Evidence-based Medicine
U.S. Surgeon General David Sackett defined evidence-based medicine (EBM) as the “conscientious, explicit and judicious use of current best evidence in making decisions about the care of the individual patient.” Physicians must understand how EBM has become politicized and how health plans and government can use narrow conceptions of EBM to control doctors and patients. Just as physicians must take back control of who defines “quality” and who defines the “performance” in “pay for performance,” we must also promote a strong, inclusive concept of EBM rather than the weak version which will be used as a weapon against consumer choice and physician autonomy.
The Center for Evidence-based Policy (CEBP) at the Oregon Health and Science University contracts for reviews of medication classes under the Drug Effectiveness Research Project (www.ohsu.edu/drugeffectiveness). The efficacy rates of different medications within a class are compared, and when the overall response rates are comparable, the drugs are considered therapeutically equivalent, unless side effects differ within the class.
What are the problems with this approach? By focusing on randomized, controlled trials (RCTs), other kinds of data can be lost. Since crossover trials are rare, CEBP reports generally conclude that drugs with similar response rates are similar, even though different patients may respond to different agents, particularly when distinct or multiple pharmacological mechanisms of action are involved. Clinicians know that individual patients respond differently to different medications, especially regarding side effects.
Patient differences are rarely addressed in RCTs, which notably fail to include multiple ethnicities, the very old, major pediatric age brackets, or the realistic range of co-morbid conditions. Studies accounting for genetic diversity in medication response, or “pharmacogenomics,” remain in their infancy. Currently, testing diverse populations is the only experimental proxy for the complex interactions of genetic diversity and acquired characteristics that shape our bodies’ responses to therapy. Available RCTs largely fail to recruit diversely, nor generally provide data on patient subgroups.
By asking very narrow questions, the CEBP reports may fail to address key clinical issues. For example, the report on medications for Alzheimer’s disease does not review the multiple studies showing delayed nursing home placement and other healthcare savings that result from the use of anticholinesterase medications. When healthcare savings or reduction of caregiver burden are not addressed, the reports can easily be misused to address drug costs in isolation from the total healthcare costs of managing certain patients.
The CEBP now sells its reports to the directors of Medicaid programs in 17 different states. While the state of Oregon uses a more inclusive process, many state Medicaid programs are using these reports without modification to justify slashing drug formularies down to generic drugs only, since all medications in a class are certified equivalent by the imprimatur of “evidence-based medicine.” Such policies are disastrous for many groups of patients. For example, people with epilepsy often have unacceptable side effects and/or inadequate control with the older, generic antiepileptic drugs. Newer agents can make the difference in seizure freedom and/or freedom from side effects.
Using RCT data in isolation creates an unrealistic, “weak” version of EBM that provides a pseudoscientific justification for cutting benefits and denying care. It takes little imagination to see how this thinking will become increasingly attractive to commercial health plans, and how this same approach will be used to restrict physicians’ ability to tailor any aspect of healthcare to the needs of individual patients. Something very similar has already occurred in California Workers Compensation, where insurers now use the ACOEM guidelines as a justification for denying care, guidelines that were developed with less scientific rigor than the CEBP analyses.
Crucially, no study addresses whether “evidence-based management” using CEBP or ACOEM guidelines results in better outcomes, better quality of life, or lower costs. While the EBM studies attempt to review medical literature with great statistical rigor, their use in patient management is not subject to any review or analysis.
When sterile guidelines are not enriched by clinical judgment, healthcare stands in danger of being commodified: straightforward competition on price without regard to genuine, patient-oriented quality. And of course, restrictions on our ability to prescribe medications or other treatments will do nothing to protect us from medical malpractice claims in the event of a bad result.
What would a “strong” version of EBM look like? Therapies would be evaluated using a variety of data, including observational studies, patient registries, and insurance claims data, as well as RCTs. Beyond response rates, treatments would be evaluated based on quality-of-life indicators and outcome data. Different management strategies would be compared using prospective and retrospective data. Then this information would be assessed in a transparent, public process, including physicians, other providers, consumers, policymakers, the life sciences industry and other major stakeholders. Gaps in knowledge would drive further research, for example, assessing outcomes with differing management strategies. And the results would then be explicitly tailored to the needs and preferences of the individual patient, including lifestyle, spiritual and other issues and values.
Recently, the tragic case of Terri Schiavo captured the national imagination. The clinical facts were not in serious dispute. All the drama and tension arose from a crisis in values between different segments of her family, with each side’s perceptions clearly colored by their points of view. The first lesson is that a patient’s or family’s value system must be considered to provide effective care. A second lesson is that values can be disguised as “facts,” illustrated by the widely differing interpretations of her sparse behaviors given by warring commentators.
In the same way, if we fail to examine the assumptions and limitations that underlie the weak version of EBM, then “values” of control and denial will be painted as objective, scientific “facts,” and a covertly political process will be used to silence patients and doctors who believe in consumer-focused care. We must insist on a strong evidence-based medicine that truly offers both the art and science of medicine.