AMA Discloses Masterfile Physician Data to Pharmaceutical Companies

Laurie Barclay, MD

July 12, 2007

July 12, 2007 — Recent protests at the American Medical Association (AMA) annual meeting in Chicago brought to the forefront the fact that the AMA has begun disclosing information in its Masterfile, or physician database, to pharmaceutical companies, as reported in a Chicago Tribune news article published June 24. Although the Prescription Project, the National Physicians Alliance, and the American Medical Student Association (AMSA) are protesting the "sales" of these data for pharmaceutical marketing purposes, the AMA calls the financial arrangement "licensure" and notes that physicians have the right to opt out of sharing their personal information.

"The Prescription Project, AMSA and the National Physicians Alliance have partnered to bring the practice of data mining into the public eye and educate physicians about the problem," Robert Restuccia, BA, MPA, executive director of the Prescription Project, told Medscape.

"We are calling on the AMA to stop the practice and on individual doctors who do not want to be complicit in this process to communicate their concerns to the AMA," Mr. Restuccia said. "We are also calling on state officials to support legislation to end the sale of prescriber level data for pharmaceutical marketing purposes." The Prescription Project is led by Community Catalyst in partnership with the Institute on Medicine as a Profession and is funded by the Pew Charitable Trusts.

"The AMA has been responsibly licensing its database for 65 to 70 years for all sorts of purposes: medical education, both continuing and graduate, and for marketing and credentialing purposes," Robert Musacchio, PhD, AMA senior vice president of Publishing and Business Services, told Medscape. "Every hospital in the country uses data from our file to verify that a physician is who they say they are. We do not sell the information, we license it, which means we can control where it goes, to whom it goes, and the manner in which it's used throughout our contract."

Since the AMA began compiling physician data in 1906, the Masterfile is now more than a century old and includes approximately 900,000 physicians, about two thirds of whom are not AMA members.

Physicians, Organizations Object to Data Sales

"As a physician, I have no knowledge or control over data about me that is sold in the AMA Masterfile," Michael Mendoza, MD, MPH, a clinical assistant professor of family medicine at the Pritzker School of Medicine, University of Chicago in Illinois, told Medscape. "I am not an AMA member, and I feel that the AMA has abused my rights to privacy by selling data about me without my consent for commercial and marketing purposes.

"This practice ultimately hurts patients because these data ultimately help pharmaceutical representatives to create highly effective tactics to persuade doctors to improperly prescribe more profitable drugs that have not been shown to be as effective as older drugs," added Dr. Mendoza, who is also a member of the National Physicians Alliance and Prescription Project. "These newer drugs are more expensive and also contribute to rising healthcare costs, a cost that is passed on to taxpayers and individual patients."

The massive database contains "information you could obtain from just about any state licensing board Web site," according to Dr. Musacchio, including the name, address, demographic data, practice type, specialty, medical school and residency training, and licensure and credentialing information on all US physicians — AMA members and nonmembers alike. However, prescribing data are not compiled by the AMA.

The AMA has issued licenses to peruse the database to about 10 different licensees, some of which have various business arrangements with pharmacies.

The pharmaceutical companies "take our data as well as [data from] several dozen other databases and combine them together with information that they receive from pharmacies, and they put together a picture of physicians' prescribing habits — by zip code, by specialty, by individual physician — and they use it for their planning and marketing purposes," Dr. Musacchio said. "It benefits physicians in that they receive targeted visits from pharmaceutical reps as opposed to broad visits, so that physicians don't receive visits concerning therapeutics that they might or might not be interested in. This is not really a patient issue, from our perspective, but I would ask how it would hurt patients if it's designed to provide physicians with information about therapeutics."

Mr. Restuccia told Medscape that health information firms, such as IMS Health Inc, match the physician-identifying information in the Masterfile with prescription data purchased from pharmacies to produce physician prescribing profiles. These profiles are then sold to pharmaceutical manufacturers to be used by detailers to tailor their sales pitches to doctors.

"Information from the AMA Masterfile should not be sold for pharmaceutical marketing purposes," Mr. Restuccia said. "The job of these pharmaceutical reps is to increase sales of their products, not to provide unbiased evidence based prescribing information to physicians. Thus we think it is counter to the mission of AMA to provide the means for the industry to influence prescribing patterns of physicians."

Mr. Restuccia added that drugs that are most heavily marketed to physicians and to consumers account for the largest increases in prescription drug costs both for patients and for the healthcare system as a whole. Furthermore, some of these drugs, such as Vioxx and Avandia, have subsequently been found to pose significant risks to patients.

