Clinical results of rotator cuff repairs using the same therapeutic exercises from an early stage after surgery were investigated and compared with the preoperative X-ray findings. 70 shoulders with rotator cuff tears treated operatively and followed by active-assistive exercises in a supine position from 1 week and active exercises standing from 3 weeks after surgery were evaluated. The average age was 63 years old and follow-up periods were 11.9 months. The operative procedures were McLaughlin's method in 41 shoulders, side to side sutures in 9, acromioplasty in 17 and a tendon transfer in 3. The preoperative X-ray findings were classified into 5 groups: group A, superiorly migrated humeral head and narrowed joint space in 4 shoulders; group B, only superior migration in 7; group C, only narrowed joint space in 3; group D, no changes in 36; group E, subacromial spurs more than 5 mm without superior migration in 20. The postoperative results were assessed according to the JOA score and the failure rate of each group was investigated. No significant differences in the preoperative JOA score were found in the 5 groups. The postoperative JOA scores were 62.9±3.6 in group A, 75.1±12.6 in B, 78.0±12.2 in C, 90.4±7.5 in D, and 83.6±11.1 in E. Especially groups D and E treated by McLaughlin's methods showed a significant higher score at follow-up. Rotator cuff repairs with the same therapeutic exercises from an early stage after surgery for groups A, B and C showed poor results. Groups A and B with a superior migration of the humeral head showed a high failure rate of 40 % compared with 4.7% in group C, D and E with no superior migration. For these cases more careful postoperative exercises were recommended.
