Meta-analysis results for moderator analyses pooling all outcomes across all tendinopathies
| Moderator | Estimate of mean (95%CrI) | Probabilities | Study VPC (75%CrI) | Outcome VPC (75%CrI) | Measurement VPC (75%CrI) | |
| Assessment duration | Short (≤12 weeks) 923 outcomes from 144 trials | 0.97 (0.85 to 1.1) | p(Medium>Short) > 0.999 | 0.71 (0.66 to 0.75) | 0.27 (0.23 to 0.32) | 0.02 (0 to 0.04) |
| Medium (13–52 weeks) 442 outcomes from 23 trials | 1.2 (1.1 to 1.4) | p(Long>Medium) = 0.980 | ||||
| Long (>52 weeks) 51 outcomes from two trials | 1.4 (1.2 to 1.6) | p(Long>Short) > 0.999 | ||||
| Symptom duration | 1 year or less 308 outcomes from 44 trials | 1.4 (1.1 to 1.7) | p(1yr>2yr) = 0.920 | 0.81 (0.76 to 0.84) | 0.17 (0.14 to 0.21) | 0.03 (0.00 to 005) |
| 2 years or less 258 outcomes from 30 trials | 1.1 (0.74 to 1.4) | p(2yr>+2yr) = 0.772 | ||||
| Over 2 years 381 outcomes from 33 trials | 0.88 (0.56 to 1.2) | p(1yr>+2yr) = 0.989 | ||||
| Supervision | Supervised 354 outcomes from 35 trials | 1.4 (1.1 to 1.7) | p(Supervised> Unsupervised) = 0.989 | 0.77 (0.73 to 0.81) | 0.20 (0.17 to 0.24) | 0.03 (0.00 to 0.05) |
| Unsupervised 914 outcomes from 112 trials | 0.99 (0.83 to 1.2) | |||||
Crl, credible interval; p, subjective Bayesian probability; VPC, variance partition coefficient.