Document Type


Publication Date



Total hip replacement, Cost of medical care -- Models, Reconstructability Analysis, Information Theory, Probabilistic graphical modeling, Multivariate analysis discrete multivariate modeling, Data mining


Legislative reforms aimed at slowing growth of US healthcare costs are focused on achieving greater value per dollar. To increase value healthcare providers must not only provide high quality care, but deliver this care at a sustainable cost. Predicting risks that may lead to poor outcomes and higher costs enable providers to augment decision making for optimizing patient care and inform the risk stratification necessary in emerging reimbursement models. Healthcare delivery systems are looking at their high volume service lines and identifying variation in cost and outcomes in order to determine the patient factors that are driving this variation and generating increased cost. One way to improve predictions is through enhanced modeling methods. Current modeling is predominantly done with logistic regression (LR). This project applied Reconstructability Analysis (RA) to data from hospital based hip and knee replacement surgery. RA is partially similar to LR, but has some unique features.

RA is a data mining method that searches for relations in data, especially non-linear and higher ordinality relations, by decomposing the frequency distribution of the data into projections, several of which taken together define a model, which is then assessed for statistical significance. The predictive power of the model is expressed as the percent reduction of uncertainty (Shannon entropy) of the dependent variable (the DV) gained by knowing the values of the predictive independent variables (the IVs). RA predictive models were then generated for the total cost of the hospital episode. RA generated continuous predictions for cost by calculating expected values. Models included novel comorbidity variables, non-hypothesized interaction terms, and often resulted in substantial reductions in uncertainty.

Predictive variables consisted of both delivery system variables and binary patient comorbidity variables. Delivery system variables (surgeon, location, and surgeon volume) were found to be the predominant predictors of total cost rather than individual patient risk factors. Results suggest that provider practice patterns have a larger effect than previously considered. Improving hospital and provider efficiency may be more strategic than cherry picking low risk patients. Risk ratios were generated as an additional measure of effect size. These risk ratios were used to classify the IV states of the models as indicating higher or lower risk of adverse outcomes. Some IV states showed nearly 25% of patients at increased risk, while other IV states showed over 75% of patients at decreased risk. In real time, such risk predictions could support clinical decision making and custom tailored utilization of services.

Future research might address the limitations of this project’s data and employ additional RA techniques and training-test splits. Implementation of predictive models is also discussed, with considerations for data supply lines, maintenance of models, organizational buy-in, and the acceptance of model output by clinical teams for use in real time clinical practice.

If outcomes and risk are adequately predicted, areas for potential improvement become clearer, and focused changes can be made to drive improvements in patient care. Better predictions, such as those resulting from the RA methodology, can thus support improvement in value – better outcomes at a lower cost. As reimbursement increasingly evolves into value-based programs, understanding the outcomes achieved, and customizing patient care to reduce unnecessary costs while improving outcomes, will be an active area for clinicians, healthcare administrators, researchers, and data scientists for many years to come.

Persistent Identifier