Symptoms of pain do not correlate with rotator cuff tear severity: a cross-sectional study of 393 patients with a symptomatic atraumatic full-thickness rotator cuff tear

Abstract

Background: For many orthopaedic disorders, symptoms correlate with disease severity. The objective of this study was to determine if pain level is related to the severity of rotator cuff disorders.

Methods: A cohort of 393 subjects with an atraumatic symptomatic full-thickness rotator-cuff tear treated with physical therapy was studied. Baseline pretreatment data were used to examine the relationship between the severity of rotator cuff disease and pain. Disease severity was determined by evaluating tear size, retraction, superior humeral head migration, and rotator cuff muscle atrophy. Pain was measured on the 10-point visual analog scale (VAS) in the patient-reported American Shoulder and Elbow Surgeons (ASES) score. A linear multiple regression model was constructed with use of the continuous VAS score as the dependent variable and measures of rotator cuff tear severity and other nonanatomic patient factors as the independent variables. Forty-eight percent of the patients were female, and the median age was sixty-one years. The dominant shoulder was involved in 69% of the patients. The duration of symptoms was less than one month for 8% of the patients, one to three months for 22%, four to six months for 20%, seven to twelve months for 15%, and more than a year for 36%. The tear involved only the supraspinatus in 72% of the patients; the supraspinatus and infraspinatus, with or without the teres minor, in 21%; and only the subscapularis in 7%. Humeral head migration was noted in 16%. Tendon retraction was minimal in 48%, midhumeral in 34%, glenohumeral in 13%, and to the glenoid in 5%. The median baseline VAS pain score was 4.4.

Results: Multivariable modeling, controlling for other baseline factors, identified increased comorbidities (p = 0.002), lower education level (p = 0.004), and race (p = 0.041) as the only significant factors associated with pain on presentation. No measure of rotator cuff tear severity correlated with pain (p > 0.25).

Conclusions: Anatomic features defining the severity of atraumatic rotator cuff tears are not associated with the pain level. Factors associated with pain are comorbidities, lower education level, and race.

Level of evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.

Fig. 1

Plot of the effects of the significant predictors in the model. The gray bars represent the 95% CI for the mean effect. The mean effect of comorbidity on pain was determined by comparing those with an SCQ score of 2 with those with a score of 10; the VAS pain score (ases3) was increased by 1.07 (95% CI: 0.31, 1.83) in the latter group. HS = high school, Grad = graduate, and Bach = Bachelor’s.

Fig. 2

Nomogram for the model predicting an individual patient’s VAS pain score. First, each variable is marked on the appropriate scale, and the number of points for each is derived from the “Points” scale at the top of the nomogram. Then the points are totaled, and the total value is marked on the “Total Points” scale. Viewing down, one then derives the predicted pain score on the scale labeled “Linear Predictor.” For ease of interpretation not all predictors included in the model are shown in the nomogram. HS = high school, Grad = graduate, Bach = Bachelor’s, and SSp = supraspinatus.