Every year, 1.5 million people in the UK visit their GP with shoulder pain, and 40% still hurt six months later. This project aims to stop that waste of time and money by building a simple screening tool that tells doctors which treatment—exercise, injection, or surgery—a specific patient is most likely to respond to. The problem is that imaging and surgery rates have risen sharply despite weak evidence they help. Currently, patients with similar symptoms get different treatments and wildly different outcomes. The research will combine data from 1,481 patients in existing trials with a new cohort of 1,000 patients tracked for three years, linking their symptoms, scans, and psychosocial factors to actual recovery. If successful, the tool would let GPs and physiotherapists immediately sort patients into low-risk and high-risk groups and match them to the right treatment from the start. That could cut persistent pain, reduce unnecessary referrals for scans and surgery, and save the NHS significant resources—without requiring any new technology or specialist equipment.
View original technical description
Background: Annually, 4% of adults will consult general practice for shoulder pain, resulting in 1.5 million consultations. Outcomes are highly variable with 40% of patients reporting persistent pain 6 months after consulting healthcare. Requests for imaging (ultrasound scans) and surgery have risen sharply, despite lack of evidence for their effectiveness. Generating high quality evidence of the value of diagnostic and prognostic information for predicting treatment outcomes will provide opportunities for improving the primary care management of shoulder pain. Aim: Improve patient outcomes and healthcare resource use by developing and evaluating a stratified care intervention, which uses a screening and decision tool based on diagnostic and prognostic information to allow more effective targeting of treatments based on patients probability to respond to specific treatments. Workstream 1: Individual participant data (IPD) meta-analysis Objective: Synthesise IPD from randomised trials to test predictors (moderators) of response to shoulder pain treatments, such as exercise, injection, or surgery. Methods: Our developmental work has identified candidate treatment moderators. Eligible trials will be identified from updated systematic reviews, their risk of bias examined, and authors approached for data-sharing (provisionally agreed for 8 trials, n=1,481). Primary outcome will be shoulder pain/disability at 6 months. A two-stage IPD random-effects meta-analysis framework will be used, which will account for clustering of patients within trials and avoid aggregation bias. Advanced methods will be used to deal with missing values and make optimal use of repeated outcome data. Key outputs: Confirmatory evidence regarding treatment moderators; estimates of treatment effects for patient subgroups. Workstream 2: Prospective inception cohort of patients with shoulder pain Objectives: (i) Develop a prediction model using information from patient history, physical examination, ultrasound scans, and additional prognostic factors, (ii) define prognostic subgroups at low versus increased risk of poor outcome. Methods: A clinical inception cohort of consecutive adults (n=1000) consulting with shoulder pain in general practices or self-referral physiotherapy services, of whom an estimated 600 will attend a research clinic for assessment (history, examination, ultrasound scan). All participants will complete a baseline questionnaire to assess clinical, lifestyle, work-related, and psychosocial factors. Follow-up includes postal/online questionnaires and medical record review over 3 years to assess patient outcomes and healthcare use. Primary outcome will be shoulder pain and disability (SPADI) at 6 months. Multivariable regression will be used to investigate the prognostic value of diagnostic and prognostic information, and develop a brief screening tool to discriminate between prognostic subgroups. Key outputs: Prediction model to estimate individuals risk of poor outcome; defined prognostic subgroups; novel data on long-term patient outcomes linked to healthcare data. Workstream 3: Development of stratified care for shoulder pain Objective: Integrate results from Workstreams 1&2 to develop a screening and decision tool supporting optimal targeting of shoulder pain treatments. Methods: (a) Semi-structured interviews with approximately 15 patient-clinician dyads to investigate their perceptions of cause and prognosis of shoulder pain, uncertainty and communication around treatment and referral, and opinions regarding the potential of stratified care. Grounded theory and constant comparative method will be used to identify themes, investigating overlaps and contrasts between individuals in the dyad. (b) Nominal Group meetings, involving patients, GPs, physiotherapists, and secondary care specialists. Informed by interview results and evidence from Workstreams 1&2, they will agree: design and format of the tool; match
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know