Abstract
Objective:
Practice patterns of the emergency physician workforce have garnered increasing attention in recent years. Our objective was to assess the clinical service volume preceding and settings practiced in following emergency physician workforce attrition.
Methods:
We performed a repeated cross-sectional analysis using 2013–2021 Medicare data. Emergency physician workforce attrition was defined as not billing for emergency department (ED) services in the year after having billed at least 50 services in the prior year. Outcomes included: 1) the quantity of ED-based clinical services in the years prior to attrition, and 2) the observed non-ED practice settings billed in after attrition.
Results:
Between 2013 and 2021, 60,140 unique emergency physicians billed Medicare for greater than 50 ED services in one or more years. Of these, 13,888 exhibited workforce attrition, with annual attrition rates ranging from 3.1% to 6.6% during the study period. Compared to those who remained in practice, those who exhibited attrition delivered 12.3% fewer ED services in the year immediately preceding attrition (602.2 vs. 687.0 services). Notably, 27.9% of those exhibiting attrition reduced their services by over 50% in the year before leaving, compared to 3.5% among non-attrition physicians. After leaving EM practice, 23.7% continued billing Medicare in non-ED settings, most commonly in urgent care and office-based settings.
Conclusion:
Emergency physicians reduce clinical service volume in the year preceding attrition from the workforce. After leaving the EM workforce, the minority transitioned to other clinical settings, with those remaining in the broader healthcare workforce commonly practicing in urgent care and office-based settings that bill Medicare.
Introduction
Background
Emergency physician practice patterns, including workforce attrition, have been well-described in the literature.1–6 The COVID-19 pandemic further exacerbated workforce departure considerably, with an increase in annual attrition to an estimated 8.2% during 2020.4,7
Importance
There is limited study of clinical practice patterns of emergency physicians before and after attrition from EM practice. As an increasing number of emergency physicians leave the EM clinical workforce, it is essential to understand the clinical practice patterns of emergency physicians before attrition to better anticipate workforce needs, adequately estimate practice transitions, and ultimately optimize how the process occurs.
This line of inquiry is critical as emergency physicians are required to provide care during the nearly 140 million annual ED visits nationally.8 While increasing clinical and non-clinical opportunities become available for emergency physicians,9 shifts in clinical practice may pose challenges in maintaining a robust emergency physician workforce. A deeper understanding of emergency physician attrition is therefore critical in potentially developing targeted interventions to sustain a robust emergency medicine workforce, and ultimately, is essential in delivering emergency care.
Goals of This Investigation
Our goal was to evaluate the clinical practice of emergency physicians in the years preceding workforce attrition from the ED setting and to identify the settings in which emergency physicians practice after leaving ED-based clinical practice.
Methods
Study Design and Datasets
We performed a pooled cross-sectional analysis using the 2013–2021 Medicare Data on Provider Practice and Specialty (MD-PPAS). At the time of analysis, the 2021 study year was the most recent data available, with this data source having been used in prior healthcare workforce analyses.10–12 This study was deemed exempt by the primary author’s Institutional Review Board and followed reporting guidelines.13
The MD-PPAS dataset includes all individual providers with a valid National Provider Identifier (NPI) who submitted Part B non-institutional claims for services, procedures, imaging, or non-laboratory testing. The MD-PPAS contains information on providers’ demographics, specialty, practice location, total number of services, place of service (POS), and annual utilization summary by care setting. Annual utilization by POS captures the percent of providers’ professional claims delivered in different settings, with POS settings including ED, office, inpatient, hospital outpatient department, nursing facility or skilled nursing facility, ambulatory surgery center, patient’s residence, retail clinic, urgent care, and other.
Data Management
We identified emergency physicians whose self-designated primary specialty was “Emergency Medicine”. Emergency physicians were required to receive reimbursement for at least 50 Evaluation and Management services to qualify as practicing in that year during the study period, consistent with thresholds in previous EM workforce analyses.3–5 Workforce attrition was defined as reimbursement for at least 50 services in one study year, followed by a year in which the emergency physician did not meet the 50-service threshold. This included emergency physicians who remained practicing exclusively in non-ED settings (as identified by the variety of POS described previously) as well as those no longer billing for Medicare beneficiaries in any clinical setting.
