | 7 | 2 | 0 | 18 | − | Downgrade- Not clear what treatment controls got |
| Englert et al27 | 7 | 3 | 5 | 3 | 0 | 18 | − | Downgrade- Not clear what treatment controls got |
| Peerbooms et al28 | 9 | 2 | 7 | 3 | 1 | 22 | ++ | Upgrade- ITT analysis showed better results than PP |
| Vang et al29 | 10 | 2 | 4 | 1 | 2 | 19 | − | |
| Yoo et al30 | 9 | 2 | 6 | 5 | 0 | 22 | ++ | |
| Everts et al31 | 8 | 3 | 5 | 2 | 0 | 18 | + | |
| Gardner et al32 | 5 | 3 | 4 | 2 | 0 | 14 | + | |
| Khalafi et al33 | 7 | 3 | 3 | 2 | 0 | 15 | + | Upgrade- Large N, propensity scoring techniques used |
| Saratzis et al34 | 10 | 2 | 6 | 4 | 0 | 22 | + | |
| Trowbridge et al35 | 9 | 3 | 4 | 4 | 0 | 20 | + | |
| Kazakos et al36 | 10 | 3 | 5 | 2 | 0 | 20 | + | |
| Spyridakis et al37 | 8 | 3 | 5 | 3 | 0 | 19 | + | |
| Hom et al38 | 10 | 2 | 4 | 2 | 0 | 18 | − | |
*Friese citation was a RCT abstract and the Saldalamacchia citation was a research letter. Neither of these two citations could be scored. Both citations are included because Cochrane would have included them in their analysis.