POINT-COUNTER POINT: SHOULD STATES REGULATE DATA MINING?

Volume 28, Issue 498 September 4, 2007

Matthew Gever and Christina Kent

The controversy over prescription drug “data mining” just got hotter, with the filing in late August of two lawsuits—one against a law (LD 4) enacted last summer in Maine, and another against a Vermont law (SB 115), also enacted last summer.

The two laws would restrict the way that prescription drug records can be used. But three of the companies that collect and sell this information—IMS Health, Verispan LLC and Source Healthcare Analytics—filed suit against the laws in U.S. District Court, contending that, among other things, they violate the First Amendment by restricting the freedom of speech.

In recent years, a lucrative “data mining” market has sprung up. Pharmacies acquire prescription data in the ordinary course of business, and data mining companies purchase the prescription data, remove patient-identifying information, add additional information, and sell the combined data (which includes the names of individual physicians) to pharmaceutical companies. The drug companies then use this information to track drug usage and improve sales.

State lawmakers have begun to consider whether to regulate this new market of data mining. The challenged Vermont law would set up an “opt in” program, allowing prescribers to choose (through licensing boards) to have their prescribing information used for the marketing or promotion of prescription drugs. The Maine law would establish an “opt out” program, prohibiting pharmacies and other organizations from selling prescription drug information that identifies a prescriber who has applied to licensing boards for confidentiality protection.

Both laws were based on legislation (HB 1346) enacted in New Hampshire in 2006 (see SHN, July 23, 2007). Similar bills have been introduced in at least 11 other states.

POINT-COUNTERPOINT

At NCSL’s recent Annual Meeting in Boston, Massachusetts, two experts squared off over the issue. Below is a transcript of their presentations, edited for space and content.

New Hampshire Representative Cindy Rosenwald, sponsor of HB 1346:

Let’s look at the first reason states are examining the option of regulating prescription data mining, and that is cost. In New Hampshire, right now, every man woman and child accounts for $868 a year of prescription drugs. This works out to more than a billion dollars per year in New Hampshire.

It is really a crippling burden for us. I believe that one of the major contributors to this growth is marketing. There are literally billions of dollars spent every year marketing prescription drugs to doctors and patients.

Now let’s look at the issue of quality in health care. For data mining, drug companies have access to a private relationship that’s not authorized by either the patient or the doctor. [Data mining] is like having a drug rep listening in the exam room door. They have the ability to influence a private relationship between a doctor and a patient without the consent of either.

Patient identity and the use of that for marketing has received much less focus in this entire debate. We did hear testimony in New Hampshire that not all marketers actually do remove the patient’s name before they release those records to the outside vendor to begin the process of data mining the prescriber’s identity and prescribing record. . . .Our pharmacy board used to hear complaints from patients (about receiving ads in the mail for drugs to treat the medical conditions they have), and after our prescription confidentiality law went into effect, the complaints stopped.

Our statute prohibited the commercial use of patient/prescriber identity. It allows collection of identity, but specifies non-commercial use—law enforcement, research, reimbursement, etc. And it allows the commercial use of aggregate data. There is precedent for doing it this way, both in the European Union and several provinces in Canada.

We were sued by two of the large health information technology organizations, and the state lost the first round in federal court. But 13 other states have been looking at this as well, and our neighbors in Vermont enacted an opt-in program so that data could be sold with the permission of that prescriber and Maine enacted an opt-out program.

Let me talk a bit about the very vigorous debate we had in New Hampshire over this issue. We heard from the industry that physician tracking actually saves money in health care by making marketing more efficient. For example, drug reps know exactly which doctors should get which drug samples.

But when we looked into it, we found that it actually contributed to spending growth. First of all, this marketing and branding of drugs is not for the cheaper generics. And with the $2 billion this industry generates, that cost is passed onto consumers and the payers.

Further, there is evidence that shows that market driven prescribing may not improve care. In fact, studies show heavily marketed hypertension drugs are over-prescribed for their effectiveness compared to generics. And several studies have shown that gifts create a sense of obligation and change prescribing habits.

This issue is not going to go away. It’s going to continue to be important for legislators. More data will be electronically available, and this data will continue to have enormous value. Marketers are already promoting electronic health records as better ways of tracking data across modes of care. As always in considering public policy, we have to weigh costs, quality and equality of health care for our constituents.

Donna Boswell, partner at Hogan & Hartson, Washington, D.C.:

Yes, drugs are marketed. It’s a commercial market. They are marketed every day. But that does not result in a prescription. It does result in information, but it does not result in a prescription. If a physician prescribes a drug based on just what a sales rep told him to do—Holy Cow!—the partners in my law firm are going to get such good business you just won’t know it.

The consulting relationship between the doctor and patient is sacrosanct. It has to be protected. But doctors are busy people. They need information about sophisticated drugs. Other talk about the research base for brand new name drugs, the ones that you like to look at because their prices are not as competitive with generic drugs. I’m very proud to work with companies that spend great amounts of time and effort and energy trying to make our medicines better. And I’ve got to tell you—the more research that gets done on medicines, the more we learn about the things we put in our bodies, everything we do to our bodies, involves changing the risk profile, changing the risk. There is a risk to everything we take. And today’s sophisticated medicines, a part of the job is to make a more effective medicine.

And I’m going to submit that the top privacy goal is how to ensure we have safe and effective medicine. The legislation has as its stated goal cost, prescriber price and quality. The quality issue is what I’m going to focus on. I think cost is the same old issue we heard about in other kinds of legislation.

And in fact, when states are doing a really dynamite job of regulating the costs of their health care programs, they are violating private prescriber privacy right, left and center in order to get them to adhere to patterns of care to chronic care management to the kinds of things they want to be prescribed in their Medicaid programs. So I think the ruse about prescriber prices is not really what we’re talking about here and certainly the court didn’t credit that, since the information about who is prescribing what is not something that is protected from other situations.

My concern is that this legislation is actually very much more another in the variation of cost containment legislation. Yes, state legislatures have an obligation to deal with their budgets and to manage their costs effectively. But in this particular one, I don’t think that restricting the prescriber data is actually going to achieve the cost containment. Because unless you show that inappropriate prescribing is resulting from all of these kinds of activities, all you have done is create one more hurdle where the safety studies can’t be done or the safety plans can’t be implemented. Our problem is that if in states where the data are prohibited we still want to market the drug, companies are going to have to get the correct data one way or another. There will be instances where this is not an option.

And the Food and Drug Administration has said that you have to sample and survey and figure out who is prescribing your drug and make sure they’re prescribing only as indicated by the label—we’ve got to do it, or our drug cannot be marketed. It’s a very interesting, important and serious public policy initiative. And like the data mining that people are talking about, I suspect it is one that will become ever more critical both to the quality and the safety of our health-care system.