Outcomes
Our primary outcome was the number of ED-based clinical services that emergency physicians were reimbursed for in the years prior to attrition compared to emergency physicians remaining in the EM workforce. Our secondary outcome was the prevalence of Medicare billing in non-ED clinical settings by emergency physicians who left ED-based clinical practice.
Statistical Analysis
We first identified the population of emergency physicians that exhibited attrition within each year, including those who left the workforce between 2014 (requiring 2013 for the year prior) and 2020 (requiring 2021 to confirm that attrition occurred). We then calculated the mean number of ED-based services reimbursed for exiting emergency physicians in the five years preceding the year of attrition. Clinical service volume in the year of attrition was not evaluated since the month of clinical practice exit was not identifiable (given annual values). We then calculated the percent reduction in the number of ED-based services for each emergency physician, including both those who left practice and those who did not, and categorized physician clinical service volume year-to-year reductions as >50% reduction, 25–50% reduction, 10–25% reduction, and <10% reduction or an increase in services compared to the prior year.
Among the population of emergency physicians exhibiting attrition, we then assessed the year following attrition and tabulated the number that did not practice in any setting capable of caring for Medicare beneficiaries and the number that practiced primarily in non-ED clinical settings (less than 50 ED-based services, with another setting as the plurality of services). Among those practicing exclusively in non-ED settings, we then characterized the settings that comprised the majority or plurality of services delivered, thereby identifying settings where emergency physicians most commonly practiced after exhibiting EM workforce attrition.
Data preparation and analyses were performed in Stata 16 (StataCorp, College Station, TX). We used GraphPad Prism (San Diego, CA) for data visualization.
Results
Characteristics of the Study Population
Between 2013 and 2021, our study sample included 60,140 unique emergency physicians practicing in the ED setting, including 17,623 (29.3%) female emergency physicians. The five most common states of practice were California (11.3%), Texas (7.1%), New York (6.7%), Florida (5.3%) and Ohio (4.6%). The mean (SD) age of emergency physicians providing ED-based clinical services in 2021 was 46.5 (11.1) years. In 2013, there were 33,571 emergency physicians in the workforce serving Medicare patients, with a slight decrease to 32,953 emergency physicians by 2020.
Main Results
Between 2013 and 2020, a total of 13,888 emergency physicians exhibited attrition from the workforce, with an annual attrition rate ranging from 3.1% to 6.6% during the study period (Table 1). When comparing ED-based clinical productivity in the year immediately preceding attrition, emergency physicians who exited the workforce had 12.3% fewer ED-based clinical services than those who remained in the workforce (602.2 vs. 687.0) (Figure 1).
Table 1.
Practice patterns of emergency physicians the year after exhibiting EM workforce attrition.
| Year | Total in workforce | Total leaving ED practice completely by following year, n | Not practicing in any setting in following year, n | Practicing in a non-ED setting in following year (as plurality), n | Annual Attrition Rate (%) |
|---|---|---|---|---|---|
| 2013 | 33,571 | 1,043 | 538 | 505 | 3.1 |
| 2014 | 34,473 | 1,064 | 599 | 465 | 3.1 |
| 2015 | 35,328 | 2,013 | 1,598 | 415 | 5.7 |
| 2016 | 35,295 | 2,247 | 1,879 | 368 | 6.4 |
| 2017 | 34,939 | 2,319 | 1,957 | 362 | 6.6 |
| 2018 | 34,248 | 2,070 | 1,750 | 320 | 6.0 |
| 2019 | 33,881 | 1,882 | 1,461 | 421 | 5.6 |
| 2020 | 32,953 | 1,250 | 809 | 441 | 3.8 |
| Total | N/A | 13,888 | 10,591 | 3,297 | N/A |
Note: The total leaving ED practice completely is reflective of those leaving all clinical practice and those practicing in non-ED settings.
