Threats to Scientific Advancement in Clinical Practice

Abstract

Study Design. 

Point/counterpoint.

Objective. 

To facilitate debate regarding conflicts of interest and ethical considerations in physician/industry relationships.

Summary of Background Data. 

We find ourselves working in a health care system in which nearly $2 trillion are spent, only to rank 37th in world health. That much money at stake piques peoples’ interest, and various parties, without loyalty to the Hippocratic principles, will find opportunities to gain financially. Clearly, current health expenditures do not always coincide with better outcomes. Yet some believe that over 20% of the gross national product will be necessary to maintain our health care system in the future. Medical liability causes a paradoxical situation: physicians are at times driven to use the newest technology and perform many diagnostic tests, for fear of litigation. This practice is often independent of the best evidence.

Methods. 

Literature search and experience.

Results. 

The nature or magnitude of conflict(s) of interest influences the risk of bias in research. It is important to acknowledge and disclose these potential conflicts while recognizing that the existence of such conflict does not necessarily adversely affect the quality of the research.

Conclusions. 

Conflict of interest need not be a conflict of the mind and is not an evil. Physicians need to disclose all relationships to industry and/or other potential conflicts in order to maintain the trust of one’s community, and in order to advance the best science.

In Brief

In the face of increasing academic and industry/business interdependency, the ethical road has only 1 signpost: disclose!

Conflicts of interest and ethical considerations in physician-industrial relationships are all about the moral imperative espoused in the Hippocratic Oath, “Do no harm.” In designing our future systems of care, it is important to frame the tenuous nature of physician-industry relationships thoughtfully, as not to do so may result in unintended and sometimes tragic consequences.

As medicine evolves, we are, at times, at risk from the newest technology we may employ for our patients. Despite the meaning of the phrase, “technological advances” are not always improvements. In our pursuit of scientific knowledge (at times coupled with the pursuit of wealth), we can be slapdash in our trial designs and methods, undermining genuine advancement. It is true that new treatments can allow for improved care, but we should not assume they automatically do. What questions do we need to be asking ourselves so we are vigilant in our pursuit of truly good (vs. simply novel) health care technologies? One significant issue is that of technologies, medications, and therapies that are introduced to the public under the compromising conditions of unintended bias and conflict of interest. Increasingly close partnerships between academia and industry, at times closing the gap between theory and application to no more than the pen strokes of a signature on a patent application, demand complete transparency of those connections to everyone involved. Most importantly, the patient must be informed of these connections if true informed consent and patient choice are to be obtained.

Case in Point

Conflicts of interest can taint the very nature of good scientific work. Genovo, Inc., is 1 of the most recent and more dramatic examples of an apparent conflict of interest. Seventeen-year-old Jesse Gelsinger died in 1999 following treatment of ornithine transcarbamylase deficiency with normal genes containing the missing enzyme imbedded in a common cold virus. Genovo was the company involved in the development of this new therapy. As it turned out, the physician and the university where Jessie was treated had some ownership and stock options in Genovo. This in and of itself was not the problem. The problem was the lack of true informed consent on the part of the physicians and institution, the least of which was failure to disclose the conflict of interest. The family was also not made aware of the serious side effects seen in early primate research, including organ failures and death, or of the 1 prior human treated with the new gene therapy who developed multiple organ system failure and died. The conflict of interest in this case may in fact be unrelated to the outcome but raises serious concerns. Whenever possible, conflict of interest must be avoided or disclosed, as part of our moral obligation to our patients. Self-interest and greed are simply not options if we are to maintain our Hippocratic principles.

