Table 3. GRADE summary of evidence.
| Population-People with frailty having surgery | ||||||
|---|---|---|---|---|---|---|
| Intervention-Exercise therapy | ||||||
| Control-No or non-standardized exercise therapy | ||||||
| Quality assessment | ||||||
| Participants (studies) | Risk of bias | Inconsistency | Indirectness | Imprecision | Overall quality | Comment |
| Postoperative function | ||||||
| 503 (4) | Moderate | Low | No serious indirectness | No serious imprecision | Moderate1 | Significant improvement in most physical performance measures in 3/4 studies |
| Postoperative health related quality of life | ||||||
| 314 (2) | Serious | Moderate | No serious indirectness | No serious imprecision | Low2 | Significant improvement in physical and mental health in a randomized trial |
| Postoperative length of stay | ||||||
| 245 (2) | Moderate | Moderate | No serious indirectness | Moderate imprecision | Very low3 | Decreased length of stay in larger observational study; none in small pilot randomized trial |
1. Downgraded as not all studies showed improvement, and 1 was non-randomized
2. Downgraded due to unclear allocation concealment and blinding in RCT, no effect in observational study
3. Downgraded due to inconsistency, positive effect was from a high risk of bias observational study