The Social Medicine Portal

New Hampshire's Law to Restrict Data Mining by Pharmaceutical Sales

Rep. Cindy Rosenwald

On December 8, 2009  Representative Cindy Rosenwald, Chair of the New Hampshire House of Representatives Committee on Health, Human Services and Elderly Affairs came to Social Medicine Rounds in the Bronx.  In her talk, entitled “Eavesdropping in the Exam Room” she discussed New Hampshire’s efforts to limit data mining by pharmaceutical companies.

It was particularly nice that Representative Rosenwald was accompanied by her husband, Dr Peter T. Klementowicz, who had been a Chief Resident, Cardiology Fellow and later cardiologist at Montefiore in the 1980′s. He now practices cardiology in Nashua at Catholic Medical Center.  He made his acting debut at Social Medicine Rounds.

Eavesdroping on doctors, eavesdropping on patients

The talk began with a short skit involving a pharmaceutical rep (played by Representative Rosenwald) who attempted to convince Dr. Klementowicz that he should be prescribing more PROFITAL to his patients instead of the competition’s brand MARGINEX. The drug sales representative asked how things had been going with the free PROFITAL coupons and samples that she had left off earlier. Dr. Klementowicz seemed to defer a bit and one sensed that perhaps he had not given out any free samples. She insisted that PROFITAL had a great safety profile, while there were some concerning side effects associated with MARGINEX. The meeting ended on an inconclusive note.

Representative Rosenwald then told us that she, in her role as sales representative, actually knew at the time of the meeting exactly how much PROFITAL and MARGINEX Dr. Klementowicz was prescribing. Not only that, she already had a pretty good idea of to whom he was giving the prescriptions. Her foreknowledge of the good doctor’s prescribing habits was the result of “data mining” a process by which commercial companies acquire information on what individual doctors are prescribing.  Knowing in real-time exactly what every doctor was prescribing was, she said “the perfect marketing tool” for the pharmaceutical companies.

What kind of data was available in these databases? She began by pointing out that – at the present time – prescription databases were exempted from federal HIPAA (privacy) legislation. This is a most curious exemption and it has allowed an industry to develop selling patient data.

What kind of patient data is available for sale?  At the very least there is information about the physician and the drug. But the data might also include PLD (patient level data) that contained some “de-identified” data about individual patients and their interactions with the health care system.  This might include simply the patient’s gender, age. But, she pointed out, in a sparsely populated state like New Hampshire, knowing that a 95 year old woman in Nashua got a medication for heart failure, might identify a single individual.

One of the companies that sell such data is SDI.  Looking at their website they state “SDI actively tracks a patient’s total healthcare experience, including:

• Filling prescriptions at a pharmacy
• Undergoing treatment by a provider in the office setting
• Entering a hospital or other facility for inpatient or outpatient treatment
• Recognizing how lifestyle contributes to healthcare decisions
• Receiving a laboratory test to evaluate disease severity
• Detailed clinical information from electronic medical records.”

Representative Rosenwald noted that the pharmaceutical companies had additional methods to learn about individual patients.  When patients are given coupons for samples or participate in patient assistance programs, they hand their personal information over the pharmaceutical company.  This is an added bonus to the “powerful marketing tool” represented by samples and the PAP.

The Players

Data mining is not a small business.  The two largest firms are IMS Health and Verispan.  To give a sense of size, Representative Rosenwald noted that IMS Health was recently sold for $ 4 billion.

An additional player in this market is the American Medical Association (AMA), which maintains a masterfile on all MD’s and DO’s in the US. The AMA leases the Masterfile to the drug companies.   By combining data in the AMA Masterfile with the prescription data from the pharmacies, the pharmaceutical companies can create comprehensive physician prescribing profiles.  The AMA has been challenged on this policy (see the Medscape Article: AMA discloses Masterfile Physician Data to Pharmaceutical Companies). In response, the AMA set up a Physician Data Restriction Program (PDRP) in 2006.  This program allows individual doctors to “opt out” of having their data shared with individual drug salespersons.  The data, however, is still leased to the pharmaceutical companies; the AMA does not allow an opt-out for this.

Why is this system troublesome?

Representative Rosenwald noted that the current system basically allows powerful corporations to find out what is going on between individual doctors and their patients.  Beyond the privacy concerns, Representative Rosenwald raised a broader question of values.  These corporations are concerned primarily with revenue, market share and profit; this is their mandate as corporations.   Data mining provides them with a powerful, perhaps an ideal marketing tool for their ends. But as a legislator her concern, which should be the concern of clinicians, is about providing affordable, safe, high quality health care.  Data mining seems to work against those ends.by encouraging the use of expensive new medications, instead of proven and inexpensive generic ones.

