High Critical Shoulder Angle Values Are Associated With Full-Thickness Posterosuperior Cuff Tears and Low Values With Primary Glenohumeral Osteoarthritis.

PURPOSE

To compare critical shoulder angle (CSA) measurements using high-quality radiographs in the following groups: Group 1: symptomatic atraumatic full-thickness rotator cuff (RC) tears; Group 2: symptomatic primary glenohumeral osteoarthritis (GHOA); and Group 3: no RC tear or primary GHOA being treated for glenohumeral instability or symptomatic labral pathology (control group).

METHODS

A prospective observational case control study with 10 shoulders in each group was performed GHOA and full-thickness RC tears were diagnosed by radiographs and magnetic resonance imaging (MRI). For these three groups, the exclusion criteria were the following: Group 1) partial thickness RC tears, traumatic RC tears, isolated subscapularis tears, and advanced cuff tear arthropathy with erosion of the superior glenoid; Group 2) secondary causes of glenohumeral arthritis; coexistent full-thickness RC tear; and Group 3) glenoid bone lesions that may affect the CSA measurement. Only shoulders with adequate radiographic quality (Suter-Henninger type A and C) were eligible. A one-way ANOVA, followed by Tukey multiple pairwise-comparisons test, was performed to compare the groups. Interobserver and intraobserver reliability was assessed using Intraclass Correlation Coefficients (ICC).

RESULTS

Mean CSA values were 37.4° ± 4.7 (RC tear group), 28.9° ± 2.4 (GHOA group), and 32.8° ± 1.1 (control group). The CSA of the RC group was higher than the control group (P = .006) and the GHOA group (P = .000). The CSA of the GHOA group was lower than the control group (P = .027). Intraobserver and interobserver reliabilities for the CSA measurement were excellent (Observer 1 [ICC]: .986 [95% CI .970-.993]; Observer 2 [ICC]: .976[95% CI .951-.989]; and Observer 1v2: 0.968[95% CI .933-.985]).

CONCLUSIONS

There is a difference in the CSA between patients with symptomatic atraumatic full-thickness RC tears (4.6° higher than the control group), symptomatic GHOA (3.8° lower than the control group), and glenohumeral instability or labral pathology with no RC tear or GHOA.

LEVEL OF EVIDENCE

Level 2, prospective observational case control diagnostic study.