Kolk 2013.

Methods	Study design: multicentre, parallel‐group, two‐arm, double‐blind, randomised placebo‐controlled trial Setting: outpatient clinics of 5 Dutch hospitals Trial time period: enrolment 2001–2003 Interventions: rESWT vs placebo Sample size calculation: sample size calculation was based on an unpublished pilot study, estimated that 35 participants per group were needed to detect a 50% difference in pain (VAS) between groups Analysis: ITT. Missing values were included in analyses using the 'last case carried forward' principle
Participants	Number of participants: screened: not reported enrolled: 94 (12 excluded as they were 'not randomised') randomised: 82 (44 in rESWT group; 38 in placebo group) included in analyses at 6 months: 69 (35 in rESWT group; 34 in placebo group) Inclusion criteria: symptoms for ≥ 6 months clinical signs of chronic tendinitis on painful arc and a positive empty can test (pain or weakness with downward pressure in 90‐degree elevation in scapular plane and full internal rotation) aged 18–67 years no treatment for cuff tendinitis for ≥ 6 weeks before study Exclusion criteria: pregnancy blood coagulation disorders systemic diseases tumours of the shoulder region presence of a pacemaker glenohumeral arthritis history of frozen shoulder (distinguished by a capsular restriction during passive exorotation or elevation, or both) rotator cuff tear (US examination if suspected) history of shoulder surgery Baseline characteristics: rESWT (44 participants):  mean (range) age: 48 (29–65) years number male/female: 12/32 number (%) radiographic findings: normal 20 (45), calcified 23 (52), acromio‐clavicular arthrosis 1 (2) mean (range) duration of symptoms: 24 (6–78) months mean (SD) Constant score: 55 (13.8) mean (SD) pain VAS: 65 (20) mean (SD) SST: 4.8 (2.9) number (%) treatment history: rest 22 (50), physiotherapy 32 (73), medication 21 (48), corticosteroid injections 37 (84) Placebo (38 participants) mean (range) age: 46 (24–67) years number male/female: 13/25 number (%) radiographic findings: normal 17 (45), calcified 17 (45), acromio‐clavicular arthrosis 2 (5) mean (range) duration of symptoms: 29 (6–180) months mean (SD) Constant score: 60.4 (14.4) mean (SD) pain VAS: 70 (16) mean (SD) SST: 5.3 (2.6) number (%) treatment history: rest 15 (39), physiotherapy 27 (71), medication 20 (53), corticosteroid injections 25 (66) Pretreatment group differences: none
Interventions	rESWT: description of modality: Swiss DolorClast radial shock wave device (EMS, Nyon, Switzerland) method of administration: participants were supine and treatment was applied at the anterolateral side of the acromion by a single physiotherapist for all participants. No pain medication or local anaesthetic was used. dose: 2000 pulses of 0.11 mJ/mm² at a frequency of 8 Hz at a pressure of 2.5 bar frequency: 3 sessions at an interval of 10–14 days co‐interventions: ice applied for 10 minutes after each treatment and participants advised to use their arm normally and continue with their usual pain medication. Placebo: description of modality: as above. A placebo probe that emitted the same sounds as the real probe was used. method of administration: as above dose: as above frequency: as above co‐interventions: as above
Outcomes	Measured at 3 and 6 months Outcomes included in review: function measured by the CMS: 0–100, higher score indicating better function pain, VAS 0–100, 100 indicating worst pain imaginable adverse events withdrawals due to adverse events, intolerance to treatment or other reasons Outcomes excluded from review: SST
Source of funding	Study supported by EMS. Although none of the authors received benefits for personal or professional use from a commercial party related directly or indirectly to the subject of the article.
Notes	Trial registration: not reported Time points included in review: 3 and 6 months Data analysis: the VAS and CMS were extracted at 6 months. The study contact was e‐mailed to gain further information on the mean duration of symptoms of the overall cohort of study participants and for the methods of allocation concealment Withdrawals: 9/44 in rESWT group (2 lack of treatment effect, 1 lack of confidence in physiotherapist, 6 unspecified) and 4/38 in placebo group (2 failure of clinician instructions, 2 unspecified). We assumed withdrawals due to adverse events or intolerance to treatment was 3/44 in rESWT group and 2/38 in placebo group Adverse events: rESWT: serious adverse events: 0/44 other adverse events: 0/44 Placebo: serious adverse events: 0/38 other adverse events: 0/38
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "An independent coordinator, who was not involved in the treatment or evaluation of the patients performed the randomisation by a closed envelope system."
Allocation concealment (selection bias)	Unclear risk	A closed envelope system was used; however, it was not reported if the envelopes were opaque.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Participants and surgeon were blinded to treatment allocation.
Blinding of outcome assessment (detection bias) Self‐reported outcomes	Unclear risk	Local anaesthesia was not used so participants may have been able to guess their allocation.
Blinding of outcome assessment (detection bias) Assessor‐reported outcomes	Low risk	Low risk of bias in assessor‐reported outcomes.
Incomplete outcome data (attrition bias) All outcomes	High risk	13/82; 9/44 (20%) in rESWT group (2 lack of treatment effect, 1 lack of confidence in physiotherapist, 6 unspecified) and 4/38 (10%) in placebo group (2 failure of clinician instructions, 2 unspecified) were lost to follow‐up.
Selective reporting (reporting bias)	Low risk	No published study protocol and trial was not registered; however, all specified outcomes were measured and reported.
Other bias	Low risk	No other biases apparent.