Working Paper: NBER ID: w29992
Authors: Keith Marzilli Ericson; Adam Sacarny; R. Annetta Zhou
Abstract: Fragmented healthcare received from many different physicians results in higher costs and lower quality, but does it contribute to dangerous opioid prescribing? The effect is theoretically ambiguous because fragmentation can trigger costly coordination failures but also permits greater specialization. We examine dangerous opioid prescribing, defined as receiving high dosages, long prescription durations, or harmfully interacting medications. Cross-sectionally, regions with higher fragmentation have lower levels of dangerous opioid prescribing. This relationship is associational and may result from unobserved patient-level confounders. Identifying the impact of healthcare fragmentation by examining patients who move across regions, we find a relatively precise null effect of fragmentation on dangerous opioid prescribing. These results cast doubt on the role of fragmentation in this phenomenon and highlight the potential role of other forces in driving it.
Keywords: healthcare fragmentation; opioid prescribing; public health; prescription drug monitoring
JEL Codes: I11; I12; L14; L22
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Healthcare fragmentation (I18) | Dangerous opioid prescribing (I11) |
| Unobserved patient-level characteristics (I14) | Healthcare fragmentation (I18) |