Opioid prescriptions following behavioral health training among primary care providers

Overdose deaths due to opioids are a major concern in the United States. Physicians often report inadequate training in chronic pain and substance use disorder management. Here, we evaluate whether a specialized program, the Train New Trainers Primary Care Psychiatry (TNT PCP) Fellowship, affected opioid prescription practices among primary care physicians.

We retrieved information from a publicly insured health program in Southern California on 11,975 patients and 180 primary care providers (PCPs) engaged in care between 2017 and 2021. Of the 180 PCPs, 38 received TNT training and 142 did not. We considered a patient as exposed to the provider's TNT "treatment" if they received care from a provider after the provider completed the 1-year fellowship. We utilized the number of opioid prescriptions per patient per quarter-year as the key independent variable. Linear regression models controlled for provider characteristics and time trends. Robustness checks included clustering patients by provider identification.

Post-TNT training, PCPs prescribed fewer than expected opioids. This result remains robust after controlling for several covariates (coef: - 0.209 ; standard error = 0.052, p < 0.001) as well as after clustering patient observations by provider.

In a large Southern California healthcare system, the TNT training program preceded a reduction in primary care providers' prescription rates of opioids. If replicated in larger samples, a low-cost provider training program has the potential to promote more judicious use of opioids for pain management. We encourage more studies to understand the program's long-term impact on physician behavior and, potentially, on patient outcomes.