Information and Patient/Consumer Decision Making in Health Care
Paper Session
Sunday, Jan. 8, 2023 10:15 AM - 12:15 PM (CST)
- Chair: Michael Chernew, Harvard University
Provider Opioid Prescribing Behaviors and Opioid Use in Medicaid
Abstract
Liberal prescription of opioids is widely believed to have contributed to the ongoing epidemic of opioid misuse and related harms within the United States and elsewhere. Policies aimed at curbing the epidemic have focused on encouraging providers to adopt stricter opioid prescribing behaviors. However, the extent to which the association between providers' opioid prescribing behaviors and their patients' opioid use reflects a causal influence of behavior versus patient-provider sorting is unclear. Using Medicaid claims data for three states from 2016-2021, we use provider exits from Medicaid to evaluate how enrollees with chronic pain are affected by a switch to a lower- or higher-prescribing provider. We find that among patients with prior opioid use, switching to lower intensity physicians leads to as much as a 75% decrease in opioid use, with evidence of increased hospitalizations. While we observe a 15% increase in opioid use among opioid-naive enrollees who switch to more intensely prescribing providers, the health effects are less clear. Our findings are similar when using an instrumental variables approach to correct for the potential endogeneity of the destination provider's prescribing intensity.Quality Disclosure, Demand, and Congestion: Evidence from Physician Ratings
Abstract
Ratings provide consumers with useful quality information, however, when ratings shift demand to highly-rated sellers, congestion might occur at the top of the quality distribution. Congestion caused by disclosure may be observed in the health care setting, where prices often cannot adjust to reflect varying quality. I study the trade-off between providing quality information for consumers and congestion using a star rating disclosure policy implemented at a large integrated health system in the United States, which requires every physician to have star ratings posted online in a standardized fashion. I identify the effects of physician star ratings on patient volume using a regression discontinuity and difference-in-discontinuity design which leverages the rounding of ratings to discrete values and the fact that I observe ratings before and after their public disclosure online. I find that an increase in a physician’s rating increases the number of new patients seen by 2.96 visits per month on a baseline of 5.48 (54% increase). I show that star ratings shift patients to physicians who more often provide medically recommended screenings, counseling, and vaccinations. However, I also show that a higher rating causes patients to wait longer for treatment. New patients wait 2.7 additional days (30.5% longer) for an additional increment of the rating scale and existing patients wait longer as well. I use these findings to compute a revealed-preference estimate of the “shadow price of a star”; I find that patients are willing to wait 3 additional days in exchange for a one standard deviation increase in physician ratings. In the absence of a price, wait times may serve as an equilibrating factor to clear the market.To enable screen reader support, press Ctrl+Alt+Z To learn about keyboard shortcuts, press Ctrl+slash
Discussant(s)
Yiqun Chen
,
University of Illinois-Chicago
Jennifer Kwok
,
University of Illinois Chicago
Nicolas Ziebarth
,
Zentrum Für Europäische Wirtschaftsforschung (ZEW)
JEL Classifications
- I1 - Health
- I1 - Health