Figure 1. EM physician clinical services performed to Medicare beneficiaries in the years prior to attrition.
Note: Red- Emergency physicians who did not exhibit attrition (Retention). Blue- Emergency physicians who did exhibit attrition (Attrition).
Note: As an example, an EM physician who left clinical practice in 2019, may have billed 615 services in 2018 (the year prior to attrition), 640 services in 2017 (2 years prior to attrition), etc. The weighted average values (point estimates) were determined by separately aggregating all annual ED service volumes for clinician leaving 1 year prior to attrition, 2 years prior to attrition, etc.
In the years prior to attrition, emergency physicians leaving the workforce gradually reduced their ED-based clinical volume, with a substantial decline in the year immediately preceding attrition (Table 2). Across all study years, 27.9% of emergency physicians reduced their clinical services by >50% if they exhibited attrition in the following year compared to just 3.5% of emergency physicians identified as reducing their clinical services by >50% if they did not exhibit attrition in the following year. Additionally, 20.2% of emergency physicians reduced their clinical services by 25–50% in the year prior to exiting the workforce, compared to 14.5% of those not exhibiting attrition.
Table 2.
Percent reduction in ED-based services one to five year prior to attrition among exiting and non-exiting emergency physicians
| Year Prior to Attrition | Among Emergency Physicians Exhibiting Attrition | Among Emergency Physicians Not Exhibiting Attrition | ||||||
|---|---|---|---|---|---|---|---|---|
| % of Emergency Physicians with… | % of Emergency Physicians with… | |||||||
| >50% reduction | 25–50% reduction | 10–25% reduction | <10% reduction1 | >50% reduction | 25–50% reduction | 10–25% reduction | <10% reduction1 | |
| -1 | 27.9 | 20.2 | 16.0 | 36.0 | 3.5 | 14.5 | 23.0 | 59.0 |
| -2 | 4.3 | 14.0 | 21.8 | 59.9 | 2.7 | 11.0 | 21.5 | 64.8 |
| -3 | 3.7 | 12.7 | 20.9 | 62.7 | 2.5 | 10.6 | 21.1 | 65.8 |
| -4 | 4.1 | 12.0 | 19.7 | 64.2 | 2.3 | 10.0 | 20.1 | 67.7 |
| -5 | 3.0 | 9.2 | 19.2 | 68.6 | 2.1 | 9.2 | 18.5 | 70.2 |
<10% reduction also includes an increase in services provided.
Of the 13,888 emergency physicians who exhibited attrition, 10,591 (76.3%) did not bill Medicare from any clinical setting the year after leaving the EM workforce, and 3,297 (23.7%) emergency physicians practiced primarily in non-ED settings after attrition (Table 1). The most common settings representing a majority or plurality of services rendered among emergency physicians exhibiting attrition were urgent care clinics (36.1%), office-based (31.8%), and hospital outpatient departments (15.1%) (Figure 2).
Figure 2. Practice patterns and non-ED settings practiced by emergency physicians after attrition.
Note: Practice setting is mutually exclusive. We included the clinical setting that comprised a plurality of services performed by an EM physician after exhibiting attrition from the ED clinical practice setting. For a small number of EM physicians that exhibited attrition, two settings may have contributed the exact same % to their overall distribution of billed services for the following year, thereby identifying two settings as the plurality. Four settings (retail clinic, patient’s residence, ambulatory surgical center, and other) were possible practice settings but were not visualizable on the graph due to negligible volumes.
Limitations
At the time of analysis, our data were limited to the 2013 to 2021 timeframe, thus the totality of effects on workforce patterns from the COVID-19 pandemic may not be fully included in our study. Among emergency physicians who exhibited attrition, some individuals retired, who may have exhibited temporary workforce attrition, or decreased their workload to part-time. Additionally, data from practice at children’s hospitals and Veterans’ Association hospitals may not be evident in Medicare data, nor can physicians who primarily care for children. Furthermore, some services cannot be reimbursed because they are not covered by Medicare or because they were not properly documented and billed for, which may affect our inclusion threshold.
