Can a $3 Million Prize Solve a $2.5 Billion Dollar Problem?

If a health plan could predict which patients will incur the most costs, theoretically it could focus on those patients without wasting resources on its healthy enrollees. According to an article in Slate:

Dr. Richard Merkin, the president and CEO of the Heritage Provider Network (HPN) in California … has put a precise value on solving this problem: $3 million. That’s what Heritage is offering to anyone who can build the algorithm that best predicts which patients will be hospitalized and for how many days over the course of a year, based on a given data set. Merkin hopes the award, called the Heritage Health Prize and kicked off today, will help make Americans more healthy and the American health care system more efficient.

The alternative is to hire a consulting firm to develop a mathematical model which may be expensive and perform rather poorly. But with a prize, rather than just one team of consultants, numerous teams from different disciplines will all simultaneously work on the project. History shows winning teams collectively tend to spend 60 percent more than the value of the prize on the solution. This is not just a theoretical exercise. The teams developing the algorithm will have access to several years worth of HPN’s claims data.

Comments (6)

Trackback URL | Comments RSS Feed

  1. Joe S. says:

    Interesting concept. It’s worked in other fields.

  2. Bret says:

    I believe John Goodman made the point that this sort of thing is rare in health care in his Heatlh Alert the other day.

  3. Erik says:

    Boondoggle at best.

  4. Devon Herrick says:

    One factor I thought of is the claims data has all demographic information removed. There, for instance, is (presumably) no record of a person’s age, gender, weight, occupation, education, marital status, etc. These variables could come in handy when estimating probabilities for various medical expenditures.

  5. artk says:

    Of course the punchline is that they will apply the algorithm when renewing policies and omit the bad risks.

  6. Surgeons classify patients based on what we call “risk factors.” These predict which patients are more likely to have complications after surgery. Complications, in general, mean a longer hospital stay or a longer period of time under medical care. More time in the hospital or under medical care, again in general, equates to more cost.

    These risk factors are already well known. Age, activity level, height/weight ratio, smoking history, diabetes, history of atherosclerosis including past heart attack or stroke, and kidney disease are the main predictors of potential complications.

    It would be very easy for a regulatory agency, e.g. the Independent Payment Advisory Board, to deny care for patients based on these criteria or risk factors.