When I asked Senior Editor Frank Diamond to do this month's cover story on risk adjustment, I was intrigued by the concept. Measure the sickness of populations — even individual patients — so that payers, plans and providers can set rates more scientifically.
I knew there was considerable dispute about how advanced this "science" was, but this was a case in which my optimism overpowered my skepticism. Noting that Medicare is about to jump on the risk adjustment bandwagon, Frank uses the Field of Dreams metaphor. "They could easily be saying something like, 'If you build it, it will come to you' — the 'it' being the technological wherewithal to make risk adjustment work."
Starting in 2000, HCFA plans to use inpatient utilization data from one year to predict the cost of all services for patients in Medicare+Choice the ensuing year. As one of our experts points out, "That's quite a jump." And even if technology advances to the point where HCFA can also use clinical encounter data from physicians' offices, there will always be some risk that risk adjustment doesn't tell the entire story. Another of our experts reminds us that, at best, we might be able to predict 25 percent of health care costs.
Yet risk adjustment, if it works as advertised, will be more equitable and should encourage more emphasis on quality as a basis of competition. The ones who stand to lose are those who benefit unfairly now by avoiding sicker members/patients.
The pain involved in making such a system work is largely self-induced. While other industries seized on "information technology," health care lagged. While patient charts are still largely handwritten, HCFA's need for encounter data might speed up the inevitable (there, I've said it) progression to computerized records. Information allows medicine to improve, and will help improve systems for financing and delivery of care as well. Let's hope the data we are collecting is good enough for Medicare's risk-adjusted system for 2000.