A Neuro-fuzzy Approach to Bad Debt Recovery in Healthcare

In the U.S. the healthcare industry is often plagued by unpaid bills, collection agency fees, and outstanding medical testing costs. All these factors contribute significantly to the rising cost of healthcare. Health care providers often have to treat patients on credit, especially in emergency and trauma cases. Unlike financial institutions health care providers do not collect financial information about their patients. This lack of information makes it difficult to evaluate whether a particular patient-debtor is likely to pay his/her bill. In recent years researchers have started to recognize the potential of data mining methods in improving our understanding of medical bad-debt, but there is relatively little research that examines the effectiveness of data mining methods in classifying bad debt in healthcare. This paper evaluates the effectiveness of an adaptive neuro-fuzzy inference system (ANFIS) in classifying bad debt in the healthcare context. The data analysis and evaluation of the performance of the ANFIS model are based on a fairly large unbalanced data sample provided by a healthcare company, in which cases with recovered bad debts are grossly under represented. Computer simulation shows that ANFIS is a viable method which produced under some scenarios good classification accuracy. More in-depth interpretation of the results, including nonlinear interaction between various factors, is provided through the analysis of the control surfaces generated by ANFIS and receiver operating characteristic (ROC) charts. Finally the paper also shows the potential of data mining models to classify unknown cases, which are a potential source of revenue recovery.