Support of that magnitude inevitably brings conflicts of interest, and both sides have supported systematic efforts at self regulation during interactions between physicians and pharmaceutical company representatives. The American Medical Association (AMA) in 1991 issued voluntary ethical guidelines on gifts to physicians by drug reps and has augmented them with clarifying addenda, and the Pharmaceutical Research and Manufacturers of America (PhRMA) in 2002 issued a voluntary industry code regarding marketing interactions with physicians.

While research on the subject of conflicts of interest is fairly extensive in the medical literature, a subset of that literature has begun to question the efficacy of self-regulation, culminating in an article published this January in the Journal of the American Medical Association – "Health Industry Practices that Create Conflicts of Interest" – which upped the ante by declaring the failure of voluntary efforts to prevent ethical transgressions in the relationships between physicians and the drug and medical device industries, which it said continue to erode medical professionalism as well as the integrity of the industry. By offering physicians even subtle inducements – small gifts, modest meals and drug samples – the pharmaceutical industry is still successfully getting physicians to prescribe the newest and often most expensive drugs when cheaper therapeutic equivalents are available, and to petition their inclusion on hospital formularies, according to the article, which included as co-authors representatives of the Association of American Medical Colleges (AAMC) and the American Board of Internal Medicine (ABIM).

This recent critique is premised on a surprising research finding: that permitting only small gifts, requiring full disclosures and making earnest efforts to avoid bias – hallmarks of the voluntary guideline approach – do not prevent nonrational prescribing behavior by physicians after exposure to promotional presentations by drug reps. The JAMA article argued that the urge to reciprocate, even when given a gift of nominal value, operates at an unconscious level which physicians cannot effectively prevent, and recommended more external restrictions on physician contact with drug reps – specifically at academic medical centers because of their leadership role in the medical profession, their ability to organize and implement new policies rapidly, and the importance of habit-formation during physician training.

Some academic medical centers and health systems, citing concerns about drug rep marketing behaviors, agree with the premises of the JAMA article and have begun to implement further institutional restrictions on drug rep interactions with physicians, in some cases banning all gifts and drug samples.

Some believe that an even tougher approach than institutional intervention is needed to curtail inappropriate influence by industry of physician prescribing, as seen in the recent passage by the Massachusetts Senate of a bill that would require drug reps to be licensed by the state and undergo training, as well as prohibit them from offering any gifts to physicians.

A related industry practice has also recently come under scrutiny. Before a drug rep meets with an individual physician, he or she has most likely had the opportunity to review that physician’s personal prescribing behavior – a practice that has resulted in a new AMA policy that seeks to curtail the sharing of that data with drug reps and their sales managers, at the request of any physician.

Critics of the AMA policy say that it requires physicians to trust pharmaceutical companies not to share with its sales force prescribing data of physicians who have "opted out," that its compliance monitoring mechanism – physicians who detect breaches and file complaints – will be ineffective, and that many physicians may not take the time to opt out.

Among those critics is the state of New Hampshire, which recently passed a law prohibiting the sharing of any health provider prescribing information for commercial purposes with pharmaceutical companies – an approach which the AMA and the drug industry both believe is too rigid and will stymie legitimate research and educational uses of the data.

"No Gifts" Movement

The "Health Industry Practices" article in JAMA was published as part of a project, co-sponsored by the ABIM Foundation, to identify behaviors that physicians should enact to honor their commitment to the foundation’s Physician Charter on Medical Professionalism – which maintains that physicians have an obligation to recognize and manage conflicts of interest that arise in the course of their professional activities so as to not compromise the trust of the public, according to Christine Cassel, M.D., president of the ABIM and its foundation.

The foundation was concerned that the pharmaceutical industry’s primary goal is a commercial one, and that its commitment is not primarily to patients or the medical profession, but to shareholders – which is potentially in conflict with a stance that is evidence-based, least-cost, and least likely to have adverse side effects, says Cassel.