Bayh-Dole

In 1980, the Patents and Trademarks Amendments Act (“Bayh-Dole Act”) allowed universities to automatically receive patents for inventions based on research supported by the federal government. Since that time, there has been an acceleration of academic and industry collaborations. This is not always bad; in fact, there are many examples wherein these collaborations were favorable and necessary, hip replacement being 1. As expected, Bayh-Dole allows for and encourages financial gain. In the Gelsinger case, the physician and institution were reported to have received millions of dollars following a legal settlement.¹

Health Care 2006

We find ourselves working in a health care system in which $1.9 trillion are spent, only to rank 37th in world health.² That much money at stake piques peoples’ interest, and various parties, without loyalty to the Hippocratic principles, will find opportunities to gain financially. Clearly, current health expenditures do not always coincide with better outcomes. Yet some believe that over 20% of the gross national product will be necessary to maintain our health care system in the future. Medical liability causes a paradoxical situation: physicians are driven to use the newest technology and perform as many diagnostic tests as possible, for fear of litigation. This practice is often independent of the best evidence.

Good Research?

However, good evidence is hard to come by. We continue to bring new technologies into practice under cover of poorly designed studies that are often underpowered and provide a false sense of security to our practicing physicians. Inadequate follow-up and failure to report questionable findings undermine the scientific process that has advanced our profession. Many new and exciting ideas are brought forward, without appropriate step-wise introduction via prospective randomized multicentered and/or register-based studies.^3–5 Several technologies are introduced without well-designed trials. The consequences are often delayed and manifest in patients who undergo these techniques, only to be judged ineffective or no better than placebo in randomized controlled trials.^6,7 Such failures create a jaundiced view of our profession and our work. In some cases, there was found to be, at the very least, unintended bias; the investigators/authors may be financially linked to the product under investigation and economically advantaged by the success of their studies.

Such faulty evidence and conflict can drive our increased exposure to medical liability and can affect reimbursement. Given how technology dependent we have become (e.g., 45% of orthopedic patients receive implants), it follows that if good trials are not performed before coming to market to ensure not only Food and Drug Administration (FDA)-required device safety but clinical efficacy without bias, our vulnerability to lawsuit is higher than necessary. In fact, by not requiring clinical efficacy data, the FDA process itself undermines the real test of new technology. Furthermore, failure to provide postmarket surveillance for new devices has untoward consequences that are clearly avoidable (e.g., bone lock cement). Beyond issues of safety, a practicing physician needs evidence of clinical efficacy to support indications and reimbursement. Raising the bar to require clear clinical efficacy would go a long way to negate class action claims. Cost-effectiveness and/or comparable effectiveness data in today’s climate would be even better.

Dissemination of Findings?

Some hundred thousand scientific and medical journals are responsible for publishing literally millions of articles each year, yet most are never cited again (Science Citation Index). Consensus analysis shows that only 10% to 14% of the information published is useful, and most articles are never quoted again except by the author who published the article.^8,9 In spine, our journals should remain competitive by publishing articles with appropriate study designs, methodologies, and interpretation of results; we should not simply publish as many articles in as many journals as the market will bear. Conflicts of interest need to be revealed without fear of reprisal and interpreted by those reading the scientific product for bias or misinterpretation.

Reimbursement Implications

Good, unbiased evidence can alleviate our reimbursement woes. Volumes outcome research from secondary data sources can also provide useful information. Randomized studies by Rorabeck et al,¹⁰ work by Brigham and Dartmouth,^11,12 recent work from the University of San Francisco on total joint arthroplasty,¹³ and previous studies have demonstrated that total hip arthroplasty is cost effective and that when you do more, you tend to have less morbidity and better outcomes. Armed with this kind of evidence-based data, negotiations with payors have recently broken out the DRG for hip replacement (209) into 2 new categories: primary (544) and replacement (545), leading to better reimbursement for a well-proven procedure.¹⁴

System Failure–Pay for Performance ≠ Pay for Good Outcomes

In an effort to encourage evidence-based practice, Medicare Demonstration Projects have introduced “Pay for performance,” in which physicians are reimbursed based on the number of proven procedures or tests that they perform. Unfortunately, this may turn out to be an inappropriate or ineffective incentive. Busy physicians may feel they are doing a good job simply by checking off a list of procedures and tests performed, and may feel that gathering local information about quality/outcomes or cost is no longer necessary, Lest we be too cynical, this is clearly an effort to assist in chronic disease management, wherein there is potential to negate or seriously change the effects of diabetes and other chronic diseases that over time are extremely costly to our nation. However, paying for “quality, evidence based care” by informed patients who are knowledgeable about their treatment choices, wherein their values and preferences matter, is the method by which reimbursement should be calculated.^15–17

Irrational Health Care

Technology is an essential partner in the provision of good health care. But rational, evidence-based care with transparency is in the best interest of all and essential in our health care equations moving forward.

