Clinical care guidelines have proliferated wildly, with the National Guideline Clearinghouse listing thousands of them. While they are supposed to be voluntary, they aren’t; doctors are often reimbursed based on their use, and failure to follow them may be evidence in a malpractice case. The rationale for guidelines is basically that the developers know more about appropriate medical care than the treating physician and that their use reduces unwarranted variation in care. In an era where medicine is supposedly becoming more personalized, this is a strange attitude. The implication is that guideline developers are capable of anticipating enough specific patient factors, however unique, to mandate a specific course of treatment to optimize outcomes. It doesn’t seem reasonable to expect that a clinician not looking at the patient and all the circumstances going into a health state can make the best decision. A paper at the National Bureau of Economic Research explores this conundrum. (NBER Paper) Using some common economic theory assumptions about how people make decisions, the paper suggests that fully knowledgeable and rational physician would in essence not need guidelines; such a physician would inherently maximize patient outcomes. But physicians have been shown by research to usually not have or gather full relevant information and to not always make the best treatment judgments that create the most patient welfare.
That leaves an opening for the utility and value of a guideline. But enforcing guidelines assumes that the guideline promulgators also act in a fully rational way and the paper finds that developers also lack maximum knowledge. One problem that is highlight is the presence of methodological issues in the research that underpins guidelines. These issues include extrapolating from study populations to patient populations and from surrogate outcomes to ultimate outcomes. The author concludes that much medical care poses an issue of decision-making under uncertainty conditions, and then discusses how to make the best decisions in that kind of environment. Reviewing several techniques to maximize the outcome of treatment decisions under uncertainty, he concludes that, contrary to the rationale for guidelines, diversification of treatment may be best, as it avoids bad errors that affect all patients, and allows for natural experiments to further study the impact of different courses of diagnosis and treatment. I have long been troubled by guidelines, especially when essentially mandated. I don’t see how a guideline gives a physician flexibility to consider all possible relevant factors. And the issues with guideline adequacy is repeatedly demonstrated by frequent changes in them even for very common diseases. Better to view guidelines as just that; a helpful hint to physicians and other clinicians about what they should consider in treating a patient; and better to ensure that physicians are well-educated and get paid for constant continuing medical education, which keeps them up-to-date on developments in medical knowledge and practice.