• The ultimate evidence would be health outcomes. For people with disabilities, those outcomes need to be determined based not just on medical results but also on the patient’s judgment of how the health intervention has affected his or her life. For example, has the health item or service improved or maintained the patient’s functional independence and autonomy (or slowed losses that would reduce such independence and autonomy)?  Most discussions of “evidence-based medicine,” however, tend  to focus on scientific clinical studies, especially double-blind studies. There are several problems with this standard, including, but not limited to, the following:

1.    Available clinical studies may have been done on patient populations with different demographics and different co-morbidities than the particular patient being seen by the Medicaid provider. How would such studies be applied in a clinical setting? Would the results of a study on young, healthy, white males be applied to treatment of a 65-year-old Latina woman with post-polio syndrome and heart problems? Would the absence of an approved study specifically on people who share the patient’s demographic and medical profile mean that only the standard (or least-expensive) treatment would be covered—even if that treatment might be ineffective or harmful for the patient?

2.    For many conditions, there are no clinical studies at all. That’s especially true for relatively rare or newly identified conditions and for those affecting relatively small populations, particularly those without much political or economic power. Will the absence of studies translate into the absence of appropriate treatment for these conditions and populations—even when there is a treatment that has a record of success based on the clinical experience of the physician and/or as reported in medical literature? If so, the system will be discriminating against many of the very people Medicaid is intended to serve—including people with disabilities.

3.    Results of various studies can and often do conflict. If “scientific evidence” is used as a basis for covering or not covering various treatments, the people who select the studies to review for that “evidence” will have tremendous power over the health-care options of Medicaid beneficiaries. How will they be selected? What kind of oversight will there be to assure that there is no bias (intentional or unintentional) in the selection of the studies used for developing “evidence-based” guidelines? How often will the “evidence” be updated or re-evaluated? Long time lags could deprive Medicaid beneficiaries of the most effective treatments for their health needs not because the evidence doesn’t exist but rather because the state hasn’t been keeping up. As long as the research is for the purpose of informing physicians—not for directing or restricting their treatment options—the system can be self-correcting and can lead to better health care. But if “evidence-based medicine” means using these guidelines as a kind of one-size-fits-all treatment grid, it will lead to poor health outcomes and potentially greater costs.

4.    Unfortunately, it is often the availability of funding, rather than the need for research data, that determines what gets studied. Thus conditions that haven’t attracted the attention of drug companies or foundations or condition-specific organizations that fund research projects are less likely to have been studied. Furthermore, in many cases, the funder of a study decides the question to be investigated. And the way the question is posed can affect the study outcome. Again, if Medicaid begins to rely on “evidence-based medicine,” how will it assure that biases—conscious or unconscious—aren’t built into the studies that are used as “evidence”?

5.    We recognize that scientific evidence can be valuable in developing treatment guidelines. However, those guidelines must be voluntary, allowing physicians and other health-care providers to work in partnership with their patient to find the most appropriate treatment. Mandatory “guidelines” would be extremely dangerous for people with disabilities, serious chronic health problems, and rare diseases, as well as for members of many racial and ethnic minority groups.

• Oregon’s approach to Medicaid cost-containment.

1.    The Oregon approach to Medicaid cost-containment has been positively cited several times in regard to Medicaid reform. And, in fact, there is  much to learn from the experience of Oregon—and from the experience of other states as well. However, we see some significant problems. For example:

• According to an Issue Brief  by the National Mental Health Association,  Oregon’s Center for Evidence-based Policy—directed by former Oregon Governor John Kitzhaber—plans to conduct a wide-ranging review of all available research to determine the "best" therapies in various medication classes based on evidence from clinical studies, comparative analyses in journals and manufacturer dossiers. Unfortunately, the brief points out, “the group’s literature also notes that when studies are inconclusive about obvious ‘best choice’ among drug therapies in a class, the cheapest option will be recommended for a preferred drug list (PDL).” 

• This can be a major problem if a PDL is treated like a restrictive drug formulary (or used to create one), thereby excluding appropriate medication based not on any indication of ineffectiveness or risk to a particular patient, but rather based solely on the absence of enough reviewed studies that show the treatment has worked substantially better than competing therapies. As noted above, “evidence” can mean different things; study results often conflict with one another, and the research simply hasn’t been done for many conditions and populations. None of those circumstances justifies limiting covered treatment to the cheapest option in all cases, regardless of the patient’s needs. (See also Point #2.) 

• If the PDL were to be used as the basis for a restrictive formulary, it would be critically important to include provisions for covering off-list prescriptions when medically appropriate. Such provisions must involve a process that is timely and easy for both the patient and the physician to use. If the patient doesn’t understand the system, he or she is unlikely to recognize that an off-list treatment still may be available. If the system is too cumbersome, the physician is unlikely to bother with it, especially in view of low Medicaid provider compensation. In either case, the result is liable to be avoidable health problems and increased costs to Medicaid down the line.

2.  The NMHA Issue Brief challenges the Drug Effectiveness Review Project process on several grounds, including a lack of meaningful public input. It also points out that the Center for Evidence-based Policy has been approaching state Medicaid directors and offering, for $100,000, to provide access to the center’s data.  If this is to be the basis for Medicaid’s Preferred Drug List and/or for any system of limiting access to prescription drugs or other health services based on “scientific evidence,” how will Medicaid protect beneficiaries for whose conditions there is little, if any, scientific research?  If the data that is provided amounts to some kind of ranking that reflects aggregated data only, how would physicians—or Medicaid administrators—even know the details of the individual studies that were reviewed? Such study-specific information might be critically important to determining the appropriateness of therapy alternatives for patients with complex or multiple health problems.

If “evidence-based” guidelines are to be developed, we believe that people with disabilities and chronic illnesses should play a major role in evaluating the effectiveness of treatments for beneficiaries with disabilities and creating those guidelines. After all, we are the people who must live with the health outcomes.

• Conclusion: Scientific evidence can be valuable if it is used—along with information on best practices, outcomes research, and clinical experience—t to inform physicians about treatment options. But cost considerations must be carefully separated from a review of scientific data to protect the reliability of the research. If not, “evidence-based medicine” is likely to become not a guideline but rather a straitjacket and an excuse for denying care that is medically necessary and appropriate.  This is something we strongly oppose.

• On the basis of the total Medicaid population
  This would be an appropriate way to determine an initial treatment choice for most conditions for most Medicaid beneficiaries.

• On the basis of a sub-set of that population.
  This would be appropriate for determining the initial treatment choice for most people with a given condition with given demographic characteristics.

• On the basis of the individual patient.
  This would evaluate treatment options in the context of the patient’s specific characteristics, including but not limited to age, gender, ethnicity and co-existing health problems. This type of cost-effectiveness must be ultimate determining factor in selecting the treatment option.