US Pharm. 2009;34(6):48-49.
In recent years, the practice of pharmacies selling data to pharmaceutical companies or their agents has become commonplace. This information, called prescriber-identifiable (PI) data, includes which physicians prescribe what drugs, with any identifying patient information deleted first. The two leading purchasers of this data are IMS Health and Verispan.1 Together they monitor at least one billion prescriptions per year. As many as 51,000 pharmacies in the United States are wired to facilitate “data mining,” resulting in revenues to the “data miners” in excess of $2 billion annually.2 This is just the tip of the iceberg. The pharmaceutical industry expends upwards of $30 billion annually on detailing, advertisements in medical journals, direct-to-consumer advertising, and drug samples, with approximately $7 billion spent on marketing directly to physicians (averaging about $9,000 per prescriber).3 Even the American Medical Association (AMA) sells data to the groups that gather this information. In one year alone, the AMA profited $44 million in income from the practice (16% of its annual budget).4
Several states have tried to ban prescription data mining practices using rights of privacy and confidentiality rubrics. Early on in these efforts, at least two federal district courts held data mining prohibitions unconstitutional as a violation of the miners’ rights of free speech.5,6 Following some recent cases, and with an accompanying plethora of additional state legislation, the tide may have turned, at least for now, in favor of protecting the privacy of prescribing patterns.
On November 18, 2008, the decision of the federal district court sitting in New Hampshire that had ruled that the state’s ban on data mining was unconstitutional was reversed by the First Circuit Court of Appeals.7 The appellate court found that the New Hampshire Prescription Confidentiality Act8 substantially furthers the state’s interest in keeping drug costs affordable by encouraging the use of multi-source generic drugs, an effort diametrically opposed by the data miners who use the information to encourage physicians to prescribe more expensive, patent-protected, brand-name drugs. Arguing in favor of upholding the state law, the attorney general of New Hampshire called the drug companies trying to obtain the questioned data “invisible intruder[s] in the physician’s examination room.” The First Circuit ruling states that the pharmaceutical data at issue is akin to an article of trade that can be regulated without running afoul of the First Amendment.9 Judge Bruce Selya, in authoring this opinion, wrote that: “The plaintiffs, who are in the business of harvesting, refining, and selling this commodity, ask us in essence to rule that because their product is information instead of, say, beef jerky, any regulation constitutes a restriction of speech. We think that such an interpretation stretches the fabric of the First Amendment beyond any rational measure.” In other words, data mining is “conduct,” not First Amendment “speech.” He concluded that: “While the plaintiffs lip-synch the mantra of promoting the free flow of information, the lyrics do not fit the tune.” IMS and Verispan jointly filed a petition asking the U.S. Supreme Court to review the First Circuit’s opinion.10 To date, the Supreme Court has not ruled whether it will hear the petition.
Maine also adopted a state prescription drug privacy law in June 2007.11 The bill prohibits the sale of prescription drug information that identifies, directly or indirectly, patients or health care providers. The amendment includes language substantially similar to the New Hampshire Prescription Confidentiality Act.12 However, the law would allow Maine doctors to opt out of participation in data mining, which would prevent a prescription drug information company from licensing, selling, or getting any value from transferring a doctor’s prescribing information.13 A federal district court judge sitting in Maine issued a preliminary injunction in December 2007 against many provisions of a state law slated to take effect on January 1, 2008.14 That action was stayed pending the First Circuit’s opinion regarding the legality of the New Hampshire law.
In another case, a federal judge in the District Court of Vermont upheld that state’s 2007 legislation banning data mining unless physicians opt-in to let drug manufacturers access their prescribing records.15 Judge J. Garvan Murtha issued an opinion on April 23, 2009, summarizing the data mining process and its implications. Pay close attention to the judge’s view of the participation of pharmacies in this practice:
In the course of filling prescriptions, pharmacies acquire prescription information. Certain information, including the prescriber’s name and address, the name, dosage and quantity of the drug, the date and place the prescription is filled and the patient’s age and gender, is purchased by third parties who, after manipulating the data, sell it to customers, principally pharmaceutical companies…The manipulated data shows, among other things, details of physicians’ prescribing patterns in terms of gross number of prescriptions and inclination to prescribe a particular drug.
