Proportion of positive, negative, and neutral abstracts in each subgroup
| Subgroup | Sample (n)* | Positive | Negative | Neutral |
| All included studies | 115 | 53 (46.1%) | 22 (19.1%) | 50 (34.8%) |
| Both general and neuraxial anesthesia | 26 | 12 (46.2%) | 7 (26.9%) | 7 (26.9%) |
| Only general anesthesia | 24 | 12 (50%) | 7 (29.2%) | 5 (20.8%) |
| Only neuraxial anesthesia | 49 | 22 (44.9%) | 5 (10.2%) | 22 (44.9%) |
| Hip arthroplasty | 27 | 12 (44.4%) | 5 (18.5%) | 10 (37.1%) |
| Knee arthroplasty | 92 | 41 (44.6%) | 19 (20.6%) | 32 (34.8%) |
| Studies with >80 patients | 55 | 17 (30.9%) | 17 (30.9%) | 21 (38.2%) |
| Studies with <80 patients | 60 | 36 (60%) | 5 (8%) | 19 (32%) |
| High risk of bias studies | 37 | 23 (62.2%) | 3 (8.1%) | 11 (29.7%) |
| Low risk of bias studies | 78 | 30 (38.5%) | 19 (24.4%) | 29 (37.1%) |
| Observational studies | 40 | 15 (37.5%) | 11 (27.5%) | 14 (35%) |
| Randomized control trials | 75 | 38 (50.7%) | 11 (14.7%) | 26 (34.6%) |
*These subgroups are not all mutually exclusive and there may be overlap between groups.