Pajareya 2004.

Methods	Design: Parallel group, two‐arm single‐blind randomised Interventions: Physical therapy (continuous short wave diathermy, mobilisation and passive glenohumeral joint stretching exercises) plus ibuprofen or ibuprofen alone Sample size calculation: 60 participants per group were estimated to be needed based upon detecting a difference in success rate (measured by improvement in a global pain and disability index) of 25% at the 5% level of statistical significance with 80% power Analysis: Per‐protocol analysis (reported that intention‐to‐treat analysis was used to test statistical significance, but outcome data presented in tables was reported as based on the number of participants completing assessments at each week) Source of funding: Department of Research Promotion, Faculty of Medicine, Siriraj Hospital, Mahidol University and partially supported by Thailand Research Fund (non‐industry)
Participants	Number of participants: 122 (61 per group) Baseline characteristics: Baseline characteristics reported for the participants who completed the week 3 assessment (N=119) Physical therapy plus ibuprofen group: Mean (SD) age = 56.3 (10.6) years; Male:Female = 14:45 Duration of symptoms: No. participants with duration <6 weeks (N=6), between 6‐11 weeks (N=20), and 12 or more weeks (N=33) Ibuprofen alone group: Mean (SD) age = 57.7 (10) years; Male:Female = 24:36 Duration of symptoms: No. participants with duration <6 weeks (N=13), between 6‐11 weeks (N=20), and 12 or more weeks (N=27) Inclusion criteria: 1. Had shoulder pain and limitation of a passive range of shoulder motion in all directions that interfered with their activities of daily living Exclusion criteria: 1. Secondary adhesive capsulitis 2. Intrinsic causes of shoulder problems such as a history of fracture, or dislocation or extrinsic causes such as neuromuscular disorders (stroke, parkinsonism), generalised arthritis, bilateral involvement, contraindication for NSAIDs 3. Bleeding tendencies
Interventions	Both groups received ibuprofen 400 mg three times daily for three weeks, and general advice (an information sheet containing advice on protection of the shoulder from vigorous activities such as pushing and pulling, and encouragement to use their arms in a normal fashion for reaching and other activities of daily life) Physical therapy plus ibuprofen (N=61) Components of intervention: ‐ Electrotherapy: Continuous short wave diathermy ‐ Manual therapy: Mobilisation. If, during the passive movements the patients felt pain before the therapist reached the end of the range, exercise was not attempted ‐ Supervised exercise: Passive glenohumeral joint stretching exercises up to the participant's tolerance, based on Cyriax ‐ Home exercise: Pulley exercises (actively assisted exercises for five minutes) and active non‐assisted exercises using a towel and wall (five minutes after applying a hot pack for 20 minutes) Dosage: ‐ Electrotherapy: 20 minutes ‐ Manual therapy: Not reported ‐ Supervised exercise: Not reported ‐ Home exercise: 10 minutes Frequency of administration: ‐ Electrotherapy: Three times a week for three weeks (9 sessions) ‐ Manual therapy: Three times a week for three weeks (9 sessions) ‐ Supervised exercise: Three times a week for three weeks (9 sessions) ‐ Home exercise: Four days a week for three weeks (on the days they did not receive the hospital‐based physical therapy program) Provider: Physical therapist Ibuprofen (N=61) See above
Outcomes	All outcomes assessed at the end of three weeks treatment (except for "success", which was also assessed at six, 12 and 24 weeks) Primary outcome: 1. "Success", measured by participants rating themselves as having disappearance of shoulder complaints or some pain/limitation which does not interfere with everyday life (on a global pain and disability index with a 5‐point Likert scale with response options "disappearance of shoulder complaints", "some pain or limitation but which does not interfere with everyday life", "minimal inconvenience to everyday life", "moderate inconvenience", and "marked inconvenience") Secondary outcomes: 2. Shoulder pain and disability index (SPADI) (0‐100 scale where a higher score indicates worse pain and/or disability) 3. Passive range of motion (abduction, external rotation, internal rotation quantified by measuring the distance between thumb and tip of C7 spine in hand behind back position) using a goniometer 4. Adverse events recorded for the physical therapy group by asking "Do you have pain that persisted more than 2 hours after treatment or more disability the next morning or not?", and by asking all patients, "Have the trial drugs and/or treatment program upset you in any way?" and examining the patient for any signs of echymosis or burn during range of motion evaluation
