Managing conflict of interest in clinical practice

TY - JOUR

T1 - Managing conflict of interest in clinical practice

AU - Camilleri, Michael

AU - Cortese, Denis A.

N1 - Funding Information: Analysis of clinical situations that present potential conflict of interest is often complex and needs to be addressed using multiple approaches. Three fictional scenarios illustrate specific challenges in identifying and managing potential conflicts of interest. Scenario 1: Selection of Medication From a Menu of Drugs With Similar Efficacy. In clinical practice, 2 or more medications or classes of medication often have similar potential for benefit. What determines the clinician's choice of a diuretic for heart failure, antiplatelet adhesiveness medication for angina, laxative for constipation, nonsteroidal analgesic for degenerative joint disease, or antiinflammatory medication for inflammatory bowel disease? These choices determine medication spending of a few dollars (eg, for aspirin or chloroquine or prednisone for a rheumatological or inflammatory bowel disease) to a few thousand dollars (eg, for a monoclonal antibody prescribed for these conditions) per month. Physicians may appropriately choose newer and more expensive medication based on either published data or personal observations that indicate that the drug is safer and/or more effective than an older inexpensive regimen. On the other hand, perhaps influenced consciously or subconsciously by personal relationships with industry, physicians may inappropriately prescribe newer, more expensive drugs that are not necessarily more effective for the specific clinical situation. Motivation or Benefit. Evidence shows that prescribing practice is altered by interaction with the pharmaceutical industry, often leading to a clinical decision that is not supported by the available literature. 13–16 The motivation for selecting one medication over another may be the personal financial benefit that results from consulting or participating in education programs funded either directly or through an “independent” education company by the manufacturer of the drug. An alternative motivation may be the receipt of a research grant that results in no personal financial benefit. Finally, the clinician may be influenced consciously or subconsciously by personal relationships with industry or a desire to appear on the cutting edge of medical advances. The Conflict and the Associated Potential for Harm. How can the patient be protected from the selective, perhaps capricious or conflicted, prescribing practices of the individual physician? What can an institution do when the anticipated efficacy of medications for the specific indication in an individual patient is similar, but the choice may expose the patient to a greater risk of adverse effects, eg, greater risk of angina or stroke with a cyclooxygenase 2 inhibitor rather than a nonselective nonsteroidal antiinflammatory drug; greater risk of persistent bleeding in a patient with an ulcer who is taking a platelet IIB/IIIA antagonist rather than aspirin; and a potentially greater risk of infection/lymphoma with maintenance anti-tumor necrosis factor α therapy than with maintenance corticosteroids? Management Options. If the clinician has a personal financial or nonfinancial interest, the potential for conflict of interest needs to be managed. Potential management options to consider are as follows: (1) de minimis (Latin for “about minimal things”) financial interest that triggers oversight of the prescribing practices of the clinician; (2) limit on personal income from industry relationships to ensure that compensation meets market standards and avoids the concern that the payment is for consciously or subconsciously inducing “loyalty” in the prescribing practice; (3) independent review of research contracts and budgets that should be auditable as a deterrent to “padding”; and (4) recusal of the involved clinician from purchasing or formulary decisions, which may ultimately influence not only the individual's practice but also that of the medical center. However, these management strategies come at a potential risk that physicians, who are sensitized by the potential for being perceived as having a conflict of interest, may select drugs that may be inferior to the more expensive, newer drug. Thus, balance in these management approaches is essential to safeguard the best interest of the patient. Scenario 2: Selection From a Menu of Devices With Similar Efficacy Projected for an Individual Patient. There are many instances in procedural and surgical clinical practice in which 2 or more devices or procedures have similar potential to benefit the patient. Indeed, situations may even exist in which the choice of an interventional approach may not provide a significantly better benefit than noninterventional therapy. A classic example is the choice for a patient with serious but stable angina among medical therapy vs one of a variety of cardiac interventions (balloon angioplasty, stent, drug-eluting stent) vs coronary artery bypass grafting. Another example is the choice among a variety of orthopedic prostheses for the patient who has an undisputed need for surgical joint replacement. The choice may be further complicated if the surgeon has contributed intellectual property that led to the academic center licensing the product and receiving a sales-based royalty. Motivation or Benefit. The motivation for choice of one treatment over another should be dictated by the nuances of the individual case, with the best interest of the patient at the focal point of decision making. However, in situations in which there are no specific benefits of one therapeutic approach over another, it is conceivable that one approach may selectively benefit the institution and, through its royalty-sharing policy, an individual (independent of any patient considerations). Thus, a physician with a potential conflict of interest (through personal gain) may be in a position to direct the care and select a specific treatment for an individual patient. The Conflict and the Associated Potential for Harm. The conflict is similar to that observed in Scenario 1 for the clinician who benefits through consulting, royalty income, or research grants from the manufacturer or by participating in educational meetings. The potential for harm exists when the pros and cons in the choice of the device or treatment are subject to influences other than the best interest of the patient. Management Options. In addition to the management options in Scenario 1, federal legislation and regulation on kickbacks in the health care setting (referred to as the Stark law 17,18 ) provide consideration in managing this kind of conflict of interest in clinical practice. Under the Stark law, the individual and institution are excluded from financial benefit associated with use of a device (on which sales-based royalty is received) when the device is implanted or used in their own clinical practice. This includes a prohibition on receiving royalties for the use of the device by the inventor. To make such policy operational, the academic medical center needs adequate tracking systems to identify inventor use of their own inventions to ensure that the sales occurring at the institution are deducted from the annual computation of sales from which the sales-based royalty is calculated. Scenario 3: Development of Practice Guidelines and Their Use by Formulary Committees. Consensus practice guidelines are intended to have significant impact