The Unique Credit Characteristics of Healthcare Patients. An Equifax Predictive Sciences Research Paper December 2003

Size: px

Start display at page:

Download "The Unique Credit Characteristics of Healthcare Patients. An Equifax Predictive Sciences Research Paper December 2003"

Rosanna Booth
6 years ago
Views:

1 The Unique Credit Characteristics of Healthcare Patients An Equifax Predictive Sciences Research Paper December 2003

2 Executive Summary As today s healthcare payment trends shift toward an ever increasing self-pay population, providers are at greater risk of losses. Due to rising costs associated with most health insurance plans, many employees are opting for reduced coverage to keep their monthly costs down. Add to that the uninsured population, and one can quickly assume that the financial risk levels in the healthcare industry are increasing. Equifax Predictive Sciences, the modeling and analytical division of Equifax, undertook this research paper to provide a tool that will aid healthcare executives in understanding the value in using credit scoring in their decisioning practices. For many years, financial institutions have utilized credit data to better understand the financial risk associated with a client, and since the mid-80s have placed more confidence in the use of credit scoring to understand the client s behavioral tendencies. The predictive power that lies within credit information is so strong that most lenders use it as the main consideration when extending credit. Credit scoring has gained huge acceptance in this process because of its non-biased nature and the speed in which a decision can be made. Due to the increasing self-pay population, the healthcare industry can benefit from the lessons learned in the financial arena by placing confidence in the predictive power of consumer credit data and implementing credit scoring strategies immediately. When discussing credit modeling, there are two types that are usually mentioned, custom and generic. A custom model is developed on customer-specific data and, in most cases, will perform better than a generic model. These custom models can be used to predict potential risk, response rates for a marketing campaign, and behavior for account management and cross-sell opportunities. Generic models are usually developed from a random sample of the overall population and predict the likelihood of someone becoming delinquent over a certain period of time. Additionally, there are industry-specific generic models that are used when the population that is being modeled is unique compared to the general population (i.e., healthcare, telecommunications, utilities, auto finance). Our research and analysis has shown that an industry-specific model generally performs better than a generic solution developed on the overall consumer population. The generic solutions seem to work fairly well on separating the good accounts (non-charged off on healthcare accounts) from the bad accounts (charged off on Equifax is pleased to provide this information for your convenience; however, it is provided with the understanding that Equifax is not engaged in rendering legal, accounting, security, or other professional advice. The information contained in these materials is believed to be reliable at the time it was written, but it cannot be guaranteed insofar as it is applied to any particular individual or situation. No endorsement of Equifax or any Equifax product is expressed or implied by the mention of any third party in these materials. 2

3 By using credit scoring in its decisioning and collection efforts, a healthcare provider will utilize a cost-effective tool that will aid in identifying its most profitable customers while setting the ground work for strategies in its collection efforts. healthcare account) due to the fact that a patient who had charged off on their healthcare account has close to a 70 percent chance in performing poorly on their other accounts. However, the industry-specific model performs better when trying to identify the charged-off dollars associated with bad healthcare accounts. Because the industry-specific model performs better in identifying not only charged-off dollars but also patient payment amounts, it can be used in setting strategies for stages in the healthcare credit lifecycle. For the point or time-of-service stage (the first stage of incurring healthcare debt is when a patient requires a service from a healthcare provider, thus point or time of service), a credit score can be used to set the upfront payment amount. For example, patients without insurance who have a high score may not be required to make a payment at the time of service because they have the highest propensity to pay. On the other hand, self-pay patients who have a low score would be required to make a payment because they have a high rate of default. Credit scoring can assist the healthcare provider in setting up payment options for patients, either at the time of service or soon after service has been rendered. Finally, for collection efforts, credit scoring can be used to segment the tactics to be used in collection of the debt. Our research in this population has yielded strong results in favor of using credit scoring to prioritize collection efforts. In one example, the collection efforts by one hospital had yielded an 18.9 percent pay ratio (the amount collected divided by amount charged off). If this hospital had used the industry-specific model to segment the population it would have identified one part of the population with a pay ratio of 40.8 percent and another with a pay ratio of 30.1 percent. The remaining population resulted in an 11.5 percent pay ratio. By utilizing the industry-specific model, this hospital could have streamlined its collection efforts toward working the population with the greater propensity to pay and possibly outsourced the remaining population, resulting in reduced collection costs. By using credit scoring in its decisioning and collection efforts, a healthcare provider will utilize a cost-effective tool that will aid in identifying its most profitable customers while setting the ground work for strategies in its collection efforts. This research has shown how unique the healthcare population is and how well an industry-specific solution works on segmenting the population. By utilizing this knowledge, healthcare providers will become more efficient in servicing their growing self-pay population. 3

