Introduction
Rotator cuff tears are among the most common disorders of the shoulder girdle requiring surgical management, and arthroscopic rotator cuff repair has proven to be an effective and cost-efficient treatment option for rotator cuff tears which have failed conservative management. The interpretation of patient-reported outcome measures (PROMs) is challenging as statistical significance does not necessarily equate to clinical relevance. The purpose of this study is to determine the Minimal Clinically Important Difference (MCID) of the OSS at 5 years post-operatively in a cohort of patients who have undergone arthroscopic rotator cuff repair.
Methodology
Prospectively collected data of 123 patients who underwent arthroscopic rotator cuff repair performed by a fellowship-trained shoulder surgeon in a single high-volume institution between 2015-2017 was retrospectively reviewed. All patients sustained symptomatic rotator cuff tears proven with either ultrasound or Magnetic Resonance Imaging (MRI). All patients were assessed by trained physiotherapists pre-operatively, as well as at 6 months, 2 years, and 5 years post-operatively. Functional outcome was assessed by the Oxford Shoulder Score (OSS). Patients’ satisfaction and expectations fulfilment were captured on Likert scales of 1-6 and 1-7 respectively.
The MCID for OSS was determined through 2 methods; an anchor-based linear regression approach as well as a distribution-based approach. Mean (standard deviation; sd) was used to summarise the OSS and the mean change in the patients’ OSS over time was estimated with 95% confidence intervals. For the anchor-based linear regression approach, the patients’ satisfaction and expectation fulfilment were used as anchors. The MCID was determined using the slope of the line of simple linear regression analysis with the change in OSS as the dependent variables against patient satisfaction and expectation fulfilment. For the distribution-based approach, the standard-deviation (SD) estimate of the MCID was determined as 0.5*SD of the 5-year OSS, with 0.5 taken to represent a moderate effect size.
Results
A total of 123 patients who underwent arthroscopic rotator cuff repair were included in this study. There were 58 male and 65 female patients, and the mean age was 56.6 (SD 10.4) years among males and 59.2 (SD 9.6) years among females (overall 58.0 (SD 10.4) years). The mean pre-operative OSS was 29.8 (SD 10.5). The mean OSS at 6 months, 2 years and 5 years post-operatively were 19.4 (8.5), 14.4 (5.5), and 14.0 (4.9) respectively. At 5 years post-operatively, the MCID of OSS identified by the anchor-based linear regression approach for satisfaction was 2.8 (95% CI 2.0 - 3.6), and that for expectation fulfilment was 2.6 (95% 1.9 – 3.3). The MCID of OSS identified by the distribution-based approach was 2.5.
Discussion And Conclusion
This study was able to determine statistically significant MCID values for OSS at 5 years after arthroscopic rotator cuff repair through 2 different statistical approaches - the anchor-based linear regression approach and the distribution-based approach. Taking the higher value to be the minimum change in score required, the proposed MCID for OSS at 5 years post-operatively is 2.8.