"We feel that the only way pharmaceutical companies could use this data for sales is to target those physicians who are not prescribing 'enough' of a particular medicine," AMSA President Michael J. Ehlert, MD, from Reston, Virginia, told Medscape. "The prescribing patterns are influenced by the specific conditions of each individual patient, and it is reckless to determine that there are quotas for how much of any given drug needs to be prescribed.... Using therapeutics is a sophisticated process that is learned through almost a decade in medical school and should be free of the influence of hired sales reps without a medical degree and no access to a patient's chart or medical history."

In the May 22 issue of the Washington Post, Marjorie E. Powell, senior assistant general counsel for the Pharmaceutical Research and Manufacturers of America, stated that prescribing data allows "the company to do more targeted marketing, which lowers the total costs of its marketing." Dr. Ehlert believes that this comment further supports the notion that pharmaceutical companies use this information to target physicians whom they feel can increase the use of the company's drugs.

Precautions to Protect Physician Privacy

According to Dr. Musacchio, precautions to protect physicians' privacy include never licensing or releasing Social Security numbers and providing physicians with several options in addition to their ability to opt out of having their data released to licensees. When a physician elects to opt out through the Physician Data Restriction Program (PDRP), the pharmaceutical companies agree to not release the physician's prescribing profile to individual detailers.

"For at least the past 25 years, physicians have had the option of checking off a box that says 'do not contact,' so that their data will never be released for marketing purposes by email, phone, fax, or mail, or 'do not release,' which means that their data is not released to anyone, in any way, shape, or form, including hospitals to which they are seeking admitting privileges," Dr. Musacchio said. "Every survey that goes out has a letter attached to it [in which] we inform the physician of their ability to do this, and recently we've been highlighting the PRDP, because the program is only a year old. Our attention is going further in the future to give physicians many more options for how they can manage their preferences."

For example, physicians can elect to receive informational materials from pharmaceutical companies only by mail but not by phone or at the office, or they can agree to visits by pharmaceutical representatives without having their data released to the pharmaceutical companies.

"Physicians can opt out by enrolling in the AMA's PDRP, by which pharmaceutical companies agree to not release the physician's prescribing profile to individual detailers," Mr. Restuccia said. "Although this program is inadequate and relies on the drug companies to police themselves, it is better than the alternative — tacit compliance. Individual doctors can [also] educate their peers and voice their concerns to the AMA and their state medical societies. "

Mr. Restuccia cited a survey by the Kaiser Family Foundation showing that only 60% of physicians were aware that the AMA is selling their information, but 74% of physicians were opposed to the practice once they were so informed.

"It is relatively difficult for individual physicians and/or medical students to do anything about this once the data are gathered," Dr. Mendoza said. "Even if physicians opt out using the AMA's PDRP, pharmaceutical companies can still purchase the data from the AMA. Broader reform is needed to truly address this problem: Legislation banning the sale of physician data for commercial and marketing is the only way to protect physicians and students from this practice."

Physician awareness of the PDRP is limited but increasing, according to Dr. Musacchio. An AMA survey in January 2007 showed that only about 10% to 15% of physicians were aware of the PDRP; a second survey in late May showed that physician awareness had increased to 33%.

Increase Physician Awareness of Opt-Out Program

"Our goal is to have 60% of the physicians aware of the program at the end of the year and shoot for 100% by the end of 2008, which is probably an unrealistic goal," Dr. Musacchio said. "We're doing everything possible to make physicians aware of this, and so far there has been a higher level of awareness than I thought there would be, but a much lower response in terms of people actually opting out."

Dr. Musacchio attributes some of the increase in physician awareness to AMA advertising of the PDRP in specialty medical journals, featuring it on the AMA Web site, and heightened media coverage. The AMA sent out about 110,000 emails to physicians who have given permission to contact them via email, and less than 1.5% responded by electing to opt out. Even with direct registered mail, the response rate has only been about 3%. More than 8000 physicians have opted out to date.

"While roughly around 60% to 65% of the physicians said [data disclosure] was an issue, most of them said it was not a major issue, and about 85% of them said that putting together a program like the PDRP would alleviate their concerns, so that's the route we chose," Dr. Musacchio said. "We personally feel it's a better route to go than legislation — we don't think the legislation will work, as indicated in New Hampshire."

In April, a federal judge in New Hampshire ruled in favor of IMS Health Inc and Verispan LLC by deeming unconstitutional a "prescription information law" advocated by the state's attorney general to curtail the compiling and use of data identifying doctor prescribing practices. The two companies had purchased prescribing data, removed patient identifiers, and combined the information with other data to be sold to pharmaceutical companies. Subsequently, Vermont passed a law allowing physicians to approve or disapprove use of their personal prescribing data to market drugs, and it appears that Maine will soon follow suit.

Legislative Reform in the Works

Groups such as the National Physicians Alliance, AMSA, and the Prescription Project have joined together to fight for legislative reform. AMSA policy documents published in 2006 state that AMSA "opposes the tracking of prescriptions by commercial entities and supports legislation to limit access to individual prescription patterns of physicians by the sales and marketing departments of pharmaceutical companies."