on clinical practice through decision making. Practice guidelines and formulary committees are designed to debate and develop recommendations that are then implemented as limits on the clinical choices of their colleagues. Thus, they can standardize care and introduce evidence-based criteria for selection of treatment. Incumbent on this process is the requirement for clear objectivity in the formulation of the guidelines. Participation in development of guidelines by individuals having conflicts of interest may introduce bias that may negatively impact patients' well-being (eg, exposure to clinical risk and acquisition of financial burden when a treatment is only partly funded by the third-party payer). Motivation or Benefit. The motivations in scenarios 1 and 2 all may apply because participants in the development of the guidelines may receive personal benefit through consulting fees, royalties, or research grants. Guideline development requires interpretation of the literature and the recommendations by one person or a group of clinicians (and other experts) from the same or different academic centers, and their judgment may conceivably be tainted or may be perceived to be tainted by conflict of interest. The Conflict and the Associated Potential for Harm. The conflict, typically represented by personal financial interest, has even greater potential for harm as guidelines are implemented by a spectrum of health care delivery systems through their formularies. What are the responsibilities of the academic medical center in managing such potential conflicts of interest? Substantial scrutiny is necessary to ensure that authors of consensus practice guidelines do not have personal conflicts of interest. 19 Of note, a Dutch group was able to develop consensus for clinical guidelines but could not develop guidelines on management of conflict of interest issues because of a lack of consensus. 20 The potential for good from consensus guidelines is significant but so is the harm if the guideline is biased and if that bias is in a direction detrimental to patient care. Such documents from governmental and national organizations or learned physicians are used extensively by formulary committees or groups entrusted with purchasing decisions. These decisions can meaningfully impact the revenues of the product manufacturer. At the same time, a biased guideline could potentially induce harm by not reflecting the best management strategy or exposing the patient to personal or financial risk. Thus, although institutions may manage individual conflict of interest by requiring employees to recuse themselves from the purchasing process, the institution is potentially vulnerable when it follows consensus guidelines developed by learned groups that did not appropriately manage their conflicts of interest. Management Options. Development of consensus practice guidelines usually occurs outside the purview of any single medical center; hence, policies developed by a single academic medical center may not affect management of conflicts of interest in consensus guidelines development. Nevertheless, we believe that the academic medical center should educate its staff regarding responsibilities when invited to participate in development of guidelines and to ensure objectivity in the institution's formulary practice. In some countries, national bodies or centralized systems such as the National Institute of Clinical Excellence in the United Kingdom are entrusted with guideline development. 21 This independent organization is responsible for providing national guidance on the promotion of good health and the prevention and treatment of ill health, as well as providing appraisals of health technology, clinical guidelines, and interventional procedures. 21 There are no uniformly accepted guidelines for management of conflicts of interest in development of practice guidelines in the United States, although the National Institutes of Health Consensus Development Program sponsors evidence-based assessments of important medical issues. As reported recently by Steinbrook, 22 these assessments include a systematic literature review prepared through the Agency for Healthcare Research and Quality, a public conference with research presentations (which may be provided by speakers who have conflicts of interest as long as those conflicts are disclosed), jurors and witnesses who do not have conflicts of interest, and a consensus statement that is disseminated widely and ultimately reflects the conclusions of the panels. In practice, most guidelines in the United States are not produced under the auspices of this program. The academic medical center may educate its staff to adopt strict criteria and self-recusal from participating in development of practice guidelines. For example, the National Institute of Clinical Excellence regulations exclude individuals from any participation in practice guidelines advisory groups for the following reasons: (1) receipt of consulting fees or conduct of fee-paid work, (2) directly held equity, or (3) receipt of expenses or hospitality provided by the manufacturer of the drug or device being evaluated. 23 “Policing oneself” is a necessary approach since learned bodies or nongovernmental groups seem little inclined to require such recusal. A systematic analysis of 52 guidelines compared the quality of the Agency for Health Care Policy and Research (now the Agency for Health Care Research) guidelines with those of subsequent “updates” authored by other experts and guidelines that referenced or were adapted from the same guidelines. The comparison was based on 30 criteria and concluded that North American guidelines developed after and in the same topic areas as the Agency for Health Care Policy and Research guidelines are of substantially worse methodological quality, with the 2 most common problems being failure to conduct a systematic review and failure to use a multidisciplinary panel. 24 The one important caveat to consider is that most qualified participants in such consensus guideline development may be those with the greatest conflicts because they are sought out by industry for their expertise. This reflects the same tension that led Dr Scott Gottlieb, who at the time was the Food and Drug Administration's (FDA) Deputy Commissioner for Medical and Scientific Affairs, to state the following: “Some of the most valuable input often comes from people who are active practitioners but also heavily engaged in clinical research and we need to make sure that we continue to have the ability to recruit top clinical trialists” to the Advisory Committees of the FDA. 25 Ultimately, there is need for a balanced approach to participation in guideline development and adoption of principles to identify more clearly the conditions under which conflict of interest waivers may be granted. The more recent FDA guideline 26 on recusal of advisory panel members with personal conflicts exceeding $50,000 seems excessively liberal, especially since it exceeds the threshold of $10,000 for definition of significant financial conflict of interest in Public Health Service regulations. 2

PY - 2007/5

Y1 - 2007/5

UR - http://www.scopus.com/inward/record.url?scp=34248154377&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34248154377&partnerID=8YFLogxK

U2 - 10.4065/82.5.607

DO - 10.4065/82.5.607

M3 - Comment/debate

C2 - 17493426

AN - SCOPUS:34248154377

VL - 82

SP - 607

EP - 614

JO - Mayo Clinic Proceedings

JF - Mayo Clinic Proceedings

SN - 0025-6196

IS - 5

ER -