New Data: Linking prescribing practices of emergency physicians with long-term opioid use

I thought I had caught him. Logging onto our prescription monitoring system, I found recently filled prescriptions for hundreds of doses of Oxycodone, Morphine, and Hydromorphone from multiple emergency departments and pain management specialists, all under my patient’s name.

“He’s not an addict, if that’s what you’re looking for” said his mother, scowling at me from across the hall. “He has real pain, and you doctors never believe him.” She had a point. Her son’s recent imaging showed severe intra-abdominal pathology treated in emergency departments, operating rooms, and pain management offices over the past year without relief. Her son had become dependent on substantial amounts of narcotics not only to treat his pain, but to treat a burgeoning addiction. Unfortunately, his story is not unique, and as an emergency provider I may be part of the problem.

Because of a systemic reliance on opioids, emergency providers are in a difficult position between treating pain and preventing addiction. In 2010, US physicians prescribed enough opioids to provide every adult 5mg of hydrocodone every 4 hours for a month [1]. Our reliance on opioids has added to the growth of overdoses which account for more deaths than motor vehicle accidents. 

A recent article published in the New England Journal of Medicine analyzes how ED physician prescribing practices impact future opioid use of our patients[2]. The study was a collaboration between the Harvard School of Public Health, Harvard Medical School, Brigham and Women’s Hospital, Massachusetts General Hospital, and the Cambridge National Bureau of Economic Research to evaluate the extent of how individual prescriber practices impact long-term opioid use. Researchers used a random sample of Medicare Part D beneficiaries to identify patients who visited an emergency department from 2008 through 2011. Patients were included in the study if they had not had an opioid prescription for at least 6 months prior to their first ED visit during the study period (called the ‘index visit’). Those with hospice claims or cancer diagnosis between 2008 and 2012 were excluded. For each patient, the emergency physician who treated them during their index visit was identified using NPI billing identifiers. Using this pool of aggregate results, the researchers identified patients who were prescribed opioids, and converted each prescription into total morphine equivalents. Physicians were then separated into high-intensity and low-intensity prescribing groups based on the median dose in morphine equivalents. The primary outcome was long-term opioid use, defined as 180 days or more of opioids within 12 months of the index ED visit, while secondary outcomes included rates of hospital encounters and repeat ED visits.

In total, 215,678 patients treated by a low-intensity prescriber and 161,951 patients treated by a high-intensity prescriber. Prescribing rates among high-intensity prescribers was three-times those of low-intensity prescribers, and those treated by a high-intensity prescriber were significantly more likely to be on long-term opioids 12 months later. Further, the authors used the odds ratio between the two groups to determine the number needed to harm and reported that every 48 opioid prescriptions led to 1 excess long-term opioid user. Additionally, those treated by high-intensity prescribers had increased rates of opioid-related encounters (9.96% vs 9.73%) and encounters for falls or fractures within 12 months after their index visit (4.56% vs. 4.28%).

By using Medicare data, this study could access a large data pool; however, because of limitations of the Part D database, the researchers were unable to precisely link opioid prescriptions to individual providers. If a patient filled a prescription within 7 days of their index ED visit, this was attributed to the physician from that visit, leaving room for some error. The study also did not evaluate for visits with pain specialists, or the potential for differences in outpatient follow-up.

Despite its limitations, this research identifies that our own individual prescribing practices are important. Our colleagues differ in prescribing patterns, some writing triple the number of opioid prescriptions than others, and for every 48 prescriptions we write, 1 patient will become a long-term user. Some physicians may be worried about increased return visits if we do not adequately treat pain, yet this study found no increase in return visits for inadequate pain control. With new opioid-free departments becoming more common and opioid overdoses continuing to increase[3], this article shows we must re-evaluate when and how we are utilizing opioid prescriptions and what effect it may have on our patients. Providing better patient care does not always mean providing more opioids.

Resources for the provider:

Physicians for Responsible Opioid Prescribing: Link Education & Protocols: Link

ACEP Opioid Resources: Link


1.              Committee, D.o.H.a.H.S.B.H.C., Addressing prescription drug abuse in the United States: current activities and future opportunities. 2014.

2.              Center, C.I. Opioid Data Analysis. 2017; Available from:

3.              Barnett, M.L., A.R. Olenski, and A.B. Jena, Opioid-Prescribing Patterns of Emergency Physicians and Risk of Long-Term Use. N Engl J Med,        

                 2017. 376(7): p. 663-673.


Evan Kuhl, MD is an Emergency Medicine Resident at The George Washington University Hospital