4 Introduction With today's healthcare trends showing an increase in the self-pay population, healthcare providers are at greater risk of losing money. Understanding the behavioral tendencies of this population is critical to the success of any healthcare risk management policy. We undertook this research to provide healthcare management with a tool to understand these tendencies and how to apply this knowledge in their everyday practices. We started our research under the assumption that patients will default on their healthcare loans at a higher rate than their other credit obligations. We analyzed how the patients other credit trades performed against the overall population. In addition to sharing these results, we will discuss how credit information can aid healthcare providers in the different phases of the credit lifecycle from point of service to collections. Methodology The information used in this analysis was a dataset that included patient information and health-care-specific performance from multiple hospitals and is summarized in Table 1. Population Total Good Bad Bad Rate Total 100,288 76,332 23, % Self-pay After Insurance 88,951 73,131 15, % Self-pay 11,337 3,201 8, % Table 1: Observed Dataset The self-pay population is broken out between patients who were insured and those who did not have insurance coverage. Good accounts are accounts that did not charge off during the performance period vs. bad accounts that did charge off during the 24-month performance period. The bad rate is the calculation of the number of bad accounts to the total accounts for that specific population. In order to perform research on this dataset, a retro-study was performed using our archived credit database where credit information was appended to the dataset received from the hospitals for new patients that visited the hospitals during a sixmonth timeframe. These hospitals supplied Equifax with the patient performance information for a 24-month period from June 2001 through June Generic scores were appended at the observation point (June 2001) and credit bureau performance criteria along with the hospital performance was appended at the performance point (June 2003). The performance criterion was appended to analyze how the patients performed on their other 4

5 credit obligations. Since most serious delinquencies are considered 90 days past due or worse, this was the bad definition used for the other trades in this analysis. For example, if a patient went 90 DPD or worse on their other credit obligations once over the 24- month performance window, they were considered bad. Analytical Results Total Population Table 2 represents the total population and illustrates how healthcare patients performed on their other trades. With a bad rate on their other trades greater than 40 percent, this population is definitely a greater credit risk than the overall credit population. Additionally, a patient that performs poorly on their other accounts is three times more likely to perform poorly on their healthcare account. Total Good Bad Bad Rate Other Good Other Bad Bad Rate 100,288 76,332 23, % 57,400 42, % Healthcare Good Healthcare Bad Good Other Bad Other Good Other Bad Other 50,176 26,156 7,224 16,732 Good Other Bad Other HC Good HC Bad HC Good HC Bad 50,176 7,224 26,156 16, % 39.01% Table 2: Total Population Multiple 3.1 Self-Pay After Insurance Table 3 represents the self-pay after insurance population and illustrates how healthcare patients performed on their other trades. Since this population represents 88 percent of the total population in Table 2, the results are similar. One interesting observation is how high the bad rate of other trades is on this population. One might expect a patient carrying insurance to be a good credit risk. This is true when comparing how they pay their healthcare accounts to the true self-pay population. However, this population has a very high bad rate on their other trades and would not be considered a good credit risk to most financial lenders. 5