Discussion
This study presents an analysis of attrition trends and practice patterns among emergency physicians. To our knowledge, our study is the first to characterize the practice patterns of emergency physicians prior to attrition and the most common practice settings practiced after attrition. Our findings showcase that emergency physicians who exhibit attrition reduced their clinical services in the years prior to attrition, with approximately 1 in 4 emergency physicians exhibiting stark (>50%) reductions in clinical volume in the year prior to attrition. The results of our study have implications for EM departments in terms of recruitment and retention strategies. The observed reduction in clinical service volume among emergency physicians in the years before attrition underscores the importance of proactively addressing the multifactorial contributors to attrition.
After ED-based clinical workforce attrition, only a minority of emergency physicians remained practicing in non-ED clinical settings. Understanding the motivations and challenges associated with transitions to non-ED settings may inform workforce planning efforts and ensure the alignment of training programs with emerging EM career pathways. Furthermore, the high proportion of emergency physicians who did not transition after attrition to other clinical settings potentially highlights the need for targeted efforts to retain emergency physicians within the specialty, especially if this reflects a younger demographic. It is critical to address the supply and demand mismatch of increasing ED visit rates nationally while the physician workforce decreases, the effects of this mismatch, and its implications on clinical care in the ED.14,15
As described by Oskvarek et. al., the 2022 American College of Emergency Physicians (ACEP) New Practice Models Task Force identified 59 emergency physician roles utilizing the EM skillset outside of the ED.9 The Task Force projected the highest demand for emergency clinical roles in rural emergency medicine, urgent care, critical care, global emergency medicine, palliative care, pain medicine, wound care, and addiction medicine, many serving as roles outside of the traditional brick-and-mortar ED.9 The identification of urgent care clinics and office settings in our study as common post-attrition practice environments suggests a shift in career trajectories among some emergency physicians.
Conclusion
In conclusion, our study contributes to a deeper understanding of attrition trends and practice patterns among emergency physicians, highlighting the potential need for targeted interventions to promote workforce retention and optimize healthcare delivery in EM. A substantial population of emergency physicians reduce their clinical workload before leaving the EM workforce, with the vast majority not subsequently providing services in other non-ED clinical settings. These findings suggest that EM workforce attrition may be difficult to predict given the abrupt departure of emergency physicians from the workforce, potentially and critically exacerbating barriers to accessing emergency physician care.
Supplementary Material
Grants/Financial Support:
Dr. Agboh is a Postdoctoral Fellow in the National Clinician Scholars Program which receives support from the Clinical and Translational Science Awards Program (TL1 TR001864) at the National Center for Advancing Translational Science (NCATS), a component of the National Institutes of Health (NIH). In the past 36 months, Dr. Jeffery has received unrelated funding from Nation Institute on Drug Abuse (NIDA), the United States Food and Drug Administrations (FDA), the Agency for Healthcare Research and Quality (AHRQ), and NCATS. Dr. Gettel receives support from the American Board of Emergency Medicine / National Academy of Medicine Fellowship and the National Institute on Aging (NIA) of the NIH (R03AG073988). The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation or approval of the manuscript.