Notes	Adverse events due to ibuprofen were not reported separately per group: "During the 3‐week period, the patients in the study group reported a total of 10 episodes of pain that persisted more than 2 hours after treatment from 4 subjects.There were no other complications recorded. Regarding NSAIDs, 15 subjects (12.6%) had gastrointestinal side effects; the number of those who had severe dyspepsia and had to stop NSAIDs was 6 (4.2%). There were 2 reports of severe oedema and 1 case with a severe headache, which rapidly subsided after the drug was discontinued" (pg 477 of trial publication). Unpublished data regarding study design (required for risk of bias assessment) provided by trialist on request.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The patients who gave informed written consent were randomly allocated to a 3‐week treatment protocol by simple randomisation using a random numbers table and allocation concealed within an opaque envelope." Comment: An adequate method was used to generate the allocation sequence
Allocation concealment (selection bias)	Low risk	Quote: "The patients who gave informed written consent were randomly allocated to a 3‐week treatment protocol by simple randomisation using a random numbers table and allocation concealed within an opaque envelope." Personal communication: "I prepared opaque envelopes before hand. Within each envelope, I put the letter "I" or "C". The series of "I" and "C" came from the random number table. I didn't remember any part of the series" Comment: An adequate method was used to conceal the allocation sequence
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: Given the nature of the interventions, participants were not blind to treatment, and may have had different expectations about the benefits of each intervention
Blinding of outcome assessment (detection bias) Self‐reported outcomes	High risk	Comment: Unblinded participants who may have had different expectations about the benefits of the intervention they received self‐reported a global pain and disability index and the SPADI
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Quote: "Moreover, at each follow‐up, an investigator, blinded to treatment modality asked all patients "Have the trial drugs and/or treatment program upset you in any way?" and examined the patient for any signs of echymosis or burn during range of motion evaluation." Personal communication: "The range of motion assessor was blinded. I had told all of the participants that "Please don't tell the assessor about the treatment you have"" Comment: Assessors of adverse events and range of motion were probably blind to treatment
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Quote: "At the end of the 3rd week, 2 subjects dropped out from the study; 1 from the control group and 1 from the study group. The total number of cases included in the analysis was 59 in the control and 60 in the study group. By the end of the 24th week, a total of 12 cases (10.1%) had withdrawn from the study (Fig. 1). All of them lost to follow‐up for unknown reasons and the investigators could not contact them." Quote: "The results were analysed by intention to treat analysis even though the treatments actually received were modified from the protocol, because it was found that the reasons for modifying the treatment were strongly related to the results of allocated interventions." Comment: It is unclear whether reasons for losses to follow‐up were related to the interventions received
Selective reporting (reporting bias)	Unclear risk	Quote: Outcome data was fully reported for all outcomes specified in the methods section of the publication, but without a trial protocol it is unclear whether other outcomes were measured but not reported based on the results
Other bias	Low risk	Quote: "About three‐quarters of the subjects of both groups received NSAIDs as prescribed. The reasons why some patients received fewer NSAIDs than the others was due to gastrointestinal discomfort, forgetting to take them or a misunderstanding about the schedule. In the study group, 7 cases (11.7%) received fewer than 6 sessions of hospital‐based PT, 5 cases (8.3%) performed the home programme exercises fewer than 6 sessions. Two cases from the control group reported that they had additional treatment; 1 had Chinese herbal medicine and 1 received analgesics from a private clinic. No patient in the control group had hospital‐based PT or home exercise therapy for their shoulder." Quote: "The deviation from the protocol in the present study might not reverse the results. On the contrary, the differences of the outcomes at the end of the study should be elicited more easily if there was no protocol deviation. Because the patients in the study group received fewer treatments than the schedule determined (six cases had fewer than 6 sessions of hospital‐based PT and 6 cases performed home exercise fewer than 6 sessions), while the subjects in the control group received more treatment than the schedule (one case had Chinese herbal medicine and 1 case had analgesics from a private clinic)." Comment: Protocol violations are unlikely to have influenced the results