Quality of included studies and assessment of bias, as evaluated by Downs and Black (28)
| Study | Year | Study design | Measure | Score | Overall quality* | |||
|---|---|---|---|---|---|---|---|---|
| Quality of reporting | External validity | Internal validity | Power | |||||
| Moses (32) | 1951 | Retrospective cohort | 2 | 1 | 1 | 0 | 4 | Poor |
| Flanc et al (33) | 1969 | Retrospective cohort | 7 | 2 | 6 | 0 | 14 | Fair |
| Miller et al (39) | 1976 | RCT | 5 | 1 | 6 | 0 | 12 | Poor |
| Prerovský et al (40) | 1988 | RCT | 6 | 2 | 6 | 0 | 14 | Fair |
| Lassen and Borris (29) | 1991 | Prospective cohort | 3 | 1 | 4 | 0 | 8 | Poor |
| Pearse et al (34) | 2007 | Retrospective cohort | 9 | 0 | 6 | 0 | 15 | Fair |
| Vioreanu et al (41) | 2007 | RCT | 7 | 3 | 6 | 0 | 16 | Fair |
| Chandrasekaran et al (35) | 2009 | Retrospective cohort | 8 | 1 | 8 | 0 | 17 | Fair |
| Sorbello et al (42) | 2009 | RCT | 10 | 3 | 7 | 0 | 20 | Good |
| Amin et al (46) | 2010 | Secondary analysis of RCT | 11 | 3 | 9 | 0 | 23 | Good |
| Frantzides et al (36) | 2012 | Retrospective cohort | 7 | 3 | 4 | 0 | 14 | Fair |
| Cassidy et al (37) | 2014 | Retrospective cohort (NSQIP) | 8 | 3 | 8 | 0 | 19 | Good |
| Bhatt et al (31) | 2017 | Retrospective cohort | 8 | 2 | 6 | 0 | 16 | Fair |
| Wang et al (43) | 2016 | RCT | 8 | 1 | 9 | 0 | 18 | Fair |
| Karic et al (30) | 2017 | Prospective cohort | 8 | 3 | 6 | 1 | 18 | Fair |
| de Almeida et al (44) | 2017 | RCT | 11 | 3 | 11 | 1 | 26 | Excellent |
| Guo et al (45) | 2019 | RCT | 11 | 2 | 11 | 1 | 25 | Good |
| Silver et al (38) | 2019 | Retrospective cohort | 9 | 3 | 5 | 0 | 17 | Fair |
Note: NSQIP = National Surgical Quality Improvement Program, RCT = randomized controlled trial.
↵* Scale for quality scores: poor: ≤ 14; fair: 15–19; good: 20–25; excellent: 26–28.