The Dartmouth Atlas provides information that health care is currently irrationally distributed and that outcomes are not necessarily better when more money is spent. A recent study by Fisher et al¹⁸ compared higher spending regions to lower for end-of-life care. End-of-life care was chosen because it is unrelated to regional differences in rates of illness. The measures used included: content of care (e.g., frequency and type of services received); quality of care (e.g., use of aspirin after acute myocardial infarction, influenza immunization); access to care (e.g., having a usual source of care); 5-year mortality; change in functional status; and satisfaction. Their conclusions were that while the higher spending regions spent over 60% more on care than the lower spending regions, there were no differences in any of the outcomes except for use of preventative measures, which was actually lower in the higher spending regions. Rates of orthopedic procedures were not related to spending. There was less total joint, spine, and arthroscopic surgery in higher spending areas.^18,19 Thus, the care received had less to do with evidence of its effectiveness than with hospital capacity and physician (specialist and subspecialist) supply. Patients in high-utilization areas had more in-patient treatment, more diagnostic tests, and increased use of specialists, but no better outcomes.

Professional Regulatory Bodies, Industry, and Government

Recent events have pushed professional regulatory bodies, industry, and government agencies toward a new call for regulation of relationships of physicians with industry, and further intensification is expected.²⁰ The pharmaceutical industry spends about $12 billion annually on gifts and payments to physicians. Today, private organizations fund 60% of biomedical research. Pharma funds about 70% of clinical trials and pays for nearly 50% of CME programs. At what point do these relationships become dubious and Genovo-like?²¹

Institutional Review Boards

Institutional review boards and conflict of interest committee policies regarding disclosures of financial conflicts of interest to potential research participants vary across the academic landscape. Weinfurt et al²² identified 123 U.S. academic medical centers, and sought their institutional review boards and institutional policies regarding financial conflicts of interest. They also assessed institutions’ guidelines for disclosure. They found that only 48% discussed disclosing financial conflicts to potential research participants, and institutional policies were varied.

Proactive Advancement

There are several things we can do to proactively advance spine health care. One is to use those treatment techniques that have been proven effective over time and through well-designed, nonconflicted, trials of efficacy and cost-effectiveness. Naturally, given their investment in development, industry supported trials are driven by the need for swift FDA approval in order to get a device or drug to market and begin to pay down the cost of that development. However, since we know that industry supported projects showed a nearly 73% positive result, whereas unfunded research had a much more even distribution of results, with positive at 46.2%, neutral at 45.6%, and negative at 8.2%, we must remain skeptical of device approvals that arise strictly from their makers’ trials.²³

However, we must also pay close attention to the affiliations of those in academia that perform research. At university campuses around the country, we are seeing a changing influence of large corporations in the naming of science centers, departments, professorships, and, ironically, schools of ethics. So, carefully reading the disclosure statements of articles reporting the results of trial is imperative to a complete understanding and interpretation of the results. We must, therefore, be cautious in adopting new technology that has not been evaluated with the rigor that affords us more than a modicum of confidence that we are doing what is best for our patients.

Spine’s Approach to Conflict of Interest

Spine has always had a strong conflict of interest policy, but the effectiveness of these policies is largely unknown.²⁴ The public has, appropriately, long held the medical profession to a high ethical standard. Disclosure can serve to disarm the often-implied effects related to conflict of interest. Lack of disclosure undermines the necessary ethos needed to ensure the public trust. To that extent, many prominent medical journals have gone to the media with plans to change their current policies related to reporting and publishing manuscripts with conflict of interest. Many of the proposed changes have been in effect at Spine for several years (Appendix A available online only through Article Plus), and the topic is revisited annually.

