Applying models of care for total hip and knee arthroplasty: External validation of a published predictive model to identify extended stay risk prior to lower-limb arthroplasty

Abstract

Introduction/AimsMultiple predictive tools have been developed to identify patients requiring an extended hospital stay following lower limb arthroplasty. Use at new sites requires verification of appropriate data coverage and evidence of validity in a new population. The aim of this study was to externally validate a previously reported model for identifying patients requiring an extended (5+ day) stay following total hip or knee replacement in a medium-sized public hospital orthopaedic department.MethodsElectronic medical records were accessed and retrospective data extracted from 200 randomly selected total hip or knee arthroplasty patients. Data fields were matched to the candidate model and organised for validation analysis. Model validation was assessed with model discrimination, calibration on both original (unadjusted) and adjusted forms of the candidate model. Decision curve analysis was conducted on the outputs of the adjusted model.ResultsThe original model performed poorly in the validation dataset, grossly overestimating length of stay. Performance improved following adjustment of the model intercept and model coefficients, although the model remained poorly calibrated at low and medium risk threshold and net benefit of the adjusted model was modest.ConclusionConsideration of predictive models to improve resource allocation for postoperative management of arthroplasty patients should include practical assessment of discrimination, calibration and net benefit of intervention at a clinically acceptable threshold prior to implementation within their local setting. This forms a clinically meaningful assessment of the accuracy, as well as costs and benefits associated with the desired change in practice. The selection of an appropriate model for quality care improvement remains challenging.