1. Stratified analyses: Pain.
| Variable | Number of studies | N of participants corticosteroids | N of participants control | Pain intensity SMD (95% CI) | Heterogeneity I2 (%) | P value* |
| All trials | 26 | 922 | 827 | ‐0.40 (‐0.58 to ‐0.22) | 68% | |
| Allocation concealment | 0.15 | |||||
| Adequate | 2 | 88 | 83 | ‐0.16 (‐0.46 to 0.14) | 0% | |
| Inadequate or unclear | 24 | 834 | 744 | ‐0.42 (‐0.62 to ‐0.22) | 69% | |
| Blinding of participants | 0.64 | |||||
| Adequate | 6 | 220 | 218 | ‐0.34 (‐0.61 to ‐0.06) | 49% | |
| Inadequate or unclear | 20 | 702 | 609 | ‐0.42 (‐0.65 to ‐0.19) | 72% | |
| Blinding of therapists | 0.45 | |||||
| Adequate | 3 | 92 | 92 | ‐0.24 (‐0.66 to 0.17) | 44% | |
| Inadequate or unclear | 23 | 830 | 735 | ‐0.42 (‐0.62 to ‐0.22) | 70% | |
| Intention‐to‐treat analysis | 0.29 | |||||
| Yes | 9 | 236 | 233 | ‐0.26 (‐0.57 to 0.06) | 59% | |
| No or unclear | 17 | 686 | 594 | ‐0.47 (‐0.69 to ‐0.24) | 71% | |
| Type of control intervention | 0.08 | |||||
| Sham injection | 19 | 614 | 526 | ‐0.50 (‐0.72 to ‐0.28) | 65% | |
| No intervention | 7 | 284 | 280 | ‐0.18 (‐0.47 to 0.11) | 63% | |
| Funding independent of industry | 0.80 | |||||
| Yes | 11 | 341 | 333 | ‐0.37 (‐0.55 to ‐0.18) | 26% | |
| No or unclear | 15 | 581 | 494 | ‐0.41 (‐0.70 to ‐0.12) | 78% | |
| Trial size | 0.05 | |||||
| ≥ 50 per trial group | 3 | 205 | 204 | ‐0.13 (‐0.37 to 0.12) | 34% | |
| < 50 per trial group | 23 | 717 | 623 | ‐0.44 (‐0.65 to ‐0.24) | 67% | |
| Trial size | 0.013 | |||||
| ≥ 100 per trial group | 1 | 103 | 103 | 0.00 (‐0.27 to 0.27) | N/A | |
| < 100 per trial group | 25 | 819 | 724 | ‐0.42 (‐0.61 to ‐0.23) | 66% | |
| Publication type | 0.93 | |||||
| Full journal article | 22 | 785 | 706 | ‐0.40 (‐0.61 to ‐0.20) | 70% | |
| Other type or unpublished material | 4 | 137 | 121 | ‐0.38 (‐0.84 to ‐0.08) | 65% | |
| Ultrasound guidance of injections | 0.71 | |||||
| Yes | 2 | 70 | 70 | ‐0.62 (‐1.83 to 0.58) | 89% | |
| No or unclear | 24 | 852 | 757 | ‐0.39 (‐0.57 to ‐0.20) | 67% | |
| Use of local anaesthetic | 0.41 | |||||
| Yes | 5 | 172 | 157 | ‐0.55 (‐0.93 to ‐0.16) | 62% | |
| No or unclear | 21 | 750 | 670 | ‐0.36 (‐0.57 to ‐0.15) | 70% | |
| Concomitant viscosupplementation | 0.08 | |||||
| Yes | 4 | 129 | 127 | ‐0.16 (‐0.42 to 0.09) | 4% | |
| No or unclear | 22 | 793 | 700 | ‐0.46 (‐0.67 to ‐0.25) | 71% | |
| Concomitant joint lavage | ≤ 0.001 | |||||
| Yes | 4 | 197 | 187 | ‐0.06 (‐0.26 to 0.15) | 0% | |
| No or unclear | 26 | 725 | 640 | ‐0.57 (‐0.78 to ‐0.35) | 72% | |
| Use of crystalline preparation | 0.82 | |||||
| Yes | 18 | 623 | 562 | ‐0.47 (‐0.69 to ‐0.24) | 72% | |
| No or unclear | 12 | 299 | 265 | ‐0.52 (‐0.90 to ‐0.14) | 76% | |
| Prednisolone equivalence dose | 0.53 | |||||
| ≥ 50 mg | 17 | 520 | 470 | ‐0.55 (‐0.85 to ‐0.25) | 80% | |
| < 50 mg | 13 | 402 | 357 | ‐0.43 (‐0.66 to ‐0.20) | 56% | |
Number of randomised comparisons are shown in "number of studies" for stratified analyses according to use of lavage as co‐intervention, crystalline preparation, prednisolone equivalence. *P value for interaction. N/A: not available.
CI: confidence interval SMD: standardised mean difference