Prospective observational study to examine health-related quality of life and develop models to predict long-term patient-reported outcomes 6 months after hospital discharge with blunt thoracic injuries

Baker, Edward; Battle, Ceri; Banjeri, Abhishek; Carlton, Edward; Dixon, Christine; Ferry, Jennifer; Hopkins, Philip; Jones, Robert; Murrells, Trevor; Norton, Christine; Patient, Lee; Rasheed, Ashraf; Skene, Imogen; Tabner, Andrew; Tunnicliff, Malcolm; Young, Louise; Xyrichis, Andreas; Lee, Gerry

OBJECTIVE: This study aimed to examine the long-term outcomes and health-related quality of life in patients with blunt thoracic injuries over 6 months from hospital discharge and develop models to predict long-term patient-reported outcomes. DESIGN: A prospective observational study using longitudinal survey design. SETTING: The study recruitment was undertaken at 12 UK hospitals which represented diverse geographical locations and covered urban, suburban and rural areas across England and Wales. PARTICIPANTS: 337 patients admitted to hospital with blunt thoracic injuries were recruited between June 2018–October 2020. METHODS: Participants completed a bank of two quality of life surveys (Short Form-12 (SF-12) and EuroQol 5-Dimensions 5-Levels) and two pain questionnaires (Brief Pain Inventory and painDETECT Questionnaire) at four time points over the first 6 months after discharge from hospital. A total of 211 (63%) participants completed the outcomes data at 6 months after hospital discharge. OUTCOMES MEASURES: Three outcomes were measured using pre-existing and validated patient-reported outcome measures. Outcomes included: Poor physical function (SF-12 Physical Component Score); chronic pain (Brief Pain Inventory Pain Severity Score); and neuropathic pain (painDETECT Questionnaire). RESULTS: Despite a trend towards improving physical functional and pain at 6 months, outcomes did not return to participants perceived baseline level of function. At 6 months after hospital discharge, 37% (n=77) of participants reported poor physical function; 36.5% (n=77) reported a chronic pain state; and 22% (n=47) reported pain with a neuropathic component. Predictive models were developed for each outcome highlighting important data collection requirements for predicting long-term outcomes in this population. Model diagnostics including calibration and discrimination statistics suggested good model fit in this development cohort. CONCLUSIONS: This study identified the recovery trajectories for patients with blunt thoracic injuries over the first 6 months after hospital discharge and present prognostic models for three important outcomes which after external validation could be used as clinical risk stratification scores.