Examination of the Validity of a Clinical Prediction Rule to Identify Patients With Shoulder Pain Likely to Benefit From Cervicothoracic Manipulation

STUDY DESIGN: Secondary analysis of a randomized controlled trial. BACKGROUND: Prognostic variables identifying patients with shoulder pain who are likely to respond to cervicothoracic manipulation have been reported; however, they have yet to be validated. OBJECTIVE: To examine the validity of previously reported prognostic variables in predicting which patients with shoulder pain will respond to cervicothoracic manipulation. METHODS: Participants (n = 140) with a report of shoulder pain were randomly assigned to receive either 2 sessions of range-of-motion exercises plus 6 sessions of stretching and strengthening exercises (exercise group), or 2 sessions of cervicothoracic manipulation and range-of-motion exercises followed by 6 sessions of stretching and strengthening exercise (manipulative-therapy-plus exercise group). Outcomes of disability (Shoulder Pain and Disability Index, shortened version of the Disabilities of the Arm, Shoulder and Hand Questionnaire) and pain (numeric pain-rating scale) were collected at baseline, 1 week, 4 weeks, and 6 months. Time, treatment group, status of predictor variables, and 2-way and 3-way interactions were analyzed using linear mixed models with repeated measures. RESULTS: There were no significant 3-way interactions for either disability (P =.27) or pain scores (P =.70) for time, group, and predictor status for any of the predictor variables. CONCLUSION: The results of the current study did not validate the previously identified prognostic variables; therefore, we cannot support using these in clinical practice. Further updating of the existing prediction rule may be warranted and could potentially result in new prognostic variables and improved generalizability. Limitations of the study were a mean duration of symptoms of greater than 2 years and a loss to follow-up of 19% at 6 months.