U.S. Health Care Costs Part VII: Reining in Doctors Who Cost Too Much

By Uwe E. Reinhardt


Uwe E. Reinhardt is an economist at Princeton. For previous posts in his series on why America pays so much for health care, click here, here, here, here, here and here.

Last week I wrote about the inexplicably high variation in per-capita health spending across the United States. The rich and highly heterogeneous reader responses have been illuminating and, in some cases, disturbing.

The ideal health systems recommended by readers ranged from government-run, single-payer systems to completely free-market systems in which individuals are responsible for their own health care. Both approaches, of course, ration health care — the first based on the capacity of the existing health care system, and the second based on individual patients’ ability to pay. A judgment of which system is better depends strictly on one’s moral values, not on economics.

I was surprised, though, that so few readers suggested that America’s health care spending issues might have something to do with our fee-for-service payment system.

Studies have shown that physicians are not impervious to the financial incentives inherent in fee-for-service payments. For example, on average, physicians who have a direct financial interest in the use of imaging services, like CAT scans or M.R.I. scans, recommend far more such services for their patients than do physicians without such financial interest (see this and this).

In recognition of the conflict of interest inherent in fee-for-service payment, there is now a worldwide movement to replace the system with so-called “evidence-based case reimbursement”. Under this approach, one single payment would cover all of the supplies and services that are needed, under best, evidence-based clinical practices, to respond adequately to well-defined medical conditions.

Not all health care, of course, can be categorized into neat, episodic bundles for this purpose. For chronic conditions, for example, payment may shift from fee-for-service to annual or monthly fees per patient, and care for such patients might be organized by clinically integrated health systems or by managed care companies procuring health care from different systems on an integrated basis. And for some patients with very complicated conditions, fee-for-service would still persist. One would think, however, that a move toward these alternative payment systems would not only reduce the geographic variations in per-capita health spending, but help control the annual growth of health spending over all.

Aside from payment reform, modern electronic health-information techniques can make doctors and hospitals more publicly accountable for their use of resources through greater transparency. Greater transparency in this realm is the sine qua non of a more cost-effective health system.

For example, it is technically feasible to capture electronically every supply and service requisitioned by every doctor in a hospital for every patient, by type of supply or service ordered. It is also feasible to electronically capture detailed information on the health status of every patient admitted to a hospital.

Combined, these two databases can be reorganized to produce cost-effectiveness profiles for every physician affiliated with a hospital, adjusted for the health status of their patients (in technical jargon, “risk adjusted”). Equipped with such risk-adjusted profiles, hospital administrators could periodically gather all affiliated physicians performing a particular procedure — e.g., coronary artery bypass grafts or knee replacements — into one room and ask those physicians who on average trigger high (risk-adjusted) costs for such procedures to justify these higher costs to their colleagues with lower costs.

The New Jersey Governor’s Commission on the Rationalization of Health Care Resources recommended that the state’s government help finance such hospital-based information systems and make them mandatory at all hospitals in the state. Each hospital administrator would then be required to certify that the physicians affiliated with the hospital have reviewed and discussed these cost profiles at least once a year. For hospitals seeking assistance from the state, these physician cost-effectiveness profiles should be accessible to the governor’s office as a condition for such assistance.

Short of revamping our entire health system, payment reform and more widespread use of electronic information systems most probably could drive even our existing system toward greater cost-effectiveness of care and, most likely, more moderate growth in health spending. There is every reason to believe that the Obama administration will pursue these avenues, as would have his rivals.