Open Access Highly Accessed Research article

Predicting response to physiotherapy treatment for musculoskeletal shoulder pain: a systematic review

Rachel Chester12*, Lee Shepstone3, Helena Daniell2, David Sweeting4, Jeremy Lewis56 and Christina Jerosch-Herold1

Author Affiliations

1 School of Allied Health Professions, Faculty of Medicine and Health Sciences, University of East Anglia, Norwich, Norfolk NR4 7TJ, UK

2 Physiotherapy Department, Norfolk and Norwich University Hospital, Norwich, Norfolk NR4 7TJ, UK

3 Norwich Medical School, Faculty of Medicine and Health Sciences, University of East Anglia, Norwich, Norfolk NR4 7TJ, UK

4 Physiotherapy Department, East Coast Community Healthcare CIC, Lowestoft Hospital, Tennyson Road, Lowestoft, Suffolk NR32 1PT, UK

5 School of Health and Social Work, Department of Allied Health Professions, University of Hertfordshire, College Lane, Hatfield AL10 9AB, UK

6 Musculoskeletal Department, Health at the Stowe, Central London Community Healthcare, 260 Harrow Rd, Greater London, W2 5ES, UK

For all author emails, please log on.

BMC Musculoskeletal Disorders 2013, 14:203  doi:10.1186/1471-2474-14-203

Published: 8 July 2013



People suffering from musculoskeletal shoulder pain are frequently referred to physiotherapy. Physiotherapy generally involves a multimodal approach to management that may include; exercise, manual therapy and techniques to reduce pain. At present it is not possible to predict which patients will respond positively to physiotherapy treatment. The purpose of this systematic review was to identify which prognostic factors are associated with the outcome of physiotherapy in the management of musculoskeletal shoulder pain.


A comprehensive search was undertaken of Ovid Medline, EMBASE, CINAHL and AMED (from inception to January 2013). Prospective studies of participants with shoulder pain receiving physiotherapy which investigated the association between baseline prognostic factors and change in pain and function over time were included. Study selection, data extraction and appraisal of study quality were undertaken by two independent assessors. Quality criteria were selected from previously published guidelines to form a checklist of 24 items. The study protocol was prospectively registered onto the International Prospective Register of Systematic Reviews.


A total of 5023 titles were retrieved and screened for eligibility, 154 articles were assessed as full text and 16 met the inclusion criteria: 11 cohort studies, 3 randomised controlled trials and 2 controlled trials. Results were presented for the 9 studies meeting 13 or more of the 24 quality criteria. Clinical and statistical heterogeneity resulted in qualitative synthesis rather than meta-analysis. Three studies demonstrated that high functional disability at baseline was associated with poor functional outcome (p ≤ 0.05). Four studies demonstrated a significant association (p ≤ 0.05) between longer duration of shoulder pain and poorer outcome. Three studies, demonstrated a significant association (p ≤ 0.05) between increasing age and poorer function; three studies demonstrated no association (p > 0.05).


Associations between prognostic factors and outcome were often inconsistent between studies. This may be due to clinical heterogeneity or type II errors. Only two baseline prognostic factors demonstrated a consistent association with outcome in two or more studies; duration of shoulder pain and baseline function. Prior to developing a predictive model for the outcome of physiotherapy treatment for shoulder pain, a large adequately powered prospective cohort study is required in which a broad range of prognostic factors are incorporated.

Physical therapy; Shoulder; Shoulder pain; Musculoskeletal; Predict; Prognosis