A survey of decision-making processes in the treatment of common shoulder ailments among primary care physicians.

To encourage consistent care for patients with musculoskeletal complaints, the AAOS developed treatment algorithms to aid primary care physicians in the management of these patients. A survey was designed to assess whether a random group of primary care physicians treated their patients in a manner consistent with these algorithms. The AAOS algorithm for shoulder pain was used to develop a questionnaire for primary care physicians. An Internet company provided access to a national base of physicians who volunteered to complete the survey. Ten questions were presented on five shoulder conditions: rotator cuff disease, fractures, instability, arthritis, and frozen shoulder The "correct" answer was based upon the AAOS algorithm. The survey was completed by 706 physicians who treated variable (one to greater than ten) numbers of shoulder patients per month. Forty-eight percent of the physicians treated acute trauma according to the algorithm, 87% treated arthritis, and 58% treated instability. Only 46% of physicians chose the correct answer for an acute rotator cuff tear and 29% for chronic rotator cuff symptoms. Forty-four percent followed the algorithm for frozen shoulder. Overall only 49% of the patients described were treated according to the AAOS algorithms. Based on the number of shoulder patients seen each month, in one month over 2000 patients could be tested by the surveyed physicians in a manner inconsistent with the treatment algorithms, suggesting the need for improved musculoskeletal education for primary care physicians.