To develop models to predict opportunities for improvement in trauma care and compare the performance of these models to the currently used audit filters.

Retrospective registry-based study.

Single-centre, Scandinavian level one equivalent trauma centre.

8220 adult trauma patients screened for opportunities for improvement between 2013 and 2022.

Two machine learning models (logistic regression and XGBoost) and the currently used audit filters were compared. Internal validation by an expanding window approach with annual updates was used for model evaluation. Performance measured by discrimination, calibration, sensitivity and false positive rate of opportunities for improvement prediction.

A total of 8220 patients, with a mean age of 45 years, were analysed; 69% were men with a mean injury severity score of 12. Opportunities for improvement were identified in 496 (6%) patients. Both the logistic regression and XGBoost models were well-calibrated, with intercalibration indices of 0.02 and 0.02, respectively. The models demonstrated higher areas under the receiver operating characteristic curve (AUCs) (logistic regression: 0.71; XGBoost: 0.74). The XGBoost model had a lower false positive rate at a similar sensitivity (false positive rate: 0.63). The audit filters had an AUC of 0.62 and a false positive rate of 0.67.

The logistic regression and XGBoost models outperformed audit filters in predicting opportunities for improvement among adult trauma patients and can potentially be used to improve systems for selecting patients for trauma peer review.

Musculoskeletal shoulder pain is a common reason for people to be treated in physiotherapy services, but diagnosis can be difficult and often does not guide treatment or predict outcome. People with shoulder pain cite a need for clear information, and timely, tailored consultations for their pain. This trial will evaluate the introduction of a personalised guided consultation to help physiotherapists manage care for individuals with shoulder pain.

This is a cluster randomised controlled trial to evaluate the clinical and cost-effectiveness of introducing a personalised guided consultation compared with usual UK NHS physiotherapy care. Physiotherapy services (n=16) will be randomised in a 1:1 ratio to either intervention (physiotherapy training package and personalised guided consultation incorporating a new prognostic tool) or control (usual care); 832 participants (416 in each arm) identified from participating physiotherapy service waiting lists aged 18 years or over with shoulder pain will be enrolled. Follow-up will occur at 3 time points: 6 weeks, 6 months and 12 months. The primary outcome will be the Shoulder Pain and Disability Index (SPADI) score over 12 months. Secondary outcomes include global perceived change of the shoulder condition, sleep, work absence and the impact of shoulder pain on work performance, healthcare utilisation and health-related quality of life (using EuroQol 5 Dimension 5 Level (EQ-5D-5L)). A multimethod process evaluation will investigate views and experiences of participants and physiotherapists, assess uptake, facilitators and barriers to delivery, and changes in factors assumed to explain intervention outcomes. Primary analysis of effectiveness will be by intention-to-treat, and a health economic evaluation will assess cost-utility of introducing the personalised consultation.

The trial received ethics approval from the Yorkshire & The Humber (South Yorkshire) Research Ethics Committee (REC reference: 23/YH/0070). Findings will be shared through journal publications, media outlets and conference presentations. Supported by patient contributors and clinical advisors, we will communicate findings through a designated website, networks, newsletters, leaflets and in the participating physiotherapy services.

ISRCTN45377604.