It is curious that most physicians believe that their practices are not influenced by direct-to-physician marketing, yet the pharmaceutical industry continues to spend enormous sums of money – billions of dollars in the US annually – on this activity (1). Clearly, the pharmaceutical companies are seeing a return on their investment in physician marketing (2), though it is in their best interest, and that of physicians, to downplay its impact.
Nearly all universities have established detailed procedures to oversee conflicts of interest. These often include a ceiling on time spent on outside activities; prohibitions against receipt of both personal income and research grants from any individual company; annual declarations of all outside activities; restriction against serving in a managerial position of a for-profit corporation; requirement that stock and income received be deposited in a university account (a portion or all of which may be returned to the faculty member); and establishment of conflict of interest (COI) committees for peer review of substantial faculty-industry relationships. A requirement for full financial disclosure may soon become federal law, as a “sunshine” bill is currently pending in Congress that will require public reporting of all gifts and other payments to physicians (3).
We have spoken in these pages about other lures of industry that may escape oversight, occurring entirely beneath the radar of the COI rules and regulations (4). These ties are mostly non-financial and thus are more difficult to monitor; however, they can greatly benefit the careers of clinician scientists. They include such things as access to emerging technologies, availability of large datasets, leadership in clinical trials, and assistance with data analyses.
Another area that has largely escaped the scrutiny of COI review is whether the prescribing habits of physicians might be influenced by financial or research relationships with pharmaceutical companies. This information is surprisingly inaccessible to the academic community. It is, however, an essential tool of the trade in daily use by each of the more than 100,000 working drug and device representatives in the US.
Prescribing data of individual physicians are organized and sold commercially by several data warehouses, the largest of which is IMS Health, a company with $2.2 billion in revenues last year headquartered in Norwalk, Connecticut. The AMA sells to IMS and other firms physician identifier information contained in its Physician Masterfile (not only AMA members are in the file!). IMS can then link the physician identifier data with prescription information obtained from pharmacies and hospitals, revealing the prescribing history of the individual. These data are then sold to pharmaceutical and biotech companies, as well as to other interested parties, including (according to the IMS website) health-care policy experts. Although there are many important potential uses of this information for the health care field, the extremely high cost required to access the data (estimated at $100,000 US for the single inquiry reported below) limits the user pool substantially.
Though difficult to access outside the industry, the data are widely available to the drug representatives who wait outside the offices of physicians, and often contain other personal information about the clinician (5,6). Providers may be characterized as “high share” users, or “uncommitted splitters”, along with their national rank by volume of all prescriptions written for the drug of interest and for the class of drugs. Other less flattering classifications are reportedly used off-line to classify providers, with terms such as “loyalists”; “spreaders” (prescribers of many agents in the class}; “sample-grabbers” (enough said); “hidden gems” (susceptible to marketing); and “drug-whores” (thought-leaders who switch loyalties) (1,6).
The Prescribing Data Restriction Program (PDRP) was enacted in 2006, permitting physicians to restrict access of their personal data in the AMA Masterfile (7). To date, approximately 2% of US physicians have elected to deny access under the PDRP. (If you would like to join their ranks, go to: http://www.ama-assn.org/ama/pub/category/12054.html ) However, the provisions of the PDRP deny access of prescriber data only to drug reps and their immediate supervisors but not to others at the companies that purchase these data (8). Sales of the Physician Masterfile are a major source of income for the AMA, and it is perhaps not surprising that the AMA appears to have functioned as an enabler to the pharmaceutical industry in this area. This situation is likely not to change in the foreseeable future. A number of states have proposed a ban on sale of prescription data, and in fact such a bill was passed in New Hampshire in 2006, but a federal court subsequently ruled the ban unconstitutional, and the appeals process is likely to be a protracted one.
A number of studies have shown that marketing efforts to physicians are quite successful, though most physicians continue to deny that their prescribing habits are adversely influenced by gifts, “informational” sessions, or the like. Given the often extensive and multifaceted research activities of physician scientists, the potential for COI that could influence clinical behaviors may be higher for those doing research than for full-time practitioners. Also, because these same researchers often lecture on treatment options, the impact of changing the behavior of a clinician scientist may go well beyond his or her own practice.
