Abstract
Introduction: Physiotherapy for persistent rotator cuff-related shoulder pain (RCRSP) uses a multi-modal approach including manual therapy, exercise prescription, physical activity, and patient education. Pragmatic trials are needed to determine the effectiveness of manual therapy in a multi-modal, individualised physiotherapy setting.
Aims: (1) To determine the feasibility of conducting an appropriately designed randomised clinical trial (RCT) comparing pragmatic physiotherapy with manual therapy (MT Group) to pragmatic physiotherapy without manual therapy (NoMT Group) for patients with RCRSP; and (2) to determine the preliminary treatment effects for patient-reported outcomes.
Methods: This was a non-inferiority two-arm pilot RCT, the Otago Shoulder Health Study (OtShoH). Sixty-eight patients (aged > 35 years) with 2 3 months’ duration RCRSP were randomly allocated into one of the two intervention groups: MT or NoMT. Both groups received pragmatic physiotherapy treatment with patient education, progressive exercise prescription, and paced physical activity, of up to eight sessions over a 3-month period. The MT group also received manual therapy for at least four sessions, based on the physiotherapist's clinical reasoning for the individual. Feasibility outcomes included participant recruitment and retention rates, and intervention adherence and fidelity. Research outcome measures included the shortened Disabilities for Arm, Shoulder, and Health questionnaire (QuickDASH), pain self-efficacy, fear avoidance, estimated physical activity, patient satisfaction, and global rating of change, assessed at baseline, and 3- and 6-month follow-up. Mixed model linear regressions were used for between-group patient outcomes.
Results: Over a 6-month period, 239 volunteers completed the online screening questionnaire, of whom 108 were invited to a physical screen. Seventy of those (65%) were allocated. Two withdrew before starting the intervention, 33 in the MT group and 35 in the NoMT group completed the baseline assessment. The groups were equivalent in terms of M (SD) age (MT: 58.5 (9.9) years, NoMT: 58.4 (10.9) years), gender, symptom duration, and prior healthcare usage. The Mdn (range) of treatment sessions was 7 (2-8) for the MT group and 6 (1-8) for the NoMT group. Most commonly, manual therapy targeted the glenohumeral joint (n = 22), thoracic spine (n = 19), and soft tissue mobilisations (n = 25). Provision of patient education and exercise prescription was similar between groups. QuickDASH between-group mean differences (95% confidence intervals) at baseline, and 3- and 6 months were -2.5 (-8.6, 3.8), 0.8 (-5.5, 7.0), 1.9 (-4.5, 8.2), respectively. Non-inferiority analysis showed that, at all timepoints, the mean between-group differences were below the a priori non-inferiority margin (11 points out of 100). The Pain Self-Efficacy Scale at 6 months suggests a potential between-group difference in favour of the MT group (-6.3, 95% Cl [-11.2, -1.4). The Fear Avoidance Behaviour Questionnaire, Patient Acceptable Symptom State, GROC, and estimated physical activity suggested no differences at all three timepoints.
Conclusion: Our results support feasibility for a full trial, and strategies will be needed to improve participant diversity. Outcomes when excluding manual therapy from a pragmatic physiotherapy intervention will need to be confirmed in a full trial.