Working Paper: NBER ID: w31785
Authors: M. Kate Bundorf; Daniel Kessler; Sahil Lalwani
Abstract: In theory, there are several reasons why physician organizational form might affect the price, quantity, and quality of physician services. In this paper, we examine the effect of three aspects of physician organizational form on opioid prescribing: the number of physicians in the physician’s group (if any); the physician’s integration with or employment by a hospital or hospital system; and the average age of the other physicians in the physician’s group. We present three key findings. First, all else held constant, group physicians prescribe far fewer opioids, and prescribe them more appropriately, than do solo physicians. Second, although physicians who are employed by a hospital or practice in a hospital-owned group prescribe fewer opioids than do independent physicians, there is evidence that this difference may be due to differences in the other characteristics of physicians who are hospital-integrated rather than a causal effect. Third, we find substantial peer effects on opioid prescribing. Physicians in groups with a higher average age (excluding the physician him- or herself) prescribe more intensively and are more likely to write inappropriate opioid prescriptions than physicians in younger groups – holding constant the physician’s own age and other characteristics of his or her group.
Keywords: opioid prescribing; physician organizational form; peer effects
JEL Codes: I11; L2
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Group physicians (I11) | Fewer opioids prescribed (Y10) |
| Hospital-employed physicians (I11) | Fewer opioids prescribed (Y10) |
| Peer age (C92) | More opioids prescribed (I11) |