The AMA maintains that the locus of ethical responsibility is the individual physician, and its "Gifts to Physicians from Industry" guidelines lay out standards of conduct for the medical profession, which individual physicians must apply to minimize conflicts of interest that may be at odds with the best interest of patients, according to Edward L. Langston, R.Ph., M.D., Chair-elect of the AMA’s Board of Trustees. Those guidelines stipulate that any gifts accepted by physicians individually should primarily entail a benefit to patients, should be related to the physician’s work (e.g., pens and notepads are acceptable), should not be given with strings attached (i.e., in exchange for a commitment to prescribe products), and should not be of substantial value (i.e., restricted to the general range of $100).

PhRMA’s Ethical Relationships with Healthcare Professionals code largely mirrors the AMA guidelines.

"Physicians can certainly make a decision whether to see a pharmaceutical rep, and recognize that they are detailing," says Langston. "They also need peer-reviewed information and FDA labeling information. There are strong guidelines regarding what is a reasonable relationship with a representative – for individuals and for organizations," Langston maintains.

The belief that small gifts from industry don’t matter is simply wrong, as solid social science research has demonstrated that gifts of any size elicit a strong impulse in people – including physicians – to reciprocate, and anything that poses a risk to physician objectivity to use the best available evidence for the benefit of patients is just cause for concern by the profession, according to Jordan J. Cohen, M.D., Immediate Past-President of the AAMC, and a co-author of the "Health Industry Practices" article in JAMA.

Cohen doesn’t blame the pharmaceutical industry for doing its job very effectively, and says the fault lies with the medical profession for mistakenly deferring to Food and Drug Administration and Office of Inspector General directives, and to voluntary AMA and PhRMA guidelines, which he believes are inadequate to constrain the promotional practices and impacts of a sophisticated industry.

There is a fairly extensive literature on pharmaceutical company influence over physician prescribing, with most studies showing that contact with drug reps does influence prescribing, although it is not known how much, according to John G. Scott, M.D., Ph.D., Assistant Professor of Family Medicine, UMDNJ-Robert Wood Johnson Medical School, who is working on a qualitative study to describe how drug reps interact with physicians. Studies in the early 1990s estimated that pharmaceutical companies spend up to $13,000 per year in direct promotion and marketing to each physician, while Scott says that the figure is closer to $84,000 if it includes money the industry spends on continuing medical education, which he regards as a form of marketing.

Research notes that, when physicians are surveyed as to whether they have been influenced by pharmaceutical company marketing, they generally say "No," while they generally answer "Yes" when asked if they believe their colleagues have been influenced, says Scott. Drug samples in physician offices also play a role in steering prescribing behavior. One researcher noted that, of the 15 most frequently prescribed medications for Medicare patients, 10 of them have less expensive, generic therapeutic equivalents, says Scott. Researchers maintain that these findings suggest that physicians are being manipulated into prescribing behavior that is not evidence-based without realizing it, and without even believing it when questioned about it.

The issue is particularly germane to primary care physician practices in the northeast, which may not be doing well financially because of practice environment pressures, and which would be particularly receptive to services from drug reps such as lunches for their staff, CME cost defrayal, and drug samples for their patients, says Scott.

Research cited in the "Health Industry Practices" JAMA article lists evidence that even modest gifts influence physician behavior: receiving gifts is associated with positive physician attitudes toward drug reps; physicians who request additions to hospital drug formularies are far more likely to have accepted free meals or travel funds from drug manufacturers; and the rate of drug prescriptions by physicians increases substantially after they see sales representatives, attend company-supported symposia, or accept samples.

The authors wrote that academic medical centers must take steps to reverse the current trend, including:

· Ban all gifts (zero dollar limit) and free meals from drug and medical device makers.

· Prohibit direct support of CME programs (including faculty and trainee travel) by industry and replace it with a central fund for pooling contributions from multiple sources.

· Prohibit direct provision of drug samples to physicians and replace the practice with a voucher system.

· Limit membership on formulary committees to individuals without financial relationships with any drug or device maker.

· Prohibit faculty from serving as members of industry speaker’s bureaus, which are components of a company’s marketing system.

· Prohibit faculty from publishing anything that is ghostwritten by industry employees.

The authors urged medical schools to build consensus around these principles and to be prepared to monitor compliance and enforce them, in order to reap several benefits: more evidence-based prescribing behavior and better patient outcomes, increased use of generics and reduced total expenditures on prescription drugs, increased sensitivity among medical students and housestaff to the values of medical professionalism and scientific integrity, reduced need for external regulation to safeguard against market-driven conflicts of interest, and public reaffirmation by the medical profession to put the interests of patients first.

