Correspondence
Who Should Determine When Health Care is Medically Necessary?
N Engl J Med 1999; 341:58-60July 1, 1999
- Article
To the Editor:
I am the author of medical policy for one of the largest health care management companies in the United States. I agree with the conclusions of Rosenbaum et al. (Jan. 21 issue)1 about who should determine when health care is medically necessary, with three important exceptions.
First, “sizable” is a vague and puzzling term to describe the number of studies that would constitute sufficient evidence. Isn't the quality of evidence far more important than the number of papers one can produce? A single, appropriate, randomized clinical trial that proves that a treatment is harmful is enough evidence to make both clinical decisions and coverage decisions. Besides, for ethical reasons, it is highly unlikely that a trial showing harm would ever be repeated. Second, physicians — not insurers — should prove that a new technology produces a positive health outcome. The standard of care for future therapies should be based on careful collection of data, rather than on opinions or good marketing techniques. Third, external reviewers should confirm that insurers follow the rules stated in the insurance contract; the reviewers should not be allowed to rewrite it. If reviewers are allowed to create arbitrary definitions of medical necessity and to change the limits of coverage, it will become impossible to price insurance accurately.
The description of inconsistency in insurers' decision making also applies to decision making by the average practicing physician. A study in which 135 practicing physicians were asked to describe treatment for a lower urinary tract infection vividly illustrates this inconsistency.2 The researchers tallied 82 different approaches among the physicians' answers, with costs of the proposed treatments ranging from negligible to $250 per case. Rosenbaum et al. argue that consistency is a difficult standard in medicine. I agree, but for the same reasons, consistency is equally challenging in reviews of medical necessity.
The issues of secrecy and financial conflict are also overblown. UnitedHealth Group's treatment policies, which document the rationale and evidence used by our physician panel, are available on request to any of our contracted physicians. Rosenbaum et al. completely disregard the physician's financial incentive to perform a service. Both insurers and treating physicians have a financial conflict of interest. A good definition of medical necessity and a fair, expedient external review process can ensure that those conflicts are kept in balance.
2 ReferencesLee N. Newcomer, M.D.
UnitedHealth Group, Minnetonka, MN 553431
Rosenbaum S ,Frankford DM ,Moore B ,Borzi P . Who should determine when health care is medically necessary? N Engl J Med 1999;340:229-232
Full Text | Web of Science | Medline2
Berg AO . Variations among family physicians' management strategies for lower urinary tract infection in women: a report from the Washington Family Physicians Collaborative Research Network. J Am Board Fam Pract 1991;4:327-330
Medline
To the Editor:
Rosenbaum et al. misrepresented my views on medical necessity. I do not believe that medical necessity should be “separated” from clinical practice. Of course, decision making about medical necessity has to include careful consideration of each patient's specific case; how else could it be done? Moreover, I do not believe that insurers should have the power to make “conclusive” decisions without permitting an appeal or review. In the reference the authors cited for my views,1 I did not even discuss these issues.
Other thoughts in this article are confusing or incomplete. For example, the authors appear to argue against the use of “generalized evidence” in decisions about medical necessity, painting it with terms such as “arbitrary,” “questionable,” “separated from clinical practice,” and “not tak[ing] into account the needs and circumstances of each patient.” They do not define “generalized evidence”; I assume they mean clinical trials. If so, trials are of course imperfect and incomplete. But are the authors saying they have no application to individual patients? Is there a better way to determine what works, on whom, or how well?
The essence of the authors' proposal is to put the burden of proof on whoever would interfere with the decision of the treating physician. Depending on what needs to be proved and how much documentation is required — neither of which is described by the authors — the burden could be insurmountable. Throughout the article, the authors talk only about insurers, which gives the impression that disagreements are always between physicians and insurers. Making “management by insurers” the villain also makes the burden the authors propose seem more palatable. In fact, however, many insurers, managed-care plans, and medical groups (which increasingly bear the financial risks, through capitation) have physicians setting the policies, designing the guidelines, and reviewing decisions about medical necessity. In these settings, the conflicts are among physicians, physician groups, and panels of experts. Do the authors intend to cripple this peer management process as well?
There is overwhelming evidence that the decisions of individual physicians produce wide, medically unexplainable variations in practice and high rates of inappropriate care. This is bad for patients. If an insurer, health plan, or medical group wants to reduce the inconsistency, unfairness, bad outcomes, and waste caused by variations and inappropriate care, how would the authors recommend they do it?