Spine endorses a policy of full disclosure and has a disclosure-based peer review process. In other words, Spine also requests that its reviewers disclose any conflicts of interest. Conflicts unto themselves will not discount an article from publication. The potential for conflict of interest may arise in the context of research. The nature or magnitude of conflict(s) of interest influences the risk of bias in research. It is important to acknowledge and disclose these potential conflicts while recognizing that the existence of such conflict does not necessarily adversely affect the quality of the research.

Definitions

Definition of a potential conflict of interest: “set of conditions in which professional judgment concerning a primary interest may be unduly influenced by (a) secondary interest(s).”²⁵

Primary and Secondary Interests:

• Primary interests of the physician, researcher, and teacher involve those of the patient, the integrity of the research, and the quality of education. Conflict(s)of interest may be related to 1 or more of the following:
• Secondary interests: for example, financial issues; professional advancement; academic pressures or prestige; personal bias.

As stated by Thompson,²⁶ “conflict-of-interest rules usually focus on financial gain, not because it is more pernicious than other secondary interests but because it is more objective and more fungible.”

Figures 1 and 2 depict Spine’s disclosure trends from its Conflict of Interest statement.

Figure 1: Number of Spine authors disclosing funding by source (2001–2005). (“Org” indicates organization).

Figure 2: Number of Spine authors disclosing benefits received (2001–2005). (“rec’d” indicates received).

Stewards of the Profession

Every system is designed to get the results it gets, and market forces vaunting unproven technologies and mostly unknown or paradoxical outcomes often drive our current health care system. Our moral compass must point in the direction of our patients, and not self-interest or greed.

How do we as physicians and surgeons reestablish ourselves as stewards of our profession? We must accept accountability and work together in collaborative learning. We must be willing to change our behavior, and we must rely on clinical data collected in our clinical practices and in clinical trials to drive that change. Clinical trials are indispensable. If in major medical dilemmas the alternative is to pay the cost of perpetual uncertainty, have we really any choice? Data collection is not a recreational project. It requires each of us to take responsibility to do what is right, not just in the form of clinical research, but, most importantly, in daily clinical practice. We ourselves must become knowledgeable about the outcomes of our daily practice and treatment of our patients. This is an obligation that we should not put aside for someone else to do. This is an obligation in which all of us should participate.

Health Care System Design

I hope that we can redesign medical and surgical practice to get the results we want to benefit our patients and our society. The challenges still remain: patient and societal expectations, the spiraling cost of health care, business of health care, politicization of health care, and loss of physician autonomy/specialization. Medical liability reform is happening, though the politicization of this issue in itself remains problematic. The solutions remain: to be clear about who we are and what we represent, to be enlightened leaders who share leadership among others, and to be provocative and proactive rather than constantly reacting. As George Santayana said, “Skepticism is the chastity of the intellect, and it is shameful to surrender it too soon or to the first comer.” He also said “Science is nothing but developed perception, interpreted intent; common sense rounded out and minutely articulated.” Simple things, but as with all things, to do them well takes focus and deliberation.

Our health care system is in need of emergency care and redesign, or it will not survive. We as physicians and surgeons can help to lead this effort because what we do works, and what we do helps our patients and our society. What will not work is burying our heads in the sand and waiting for others to come to the rescue. Being transparent in our mission, vision, and research is the best medicine we can give to society. This design will give our profession and our patients the results we both want.

Key Points

• Greater technological advancement in medicine inevitably leads to physicians being increasingly involved in relationships with industry and others.
• Physicians should disclose any and all relationships with all partners in research, leaving the reader to decide if there is a potential conflict.
• Gathering clinical data is not just for research anymore but can be very helpful in daily practice.

Appendix available online through Article Plus.