Overview of Grading of Recommendations, Assessment, Development and Evaluation (GRADE) results for group comparisons with manual therapy techniques
| Outcomes | Intervention and comparison intervention | Trials | Effect size for intervention group | No. of participants (studies) | Quality of the evidence (GRADE) | Result |
|---|---|---|---|---|---|---|
| Pain | ||||||
| Manipulation ×placebo (sham or placebo therapy) | Atkinson et al18 Haik et al19 Kardouni et al91 | From 0.31 to 0.81 lower | 155 (three studies) | ⊕⊝⊝⊝ very low*†‡ due to inconsistency, indirectness, imprecision | No difference between groups | |
| Mobilisation associated with exercises ×exercises | Bang and Deyle17 Conroy and Hayes86 Cook et al88 Kachingwe et al74 Kromer et al88a | From 1.15 to 1.99 lower | 264 (five studies) | ⊕⊕⊕⊕ high§ due to large effect | Pain decrease in the experimental group | |
| Combined therapy (manual therapy, exercises, scapular control)× placebo or no treatment | Bennel et al84 Kachingwe et al74 | From 0.89 to 1.7 lower | 152 (two studies; three comparisons) | ⊕⊕⊕⊝ moderate‡§¶ due to imprecision, publication bias, large effect | Pain decrease in the experimental group | |
| Manual therapy×corticosteroid injection | Rhon et al83 Winters et al87 | From 0.32 to 2.96 lower | 183 (two studies) | ⊕⊕⊝⊝ low*‡ due to inconsistency, imprecision | No difference between groups | |
| Function | ||||||
| Mobilisation associated with exercises ×exercises | Bang and Deyle17 Cook et al88 Kachingwe et al74 Kromer et al, 2013 | From 0.93 to 2.1 higher | 250 (four studies) | ⊕⊕⊕⊝ moderate† due to indirectness | No difference between groups | |
| Combined therapy (manual therapy, exercises, scapular control)× placebo or no treatment | Bennel et al84 Kachingwe et al74 | From 0.85 to 1.3 higher | 152 (two studies; three comparisons) | ⊕⊕⊕⊝ moderate‡§¶ due to imprecision, publication bias, large effect | Functional improvement in the experimental group | |
| Function, long term | ||||||
| Combined therapy (manual therapy, exercises, scapular control)× placebo or no treatment | Bennel et al84 Dickens et al90 | From 0.67 to 1.19 higher | 205 (two studies) | ⊕⊕⊕⊕ high§ due to large effect | Functional improvement in the experimental group | |
| Range of motion, total | ||||||
| Combined therapy (manual therapy, exercises, scapular control)×placebo or no treatment | Kachingwe et al74 | From 2.49 lower to 4.28 higher | 32 (one study, two comparisons) | ⊕⊕⊝⊝ low‡§ due to imprecision, publication bias | No difference between groups | |
| Range of motion, flexion | ||||||
| Mobilisation associated with exercises×exercises | Conroy and Hayes86 Kachingwe et al74 | From 2.49 lower to 4.28 higher | 48 (two studies) | ⊕⊕⊝⊝ low†‡ due to imprecision, indirectness | No difference between groups | |
| Range of motion, abduction | ||||||
| Mobilisation associated with exercises×exercises | Conroy and Hayes86 Kachingwe et al74 | From 0.16 to 4.53 higher | 48 (two studies) | ⊕⊕⊝⊝ low†‡ due to imprecision, indirectness | No difference between groups | |
*Inconsistency: there was statistical or effect size heterogeneity between trials.
†Indirectness: there was clinical heterogeneity between trials.
‡Imprecision: there was sparse data with <200 participants for the comparison.
§Large effect: a quality point was added when >75% of the trials presented large effect (between 0.21 and 0.79) for the experimental group and not for the comparison group.
¶Publication bias: there were two comparisons from the same research group.
Bold terms refer to the level of evidence according to GRADE.