Self-assessment of range of motion in patients undergoing surgery for a unilateral shoulder condition.

Objective: This study aimed to investigate the agreement of patient-assessed and researcher/physician-assessed measurements of the difference in range of motion between the unaffected and affected shoulders in 55 patients undergoing arthroscopic surgery for a unilateral shoulder condition.

Methods: The investigation included 55 patients (17 women and 38 men; median age=53 years; range=26-74) with a symptomatic unilateral shoulder condition and a surgically treatable diagnosis. Images of a model/researcher performing active shoulder abduction, flexion, external rotation, and internal rotation were created. Each image was paired with a degree diagram or a level system (for internal rotation) on the back for the patient to accurately self-evaluate and record. Each patient was instructed to attentively examine the figures and perform the movements with the same posture as depicted. On the day of surgery, prior to the procedure, 2 independent researchers who were not involved in the patient's care used a standard goniometer to assess the same active movements that the patient had previously self-assessed. For agreement analyses, the intraclass correlation coefficient and Bland-Altman plots were calculated for continuous data (abduction, flexion, and external rotation), and Cohen's weighted kappa was calculated for ordinal categorical data (internal rotation).

Results: The intraclass correlation coefficient for abduction, flexion, and external rotation was 0.93 (excellent) 95% CI (0.87, 0.96), 0.89 (good) 95% CI (0.81, 0.94), and 0.72 (moderate) 95% CI (0.52, 0.84), respectively. Cohen's kappa for internal rotation (measured as reaching levels on the back) was 0.63 (moderate).

Conclusion: We believe that patient-assessed measurements of abduction (intraclass correlation coefficient 0.93) and flexion (intraclass correlation coefficient 0.89) can be used as a valid substitute (for measurements by a clinician or researcher). Patient-assessed measurements for external rotation (intraclass correlation coefficient 0.72) and internal rotation (kappa 0.63) are in moderate agreement and should be used more cautiously as substitutes.

Level of evidence: Level II, Diagnostic Study.