Prescription data

We thank David Zitner and Samuel Shortt for their observations concerning the shortcomings of present health care databases.¹

We do not wish to argue semantics with Stuart MacLeod; however, the remarks he has deemed to be pejorative and aspersions were neutral descriptions. Where we differ substantively is with regard to the need for prescription data to be collected and sold in accordance with CMA principles.² These principles essentially boil down to informed consent. Implicit in MacLeod's arguments is the view that informing physicians about prescription data mining activities and seeking consent would negatively affect IMS HEALTH's databases and related medical research. Similar arguments could be used against the general requirement of informed consent in medical research. Although IMS conducts prescription data mining in accordance with self-serving ethical standards,³ university research ethics boards would not likely approve a study that used IMS methodology. Academic physicians using IMS data might consider the ethical standards under which the data were compiled and encourage IMS to collect physician prescription data with informed consent.

Roger Korman's central thesis is that our article does not reflect the current reality of IMS business practices. We endeavoured to present a balanced portrayal of prescription data mining; we even provided drafts of the article to IMS and incorporated many of their suggestions. Our research led us to conclude there should be independent regulation of the industry.

IMS lauds their aggregation of prescription data sold to pharmaceutical companies. However, aggregation with physician identification does not preclude the generation of individual physician prescribing profiles. Each physician's identification number is part of such reports and can be linked to contact information and other reports on their prescribing practices. The value of prescription data sold to pharmaceutical companies would be greatly depreciated if it were not possible to link physicians to their prescribing practices.

The pharmaceutical companies pay the freight and it is principally their interests that are being served. Any benefits accruing to researchers, medical educators, physicians, politicians, policy analysts and the public are secondary and offered in exchange for allowing IMS's business practices. The value of these data does not negate the obligation to collect and sell this information with the informed consent of physicians.

Korman points out that IMS's practices have been approved by their own Health Information Advisory Board. This is peculiar, as IMS's data mining operations are specifically excluded from the board's mandate.⁴ Although IMS may be certified by the Canadian Standards Association, informed consent is a tenet of the Canadian Standards Association privacy code except where deemed inappropriate.⁵ A note in the code defining “inappropriate” cites legal, medical and security reasons, but not business concerns.

We do not think the mailings to physicians in Ontario in 1996⁶ and in Quebec in 1999,⁷ which did not include consent-response forms and did not publicize the relevant Web sites, are adequate to inform Canadian physicians about IMS's business practices. We recommend that affirmation of informed consent be sought on a regular basis from all physicians across Canada.

The current reality is that prescription data mining practices are at variance with all 5 of the CMA's principles for the sale and use of physician prescription data.² The conduct of IMS bespeaks the current reality of the need for independent regulation of this industry.

Signatures

Dick E. Zoutman
Department of Pathology Queen's University Kingston, Ont.

B. Douglas Ford
Department of Pathology Queen's University Kingston, Ont.

Assil R. Bassili
Department of Pharmacology and Toxicology Queen's University Kingston, Ont.