Table 2.
Estimated Regression-to-the-mean
| WOMAC sub-score | SD | σ2b | σ2w | Estimated RTM effect from BL to Y2 (%) | Pain change from BL to Y2 adjusted for estimated RTM effect (%) |
|---|---|---|---|---|---|
| Question 1: Walking on flat surface | 56 | 1602.5 | 1533.5 | 32.6 | 0.9 |
| Question 2: On stairs | 38 | 735.0 | 709.0 | 26.0 | −10.9 |
| Question 3: Lying in bed | 80 | 3872.0 | 2528.0 | 33.7 | 11.2 |
| Question 4: Sitting or lying | 69 | 2551.9 | 2209.1 | 35.6 | 6.3 |
| Question 5: Standing | 57 | 1614.8 | 1634.3 | 33.9 | −0.1 |
| Questions 1, 2, and 5: Weight-bearing | 39 | 734.6 | 786.4 | 27.8 | −15,4 |
| Questions 3 and 4: Non-weight-bearing | 65 | 2433.6 | 1791.4 | 31.2 | −4.6 |
The percentage of RTM of the Target knee as estimated using the method of Barnett et al. [24]. Calculation of σ2w was based on within patient data from screening visit and subsequent baseline visit, between 1 and 3 weeks apart. σ2b: Between patient-variance. σ2w: Within patient-variance. BL Baseline. Y2 Year 2, RTM Regression to the mean