Schmitt 2001.

Methods	Study design: parallel‐group, two‐arm, single‐blind, randomised, placebo‐controlled trial Setting: not reported Trial time period: enrolment from March 1999 to February 2000 Interventions: ESWT vs sham ESWT Sample size calculation: an a priori analysis gave a total sample size of 16,818 participants for a given power of 95% was needed to prove the study effect of ESWT Analysis: ITT
Participants	Number of participants: screened: not reported enrolled: 40 (1 withdrew consent) randomised: 39 (19 to sham group; 20 to ESWT group) analysed at 6 weeks: 37 (18 in sham group; 19 in ESWT group) analysed at 12 weeks: 38 (18 in sham group; 20 in ESWT group) Inclusion criteria: clinical diagnosis of chronic tendinitis of supraspinatus absence of calcification free ROM or abduction of ≥ 90 degrees and free rotation failed conservative treatment: minimum of 10 sessions of physiotherapy, ≥ 2 subacromial injections, intake of NSAIDs no treatment in last 4 weeks Exclusion criteria: glenohumeral or acromioclavicular arthritis tear of the rotator cuff allergy to mepivacaine former operations to the treated shoulder local tumours of infections aged < 18 years neurological disorders acute bursitis of the shoulder Baseline characteristics: ESWT group (20 participants): mean (SD) function: Constant score: 40.70 (13.29) mean (SD) pain: VAS 0–10 score at rest: 5.35 (2.54) Sham group (19 participants): mean (SD) function: Constant score: 42.20 (13.04) mean (SD) pain: VAS 0–10 score at rest: 5.40 (3.00) Pretreatment group differences: none
Interventions	ESWT: description of modality used: ESWT using shock wave generator Storz Minilith SL 1 (Storz Medical AG, Kreuzlingen, Switzerland) method of administration: the origin of the supraspinatus tendon was localised with US and 10 mL of mepivacaine was given as subacromial local anaesthesia dose: 2000 impulses of an EFD of 0.11 mJ/mm² (measured by a PVDF‐Hydrophone, equivalent to 0.33 mJ/mm² measured by a fibreoptic‐hydrophone) at 120 impulses per minute frequency: 3 sessions at 1‐week intervals co‐interventions: none Sham ESWT: description of modality used: as above method of administration: the origin of the supraspinatus tendon was localised with US and 10 mL of mepivacaine was given as subacromial local anaesthesia. A foil was placed between the participants and the water cushion to prevent the shock wave from reaching them. dose: as above frequency: 3 sessions at 1‐week intervals co‐interventions: none
Outcomes	Measured at 6 and 12 weeks Outcomes included in review: function measured by Constant score 0–100, higher score indicating better function pain at rest measured by VAS 0–10, 10 indicating maximum pain success 12 weeks after last treatment, defined by the increase in age‐corrected Constant score of ≥ 30 points or an absolute score of 80% of the normal value adverse events withdrawals due to adverse events, intolerance to treatment or other reasons Outcomes excluded from review: pain during activity measured by VAS 0–10 with a higher score indicating worse pain
Source of funding	Not reported
Notes	Trial registration: not registered Time points included in review: 6 and 12 weeks Data analysis: function and pain at rest extracted at 6 and 12 weeks and treatment success and adverse events was extracted at 12 weeks. Although 12‐months follow‐up data were reported, participants who reported no improvement at 12 weeks were told of their treatment group and allowed to cross‐over to the ESWT treatment if they had placebo previously; we considered this part of the trial no longer randomised and did not include the 12‐month data. It is possible that treatment success was possibly added post‐hoc Withdrawals: 0/20 in ESWT group and 2/20 in placebo group (1 loss to follow‐up, 1 withdrew consent just after randomisation). We assumed no withdrawals in either group due to adverse events or treatment intolerance Adverse events: ESWT: serious adverse events: 0/20 other adverse events: 0/20 Sham: serious adverse events: 0/20 other adverse events: 0/20
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Treatment allocation was done using random permutated blocks through telephone hotline.
Allocation concealment (selection bias)	Low risk	Participants were centrally allocated via a telephone hotline.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Both participants and study personnel were blinded to group allocation.
Blinding of outcome assessment (detection bias) Self‐reported outcomes	Low risk	Participants were likely unaware of treatment, thus there was low risk of detection bias in reporting of function and treatment success.
Blinding of outcome assessment (detection bias) Assessor‐reported outcomes	Low risk	An 'independent observer' who was unaware of treatment measured other outcomes.
Incomplete outcome data (attrition bias) All outcomes	Low risk	2/40; 0/20 in ESWT group and 2/20 in sham group due to loss to follow‐up.
Selective reporting (reporting bias)	Unclear risk	There was no published study protocol, and all outcomes were reported; however, it is possible that treatment success was possibly added post‐hoc.
Other bias	High risk	At 12 weeks, 16 participants reported they were not satisfied with treatment so they were unmasked and informed of their treatment group, and participants in placebo group were offered ESWT, effectively ending the randomised part of the study.