Margit Kooijman

120 musculoskeletal pain is rising 20 , early assessment of patients at risk for downstream healthcare use is important to focus limited health care resources towards patients who are most in need and identify those who require coordination of multiple providers 19,21,22 . In other words, modern clinical practice needs a way forward that reduces suffering and costs. Traditionally, diagnosis was considered the key component and guide for treatment. Recent research points in the direction of the influence of a multiplicity of biological, clinical and psychosocial factors that inform the likelihood of future outcome for each patient. Prognostic classification provides a framework that incorporates this individuality 23 . Musculoskeletal ultrasound Musculoskeletal ultrasound could possibly contribute to this framework. Our study on the uptake of musculoskeletal ultrasound showed that one in six physiotherapy practices in the Netherlands offer this possibility (chapter 4) . Of the questioned physiotherapists working with musculoskeletal ultrasound, 71% indicated they focused on patients with shoulder pain. More specifically, in this group they used it for detecting tissue damage and for making a diagnosis. These findings emphasize the professional need for help in ‘diagnosing’ patients with shoulder pain. A recent study explored the feasibility of musculoskeletal ultrasound to classify treatment strategy based upon the therapeutic consequences instead of traditional diagnostic labels 24 . It showed that the agreement between physiotherapists and radiologists was indeed higher in stratifying patients with shoulder pain into these new treatment related categories. This was especially true for ‘referral to secondary care’; the categories ‘indication for physiotherapy’ and ‘watchful waiting’ showed moderate agreement between professions. It was concluded that in future musculoskeletal ultrasound might be of value at first consultation to facilitate keep/refer decision-making but that further research to assess the consequences of this new subgrouping for clinical care is required first. Prognostic factors and research In our systematic review on prognostic factors of outcome in patients with shoulder pain, strong evidence was found that higher pain intensity, concomitant neck pain and a longer duration of symptoms at baseline predict poorer outcome in primary care settings (chapter 5) . Evidence now suggests that on average, musculoskeletal conditions including shoulder pain share a comparable clinical pathway despite differences in aetiology and presentation 17,25,26 . In most trials, symptom improvement mainly occurred over the first three to six months and specific treatments used were only modestly and equally effective 17,26-28 . There too is view that some factors predicting outcome may be similar regardless of pain site 17,19 . Multisite pain, high pain intensity, high functional disability,