Predictive Modeling to Analyze Hospital Charges Versus Reimbursements

The purpose of this study is to examine the use of statistical and data mining tools to investigate the funding mechanism for healthcare providers. The specific objective is to examine the relationship between the total charges billed by a hospital compared to the payments received for patient care. Currently, hospitals receive a negotiated payment for a particular diagnosis. The payment is based upon an average cost with local conditions taken into consideration. However, some patients who are severely ill will cost considerably more than average. The question becomes whether hospitals can afford to care for such patients, or whether they are forced to cost-shift to stay in business. Simulations about the total charges from different distributions were examined to investigate the payment structure. Predictive modeling was used to examine the different distributions of the total charges compared to reimbursements. The results were then applied to actual payment data to investigate the cost mechanism. Results show that the distribution of total charges and reimbursements are exponential, or come from an exponential family. Also, there is a significant shift between the two distributions. A number of predictors were extracted from the data that might explain the shift of cost and predict the likelihood of future payments. INTRODUCTION According to the US Census Bureau statistics for the year of 2004, about 45.8 million people in the United States are without health insurance coverage, which represents 15.7% of the total population 1 . These uninsured and underinsured individuals have access to many health care facilities even when they are unable to afford the health services cost. The Emergency Medical Treatment and Labor Act (EMTALA) guarantee that any individual who comes to the emergency department of a hospital with a request for treatment must be accommodated with an appropriate medical screening examination 2 . The health care provider does not have any guarantee of any reimbursement from uninsured or underinsured patients. Government reimbursement programs, such as Medicare and Medicaid, generally pay hospitals less than the cost of caring for the people enrolled in these programs, and insurance companies negotiate deep discounts with hospitals 3 . This situation has caused many health care providers to lose money; some of them have cut their budgets and others have closed their emergency departments. The question becomes whether hospitals can afford to care for such patients, or whether they are forced to cost-shift to provide for them. Simulations about the total charges from different distributions were examined to investigate the payment structure. Predictive modeling was used to examine the different distributions of the total charges compared to reimbursements. The data used in this project were obtained from the 2002 Medical Expenditure Panel Survey (MEPS). MEPS provides nationally representative estimates of health care use, expenditures, sources of payment, and insurance coverage for the U.S. civilian non-institutionalized population. MEPS is co.-sponsored by the Agency for Healthcare Research and Quality (AHRQ) and the National Center for Health Statistics (NCHS) 4 . METHOD In order to show the relationship between reimbursements received by the healthcare provider compared to charges incurred when the probability distribution of the total charges is different from that of reimbursements, we used simulations about the total charges from normal, exponential and gamma distributions. We calculated the reimbursements assuming that the payments are based upon an average cost. Simulations are used to approximate real cases. Since the contractual relationship between hospital and insurer is proprietary, it is extremely difficult to get actual data. However, the simulations will be compared to the information provided in the MEPS database. Kernel Density estimation is a data visualization technique to estimate actual population distributions. It was used on the MEPS data to show that both the distribution of the total charges and the reimbursement for hospital outpatient services are not normal. Kernel Density provides a very useful means of investigating an