Tuesday, December 30, 2008

Selling Rx Data: Recent Sleaze from the AMA

As I liked to fancy myself as something of a 60s student radical, on becoming a medical student in the early 1970s I vowed not to join the AMA--mostly because that that time it stood squarely against any efforts to provide universal access to health care. More recently, some of my AMA member friends periodically try to persuade me that I could do more good as a critic inside rather than outside the organization and that I should join.

This is one of those times when I am very pleased not to be an AMA member.

Dr. Joseph M. Heyman, chair of the AMA Board of Trustees, has recently spouted off on the issue of selling physicians' prescribing info to the drug companies:


What apparently prompted this comment was the recent action of the US 1st Circuit Court of Appeals, reversing a lower court ruling that prevented implementation of the New Hampshire law that would ban the sale of physician prescribing information. Dr. Heyman personally has no trouble sharing his prescribing info with Pharma, but admits that some of his fellow docs see the issue differently. He therefore thunders against the court ruling, and the New Hampshire law, as threatening "physician choice."

The commentary is basically a long praise of the AMA's Physician Data Restriction Program, begun in 2006. The PDRP is supposed to allow any physician (AMA member or not) to opt out of the organization's current practice of selling identifiers.

Some background--Medical data mining firms currently buy prescription data from pharmacy chains, pharmacy benefit managers, and who knows what else. They take out all the patient identifying data to comply with privacy laws. They then sell this information to drug companies to allow them to carefully target their sales pitches to docs by letting the drug reps know exactly what and how many prescriptions that doc has recently written (for instance, does he prescribe a lot of your company's drug; or does he prescribe a lot of the competitor drug; and when asked by the rep what he prescribes does he tell the truth). These data, however, are useless unless you can turn the code number that identifies the doc who wrote the Rx into a name and address. That's where the AMA comes in-- they maintain a master file of all those code numbers and they sell that list to the data mining companies for many millions of bucks annually, making up a tidy portion of the AMA's annual revenue.

Till 2006, individual docs (besides having no clue that all this was happening behind their backs) had no say in whether the AMA sold their identifying information. With the start of the PDRP, the AMA quite grudingly offered the choice of opting out. Reports from neutral parties were that very few physicians exercised that right to opt out and most docs were unaware that the PDRP was available.

Why do I accuse Dr. Heyman of sleaze? Mainly because of this statement: "The AMA does not collect, license, sell or have access to physician prescribing data. Health care information companies collect and compile physician prescribing data and sell it to pharmaceutical firms. Through licensing agreements with these companies, the AMA can exert influence to offer programs such as PDRP."

This is a pretty slippery assertion. The impression it seems designed to leave with the reader is that the AMA does not profit from the sale of prescribing data and that it can influence the companies that do sell the data only indirectly; but the AMA is so influential that somehow the PDRP works and those companies agree to abide by the request not to release some individual docs' data. Which of course is not quite true. The truth is that the AMA does not directly sell the prescribing data, but they do sell the secret code that you need to understand the prescribing data, and they make a bundle off selling that. (There is a relevant scene in the Marx Brothers movie, A Day at the Races, but that would take me too far afield.) The PDRP works if it does because the AMA has direct control over the data that it does sell, not because it can "exert influence" over the data miners. (Or so all previous accounts from neutral sources have explained.)

Dr. Heyman refers his readers to the AMA's website that describes PDRP in more detail, and that website directs you in turn to a glitzy brochure:


The brochure (almost all photos and hardly any text) seems designed to persuade you of two things--one, physician choice is absolutely essential here and it would be an ethical outrage if that choice were ever taken away; and two, there are tremendous public health benefits to selling the prescribing data. The AMA further promises to launch a new information service to docs allowing them to compare their own prescibing practices to that of their peers. (Which important service you could of course not have access to, if you were so dim-witted as to have opted out via the PDRP.)

AMA critics point out that hardly any legitimate (that is, non-commercial) researcher uses these data as they are simply too expensive to purchase. I would add that in my last 10 years or so in practice, I had virtually all the data I could want on how my prescribing practices compared to other docs in my community through the kindly managed care plans that wanted me to keep my costs down.

I find it interesting that the AMA has elected to identify the sole ethical question related to this issue as physician's freedom of choice. The ethics of assisting the drug industry to make drugs more costly for patients by aggressively marketing physicians to prescribe only expensive brand-name drugs seems irrelevant to them--as does the ethics of the AMA itself profiting handsomely from this practice.


Anonymous said...

"What apparently prompted this comment was the recent action of the US 1st Circuit Court of Appeals, reversing a lower court ruling that prevented implementation of the New Hampshire law that would ban the sale of physician prescribing information."

I always find double negatives confusing. What happened is (1) NH passed a law banning the sale of the information, (2) the lower court found the law unconstitutional, and (3) the First Circuit reversed. In other words, the First Circuit said the law WAS constitutional, and therefore the implementation of the law banning the sale of prescribing information could go forward.

I'm not a constitional scholar, but it doesn't particularly surprise me that the First Circuit reversed. The state of New Hampshire has a strong public interest in preventing the sale of the information.

Anonymous said...

There seems to be the mistaken impression that the AMA has some secret formula that allows prescribing data to be useful. The fact is there are all sorts of physician databases with unique identifiers including state license numbers, DEA numbers abd the NPI. The AMA could go away and the practice would still exist.

Dan said...

Pharmacies sell prescribing data, also known as prescriber-identifiable data, to the pharmaceutical industry’s providers of this data.

The two largest data mining companies are IMS and Verispan in this 2 billion dollar a year data mining industry.

The data on individual prescribers is purchased from the American Medical Association. As a result, this association receives between 40 and 50 billion dollars a year, which is nearly 20 percent of the AMA’s budget.

Over 1 billion prescriptions are monitored every year due to this process.

Advocates of the pharmaceutical industry have continuously told the public that this prescriber data is to facilitate the care the prescriber gives the patient.

Nothing could be further from the truth. In 10 years, I worked as a drug representative for 3 of the largest pharmaceutical companies in the world. With each employer, the prescribing data that we had on the prescribers we constantly tried to influence to benefit these employers was used to manipulate these prescribers, and not to facilitate their efforts in any way.