Pharmaceutical manufacturers collectively spend close to $8 billion a year to market drugs directly to prescribers, employing thousands of sales representatives. The estimated total cost of marketing to Vermont prescribers approximates $10 million, not including samples or direct-to-consumer advertising. Sales representatives provide “details” regarding the use, side effects, and risk of interactions of the drug they are selling. For this reason, sales representatives are called “detailers.” In addition to “details” and samples, representatives distribute medical literature and give small gifts such as pens, notepads, or lunch. Prescribers often rely on information provided by detailers because keeping current with the changing landscape of prescription drugs is time-consuming.16
The judge found that the state has a substantial interest in saving health care costs and that the legislation in question constitutionally furthers that interest. He wrote:
Only new, branded drugs are detailed because the introduction of generic bioequivalents into the market renders detailing no longer cost effective. The Legislature found that, coincident with the phenomenon of “data mining,” the pharmaceutical industry increased spending on direct marketing to doctors by over 275%. The data provides detailers with specific information about doctors’ prescribing practices, enabling them to target certain prescribers for their marketing efforts and to tailor presentations to individual prescriber styles, preferences, and attitudes. This information amplifies the influence and effectiveness of detailing, but does not add to its purported educational value. Detailers can provide medical literature and information regarding the drugs they are promoting without the benefit of PI data. The Vermont Medical Society has stated tailored marketing using PI data “is an intrusion into the way physicians practice medicine” and it creates the “possibility that representatives could exert too much influence on prescription patterns.” Detailing leads to increased prescriptions for new drugs over generic alternatives, which are often more cost-effective.
Research shows doctors are influenced by the marketing efforts of pharmaceutical companies. For example, doctors who attend talks sponsored by a pharmaceutical company often prescribe that company’s drug more than competitors’ drugs. Though plaintiffs attempted to show that doctors are not influenced by marketing practices, that point is belied by the nature of the industry, plaintiffs’ own documents, and scientific research. The main purpose of detailing is to increase the number of prescriptions written for the drug being promoted. The billions spent each year by pharmaceutical manufacturers on detailing is evidence of its success. Pharmaceutical manufacturers are essentially the only paying customers of the data vendor industry. This is the strongest evidence of the important role of PI data in pharmaceutical detailing. Put simply, if PI data did not help sell new drugs, pharmaceutical companies would not buy it.17
In an unrelated case, but one with similar implications for privacy rights, a federal District Court in the District of Columbia held that the public has a right to access Medicaid Pay for Performance program data.18 In January 2009, the U.S. Court of Appeals for the District of Columbia reversed that ruling by finding the public interest in prescribing data by physicians to be nonexistent, reasoning that release of information about which doctors prescribe what drugs would be an “unwarranted invasion of the personal privacy” of physicians.19 Information made public under the plaintiff’s request could be misleading if it were used to judge the competency of an individual physician based on the numbers and kinds of medications prescribed without taking into account other more important factors not considered in these records. Further, when combined with other information already in the public domain, the data sought in this case could be combined to determine a physician’s total income from Medicaid payments. Again, the appellate court did not find a compelling reason to disclose this otherwise confidential information.
Pharmacists should be aware that at least 15 states have adopted some type of privacy law preventing disclosure of PI data.20 Among others, including those discussed in Maine, New Hampshire, and Vermont, bills restricting access to prescriber data for marketing uses are pending in California, Kansas, Hawaii, Maryland, Massachusetts, Nevada, New York, and Washington State.21,22 Pharmacists should check with their state pharmacy associations or attorneys well versed in pharmacy law to avoid running afoul of any of these restrictions.