6 Total Good Bad Bad Rate Other Good Other Bad Bad Rate 88,951 73,131 15, % 54,418 34, % Healthcare Good Healthcare Bad Good Other Bad Other Good Other Bad Other 48,718 24,413 5,700 10,120 Good Other Bad Other HC Good HC Bad HC Good HC Bad 48,718 5,700 24,413 10, % 29.31% Multiple 2.8 Table 3: Self-Pay After Insurance Self-Pay The true self-pay population represents the greatest risk to healthcare providers because this population not only performs poorly on their healthcare accounts but also on their other trades. Table 4 represents the self-pay population and illustrates how the healthcare patients performed on their other trades. Total Good Bad Bad Rate Other Good Other Bad Bad Rate 11,337 3,201 8, % 2,982 8, % Healthcare Good Healthcare Bad Good Other Bad Other Good Other Bad Other 1,458 1,743 1,524 6,612 Good Other Bad Other HC Good HC Bad HC Good HC Bad 1,458 1,524 1,743 6, % 79.14% Multiple 1.5 Table 4: Self-Pay As observed in the analysis, the healthcare self-pay patients (both with and without insurance) have higher delinquency rates on their other trades than the overall credit population. With these high delinquency rates, it is obvious that a credit solution would provide an effective way to segment the population between good and bad accounts (i.e., pay vs. no pay). It is also clear that this population is unique from the general population and, therefore, generic solutions developed on the overall credit population would not be as effective as an industry-specific model. 6

7 To provide examples of how effective Equifax s Payment Predictor for Healthcare model is on the healthcare population, the validation results for one of the hospitals is included in this analysis. The identity of the hospital has been removed, but the results are actual and provide a clear picture as to the effectiveness of an industry-specific scoring solution. In Figure 1, the industry-specific model slightly outperforms the generic score and the bankruptcy score in both KS and in identifying the percentage of accounts. In Figure 2, however, Payment Predictor for Healthcare clearly is the strongest score in identifying dollars charged off. If this hospital would have used this model in June 2001 to score this population, it would have identified 65 percent of the accounts that would have charged off, but more importantly, almost 84 percent of charged-off dollars in the bottom-scoring 40 percent of the population. This clearly exemplifies how utilizing a credit-scoring solution can help identify the most risky healthcare accounts. DCM % Payment Predictor Generic Bankruptcy Collection Pop for Healthcare Risk Score Score 10% 18.0% 17.0% 16.6% 17.8% 20% 35.15% 34.8% 31.9% 33.3% 30% 52.0% 49.5% 46.8% 43.7% 40% 65.1% 64.0% 61.2% 50.5% KS Figure 1: Healthcare Charged-Off Accounts DCM % Payment Predictor Generic Bankruptcy Collection Pop for Healthcare Risk Score Score 10% 25.9% 20.1% 12.7% 31.4% 20% 48.7% 42.3% 32.4% 51.4% 30% 70.9% 59.3% 52.0% 61.1% 40% 83.7% 78.2% 70.1% 66.6% Figure 2: Healthcare Charged-Off Dollars Explanation of Charts Dcm % - Decumulative percentage of total population (for example, 10 percent represents the bottom-scoring 10 percent of the total population). Payment Predictor for Healthcare - Equifax's scoring solution that predicts the probability of a healthcare account charging off within 12 months. Generic risk score - Generic risk score predicting the probability of serious delinquency (90 days past due or worse) within 24 months. Bankruptcy score - Generic score predicting the probability of bankruptcy within 24 months. 7

8 Collection score - Generic score predicting the likelihood of payment within 6 months of placement. Kolmgorov-Smirnov (KS) value - A measure of the separation found in the model. Calculated as the absolute value of the maximum difference between the cumulative percent of the goods and the cumulative percent of the bads. % under each score represents the percentage of that chart type (charged-off healthcare accounts or dollars) captured in the particular percentage of the population. (For example, under charged off accounts under the health-care-specific model, 18 percent of charged off healthcare accounts were captured in the bottom scoring 10 percent of the population.) Utilizing Credit Scoring There are two main uses of credit scoring on a healthcare population that will be discussed in detail. First, a point or timeof-service scoring strategy would help with understanding the through-the-door population. Second, a collection strategy would aid in determining which delinquent population has the highest propensity to pay. The strategies below were formed using Equifax's Payment Predictor for Healthcare and are shown to illustrate the effectiveness of using an industry-specific generic score. Point-of-Service Strategy As noted earlier, a healthcare provider has its highest risk of loss with a non-insured patient. Setting a scoring strategy for this population at the point of service will provide additional protection against potential losses. In the example below, the hospital would have increased their amount collected by requiring a $100 payment at point of service from the bottom scoring 65 percent of the population. Rank Range Low High low-374 No Payment at POS, 94% of patient payment amounts Collecting $100 at POS would increase dollars received by 106% 8

