If shared and informed decision-making between patient and treating physician will lead to better outcomes and more appropriate and cost-effective care, when the doctor is the patient we should see a clear movement toward those goals. That is the question explored in a new National Bureau of Economic Research paper. (NBER Paper) In general, prior research has shown that most patients lack any substantial information on their health conditions or the treatment options related to those conditions. This lack of data would obviously impair good decision-making and means that patients often default to whatever a clinician recommends. The authors’ theory is that since physicians are uniquely well positioned to be informed decision-makers about the health care they receive, their behavior sets an upper bound on what we might expect from non-physicians. It might also give us a sense of whether the issue in regard to health care decisions is lack of information or more related to either inability to use the information or non-rational factors that override factual information. The research used ten years of data from the military health system and compared treatments received by physicians and non-physicians. The comparison focused on supposedly objectively-determined high value or low value services. The analysis included a number of adjustments to account for potential confounding differences between physicians and non-physicians.
The care examples used were Cesarean section delivery, which often is viewed as low-value; pre-operative testing before low-risk surgeries, specifically cataract removal and hernia repair, also viewed as low-value; indicia of good diabetes care; medication adherence for hypertension or high lipid levels and dependents’ vaccination rates, all designated as high-value care. In the case of more discretionary Cesarean section delivery, physician mothers had a slightly lower rate, about three percentage points, compared to non-physician mothers; and dependent spouses of physicians showed no difference in this category of Cesarean. Other than a slight decline in use of chest x-rays, there was no significant difference in the use of testing before low-risk surgeries between physicians and non-physicians. Similarly, in regard to the high-value care, there was no significant evidence that physicians were more likely to comply with diabetes care guidelines or to be more medication-regimen adherent. There were also very modest, at best, improvements in immunization rates. The results clearly suggest that better information alone isn’t going to change health care decision-making in what is perceived to be a positive direction. But this may be an artifact of our health system, including the military one. When people don’t directly pay for the care they receive they have no incentive to be cautious, in fact, people may perceive receipt of any health service as reducing their risk of a bad health outcome, so since it doesn’t appear to cost them anything, why not do it. Physicians are as likely to exhibit this behavior as any other patient. So information likely needs to be coupled with some financial incentive to actually change behavior in regard to low-value care. The explanation for why physicians don’t receive more high-value care is puzzling. Perhaps they don’t think it is really all that valuable or makes that much difference it outcomes. And general levels of receipt of this care may be high enough that it is hard to see differences between groups.