/ Print /

  • linkedin
  • Increase Font
  • Sharebar

    Clamping down on variation


    Stephen Hawking once said, "In the practical use of our intellect, forgetting is as important as remembering."

    Human minds don't operate much differently from computers: Our RAM allows us to concentrate on the tasks in front of us, while our ROM runs in the background. But also like a computer, we occasionally need to take a look at those unconscious functions so we can analyze them and ultimately clean out the things that are no longer needed-or better yet, re-learn a new and better way of doing something.

    The research done by John E. Wennberg, MD, MPH, and his colleagues at Dartmouth's Center for The Evaluative Clinical Sciences, serves exactly that purpose. The American healthcare delivery system has run on autopilot for so long, everyone took for granted that "the way things were done" was the best possible way.

    Not so, he says-not by a long shot. Working in 1967 in the Regional Medical Program created with a $350,000 grant from President Lyndon B. Johnson, Dr. Wennberg was analyzing Medicare data to determine how well hospitals and doctors were serving their communities.

    "Our results were fascinating, because they ran completely counter to what conventional wisdom said they would be," he told me during the interview in his cozy New Hampshire home. "Everyone expected that we would clearly see underservice in the rural hospital service areas remote from academic medical centers. But when we looked at the data, we found tremendous variation in every aspect of healthcare delivery, even among communities served by academic medical centers. We found the same thing when we compared healthcare in the Boston and New Haven communities served by some of the finest academic medical centers in the world. The basic premise-that medicine was driven by science and by physicians capable of making clinical decisions based on well-established fact and theory-was simply incompatible with the data we saw. It was immediately apparent that suppliers were more important in driving demand than had been previously realized."

    Unwarranted variation, according to Dr. Wennberg, is variation on use of medical care that cannot be explained on the basis of illness, medical evidence or patient preferences. Over the years, Dr. Wennberg has categorized four types of variation: 1) the underuse of effective care; 2) variations in outcomes attributable to the quality of care; 3) the misuse of preference-sensitive treatments; and 4) the overuse of supply-sensitive services.

    The Dartmouth Atlas Project, which includes a national sample of utilization data from Medicare, provides an assessment of healthcare performance for about 3,500 hospital service areas stretched across the United States. "We're only just now getting an idea of our situation, as a nation," he says. "It's a huge issue, and there's nothing easy about the things we are trying to do here."

    Identifying and measuring variation The first step is to understand the four major categories of variation:

    I. The underuse of effective care. Dr. Wennberg defines "effective" care (also known as evidence-based) as a treatment or intervention that has been shown in clinical studies to improve health status or quality of life.

    "To put an even further restriction on it," he adds, "its benefits must outweigh its risks to the degree that virtually all informed patients would want to have the treatment."

    The interventions on such a list are smaller than most people think. The examples he gave included the use of beta blockers after heart attacks, pneumococcal vaccinations and diabetic eye exams.

    "No region in the country really does excellent work in those areas, although some regions are better than others. There is consistent underservice," he says. "And equally interesting is the fact that good performance in one of those measures does not correlate to good performance in the others. The only consistency is a lack thereof."

    II. Patient safety. This area has received plenty of attention since the Institute of Medicine's report. The best example in this category is mortality following surgery. The Leapfrog Group has stepped in on this topic and helped identify what aspects of a surgical procedure correlate to the variations in mortality, and made those measures a criterion for its Centers of Excellence. But much more needs to be done to rationalize the production of surgery.

    "We do know that unexplained variation can be reduced," he says. "If you put into place a process of learning and evaluating, such as my colleague Dr. O'Connor's Northern New England Cardiovascular Network, you can benchmark the quality of care between healthcare organizations and learn from experience."

    III. Variation in preference-sensitive treatments. Preference-sensitive treatments are discretionary; they involve trade-offs and not all informed patients choose the same way. The choice of treatment should be based on patient preferences, as in the choice to use hysterectomy as a means to treat fibroids. Hysterectomy is one option; hormone treatment is another. According to Dr. Wennberg, the choice often comes down to the physician's personal preference rather than the patient's.

    "The solution for unwarranted variation in preference-sensitive services is shared decision-making-the active involvement of the patient in choosing. Numerous clinical trials have shown that the patient decision-support programs, such as those available from Health Dialog and the Foundation for Informed Medical Decision-Making, result in better decisions and often a reduction in utilization. But implementation isn't easy," he says. "We need to find a way to encourage and compensate physicians for the time they spend on educating and discussing things with patients.

    "Our goal is to make the patient an equal partner in the decision-making process. Until we do, decisions will be based on supply rather than a patient's needs and preferences. The current system often encourages more treatments and procedures, more risk, higher costs and a lower chance of a good outcome for the patient. A 10-minute interview is just not enough time to make decisions on what are often life-defining choices."

    IV. Variation in supply and utilization. In the traditional supply-and-demand theory of business, the more of a given product or service there is in a market, the lower the cost. As demand increases and outpaces supply, price increases.

    In healthcare, however, the prices are typically stable, especially in federal programs like Medicare. As a result, having an oversupply of a given type of healthcare provider or service doesn't reduce the price-it just increases the likelihood of that service or treatment being provided. In such situations, it is clear to Dr. Wennberg that the tail is sometimes wagging the dog.


    Follow Us On Twitter

    Find us on Facebook

    Latest Tweets Follow