Hospital Response

Some hospitals and health systems have issued new policies further constraining the relationship between drug reps and physicians, some are examining whether or how to do so, and others are skeptical about the need for further reform.

The JAMA "Health Industry Practices" article created a significant amount of discussion at Cooper University Hospital about how to preserve legitimate aspects of that relationship which benefit patients, and how to implement further restrictions to curtail conflicts of interest, according to Carolyn Bekes, M.D., Senior Vice President of Academic Affairs. The hospital follows guidelines on unrestricted grants from industry promulgated by the Accreditation Council for Continuing Medical Education, says Bekes – for example, a commercial supporter cannot direct the content of CME. Without industry’s financial support, she acknowledges, the university would not be making advances in medical research, and she notes that research grants are reviewed by university oversight committees, with money going into a grant account that is governed by strict federal and international standards of propriety and transparency. Most academic medical centers abide by ACCME guidelines and follow those protocols.

While education and research funding are important forms of industry support that directly benefit patient care, other areas are more blurry, says Bekes. Without free drug samples to physician offices, she says, low-income patients might not have access to them and very sick patients would go without starter packs – although their availability encourages the use of certain medications and has the potential to steer physicians to prescribe them. To address the issue, Cooper encourages physicians to use drug company vouchers and drug company 1-800 telephone numbers for low-income patients to obtain free or discounted medications, says Bekes.

Other interactions require more judgment by physicians, says Bekes. During speaker’s bureau activities when medical residents are present, Cooper requires full disclosure that a lecture is being sponsored by a particular drug company. A drug rep may not be present in the room unless one of Cooper’s faculty members is also present in a supervisory capacity to ensure full disclosure of salient details of a presentation – a voluntary guideline that Bekes says Cooper issued within the last few months. During a presentation to residents, for example, if a drug rep declares that a drug’s "complication rate is low," the faculty mentor would ask that the rate be revealed, might note that cheaper alternatives are available, and would be available for questions from the residents in the audience, adds Bekes. The new voluntary rule promotes full disclosure, while the pharmaceutical companies probably appreciate that Cooper has not just made a blanket rule banning such presentations, Bekes says.

The policy does not apply to non-accredited interactions, such as physician detailing, but Bekes notes that drug reps may not sponsor purely social events and must declare their presence in the hospital.

Another new guideline governs drug companies’ brochures, many of which Bekes says have been informative and helpful for patients. Drug reps may no longer place the brochures in physician waiting rooms, but must first submit them for the physician’s approval – a policy which Bekes says impacts 400 employed physicians at Cooper’s 60 practice sites in the South Jersey area, as well as another 300 to 400 physicians who are on Cooper’s medical staff.

"Our guidelines are the beginning of a educational and awareness effort regarding the influence of pharmaceutical company reps – the good things, as well as the potential conflicts," says Bekes.

RWJ Medical School has also tightened its policies in recent months, and Scott says he is no longer permitted to accept drug samples, lunches or gifts of any kind in his primary care academic clinical practice.

The Hospital of the University of Pennsylvania (HUP) found that requiring drug reps and other industry vendors to register and make appointments to meet with staff – a policy it implemented several years ago – was difficult to monitor and did not stop "cold calls" without appointments, according to P.J. Brennan, M.D., Chief Medical Officer and Senior Vice President, University of Pennsylvania Health System. PhRMA’s voluntary code on relationships with health care professionals, which still permits "modest" meals and gifts, also fueled a growing awareness by Penn of the large number of drug reps who come, and the influence they could bring, says Brennan. A group from Penn consulted with the authors of the JAMA "Health Industry Practices" article a year before its publication to discuss the nature and types of industry influence, including the principle that gifts of any size create the impulse to reciprocate. Penn ultimately decided that it needed to review and revise its policy, "not to demonize pharmaceutical companies, but to bring their interactions with physicians closer in line with evidence-based content, rather than marketing," says Brennan.