In a very small convenience dataset, we sought to examine possible associations between the prescribing habits of academic physicians and outside income or research support from the manufacturer. We chose a sample of 9 university-based senior academic physicians who are multiple sclerosis (MS) experts, and evaluated prescribing patterns for the four first-line drugs for relapsing MS, all of which are roughly equivalent in efficacy and price (Figure). On the left side of the figure, the percentage of prescriptions (Y axis) by provider (X axis) is shown for Avonex, Betaseron, Copaxone and Rebif, along with the overall US market share. Physicians receiving personal income from the manufacturer are highlighted in yellow. On the right, similar data are shown with providers highlighted in yellow based upon research support rather than personal income.
Figure. On the left side of the figure, the percentage of prescriptions (Y axis) by provider (X axis) is shown for Avonex (A), Betaseron (B), Copaxone (C) and Rebif (D), along with the overall US market share (gray bars). Physicians receiving personal income from the manufacturer are highlighted in yellow. On the right, similar data are shown with providers highlighted in yellow based upon research support rather than personal income.
Overall there is a very wide variance in prescribing habits, even within this highly interactive and collaborative group located in a single medical center. There is also a clear suggestion that a correlation exists between prescribing a product and receiving research support from the manufacturer. Faculty code of conduct guidelines at this institution prohibit personal income received from companies that also provide research funding. Thus, for senior academic physicians, it appears that more lucrative grant relationships may be more strongly associated with prescribing practice than generally modest speaking or consultation fees from companies.
The lessons here are several. First, the possibility that industry-sponsored research might bias prescribing practices requires deeper exploration. although , it is quite possible that physicians who strongly believe in the benefits of a particular drug are more likely to prescribe it and also to study it, further studies should also evaluate research funding as a motivator of prescribing practice. Second, easier access to the prescribing practices of physicians is clearly a desirable goal, and at least in theory, this is one of many advantages of a national electronic medical record. Recent reports in the lay media have highlighted the serious consequences of financial biases that may have led to overuse or advocacy for particular drugs or devices. Such examples seriously compromise the position of absolute trust that the biomedical community must maintain in order to be successful. We have and will continue to argue for creation of deeper research partnerships between academia and industry (9), but as we do so the full impact of these relationships on our clinical behavior should not escape scrutiny. None of us is free from the potential influence of the highly trained, well honed, and incredibly well funded influence of the pharmaceutical and device industry, and as scrutiny increases on obvious financial ties and public presentations, influence will be turned elsewhere.
Stephen L. Hauser MD and S. Claiborne Johnston MD, PhD
Editors
References:
- Fugh-Berman A, Ahari S: Following the script: how drug reps make friends and influence doctors. PLoS Medicine 4:0621, 2007
- Brennan TA, Rothman DJ, Blank L et al: Health industry practices that create conflicts of interest: a policy proposal for academic medical centers. JAMA 295:429, 2006
- Steinbrook R: Disclosure of industry payments to physicians. N Engl J Med 359:559, 2008
- Hauser SL, Johnston SC: Of ghosts and sirens: the subtlest lures of industry. Ann Neurol 61:A11, 2007
- Greene JA: Pharamaceutical marketing research and the prescribing physician. Ann Intern Med 146:742, 2007.
- Elliott C: The drug pushers. Atlantic Monthly, April 2006.
- Fugh-Berman A: Prescription tracking and public health. J Gen Intern Med 23:1277, 2008.
- Steinbrook R: For sale: physicians’ prescribing data. N Engl J Med 354:2745, 2006.
- Johnston SC, Hauser SL: A status report on neuroscience research, without grade inflation. Ann Neurol 60:A9, 2006.
Filed under: Message From the Editor | Tagged: conflict of interest, pharmaceutical marketing, direct-to-physician marketing, financial disclosure, sunshine bill, prescribing data, Prescribing Data Restriction Program, AMA physician masterfile, research grants, Avonex, Betaseron, Copaxone, Rebif, industry-sponsored research, bias
