Unexplained clinical variance: are health plans (partly) responsible?

The perplexing phenomenon know as "unexplained clinical variance" has drawn considerable attention.

The Dartmouth Atlas of Health Care (www.dartmouthatlas.org) compiled extensive evidence documenting marked variance in the delivery of medical services from one geographic region to another.

For example, the rate at which four common orthopedic procedures are performed on Medicare beneficiaries varies substantially across the country (Fig. 1).

[FIGURE 1 OMITTED]

Though referred to as unexplained clinical variance, one observation is consistently drawn from these data: the utilization of specific medical services is correlated with the number of physicians providing the service.

For example, the number of visits to cardiologists within a given geographic region is strongly correlated with the number of cardiologists (Fig. 2). This is not meant to single out heart specialists for criticism; the same phenomenon is seen across all specialties.

But if this unexplained clinical variance were only due to regional differences in specialist supply, there should be less variance within a given area. Yet, even in regions of the country with a similar supply of specialists, striking amounts of clinical variance are found (Fig. 3)

[FIGURE 2 OMITTED]

Clearly, there must be more to unexplained clinical variance than simply regional differences in physician supply.

As a health plan medical director, I confront equally variant utilization rates within the same community! Clinical variance doesn't occur only at the regional level. Community-level variance is also well-documented.

Other factors must play a role. Inconsistent adoption and use of guidelines, personal preference and practice "style" may be important contributors. Similarly, certain areas of medicine lack a strong evidence base upon which to standardize care.

But could there be another, an as yet unrecognized factor, contributing to this variance?

A frequent physician complaint is that health plans vary greatly in what services they approve or deny. Further, formularies differ among health plans. For a physician contracted with multiple plans it can be nearly impossible to remember which plans require prior authorization for a given procedure and it is time-consuming to have to refer to a dozen or more formularies every time a prescription is written.

Medical necessity determinations

As a health plan medical director, I was curious how others go about the process of medical necessity determinations. After all, if health plan medical directors rely upon differing methods when making decisions, perhaps our decisions differ. If so, then could we be driving some of this variance?

To answer this question, I set about to informally survey a number of health plan medical directors across the country to learn how they go about medical necessity decision making and what tools they use in doing so. I contacted eight medical directors from a variety of health plans, large and small, from every region of the country. To my surprise, I found the following:

[FIGURE 3 OMITTED]

* There is no consistency among health plan medical directors with regard to medical necessity decision making.

* Most use online evidence-based databases, medical library resources, peer consultation and outside physician consultations.

* Some use Web-based consultations.

* Some who work in large health plans rely primarily upon internal clinical policy bulletins and technology evaluation centers.

* Some rely upon the National Institutes of Health (NIH) and other federal agencies including the Agency for Healthcare Research and Quality (AHRQ).

The eight medical directors surveyed identified 24 Web sites as being particularly useful.

Although all health plan medical directors surveyed strive to use quality evidence to guide decision making, there was little consistency with respect to the sources they rely on, what constitutes "quality" evidence, how current the evidence is and what internal processes are used to be sure the evidence is up to date.

Based on this simple, informal survey, it is reasonable to postulate that at least some unexplained clinical variance is actually be driven by health plans. Using widely disparate methods and processes for medical necessity decision making and an even wider variety of online references, it is not so surprising that providers would perceive that health plans differ in what treatments they approve.

As one physician observed, how can one health plan determine that a given treatment is medically necessary and another come to the opposite conclusion for a similar patient?

If a handful of medical directors is using a wide variety of approaches to medical necessity decision making and dozens of different websites, what would a similar survey of the hundreds of health plans and thousands of medical directors reveal?

Clearly, the phenomenon of unexplained variance and to what extent, if any, health plans contribute is an area of interest for future research.

Reduce variance

The need to reduce clinical variance is unquestioned. In order to do so we need to encourage physicians to rely upon best practices as identified in the current medical literature.

The move to a consistent standard for evidence-based medicine is gaining momentum. Similarly, it is imperative that health plans develop some degree of consistency and standardization in their medical necessity decision making.

How can we realistically expect physicians to decrease clinical variance when we demonstrate "medical necessity determination variance?"

I would propose a conference of health plan medical directors be convened to discuss this issue and develop an industry-wide approach to medical necessity decision making.

Unraveling the mystery

In my 20 years of clinical practice I was sometimes chagrined and occasionally astonished at the decisions made by some health plan medical directors. How did they come up with that?

I began to wonder if variance in medical necessity decision making could, at least in part, explain variance in clinical practice. Now that I face having to make similar difficult medical necessity decisions, I am committed to certain core values to make the task accurate, fair, objective, transparent and consistent.

I rely upon the best available evidence from the medical literature as a guide. Armed with online subscriptions to a variety of sources of current best evidence, I communicate with our providers and explain how I go about medical necessity decision making and specifically what online resources I use. Here is what I rely upon:

* UpToDate (www.uptodate.com) If you haven't yet had the opportunity to use this service, you are missing out on what is almost certainly the single best source of current medical knowledge. The editors of UpToDate continually monitor 330 medical journals and other sources.

* The New England Journal of Medicine--Online (http://content.nejm.org). You read it in medical school and hopefully never stopped. The search function is very user-friendly, allowing quick retrieval of definitive articles published in this, the world's most respected medical journal.

* The American College of Radiology--Appropriateness Criteria (www.acr.org). Imaging has gotten so complicated that even radiologists have a hard time keeping up with the latest advances. That's why the ACR developed the Appropriateness Criteria--evidence-based guidelines to guide the selection of the most appropriate imaging procedure.

* The Annals of Internal Medicine (www.acponline.org) Almost as good as NEJM.

* The Sanford Guide to Antimicrobial Therapy (www.sanfordguide.com). Under the direction of lead editor and a former colleague in practice, David Gilbert, the Sanford Guide is the most relied upon resource for antimicrobial decision making worldwide.

* The Centers for Disease Control (www.cdc.gov). With new infections emerging seemingly daily, this is an indispensable site.

* The National Guideline Clearinghouse (www.guidelines.gov). A good site, though one must be careful. Not all guidelines are evidence-based. This site is a repository for the good, the bad and the ugly.

Michael A. Patmas, MD, MMM, FACP, CPE, FACPE, is clinical assistant professor of medicine at Oregon Health Sciences University and vice president of medical affairs and medical director of Clear Choice Health Plans in Bend, Oregon. He can reached by calling 541-330-8113 or by email at [e-mail]

[ILLUSTRATION OMITTED]

By Michael A. Patmas, MD, MMM, FACP, CPE, FACPE