Footnotes
Declaration of Conflict of Interests:
Dr. Venkatesh and Dr. Gettel receive support for contracted work from the Centers for Medicare and Medicaid Services to develop hospital and healthcare outcome and efficiency quality measures. Dr. Pines has received funding from CSL Behring and Abbott Point-of-Care for unrelated work.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Darbyshire D, Brewster L, Isba R, Body R, Goodwin D. ‘Where have all the doctors gone?’ A protocol for an ethnographic study of the retention problem in emergency medicine in the UK. BMJ Open. Nov 30 2020;10(11):e038229. doi: 10.1136/bmjopen-2020-038229 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gettel CJ, Courtney DM, Agrawal P, et al. Emergency medicine physician workforce attrition differences by age and gender. Acad Emerg Med. Nov 2023;30(11):1092–1100. doi: 10.1111/acem.14764 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Marco CA, Courtney DM, Ling LJ, et al. The Emergency Medicine Physician Workforce: Projections for 2030. Ann Emerg Med. Jul 2022;80(1):92–93. doi: 10.1016/j.annemergmed.2022.03.017 [DOI] [PubMed] [Google Scholar]
- 4.Gettel CJ, Courtney DM, Bennett CL, Keim SM, Camargo CA Jr., Venkatesh AK. Attrition From the US Emergency Medicine Workforce During Early Stages of the COVID-19 Pandemic. Ann Emerg Med. Aug 2023;82(2):234–236. doi: 10.1016/j.annemergmed.2023.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gettel CJ, Courtney DM, Janke AT, et al. The 2013 to 2019 Emergency Medicine Workforce: Clinician Entry and Attrition Across the US Geography. Ann Emerg Med. Sep 2022;80(3):260–271. doi: 10.1016/j.annemergmed.2022.04.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ginde AA, Sullivan AF, Camargo CA Jr., Attrition from emergency medicine clinical practice in the United States. Ann Emerg Med. Aug 2010;56(2):166–71. doi: 10.1016/j.annemergmed.2009.11.002 [DOI] [PubMed] [Google Scholar]
- 7.Bennett CL, Espinola JA, Ginde AA, Camargo CA Jr., Attrition from emergency medicine clinical practice in the United States, 2020. Acad Emerg Med. Feb 2023;30(2):139–143. doi: 10.1111/acem.14640 [DOI] [PubMed] [Google Scholar]
- 8.Prevention CfDCa. National Hospital Ambulatory Medical Care Survey. Services USDoHaH; 2021. https://www.cdc.gov/nchs/data/nhamcs/web_tables/2021-nhamcs-ed-web-tables-508.pdf [Google Scholar]
- 9.Oskvarek JJ, Blutinger EJ, Pilgrim R, et al. Beyond the Four Walls: The American College of Emergency Physicians 2022 New Practice Models Task Force Report. Ann Emerg Med. Mar 2024;83(3):250–271. doi: 10.1016/j.annemergmed.2023.08.488 [DOI] [PubMed] [Google Scholar]
- 10.Hu X, Jiang C, Fan Q, et al. US Medicare Hospice and Palliative Medicine Physician Workforce and Service Delivery in 2008–2020. J Pain Symptom Manage. Mar 9 2024;doi: 10.1016/j.jpainsymman.2024.03.006 [DOI] [PubMed] [Google Scholar]
- 11.Tsai TC, Jacobson BH, Benjamin EM, Figueroa JF. Comparison of General Surgical Practice Size and Setting in 2017 vs 2013 in the US. JAMA Netw Open. Apr 1 2021;4(4):e216848. doi: 10.1001/jamanetworkopen.2021.6848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Timbie JW, Kranz AM, DeYoreo M, et al. Racial and ethnic disparities in care for health system-affiliated physician organizations and non-affiliated physician organizations. Health Serv Res. Dec 2020;55 Suppl 3(Suppl 3):1107–1117. doi: 10.1111/1475-6773.13581 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. Oct 16 2007;147(8):573–7. doi: 10.7326/0003-4819-147-8-200710160-00010 [DOI] [PubMed] [Google Scholar]
- 14.Venkatesh AK, Janke A, Rothenberg C, Chan E, Becher RD. National trends in emergency department closures, mergers, and utilization, 2005–2015. PLoS One. 2021;16(5):e0251729. doi: 10.1371/journal.pone.0251729 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Darbyshire D, Brewster L, Isba R, Body R, Basit U, Goodwin D. Retention of doctors in emergency medicine: a scoping review of the academic literature. Emerg Med J. Sep 2021;38(9):663–672. doi: 10.1136/emermed-2020-210450 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.