Case Law Update: The June 2009 edition of this column featured a story entitled, “Pharmacists Beware: Data Mining Unlawful.” The gist of the article was that some courts had recently determined that the practice by pharmacies of selling prescribing information to “data miners” could be banned under various state laws prohibiting the practice. The primary case involved New Hampshire legislation that restricted data mining activities.8 On November 18, 2008, the First Circuit Court of Appeals reversed a decision of the district court that had determined the data mining ban unconstitutional as a violation of the First Amendment right to freedom of speech.7 The Court of Appeals reasoned that the law substantially furthers the state’s interest in keeping drug costs affordable by encouraging the use of multisource generic drugs, an effort diametrically opposed by the data miners, who use the information to encourage physicians to prescribe more expensive, patent-protected, brand-name drugs. The data miners filed a petition with the U.S. Supreme Court seeking reversal of the Court of Appeals decision and reinstatement of the federal district court opinion authorizing data mining.10 On June 29, 2009, the Supreme Court denied the petition, thereby allowing the New Hampshire law to take effect.23 By virtue of this action, a similar law in Vermont was also allowed to move forward. Other states are expected to follow with comparable legislation.
2. New Hampshire’s data mining case heads to the Supreme Court. Campaign for Prescription Privacy. April 1, 2009. www.patientprivacyrights.org/
3. Physician supply and demand: projections to 2020. Health Resources and Services Administration. http://bhpr.hrsa.gov/
4. Restuccia R, Vaias L. Prescription mining raises millions for doctors’ group. San Francisco Chronicle. July 25, 2007. www.sfgate.com/cgi-bin/
5. Vivian JC. Mining ban unconstitutional: freedom of speech wins. US Pharm. 2007;32(6):67-70. www.uspharmacist.com/content/
www.wlf.org/Litigating/ casedetail.asp?detail=462. Accessed May 2, 2009. This decision reversed the opinion discussed in the June 2007 edition of this column.
6. IMS Health v. Rowe, Slip Op No CV-07-127-B-W (January 2, 2008). See also: Maine prescription privacy law struck down. Electronic Privacy Information Center. http://epic.org/privacy/
7. IMS Health and Verispan v. Ayotte, Slip Op No 07-1945 (November 18, 2008). www.ca1.uscourts.gov/cgi-bin/
8. HB 1346 Final Version (06-24040). February 22, 2006. www.gencourt.state.nh.us/
9. Mauro T. Data mining case heads to the Supreme Court. Law.com. March 30, 2009. www.law.com/jsp/article.jsp?
10. Data mining case heads to the Supreme Court. March 30, 2009. (Summary statement of article referenced in note 5, supra.) www.patientprivacyrights.org/
11. MRSA 1711. An act to amend the prescription privacy law. State of Maine Legislature. June 18, 2007. http://janus.state.me.us/
12. Maine passes new prescription privacy legislation. Electronic Privacy Information Center. June 7, 2007. http://epic.org/privacy/
13. Qualters S. Drug ‘data mining’ digs up suits. National Law Journal. January 28, 2008. www.saul.com/common/pressroom/
14. IMS Health Inc. v. Rowe, Slip Op No 07-127 (DC Me).
15. Weiss-Tisman H. Judge upholds ban on data mining. Brattleboro Reformer. April 29, 2009. www.reformer.com/localnews/ci-
16. IMS Health v. Sorrell, Slip Op No 1:07-CV-188 (April 23, 2009), USDC Vermont.
18. Medicaid pay for performance programs slow to gain traction. University HealthSystem Consortium. May 24, 2007. www.naph.org/naph/
19. Consumer’s Checkbook Center for the Study of Services v. US Department of Health and Human Services, Slip Op No 07-5343 (January 30, 2009), DC Cir. See also: Field R. Physician rights to privacy of data prevail in two major court tests, but new questions lie ahead. Pharm Ther J. 2009;34:193-195.
20. O’Reilly KB. New Hampshire ban on sale of prescribing data upheld. American Medical News. www.ama-assn.org/amednews/
21. See note 11, supra.
22. Fact sheet. Prescription data mining. The Prescription Project. November 19, 2008. www.prescriptionproject.org/
23. Courts refuse to block state laws restricting Rx data-mining firms. iHealthBeat. June 30, 2009. www.ihealthbeat.org/Articles/
To comment on this article, contact email@example.com.