What New Hampshire did:

The spark for action in New Hampshire was the ever more expensive Medicaid program.  Prescription costs were a major part of that growth and from 2001-2005, they represented an expense of about $800 per person.  The legislators reasoned that limiting access of pharmaceutical reps to physician level data might help to encourage the use of generics and bring down prescribing costs.  On May 11, 2006 they passed the Prescription Restraint Law banning the use of prescriber-identified data in sales.  This was signed into law by the Governor on June 30^th and was set to into effect in July.  Under the law, pharmaceutical representatives were still allowed the use of zip-code identified data.

The law was immediately contested by Verispan and IMS; interested readers can consult their brief to the judge for their side of the controversy at this link (http://epic.org/privacy/imshealth/ims_facts.pdf).   They see the law as a restriction on their first amendment rights of freedom of speech.  Representative Rosenwald characterized their argument as being built around 3 myths: 1) that the data mining brought down costs by allowing them to target their sales efforts; 2) that it saved physicians time again by targeting company efforts to educate doctors; and 3) that it improved patient safety.  Representative Rosenwald characterized the latter argument as, perhaps, the most nefarious of all.  She pointed out that the use of databases to track errors and problems with drugs should be in the hands of a public authority and not the people selling and promoting the drugs.

The Court Case(s)

New Hampshire lost the first round of the legal battle when the law was struck down by a New Hampshire judge. At that point, New Hampshire was presented with a bill from Verispan and IMS Health for the $1.3 million in legal fees the companies had incurred. She felt that not only did this show the enormous efforts made to defeat the bill, but was also an attempt to intimidate a small state government for whom $1.3 million is an enormous sum.

New Hampshire appealed the case and the original decision was overturned in November of last year by the First Court of Appeals in Boston;  the US Supreme Court declined to hear a final appeal by the companies.  The First Court’s opinion is available at this link and includes the following:

In the pages that follow, we explain why we are not persuaded that the regulated data transfers embody restrictions on protected speech. In our view, the portions of the law at issue here regulate conduct, not speech. Unlike stereotypical commercial speech, new information is not filtered into the marketplace with the possibility of stimulating better informed consumer choices (after all, physicians already know their own prescribing histories) and the societal benefits flowing from the prohibited transactions pale in comparison to the negative externalities produced. This unusual combination of features removes the challenged portions of the statute from the proscriptions of the First Amendment.

There is a second basis for our decision. Even if the Prescription Information Law amounts to a regulation of protected speech — a proposition with which we disagree — it passes constitutional muster. In combating this novel threat to the costeffective delivery of health care, New Hampshire has acted with as much forethought and precision as the circumstances permit and the Constitution demands.

So the New Hampshire Law now stands.  This, however, is not whole story.

Subsequent to the passage of New Hampshire’s law, both Vermont and Maine passed similar types of legislation.  Vermont created an “opt in” system whereby physicians would have to choose to share their information with pharmaceutical companies.  Maine created an “opt out” program where (like the AMA program) physicians would have to chose not to participate.  Both programs were challenged in court.  A successful appeal to one these laws at the Appeals level could trigger review by the Supreme Court.

David versus Goliath

IMS Health was so upset with New Hampshire that, according to Representative Rosenwald, they stopped collecting data on New Hampshire doctors. This was entirely consistent with what was at stake.  In terms of prescription sales, states like New Hampshire, Vermont and Maine are relatively unimportant.  But they set a precedent that larger, far more (economically) important states like New York or Florida may follow. In fact, Senators Kohl and Durban introduced an amendment to the health care bill currently being debated in Congress that would impose similar restrictions.

So on the one hand we have a few courageous, publicly minded legislators for whom $1.3 million is a frightening sum.  And on the other hand we see the full force of organized medicine (ie. the AMA), the multi-billion dollar data mining industry and the multi-multi-billion dollar pharmaceutical industry.  It is really encouraging to see that so far the democratic process has worked in at least three states.  But it is also clear evidence of the fact that health care is no longer about providing “affordable, safe, high quality care” but seems more about protecting the interests of powerful corporations.

Lessons Learned

Representative Rosenwald concluded by noting some of the lessons she has learned through this process:

1. It is important to have the active help and support of your partners.  In this case, the law was supported by AARP, the New Hampshire Medical Society and the local Department of Health and Human Services.

2. Pharma and the data mining companies will fight using all the weapons they have: threats, intimidation and out spending.

Some practical steps forward.

Representative Rosenwald encouraged us to contact our Representatives to see about sponsoring a similar bill in New York.  Although she did not mention it, it might also be a good idea for physicians to use the limited AMA opt out.by logging on here.  Unfortunately, the process requires creating an AMA account. You can also call the AMA at (800) 621-8335.

And finally…

Keep up the good work New Hampshire!

Posted by Matt Anderson, MD