An analysis of the variation in billing charges of medical providers: causes and implications

Abstract

The purpose of this paper is to review the billing methodology of physicians that participate in fee-for-service plans, and, using a data set of billing charges, determine the significance of the variation in fees by providers of certain kind of procedures. Providers use cost-plus pricing and take into account the medical aspects of the services and market forces. Using the Current Procedural Terminologies (CPT) established by the American Medical Association, the providers define the medical services by the six digit CPT code, and bill accordingly. The statistical evidence shows that there is some evidence that fees of providers who bill under the ‘medicine’ codes tend to exhibit more variation in charges than other procedures.

Notes

 If physicians are profit maximizers this policy could be successful. Brown (1994) has shown that the supply of physicians’ services is negatively sloped (they are utility maximizers) and the income effect of a price (and thus wage) increase outweighs the substitution effect. This means primary care doctors will not produce more as their fees are reduced, while specialists will produce more as their fees are reduced.

 Lynk's (1988, p. 104) analysis was performed on the data distribution of claim charges that are similar to this study. As was indicated, ‘the observation … consists of one year's history of all charges submitted to Blue Shield for each of the medical procedures. From this it is possible to derive the full distribution of charges, i.e. each dollar amount that was charged at least once, and the number of times it was charged.’

 Many commercial plans base physician reimbursement on the lower of (1) the physician's most frequent charge for the service, (2) the mean charge made by the physicians in the area or (3) the actual charge appearing on the claim form for the service (Biller's Guide, 1992). The methodology for determining the payment is called Usual, Customary and Reasonable (UCR) and is also referred to as the ‘prevailing rate’ in the community. Insurance companies set the reimbursement at a percentile higher than the median charge from the distribution of claim data. The community is defined as a geographical area where the availability and costs of providing the service are similar. Charges are considered to be reasonable if they are consistent with the usual fees charged by the majority of similar medical providers in the geographic area in which the expenses were incurred.

 In this study, physician fees, charges, claims, and bills are all used interchangeably.

 There are six categories of CPT codes. In addition to the three that are used in this study, the others are Anesthesia, Surgery and Pathology. An Evaluation & Management (E & M) code such as 99201 is an office visit with straight forward medical decision making. An increase in the 5th digit, say to 99202, represents a longer face-to-face visit with the doctor and problems are of moderate severity. Typically, the larger the 5th digit, the higher the charge for the services.

 Socio-economic statistics are not provided for these areas because the focus of the paper is not on the variation in charges among the areas but on variation in charges among the procedures. These areas were delineated by the workers’ compensation commission as health service areas. As can be seen in the Appendix, there are CPT codes that are duplicates because they are determined from the different geographic areas. To address the role of market forces in the three areas, a dummy regression model was developed using the three areas as independent variables and the response variables of mean charge, median charge, index of skewness, percentage above two standard deviations, and coefficient of variation. It was found that none of the geographic areas were statistically significant in explaining variation in the response variables.

 Radiology studies are what is referred to as ‘intensity of service.’ Greater intensity of service now accounts for almost one-half of the annual increase in health care costs (Brown, 1993). With the guard against malpractice suits, tests are conducted for everything.

 According to ADP Medical Solutions, of the top ten CPTs nationally in terms of total charges for property and casualty claims, eight of them were the ‘medical’ procedures. The top CPT code is Spinal Manipulation followed by Hot and Cold Packs.

 Most large medical providers subscribe to services that provide data on the prevailing charges in their area. Examples are databases collected by the Health Insurance Association of America (HIAA). Other commercial companies that collect and analyse medical charge data are Medicode and Medirisk. These companies rely on an extensive database of claims from national insurance companies and other third party payers.

 Even though utilization and proliferation of treatments are not addressed in this study, it is evident by observing the average number of charges for the physical therapy procedures, one may hypothesize that a physician prescribing massage therapy (CPT 97124) would be as prevalent as the doctor telling the patient, ‘Go home, take two aspirins, go to bed, and call me in the morning.’ Even though the median charge for CPT 97124 appears to be modest, that charge is just for the first 15 minutes of therapy.