Journal of Rehabilitation Medicine 51-8 | Page 15

Non-surgical interventions for primary frozen shoulder: systematic review Table II. Cont. Domains Effective blinding Selective a) participant Intervention b) providers c) objective outcomes integrity Incomplete outcome data Outcome country Sequence Allocation generation concealment reporting Other sources Our evaluation of bias of risk of bias Park et al. 2013 (78) Adequate Adequate Unclear Unclear GC, IFU, Unclear Adequate IFU, USS, GC, High Inadequate Adequate BLD, GC, High Unclear Unclear GC Moderate Unclear Adequate FSD, USS, UIV, GC, High Inadequate Adequate GC, OMC, UIV, High USS, Unclear Unclear OMC, GC, IFU, High UIV, USS, Adequate Unclear GC, BLD,IFU High Unclear Adequate GC, IFU, Low Study, year publication, Unclear Korea Park et al. 2014 (79) Unclear Unclear Adequate South Korea Unclear Shin et al. 2013 (81) Adequate Unclear Inadequate c) Adequate a) Not possible c) Unclear a) Not possible b) Not possible Adequate Unclear South Korea c) Inadequate a) Not possible b) Not possible Unclear Unclear Adequate Japan c) Adequate a) Not possible b) Not possible Vahdatpour et al. 2014 (83) Unclear Unclear Adequate Iran c) Adequate a) Adequate b) Not possible Taiwan c) Unclear a) Mixed: not possible/ adequate Yang et al. 2012 (85) c) Primary outcome: adequate, Secondary: unclear a) Not possible Wu et al. 2014 (84) Inadequate Unclear Inadequate b) Not possible Inadequate Inadequate Adequate b) Not possible c) Adequate a) Adequate Yoon et al. 2013 (32) one week, twice a day for 2–3 weeks, and three 30 min sessions per day in week 4. Follow-up assessment occurred at 4, 12, 24 and 52 weeks and at all follow- up time-points the findings favoured the progressive static stretching and multi-modal group for the primary outcome measures of interest. For the primary outcome measures at 1 year there were significant improvements in the progressive static stretching and multi-modal th- erapy group, for passive external rotation (87° vs 39°), for passive abduction (178° vs 133°), and for active abduction (178° vs 84°). For the secondary outcome measures, the results also favoured the progressive static stretching group for DASH (at 12, 24 and 52 weeks) and for VAS (pain) at 24 and 52 weeks). At 52 weeks the DASH scores were 1.5 vs 55.3 and VAS (pain) were 1.1 vs 3.1 (0  =  no pain and 10  =  worst imaginable pain). The participants were taught how to use the device by the principal investigator, a potential source of bias, and data was collected at a single cen- tre. Compliance data were collected, but not reported. South Korea High b) Not possible Schydlowsky et al. 2012 (80) Denmark Tanaka et al. 2010 (82) a) Not possible b) Not possible Taiwan 551 Adequate Adequate Adequate b) Adequate c) Adequate Key: adequate, low risk of bias; inadequate, high risk of bias; unclear, potential risk of bias uncertain; partial, high/unclear risk to some procedures or outcomes. BLD: concerns re: baseline differences (unclear/partial); COI: concerns re: conflict of interest (first author providing intervention/authors invented device); FSD: diagnosis of frozen shoulder unclear or diagnosed by symptoms with no imaging; GC: generalizability concerns (e.g. single-site/treatment provider, choice of inclusion/exclusion criteria); IFU: inadequate follow-up (pre-post-intervention or ≤12 weeks); USS: unjustified sample size (for example- not mentioned, no a priori calculation, insufficient detail); OMC: outcome measure concerns (unclear measures/no functional outcome included); UIV: unclear intervention (e.g. lack of detail/ varying durations/pre-trial treatment). Comparison of different doses of intra-articular corticosteroid injections In a population (n  =  53) of people in the initial pain (freezing) stage of FS Yoon et al. (32) investigated whether a single intra-articular injection of high-dose (4 ml of 10 mg/ml triamcinolone acetonide and 1 ml of 1% lidocaine) (n  =  20) improved pain and function in patients with FS more than a low dose (2 ml of 10 mg/ ml triamcinolone acetonide and 3 ml of 1% lidocaine) (n  =  20) or a placebo (5 ml of 1% lidocaine) (n  =  13). Participants were described as having stage 2 of adhe- sive capsulitis (freezing stage according to Hannafin & Chiaia (16)) with at least one month of pain duration, and mean pain intensity during a day defined as a score of 3 points or more on a 10-cm visual analogue scale rated from 0 (no pain) to 10 (worst imaginable pain). The injections were performed in a hospital setting and under ultrasound guidance. After the procedures the participants were given a home programme that involved stretching, Codman exercises, wall-climbing J Rehabil Med 51, 2019