After holding public symposia and inviting medical staff input on the subject, Penn tightened its constraints on drug rep activity and on the behavior of its physicians at HUP, the university’s clinical practices, Penn’s Clinical Care Associates physician practices, and Penn’s Radnor facility. As of July, no gifts of any size – including meals, pens, pads and stethoscopes – may be offered to Penn physicians and, if a drug rep violates the policy or fails to make appointments before visiting physicians, that rep’s company will lose the privilege to visit, says Brennan.

Industry can make gifts to departments to support educational programs, while the money is disbursed at the discretion of department chiefs and chairs, says Brennan – a step that the JAMA article recommended to insulate the programs from industry influence.

Brennan says Penn had prohibited drug samples several years ago because of patient safety concerns – samples are occasionally dispensed without adequately checking patients for allergies and interactions with other medications – and implemented a voucher program as an incremental way of reducing the degree of influence that samples have on prescribing behavior. Companies now provide vouchers for free drugs, which physicians dispense to patients in the same way they would for any other prescription, along with complete documentation, says Brennan. Violating the "no samples" policy now will also forfeit the privilege of the rep’s drug company from coming on campus, and while Brennan says that Penn has not yet had to face the question of how long the forfeiture would last, he notes that some companies have well over a dozen reps who come to the campus, and a loss of privileges would have considerable impact on the company.

Physician response to the new policies has been mixed: some agree with all of them, some physicians disagree with some aspects, and some disagree with all of the reforms, says Brennan, noting that absence of the free meals, in particular, has provoked objections. "The nature of marketing and the influence it brings is very subtle, and everybody believes they are not influenced. PhRMA believes what we’re doing is an overreaction, saying that ‘It’s just pads and pens,’" Brennan says.

Penn also stopped reporting physician prescribing data to wholesale suppliers several years ago, out of concern that the information would be obtained by drug companies and would be used inappropriately for marketing purposes, according to Brennan. To stop reporting such data is probably uncommon among hospitals, Brennan believes, given that the move triggered several visits from industry reps trying to dissuade Penn from cutting off prescribing data, for fear that it would become a more widespread practice among hospitals.

Penn hopes its policies on drug rep activity and physician behavior will set a rigorous standard, while monitoring them requires an element of trust in its leadership and staff, says Brennan. The new policies will also be considered by the governing bodies of Pennsylvania Hospital and Presbyterian Medical Center in the fall, he adds.

Not all medical centers have embraced further restrictions on drug rep-physician interactions, relying instead on institutional conflict-of-interest policies and physician judgment. Drexel University College of Medicine, for example, has adopted the AMAethical guidelines on gifts to physicians, although in the last few years it has required drug reps to make appointments – to make sure they are not walking in areas where they could interrupt patient care activities, according to Richard V. Homan, M.D., Senior Vice President for Health Affairs, and Annenberg Dean. "While trying to reduce the risk of bias a laudable goal," says Homan, "you can’t ban everything. Physicians have to be the fiduciaries of their patients."

West Penn Allegheny Health System similarly relies on its institutional conflict of interest policy, and limits industry gifts to physicians to a $50 value, and a $300 per year aggregate value, while meals during informational presentations are limited to $150 per person per event, and the health system allows drug samples, according to Robert Michalski, West Penn Allegheny’s compliance officer. West Penn Allegheny also requires drug reps to make appointments so they do not interfere with patient care, Michalski says. As for modest meals and gifts, he says, "We don’t have any angst about those trinkets, unless it amounts to something more formal."

Neither institutional policies to restrict the drug rep-physician relationship, nor voluntary professional ethical guidelines are adequate to guard against inappropriate influence during drug rep-physician encounters, according to Massachusetts State Sen. Mark Montigny (D-New Bedford), who says "the line ought to be drawn more brightly, broadly and deeply," and "everybody should be accountable to the same rules." Montigny is pushing a bill to license drug reps and ban all gifts, meals, entertainment and travel to physicians, health care facilities and public officials. "Given the complexity and speediness of medical developments, the last people on earth who ought to be educating physicians is the pharmaceutical industry, from whom billions of dollars have been recovered for fraudulent business practices," says Montigny, whose bill would monitor drug reps by requiring each of them to pay $500 to be licensed in the state, and to undergo training as determined by the state department of health and pharmacy board. The bill passed the Massachusetts Senate, but died in conference committee, and Montigny says he plans to refile it in January.

