Milano 2007.

Methods	Design: a single‐centre randomised (1:1) controlled trial Setting: a university hospital in Italy Timing: not reported Interventions: rotator cuff repair with acromioplasty versus rotator cuff repair alone. Sample size: authors defined type I error equal to 0.05; power of 0.80, and assumed effect size for the DASH and Constant score 0.7, which gave 32 participants per group. To allow loss of follow‐up, the authors recruited 80 participants Analysis:nNot reported but it appears that authors performed ITT analysis.
Participants	Number of participants Number screened not reported 80 participants recruited (40 to repair with acromioplasty and 40 to repair alone) Data available for 34 (85%) in repair with acromioplasty group and 37 (93%) in repair alone group Inclusion criteria A repairable full‐thickness rotator cuff tear and a type 2 or 3 acromion Exclusion criteria Partial‐thickness or irreparable full‐thickness tear Labral pathology amenable for surgical repair Type 1 acromion Os acromiale Degenerative arthritis of the glenohumeral joint Symptomatic arthritis of the acromioclavicular joint Rotator cuff arthropathy Previous surgery in the same shoulder Workers’ Compensation claims Baseline data Repair with acromioplasty Mean (SD) Age, years 61 (7.0) Number (%) females 14 (41) Number (%) dominant shoulder involved 23 (68) Mean (SD) area, mm2 398 (464) Number (%) with fatty degeneration grade I 9 (26%); grade II 12 (35%); grade III 7 (21%); grade IV 6 (18%) Number (%) subscapularis tear 5 (15) Number (%) acromion type 2: 24 (71); type 3: 10 (29) Repair without acromioplasty Mean (SD) Age, years 59.7 (9.7) Number (%) females 18 (49) Number (%) dominant shoulder involved 24 (65) Mean (SD) area, mm2 356 (423) Number (%) with fatty degeneration grade I 7 (19%); grade II 15 (41%); grade III 11 (30%); grade IV 4 (11%) Number (%) subscapularis tear 4 (11%) Number (%) acromion type 2: 24 (65); type 3: 13 (35)
Interventions	The surgeons performed operations arthroscopically Rotator cuff repair with acromioplasty The authors performed arthroscopic rotator cuff repair and subacromial decompression, consisting of anterior‐inferior acromioplasty, release of the coracoacromial ligament, and subacromial bursectomy. Rotator cuff repairs were performed with 3 different techniques, according to tear pattern: 1) tendon‐to‐bone repair with metal suture anchors (Corkscrew, 5.0 mm; Arthrex, Naples, FL) loaded with doubled polyester braided No. 2 sutures in the crescent‐shaped tears; 2) side‐to‐side repair with polyester braided No. 2 sutures in the more retracted U‐ and V‐shaped tears, and 3) a combined repair technique (tendon to bone and side to side) in the other cases. Pathology of the long head of the biceps (LHB) was treated by shaving when the lesion involved less than 25% of the tendon; more severe biceps lesions or tendon instability was treated depending on the patient’s age20: in patients aged over 50 years, we performed a biceps tenotomy; in the other cases we performed a tenodesis with 2 metal suture anchors (Corkscrew, 5.0 mm). Rotator cuff repair without acromioplasty Only subacromial bursectomy was performed, in addition to rotator cuff repair. The repair technique was same as in the other group. Both groups After surgery, a sling was applied to the operative limb and was maintained for 3 weeks; after this period, all patients underwent the following rehabilitation program: First phase (fourth through eighth week after surgery): range‐of‐motion exercise program (passive, active‐assisted, and active). Second phase (ninth through twelfth week after surgery): muscle‐strengthening program by closed kinetic chain exercises for rotator cuff, subscapularis, biceps, deltoid, pectoralis major, and scapular stabilisers. Third phase (thirteenth through sixteenth week after surgery): open kinetic chain exercises, proprioceptive and plyometric exercises, and postural rehabilitation of the kinetic chain (lumbopelvic, thoracolumbar, and scapulothoracic regions). No differences in the rehabilitation program were considered according to the extent of rotator cuff tear or involvement of the biceps and subscapularis.
Outcomes	The outcomes were assessed at two years. Authors also reported examining participants every two weeks for the first three months and then once per month until the sixth month after surgery but probably this examination did not include outcome assessment Primary outcome Not reported Outcomes Constant score (scale the authors used is unclear, higher indicates better function) Disabilities for Arm, Shoulder and Hand (DASH) score (0 to 100,. higher score indicates worse function) DASH work (0 to 100, higher score indicates worse disability) Outcomes used in this review Mean function; Constant score
Source of funding	Source of funding not reported. The authors report no conflicts of interests.
Notes	Trial registration: No registration found Data analysis: The authors report Constant score over 100 (103) for acromioplasty group. We assume the error is systematic and applies to the other group also and thus to the difference between the groups. Withdrawals: The authors report three drop outs in repair with acromioplasty group and six in repair only group but in table 1 it seems that the numbers are vice versa. We assumed the table gives the correct numbers. Cross overs or re‐operations: No cross overs or re‐operations reported Adverse events: Not reported, unclear if measured. Serious adverse events: Not reported, unclear if measured.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote:"Randomization was performed with statistical software through a random selection of 50% of cases"
Allocation concealment (selection bias)	Low risk	Quote:"The randomisation list was kept by an independent researcher (not involved in the study), and the code for assignment of each patient to one of the two groups was revealed to the surgeon at the time of surgery."
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	The blinding of participants and personnel not described. Probably no blinding.
Blinding of outcome assessment for self‐reported outcomes including pain, function and global assessment (detection bias)	Unclear risk	The blinding of participants or assessors not described. Probably authors did not attempt blinding.
Blinding of outcome assessment for incidence of full‐thickness tears at follow‐up (detection bias)	Unclear risk	The authors did not report this outcome.
Incomplete outcome data (attrition bias) All outcomes	Low risk	6/40 (15%) participants were lost to follow‐up in repair with acromioplasty and 3/40 (8%) participants in repair only group. The authors did not disclose reasons. Likely does not bias the results.
Selective reporting (reporting bias)	Unclear risk	No protocol available. Authors report all outcomes defined in the methods. It is unclear if adverse events were measured as the paper does not report them.
Other bias	Unclear risk	The authors report Constant score over 100 points although the scale of the measure is 0 to 100. The reason is unclear and the authors did not respond to e‐mails. It is likely that authors used similar method of calculation with both group.