The Prescription Project has provided testimony on legislation recently signed into law in Vermont and soon to be signed in Maine limiting the practice of disclosing physician data to pharmaceutical companies and giving physicians a choice regarding disclosure of their prescribing information. They have also provided technical assistance to uphold a New Hampshire ban on the release of prescriber-specific data and are working in other states such as Massachusetts and Washington on similar policy changes.

Some critics of the recent move by the AMA to include pharmaceutical companies in its range of Masterfile licensees have asked why physicians were not automatically opted out of the Masterfile and then given the opportunity to opt in, rather than automatically being opted in and then having to opt out if they wished to prevent their personal data from being disclosed.

"When we started the Masterfile about a hundred years ago, it was designed to protect the public from fraud and abuse," Dr. Musacchio said. "During the last century, it was very difficult to determine who was a physician, and that was our mandate then — to start collecting data. We actually have information on every physician that has practiced, been licensed, or gone through either a US medical school or an [Educational Commission for Foreign Medical Graduates (ECFMG)] pathway going back to 1906."

Dr. Musacchio explained that even if physicians opt out of having their data released to third parties, the AMA still compiles and maintains the data to protect the public interest. For the same reason, it collects data on all US physicians even if they are not AMA members. This includes not only those physicians who are educated in US medical schools but also those who are licensed through the EFCMG pathway.

Database Has Other Uses

"The database is used by the government to protect against fraud: the FBI has used it, it's been used in Hurricane Katrina to validate physicians' licenses, and the State Boards have used it to validate information about a physician coming from another state," he said.

Potential benefits from releasing physician prescribing data may also include facilitating medical research and clinical trials, heightened drug surveillance, and reconstruction of patient records after natural disasters such as Hurricane Katrina.

Revenue gained by the AMA from licensing data in the Masterfile supports the activities of the organization, including its professional standards group, its medical education group, operations such as maintaining the Masterfile itself, and subsidizing many of the AMA's publications that are not profitable. The total revenue is approximately $45 million, or about one fifth of the association's overall budget, according to Dr. Musacchio, and that includes licensing fees for all uses of the Masterfile such as credentialing and continuing medical education (CME).

"We don't break out the amount generated by licensing to pharmaceutical companies because there are only about 4 of these licensees; it would be pretty obvious to each of them how much business they do with us; and we feel that that's proprietary," Dr. Musacchio said.

In addition to the financial benefit to the AMA and the benefit to the public provided through verifying physician credentials, Dr. Musacchio believes that allowing pharmaceutical companies to access the Masterfile can also benefit physicians. The AMA has developed a CME initiative called Therapeutic Insights, which allows physicians to see prescribing patterns similar to those viewed by the pharmaceutical companies. Through a newsletter and Web site, physicians can compare their own prescribing patterns with those of their specialty and their region as a whole.

"The feedback we're getting from this initiative is phenomenal," Dr. Musacchio said. "Physicians were concerned that the pharmacies had more information than they had, and now they are saying that this is really good information. It's free of charge — purely an educational venture — and we believe in data transparency."

Citing concerns of professionalism and of the integrity of medical education, AMSA has since 2001 declined any funding or advertising from the pharmaceutical industry and has opposed direct-to-physician advertising of drugs through sales staff, research projects, or CME lectures.

Responsible System to Allocate Resources Needed

"It is physicians who allow these sales reps into their offices, often under the guise of free samples and free lunch," Dr. Ehlert said. "This underscores the real problems in American medicine: The price and availability of drugs is such that practicing doctors feel they must rely on giveaways from companies to treat their patients. If we had a responsible system that allocated resources to each patient and regulated the pharmaceutical industry, much like the VA system, we could afford to treat each patient we saw."

AMSA has released a "pharm-free score card" ranking each medical school according to their policy about drug representatives interacting with medical students. At present, only 6 schools have implemented policies to prevent industry representatives from talking with students, but dozens are in the process of crafting similar policies.

To facilitate access to unbiased sources of drug information, AMSA is now offering free subscriptions to The Medical Letter, which has no corporate advertising, to all of its clinical-year medical student members.

"By not teaching students about the conflict of interest inherent in a paid company rep teaching about their product, we are teaching that this relationship is okay," Dr. Ehlert concluded. "It is not okay, and it should be removed from America's training institutions, and certainly [from] medical schools altogether.... It is being recognized by our educators that it is unacceptable, and hopefully America's practicing doctors will step back from the drug companies...and reach out to unbiased sources of drug information."

Requests for comments from other sources, such as the chair of the AMA Board of Trustees, the vice speaker of the AMA House of Delegates, and the AMA president-elect, were either not acknowledged or not granted.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.