9 The patients in the top scoring 35 percent would benefit from credit scoring by not making a payment a point of service since historically, they represent 94 percent of all patient payments. By understanding this through-the-door population, healthcare providers place themselves in an advantageous position and can make more proactive decisions. For example, some patients may respond well to the hospital offering financing options. By working with a third-party financing company in setting a score cut-off strategy, financing could be offered soon after the healthcare debt has been incurred. Collections Strategy For healthcare providers who are more concerned with an afterservice strategy, identifying the probability of payment once an account has charged-off is critical to any collection efforts. In the example below, the hospital would have benefited greatly from having a collections strategy in place. Instead of pursuing all patients, it would have been able to isolate the populations that represented the greatest probability of repayment. Rank Range Low High low-374 Moderate 40.8% pay ratio Aggressive 30.1% pay ratio Sell-off Only an 11.5% pay ratio Pay ratio for Total Charged-Off dollars = 18.9% The top-scoring 20 percent had a pay ratio of 40.8 percent, which means that for every dollar charged off in this scoring band, $0.408 was collected. This more than doubles the overall pay ratio of 18.9 percent and makes the decision of which accounts to collect on easier. In contrast, the bottom scoring 35 percent of the population represents only an 11.5 percent pay ratio. Instead of holding on to these accounts for long periods of time before selling them to outside collection companies, the hospital could sell them immediately and get a higher premium for these accounts. For example, if this hospital received $0.15 for these accounts, they would have increased the amount received on this population by 30 percent. 9

10 Validation Healthcare providers who are interested in researching the benefits of credit scoring should perform a historical validation on their accounts. Historical validations are performed to show how the population would have been distributed if a score was used in the past at a specific point in time. That point in time is referred to as the "observation point," and multiple generic scores will be appended for evaluation. The performance window is usually 12 to 24 months, and the performance measurements are supplied by the healthcare provider. When the validation is complete, a summary of findings will be provided along with performance tables called "gains charts." Once a validation has been presented, healthcare providers should have a clear picture of how a scoring solution should work on their portfolio. If the results favor credit scoring, the next step should be to set a scoring strategy and implement a test plan to observe the expected benefits. With the trends suggesting a continued increase in the self-pay population, implementing a credit scoring strategy is critical to the long-term financial success of healthcare providers. Conclusion The original basis for this research paper was to confirm that patients pay their other credit obligations better than their healthcare debt. The surprising result is that while this was confirmed, the level at which their other debts were delinquent was very high. The bad rates of their other debts are much greater than the overall population, and therefore make using a generic scoring solution less effective than an industry-specific scoring solution. With the trends suggesting a continued increase in the self-pay population, implementing a credit scoring strategy is critical to the long-term financial success of healthcare providers. By using credit scoring in the point-of-service or collection strategies, healthcare providers can make use of a non-biased approach that can streamline the decision process. Contact Information Equifax Inc. is a global leader in information technology that enables and secures global commerce with consumers and businesses. We are one of the largest sources of consumer and commercial data. Utilizing our databases, advanced analytics and proprietary enabling technology, we provide real-time answers for our customers. This innovative ability to transform information into intelligence is valued by customers across a wide range of industries and markets. Headquartered in Atlanta, Georgia, Equifax employs approximately 4,700 people in 13 countries throughout North America, Latin 10

11 America and Europe. Equifax was founded 107 years ago, and today is a member of Standard & Poor's (S&P) 500 Index. Our common stock is traded on the New York Stock Exchange under the symbol EFX. Equifax Inc Peachtree Street, N.W. Atlanta, Georgia Equifax is a registered trademark of Equifax Inc. Payment Predictor for Healthcare is a service mark of Equifax Inc. All other trademarks not owned by Equifax or its affiliates are the property of their respective owners. Copyright 2006, Equifax Inc., Atlanta, Georgia. All rights reserved. Printed in the U.S.A. 11

Identifying High Spend Consumers with Equifax Dimensions

Identifying High Spend Consumers with Equifax Dimensions April 2014 Table of Contents 1 Executive summary 2 Know more about consumers by understanding their past behavior 3 Optimize business performance