Restricting Prescribing Data Sharing

While institutional restrictions on drug rep activities is being debated in the medical community, action on a related front would restrict the information that drug reps have about physicians before meeting with them: data revealing exactly which drugs, and in what quantity, a physician has prescribed. A new AMA policy seeks to prevent drug reps from obtaining that data, for physicians who "opt out" of an AMA data sharing agreement with drug companies.

The AMA licenses its Physician Masterfile data – which it says is the most reliable and comprehensive collection available of physician demographics, specialties, licensure status and office locations – to data mining companies known as Health Information Organizations (HIOs), which obtain physician-specific prescribing data from various sources – including clinics, pharmacies and hospitals – and combine it with the Masterfile data to make it marketable for educational, research and commercial purposes, including pharmaceutical company use.

While data verification and authentication are the primary purposes for the Masterfile, the AMA says it has also provided a revenue source to support the association’s advocacy and services. In 2005, the AMA received $44.5 million in revenue from the sale of database products, representing 16 percent of its total revenue, according to the AMA’s 2005 annual report.

Since the AMA controls the Masterfile license agreements with the HIOs, it has contractual leverage over the use of the Masterfile data. After receiving requests a couple of years ago by several state and national medical societies – led by the American College of Physicians – to prohibit the release or sale of physician prescribing information, the AMA surveyed its membership on the matter and determined that that there was sufficient concern to offer physicians the option to prevent disclosure of their personal prescribing patterns to drug reps. This July, the AMA launched such a firewall in the form of its Prescribing Data Restriction Program, a registry intended to shield physician-specific prescribing data from front line drug reps and their sales managers.

According to Langston, pharmaceutical companies need Masterfile data as part of their purchase from HIOs to learn which physicians practice in what area, to target specific drugs to specific specialties, and to communicate drug recall and FDA labeling alerts to physicians. The problem, says Langston, has been when drug sales reps have leveraged the data by using heavy-handed sales merchandising tactics when visiting some physicians in their offices.

Physicians who want their prescribing profiles withheld from drug company employees who have direct contact with physicians – drug reps and their first-line sales managers – can ask the AMA to put their name on a confidentiality list to be included in the AMA’s license agreements with HIOs who purchase Masterfile data. HIOs need Masterfile data to make their database products useful to purchasers – such as drug companies – and the AMA, as the Masterfile licensee, can set conditions under which the combined database may be sold, Langston explains.

Drug companies would still have access to the prescribing data of physicians who have "opted out" and could still use it for marketing, compensation and research, but would be required to police their own sales forces to keep the information from front line drug reps and their managers. If a drug company is found violating the agreement – as reported by physicians through an AMA complaint line – the AMA can modify its Masterfile license agreement with HIOs and restrict to whom they can sell the information, and the drug company will lose its Masterfile data privileges, says Langston.

By giving physicians the "opt-out" choice of the registry, Langston argues, the AMA preserves its leverage over how physician prescribing data will be used, by forcing the drug industry to be sensitive to physicians’ privacy choices in order to retain convenient access to prescribing information augmented by crucial Masterfile data.

The AMA has for years also offered a much stricter, all-or-nothing "Do Not Release" policy option, which would prohibit the release of a physician’s Masterfile information to all entities and their affiliates except in national emergencies, but it also would curtail the ability of pharmaceutical companies to tell physicians about health hazard warnings or drug recalls, as well as the ability of hospitals to verify credentials unless the physician specifically authorizes it. A less stringent "No Contact"option would allow physicians to receive the warnings and hospitals to verify credentials, while curtailing the use of Masterfile data for telephone calls, faxes and promotional mailing, including professional journals from publishers and offers to participate in research studies.

Langston warns physicians to think carefully before considering those options, and he maintains that there are important and appropriate uses of physician prescribing data in medical research involving patterns of prescribing. Without specific prescribing data, pharmaceutical companies would have to rely on a physician’s specialty and geographic location to market products and deliver drug samples, which the AMA says will result in irrelevant sales calls and product samples.

The AMA also hopes to demonstrate that HIO products can be of value to physicians, and is engaged in a pilot project with the California Medical Association that Langston says will reveal important details about a physician’s own practice – including data on patient compliance and persistence with prescribed medication, patient use of generics, and peer benchmarking of physician prescribing practices.

