Table 3.
Effect of strategy on self-reported fatigue by intervention type.
| Outcome | No. studies | No. patients | Effect measure | Effectb | 95% CI | I2 (%) | P value |
|---|---|---|---|---|---|---|---|
| Physical activity | |||||||
| All physical activity interventions vs. any control | |||||||
| All scales | 8 | 607 | SMD | −0.44 | −0.64 to −0.24 | 24% | <0.0001 |
| Fatigue scale (Chinese version) | 3 | 453 | WMD | −4.11 | −6.97 to −1.25 | 80% | 0.005 |
| PedsQL multidimensional Fatigue scalec | 3 | 91 | WMD | −2.07 | −7.12 to 2.97 | 0% | 0.42 |
| All physical activity interventions vs. non-active controla | |||||||
| All scales | 7 | 560 | SMD | −0.48 | −0.69 to −0.27 | 23% | <0.0001 |
| Fatigue scale (Chinese version) | 3 | 453 | WMD | −4.11 | −6.97 to −1.25 | 80% | 0.005 |
Abbreviations: SMD, standardized mean difference; WMD, weighted mean difference; CI, confidence interval.
a
Non-active control defined as usual care, attention control, waitlist control or sham control.
b
A negative SMD or WMD suggests that the intervention is better than control at reducing fatigue severity.
c
Directionality of this scale was inverted for synthesis to be consistent with the other scales for WMD and to permit calculation of SMD.