Comparative Effectiveness Research: An Introduction

On February 17, 2009, within a month of his inauguration, President Barack Obama signed into law the American Recovery and Reinvestment Act (ARRA) of 2009, the $787 billion economic package meant to stabilize and stimulate the nation’s economy. Contained within the Act was $1.1 billion dollars designated for Comparative Effectiveness Research (CER). CER looks to improve healthcare by development of evidence-based practice guidelines based on systematic reviews and synthesis of existing basic and clinical research, through data mining of registries and cohort studies to understand the natural progression of diseases and factors which influence their clinical outcomes, among other approaches. This large infusion of funds into an already evolving area of research placed CER at center stage of the push towards national healthcare reform and thus evoked a great deal of controversy. Important stakeholders immediately started weighing in on the issue, either supporting it as a path to more rationale and cost-effective therapeutics or condemning it as antithetic to individualized, person centered care. Each group’s arguments often seem marked by hyperbole. How the CER is structured, what questions are asked, and how its findings are engrafted into clinical decision making will be of the utmost importance to physicians, including psychiatrists, other providers, as well as to patients/ consumers and their families.

No one should fear well done research. Rather, it should be embraced as invaluable in informing clinical care. However, concern should arise when research findings are used to dictate clinical decisions rather than to inform them. This is so because medical science is complicated, difficult to control and dynamic. This observation is especially true for clinical science due to the multiplicity factors impacting on each individual’s healthcare outcomes, such as their genetic makeup, environment, co-morbidities, etc. In addition, the complexity of research methodologies makes it notoriously easy to misinterpret data and misapply them to practice and policy. Individuals will need to be better informed in order to advocate about their own care while stakeholders use information to advocate with government and other powerful organizations, such as health insurers, that increasingly decide which therapies will be accessible and covered. In an age marked by calls for adherence to “best practices” and “evidence-based medicine” it is important to be aware of the shallow and narrow nature of many of those clinical guidelines. Skepticism and humility about clinical research is warranted. We would do well to remember an interchange about Miles, the protagonist in Woody Allen’s 1973 movie “Sleeper”. A doctor observing Miles’ behavior notes his request for “wheat germ, organic honey and Tiger’s milk”. Another informs that, “Those are the charmed substances that some years ago were thought to contain life-preserving properties.” The first continues, “You mean there was no deep fat? No steak or cream pies …?” The response was, “Those were thought to be unhealthy … precisely the opposite of what we now know to be true.”

That humorous yet cynical interchange draws attention to the often-tenuous basis of what we think we know. While there are undoubtedly domains in which CER is likely to yield clear guidance, perhaps with medical devices, comparative therapeutics, etc., its ability to provide clear pathways is not always evident. Many recent clinical examples support the need for skepticism but not cynicism. Recent examples of radical reversals of policy within relatively short time frames have emerged from clinical research in the fields of cardiology, hypertension, endocrinology, and psychiatry among others. They have involved positions of federal agencies, the organization which certifies hospitals, and a medical specialty society. For example, as evidence mounted that antidepressants, when prescribed for children and adolescents, might evoke suicidal thoughts, the FDA in 2004 required that a “black box” warning be added to the medication package inserts. It now seems that the FDA’s warning resulted in a decline in the prescribing of antidepressants for the identified groups and, inadvertently, resulted in the reversal of what had been a declining rate of youth suicide. The takeaway message from these episodes is of the overriding power of government, accreditation organizations, and large health insurers’ policies to influence care on a vast scale, for better or worse.

Why choose to raise the matter of CER for the readers of Mental Health News? First, ARRA directed the Institute of Medicine (IOM) to identify national priorities as research questions to be addressed by CER using ARRA funds. The IOM, which had no psychiatrist or other mental health expert on its Committee on CER Priorities, published its recommendations on 6/30/09 which included many addressing a wide array of mental health concerns. (The American Psychiatric Association was invited to present very brief comments at one of the Committee’s listening sessions.) It is unclear whether the questions brought forward in the report were ones those in the field would have posed and whether if translated into research they will provide direction to the field’s critical concerns. (The IOM Report is easily reviewed on the web.) I suggest that Mental Health News readers familiarize themselves with the report and the CER process as envisioned.

Second, as the IOM research topics and others are undertaken as CER is pursued, we must wonder how and by whom they will be applied to public policy and individual plans of care. Let me provide an example from the arena of psychopharmacology. What psychopharmacologic treatments will be considered acceptable and paid for in caring for persons with serious and persistent mental illness such as chronic schizophrenia remains an important and exemplary question. It is a fact that to date there has been no research which shows a “significant” benefit from using of two atypical antipsychotics simultaneously. Anecdotally such cases exist and there is some support for using 2 atypicals in the clinical literature. These medications are very expensive and costly to state Medicaid programs or private insurers. Given this background, the question then is, what latitude will clinicians retain when treating persons who have remained refractory to the evidenced based care laid out in clinical guidelines? The New York State Office of Mental Health will be using its PSYCKES data base to identify cases for which 2 “atypicals” are being prescribed as part of its quality indicator study. The NYS Office of the Medicaid Inspector General, charged with rooting out fraud and abuse, planned to review such cases as part of its 2008 work scope but deferred pending the results of the OMH initiative. The question is should all such use be adjudged improper? Scientifically the resounding answer should be “no”. As a matter of policy, the answer is less certain. From a statistical perspective, the absence of a finding of significance does not mean there may not be meaningful subgroups or individuals which would benefit. The criteria of large clinical trials are so broad as to obscure what might be important treatment effects for subgroups. As statisticians know, the absence of evidence is not evidence of absence (of an effect). Once government focuses on a practice, such as the “inappropriateness” of prescribing 2 atypical antipsychotics at once, doctors become far less likely to treat a patient with the “targeted” combination, even when they have exhausted standard options, due to their fear of being sighted. Likewise, if time consuming hoops are created as barriers to such prescribing by insurers, will time pressed doctors make the necessary effort to clear the hurdles? Will clinics, whose quality may be judged on such criteria, discourage their physician employees from going beyond the “approved” prescribing practices when they perceive a need? Can we be comfortable that an insurer’s policy is clinically not financially based given that industry’s problematic record? Will large governmental agencies or insurers be nimble enough to adapt their policies to clinical science’s rapidly changing landscape? (Perhaps, going forward, the use of data mining on vast data sets such as Medicaid’s, Medicare’s, or private insurers may help identify those for whom less recognized approaches to care such as the 2 “atypicals” is beneficial, if there are enough cases to provide the statistical power to perform the analysis.)

Alerted to CER and what we may expect over the coming years as this process gains momentum, how might readers of MHN think about this movement? Readers should welcome the use of CER and other avenues to improved treatment and recovery. However, we should do so with open yet skeptical minds. Dr. Jerry Avron, Professor of Medicine at Harvard and Director of the Harvard Interfaculty Initiative on Medications and Society, in a recent article in the New England Journal of Medicine, asks, “What is the moral responsibility of the physician to care for a patient for whom the best therapy may not meet the conventional standards of cost-effectiveness?” He continues, “These aspects of the debate will need to continue as we begin to implement CER with this vital new funding.” While all cannot become statisticians or methodologists, we can inform ourselves and raise questions and concerns with policy makers, both governmental and private sector, when they promote or act to circumscribe access to a variety of approaches to care.

Dr. Perlman is a past President of the New York State Psychiatric Association & past Chair of the New York State Mental Health Services Council.

Have a Comment?