The Influence of Drug Distribution Data on Diabetes Care in the Community

CLINICAL DIABETES
VOL. 17 NO. 4 1999

These pages are best viewed with Netscape version 3.0 or higher or Internet Explorer version 3.0 or higher. When viewed with other browsers, some characters or attributes may not be rendered correctly.

THE BUSINESS OF DIABETES

Steven B. Leichter, MD, FACP, FACE

Pharmaceutical companies exert an important and often constructive influence on diabetes care, whether viewed nationally, regionally, or locally. A previous article¹ evaluated their usual approach to selling products to providers. That article suggested that companies' sales approaches are based on more than the hard economics that drives their efforts.

To better understand the motivations and concerns that drive corporate efforts on a community basis, one must know the data that describe their success or failure in each community. More importantly, it is essential to become familiar with the implications of these data on the efforts of vendor companies in a given community. Most providers are unaware of these concepts or of the data that constitute their underpinning: drug distribution data. An overview of these issues is presented here.

What Is Drug Distribution Data?
Vendor companies purchase by contract data about product sales from three sources. The most important of these is IMS Health, of Westport, Conn. IMS and two competitors buy data on the sales of specific drugs or vendor products from pharmacies, chains, and product wholesalers. These data may demonstrate prescription patterns by zip code, by physician, or by other categories. Which data set a drug company or vendor company receives depends on its contract, but zip-code data and physician data are perhaps the most common sorts of data provided. These data may be provided for an individual product or for one brand versus others in a product class. Data on wholesaler shipment of products is also a common form.

When viewed by community, these data sources give vendor companies a picture they interpret as an authoritative representation of prescribing patterns for their products by specific areas (zip codes) and by individual physicians in the community. Until now, these data have not included other providers. However, other providers working with physicians would have their prescribing data attached to those of the physician listed on the prescription pad.

The data tell companies how big their present market and market share is in the community, and the data for the class of products show how much potential there is in the community. For example, data for a given brand of insulin might show how many prescriptions have been written for that brand of insulin. The data may also describe the total number of all insulin prescriptions written in that geographic area, and the percentage of the total that the individual brand comprised. Drug distribution data also show companies how many prescriptions an individual physician was reported to write for a given brand of a product such as insulin, the total number of insulin prescriptions that physician wrote, and even the total number of diabetes-related prescriptions that the physician wrote in a given period of time.

One interesting aspect of this type of data is that physicians are not supposed to know their data or the data for their communities. This is a standard part of all contracts that IMS writes with vendor companies. It is the responsibility of the vendor company to protect the data from physicians. This is a peculiar restriction, since the data not only relate to physicians themselves but also may affect their practice and community situation. Whether this restriction is legal is a controversial point.²

Impact of the Data on Individual Physicians and Their Communities
Companies develop "target" lists of physicians to cultivate for specific products. These lists are initially compiled from prior prescribing behaviors, as depicted by the drug distribution data. Physicians thought to show an active prescribing pattern for a class of drug or product will attract much more attention, especially early in the product cycle, than those who do not. Companies may decide to provide such physicians with enhanced knowledge levels about their product and the disease area, speaking opportunities, or other advantages. Physicians in that situation may at times exert influence on the company, its regional and local policies, and its sales approach for the product.

Physicians in that situation may also engage the company as partners in the development of medical education and care in their communities. Companies tend to allocate more local and regional marketing resources where they perceive there is a good or potentially good market. Resources may yield enhanced continuing education programs, educational resources for the community at large, or resources to develop local models of care.

Impact of the Data on Diabetes Care in the Community
In at least some circumstances, these data may have a profound impact on diabetes care in the community. Communities perceived to have large potential markets for a diabetes drug or product may gain increased focus from relevant companies. This focus may yield resources for community programs, physician education, or the development of new models of care. In the absence of recognition of market potential or a sense on the part of the company that it needs to invest in that community to grow or maintain market share, these resources are allocated elsewhere.

Often the physician or physicians who govern substantial market shares must be the catalyst to funnel these resources in partnership with other active elements in diabetes in the community. For this to occur, the physicians need to recognize the relationship between potential market size, their own individual influence in the community, and the possibility that a constructive partnership may be developed with the vendor company for the betterment of the community.

Potential Errors in the Data
This system is not foolproof. There are potential errors in the reporting mechanisms. Companies vary in the degree to which they accept that their data may be flawed.

The most common error is misreporting of physician location. Physician location is ascribed by IMS and others according to the Medical Education number (ME number) assigned to the physician by the American Medical Association. There may be a considerable lag between the time a physician establishes the original location for a medical career and the time that IMS or its competitors begin to report data on that physician. There may be a similar lag between the time a physician relocates practice and the time data is reported for the correct, new location. Nor is this problem necessarily uniform among all pharmaceutical companies. After establishing a new location, a given physician might be accounted by one vendor company far earlier than by another.

This is not an inconsequential problem to local pharmaceutical representatives. If a prescribing physician in their location is misassigned to another location, credit for all prescriptions written for that physician goes to the representative in the wrong location. Because bonus payments to these representatives are usually based on local sales and sales trends, the incorrect location may affect the annual earnings of the representative.

A second source of error is the absence of data sources from these databases. IMS and its competitors are currently hampered by the absence of two large pharmacy chains from participation in these databases. Neither the Wal-Mart nor the CVS retail chain provides such data at present. Consequently, IMS has compensated by devising statistical formulae to cope with these gaps. However, the greater the market share for either Wal-Mart or CVS in a community, the greater the probable error is in the IMS calculations. In communities such as ours, the presence of a large military population may skew the data if companies fail to account prescribing patterns at the military facility to relevant specialists in the community. Communities with such a situation may show inaccurately low market potential and not be viewed as favorable by vendor companies.

The third error is a failure on the part of vendor companies to recognize market influence by a given physician: they interpret the data too literally. Thought leaders may not only write a significant number of prescriptions for a given product; their prescribing patterns may influence the prescribing habits of others. Thus, a company might be more inclined to provide support to one physician who writes more prescriptions than a second physician, whereas the second physician influences more market share than the first physician. Another example of this has been seen in the market for weight-loss drugs. At least two pharmaceutical companies failed to achieve sales goals with these agents, in part because the physicians they identified by prescribing history were not the physicians with community thought-leadership.

Conclusion
Drug distribution data profoundly determine how pharmaceutical companies approach given physicians and communities. These data are often an important driving force behind vendor approach to physicians in a geographical area. An understanding of these data and their meaning to vendor companies on the part of physicians can create positive opportunities in the market for the companies and in the community for the physicians and the people of their geographic area.

REFERENCES

¹Leichter SB: The selling of diabetes products. Clinical Diabetes 16:186-88, 1998.

²Personal communication: May 19, 1999. Marc Calhoun, attorney-at-law. Page, Scrantom, Sprouse, Tucker, and Ford, Columbus, Ga.

Steven B. Leichter, MD, FACP, FACE, is the managing director of the Columbus Health Foundation in Columbus, Ga., and a professor of medicine at Mercer University School of Medicine in Macon, Ga.

Return to Issue Contents

For Technical Issues contact webmaster@diabetes.org