Prescribing data can also be put to good use to improve patient safety, and there are several programs administered by the FDA that depend on the ability of pharmaceutical companies to contact individual physicians, according to PhRMA’s Senior Vice President Ken Johnson. "For example, FDA relies on pharmaceutical companies to give immediate notice to physicians of new safety information. In the event of a drug shortage, an instance of suspected drug counterfeiting or tampering, or a drug recall, the federal system depends on the ability of the manufacturer of the product to communicate directly and promptly with prescribing physicians. Appropriate access to prescribing data enables companies to quickly identify and reach out to physicians who prescribe the specific products at issue," says Johnson.

Both the AMA and the drug industry hope the new opt-out registry deflects states away from pursuing legislation to keep all prescribing data away from salespeople – a move that Langston said would force the industry to obtain the data from other sources, some of which may not be reliable, making things even more uncomfortable for physicians. According to an article in the May issue of Pharmaceutical Executive, co-authored by a representative of the AMA and an official of a data-mining company, if drug companies abide by the rules of the AMA registry, "legislators will turn their attention elsewhere, and the industry can hang onto one of its most valuable data sources."

This June, New Hampshire became the first state to enact such a ban, declaring that "records relative to prescription information containing identifiable patient and prescriber data shall not be transferred or sold by any pharmacy benefits manager, insurance company, electronic transmission intermediary, retail, mail order, or Internet pharmacy for any commercial purpose," including advertising, marketing, promotion or any activity that could be used to influence sales or market share of a pharmaceutical product, influence or evaluate the prescribing behavior of an individual health care professional, or evaluate the effectiveness of a professional pharmaceutical detailing sales force, according to the bill’s sponsor, New Hampshire State Rep. Cindy Rosenwald (D- Nashua). The law will be monitored through complaints to the state’s attorney general’s office and pharmacy board, and enforced through the state’s unfair trade practices act, she adds.

Rosenwald says her bill is primarily intended to protect patient privacy and put a check on the rising costs of prescription drugs. While drug reps can be helpful to physicians, she says they are primarily a sales force promoting a very expensive product. She maintains that the AMA opt-out registry is inadequate because, although it requires drug companies to shield prescribing data of registry physicians from drug reps and their first-line managers, it does not shield the information from high-level company managers, who may still use it to make compensation decisions for drug reps, and it requires companies to be on their honor that they will not share restricted data with the field reps – which Rosenwald calls a "wink-wink arrangement." Rosenwald’s bill also covers non-physicians, including nurse practitioners, physician assistants and dentists.

The bill was supported by the state attorney general’s office, the state health department, AARP and the New Hampshire Medical Society, which Rosenwald says represents three-quarters of the state’s physicians. "It’s up to the pharmaceutical industry to support the AMA registry, and we must trust them," according to New Hampshire Medical Society President Marc Sadowsky, M.D., who says he can’t see how a drug company supervisor can reward successful sales reps without the drug reps knowing which drugs are being prescribed by physicians more heavily. Sadowsky also believes the AMA’s opt-out approach won’t be effective because many physicians may not take the initiative to do so. The Society agreed that Rosenwald’s bill was in the best interest of the citizens of New Hampshire, and was the best mechanism to "modulate the marketing leverage that drug companies have over physicians" to induce them "to prescribe the high-cost, trade name medicine."

Two Philadelphia-area data-mining companies, IMS Health Inc. and Verispan LLC, have filed a complaint in U.S. District Court in New Hampshire asking that the law be declared unconstitutional and maintaining that it violates free speech rights, endangers public health and impedes research.

While PhRMA has not taken a position on the bill, "We continue to be concerned that restricting the use of physician prescribing data could potentially damage continuing medical research and stymie efforts to communicate clearly and effectively with physicians regarding the appropriate use of medicines," says Johnson. "Eliminating pharmaceutical research companies’ access to this data would severely hamper companies’ ability to provide current information about drugs to physicians, including important information about new indications and side effects. It also would complicate efforts to conduct evidence-based medical research, stop drug diversion and provide drug samples to physicians whose patients are most likely to need them," he notes. Patients’ personal health information is already protected under federal law, and prescribing data provided to pharmaceutical companies and others does not contain individual patient information, Johnson adds.