DO CUFF TEARS AFFECT FUNCTIONAL OUTCOMES IN SYMPTOMATIC SHOULDERS? A RETROSPECTIVE STUDY OF OVER 400 CASES

Abstract

Data were collected on patients undergoing subacromial decompression (SAD) in our routine practice from 1998 to 2004. All patients had clinical signs of subacromial pain and had failed conservative treatment. Tears were not repaired. Data on age, gender, arm dominance and presence of cuff tear at operation were recorded. At six months post-operatively, patients were assessed with the Constant Score (CS). Data were analysed using multiple linear regression. Data on 427 patients were collected, 168 having rotator cuff tears. The mean age was 55.41 years (SD 12.00). There were 233 women and 191 men. There were no significant gender differences between the two groups (x2=2.34, df=1, p=0.13). The group with cuff tears were significantly older by 10.24 years (p< 0.001, 8.10 to 12.38). Power of the study was 90% at the 5% significance level. Multiple linear regression showed that gender, age and presence of a cuff tear all had a significant effect on the CS. Gender B = 5.52 (1.99 to 9.06) p = 0.002 Age B = −0.31 (−0.48 to −0.15) p < 0.001 Cuff tear B = −5.51 (−9.48 to −1.55) p = 0.007 Hand dominance and the side operated upon had no significant effect. We found that in patients with symptomatic shoulders, who had failed conservative treatment and undergone an SAD, the CS at six months follow up was lower in those who had a cuff tear. There was increased tear prevalence in slightly older patients, and these scored on average 5.5 (1.5 to 9.5) points less than expected on the CS, p = 0.007, after allowing for age and gender. We believe that the surgeon should discuss with the patient whether a 5.5 points lower score is enough of a difference to warrant a formal rotator cuff repair rather than debridement and SAD alone, and a joint decision should be made.