Can medical decisions be standardized? Should they be?

After a variety of regulatory and payment schemes have failed to stem the rising tide of health care cost, the Omnibus Budget Reconciliation Act of 1989 mandates the creation of a system of decision rules called practice parameters for appropriate medical action in many circumstances. A large body of practice guidelines already exists, but lacks the internal coherence required of a policy tool. Professional organizations therefore have developed attributes to achieve uniform style. However, little has been said or published about the constraints that might be imposed on the structure and content of an efficient and coherent system. The arguments presented here lead to the following conclusions: (a) process control is an inefficient tool to manage outcome--standards should reflect product control; (b) guidelines that proscribe are more likely to be supported by scientific data and consensus than those that prescribe; (c) the decision thresholds contained in such directives are policy choices rather than scientific imperatives; (d) neither decision analysis nor artificial intelligence is likely to readily influence medical decisions; and (e) as suggested by operations research, the development of practice parameters should concentrate on issues of therapeutic management in preference to issues of diagnostic activity.