Can patient reported outcomes measures (PROMs) be used in emergency admissions? Comparison of retrospective and contemporaneous PROMs after hip and knee replacement: a cohort study

Abstract

BackgroundThere is increasing, routine use of patient reported outcome measures (PROMs) to assess the effectiveness of health services. However, collecting a pre-event baseline measurement for emergency admissions is not feasible. We investigated the association between retrospectively and contemporaneously collected PROMs and the effect on this association of patients' age and socioeconomic status. MethodsPatients undergoing hip or knee replacement surgery in four hospitals who had completed a preoperative questionnaire were invited to recall their preoperative health status shortly after surgery. The questionnaires included a disease-specific PROM (Oxford Hip Score, Oxford Knee Score) and a generic PROM (EQ-5D-3L). Levels of association (intraclass correlation coefficient [ICC] consistency [C,1]) and agreement (ICC absolute agreement) between contemporary and retrospective reports were investigated. Systematic differences were assessed with Bland-Altman plots. Linear regression analysis was conducted to explore whether retrospective PROM scores could predict contemporary PROM scores. Findings520 patients completed a retrospective questionnaire (276 knee replacements, 244 hip replacements). We successfully linked retrospective and contemporaneous questionnaires for 484 patients (93%). Retrospective assessment of preoperative health status was similar to contemporaneous reports. Absolute agreement and consistency were strong and very strong as shown by the ICCs. Absolute agreement for disease-specific PROMs (ICC 0·82) was stronger than for the generic PROM (ICC 0·62). The strength of association and of agreement was consistent irrespective of the severity of a patient's primary condition. Patients' age and socioeconomic status had no significant influence on the association. Mean retrospective disease-specific PROMs for groups or populations of patients could reliably predict contemporary reports of PROMs. InterpretationRetrospective disease-specific PROMs can be used to obtain a baseline assessment of health status when contemporary collection is not feasible, and they offer a means of assessing effectiveness in unexpected emergency admissions. FundingThe research and EK are funded by an Economic and Social Research Council doctoral fellowship (grant ref ES/J500021/1). The research was supported by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care North Thames at Barts Health NHS Trust.