1 ReferencesDavid M. Eddy, M.D., Ph.D.
2435 California St., NW, Washington, DC 200081
Eddy DM . Rationing resources while improving quality: how to get more for less. JAMA 1994;272:817-824
CrossRef | Web of Science | Medline
To the Editor:
To purchase effective health care services for their members, health insurers follow the “middle” course endorsed by Rosenbaum et al. Insurers use clinically derived professional standards of care, scientific evidence when available, and expert consensus to determine medical necessity and coverage for new treatments and for the manner, intensity, and setting of established treatments. Recognizing the complexity of clinical decision making and the role of the treating physician's clinical judgment, health insurers use committees of physicians to examine the available scientific evidence and to seek expert consensus when evidence is lacking. Of course, no guideline can anticipate every circumstance concerning individual patients. Requests that do not meet coverage guidelines are individually reviewed through reconsideration processes in which the treating physicians discuss with health plan physicians their patients' unique factors and their clinical rationale for treatment.
Contrary to the authors' implication, “broad standards based on evidence involving large groups of patients” are frequently used by physicians to support medical decision making for individual patients. The assertion that clinical evidence “is of insufficient quality to displace clinical judgment” is not supported. The authors ignore the extensive body of published literature that documents huge, nationwide variations in medical practice — including the use of ineffective or unsafe treatments, underuse of life-saving treatments, and overuse of services.
More troubling, the legislative solution proposed would set back any rational attempt to address a pressing national health care problem: the lack of health insurance coverage among large segments of the population. The reason: it would shift the burden of proof for medical necessity and diminish physicians' accountability for clinical effectiveness. Insurers would have the difficult task of proving the negative — that proposed services are unsafe, inappropriate, or ineffective — when making coverage determinations. Invariably, as in the days of double-digit increases in premiums, insurers would cover many more inappropriate or ineffective tests and services, thus boosting costs while diminishing quality. Rising premiums would drive more people from the health insurance market. Thus, some people would lose access to coverage so that others could have broader coverage for unnecessary or ineffective care.
Scott P. Serota
BlueCross BlueShield Association, Chicago, IL 60601-7680Author/Editor Response
The authors reply:
To the Editor: Eddy, who is known for his scholarship on the use of practice guidelines and his leadership in guideline development,1 has clarified his views that a patient's individual circumstances must be considered even when valid scientific guidelines exist and that insurers should not make conclusive determinations about treatment on the basis of guidelines alone. However, we disagree that shifting the burden of proof would “cripple” the peer management process, put an “insurmountable” burden on insurers, disrupt peer processes of review or education, or shield individual clinical errors. Insurers refuse to disclose the information they use in their decisions and to which they have sole access and have vociferously opposed any such requirement for disclosure. If they want to shield evidence, they should be required to prove their case.
We agree with Newcomer that a single study could dominate the field, so that no additional evidence of appropriateness is necessary. Reliance on such a result would be consistent with our views, because that result would represent substantial evidence if offered by an insurer.
Serota claims, without factual support, that fair and scientifically grounded treatment-review procedures will increase the number of uninsured Americans. It is ironic that after helping to destroy or undermine efforts to insure Americans, such as health reform and portability, some industry members would now use the uninsured population to justify their opposition to standards of fair coverage.
A 1999 federal court case, Jones v. Kodak, 2 suggests how reliance on treatment guidelines alone may ultimately affect the quality of health care in the United States. In Jones, an employee benefit plan, relying on undisclosed guidelines, denied an employee treatment for alcoholism in a rehabilitation facility in Arizona, even though the company's own medical consultant concluded that the plan guidelines were inappropriate in this particular case. The 10th Circuit Court of Appeals held that the plan had total decision-making authority, not subject to review, because the guidelines were part of the terms of the plan and because employers have full discretion to buy substandard medical treatment. Although Jones is the first case of which we are aware that deals directly with practice guidelines, the decision is consistent with previous judicial holdings with regard to employer discretion under the Employee Retirement Income Security Act. The message is that even in the absence of disclosure, employee benefit plans can contract for fixed, unyielding, and unscientific standards of care that eviscerate the individual decision-making process that is the hallmark of professional medicine. Bluntly put, the decision makes a mockery of efforts to improve the quality of health care.
2 ReferencesSara Rosenbaum, J.D.
George Washington University Medical Center, Washington, DC 20037David M. Frankford, J.D.
Rutgers University School of Law at Camden, Camden, NJ 08102Brad Moore, M.D., M.P.H.
George Washington University Medical Center, Washington, DC 200371
Eddy DM . Rationing resources while improving quality: how to get more for less. JAMA 1994;272:817-824
CrossRef | Web of Science | Medline2
Jones v. Kodak, 1999 WL 111147, 10th Cir.
- Citing Articles (1)
Citing Articles
1
Robert D. Eden, Robert D. Eden, Ann Penka, David W. Britt, Ellen J. Landsberger, Mark I. Evans. (2005) Re-evaluating the role of the MFM specialist: Lead, follow, or get out of the way. Journal of Maternal-Fetal and Neonatal Medicine 18:4, 253-258
CrossRef
