Development of Multipayer Claims-Based Emergency Department Episodes of Care

ABSTRACT

Objectives: To describe the design of multipayer claims-based episodes of care initialized by visits to the emergency department (ED) and to demonstrate the utility of this framework for health care quality improvement work.

Study Design: A retrospective analysis of paid medical insurance claims organized into 30-day episodes of care initialized by ED visits for Michigan adult residents with private insurance, Medicare, and Medicaid.

Methods: Thirty-day claims-based episodes of care initialized by ED visits at Michigan hospitals were constructed for 15 medical conditions. Mean episode payments were price standardized and risk adjusted. Analyses described episode payments and postacute care utilization primarily across 6 conditions: abdominal pain, cellulitis, chronic obstructive pulmonary disease, congestive heart failure, nonspecific chest pain, and urinary tract infection.

Results: A total of 2,657,818 ED-based episodes of care for 15 conditions were identified for commercially and government-insured adult patients receiving ED care at 105 Michigan hospitals. Total payments across a 30-day episode of care and utilization of postacute care services varied substantially by condition and across the state, with mean risk-adjusted, price-standardized 30-day total payments for a given episode ranging from $3262 for abdominal pain to $17,941 for congestive heart failure.

Conclusions: Episodes of care created from multipayer claims data can be used to provide insight into opportunities for collaboration and improvement of patients’ care continuum following an ED visit.

Am J Manag Care. 2025;31(2):In Press

Takeaway Points

Multipayer insurance claims data organized into episodes of care initialized by emergency department (ED) visits can inform opportunities for targeted quality improvement at hospitals in Michigan and worldwide.

• More than 2.6 million 30-day episodes of care initialized by ED visits billed on insurance claims were identified for 15 high-volume conditions among ED patients at 105 Michigan hospitals.
• ED-based episodes of care provide information on patient health care utilization and cost during and following ED visits.
• Episodes were created and reported on by 2 statewide collaboratives working in collaboration with hospitals and physician organizations across Michigan and using a model that can be adapted elsewhere.

Emergency departments (EDs) serve a critical role in US health care for stakeholders including patients, providers, payers, and health systems,1 offering around-the-clock services to any patient seeking care and increasingly providing primary care–adjacent services to select patient populations.2,3 More than 10% of health care expenditures in the US are spent on care received during and directly following ED visits,4 and ED use has risen over the past decade.5,6 Despite the growing role of ED use in the US health care system and an increased emphasis on value-based care, data-driven approaches to improving the quality and costs associated with ED care have been limited.7

An episode-of-care approach may provide opportunities to improve the value of ED-based care. Commonly used in hospital-based settings, an episode of care is defined as a period of time encompassing care received during and after an index event such as a hospitalization or a procedure. Episodes summarize a patient’s health care utilization up until a set number of days following the index event, such as 30 or 90 days post discharge.8-10 The episode-of-care approach is the foundation of value-based payment programs including the Bundled Payments for Care Improvement initiative.11 Episodes of care based on ED admissions have been developed previously, but they are often limited to single insurance providers or based on noncomprehensive survey-based data.4,12,13 Robust risk- and price-standardized ED-based episodes-of-care data could provide insights into ED-based care delivery to a wide range of providers, institutions, and quality improvement groups.

The objective of this study was to describe the development of ED-based episodes of care based on multipayer administrative claims data in Michigan. Specifically, this article outlines the methodological approach for creating ED-based episodes of care for select medical conditions and health care spending and utilization measures therein. Additionally, descriptive analyses detail the frequency of episodes of care, mean risk-adjusted and price-standardized payments for episodes of care, and health care utilization metrics for ED-based episode-of-care measures.

METHODS

Collaborative Quality Initiative Setting

Specifications for the ED-based episodes of care were developed by 2 statewide collaborative quality initiatives in Michigan funded by Blue Cross Blue Shield of Michigan (BCBSM): the Michigan Value Collaborative (MVC) and the Michigan Emergency Department Improvement Collaborative (MEDIC). MVC is a statewide collaborative representing more than 100 nonfederal acute care and critical access hospitals and 40 physician organizations in Michigan. The purpose of MVC is to improve the health of Michigan residents through sustainable high-value health care. MVC produces robust data outputs and maintains an interactive claims-based registry that provides insights from hospital-based episodes of care using comprehensive medical claims for Michigan residents insured by commercial and government plans.8 MEDIC partners with more than 40 hospitals across Michigan on ED quality improvement initiatives to establish clinical best practices for both adult and pediatric ED care. MEDIC combines data collection and measurement, audit and feedback, knowledge translation of evidence-based care, and ED clinician engagement within a learning health model of practice improvement.14

Data Sources

MVC receives multipayer medical insurance claims data for Michigan residents insured by BCBSM preferred provider organization commercial and Medicare Advantage plans, Blue Care Network (BCN) health maintenance organization plans, Medicare fee-for-service (FFS), and Michigan Medicaid. ED-based episodes of care were created from each of these data sources using facility and professional medical claims along with patient enrollment files. Additional data sources included the publicly available monthly National Plan and Provider Enumeration System files from CMS and Medicare Severity Diagnosis Related Group (DRG) classification tools.

Creating 30-Day ED Episodes of Care

Each ED-based episode of care was initialized by a facility claim containing a revenue center code indicative of ED care (0450, 0451, 0452, 0456, 0459); this was designated the index ED claim. Admission and discharge dates on the index ED claim were used to define a given index ED event, some of which were inclusive of an inpatient hospitalization directly following the ED visit. Organization National Provider Identifier numbers identified the facility at which an index ED event took place, and episodes were created exclusively for index events attributed to Michigan hospitals.

To be eligible for an episode, the index ED claim was required to have a primary diagnosis code matching criteria for 1 of 15 high-volume ED-related conditions for which clinical management variation is to be expected: abdominal pain, asthma, atrial fibrillation, cellulitis, chest pain (nonspecific), chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), deep vein thrombosis, diabetes long-term complications, diabetes short-term complications, gastrointestinal bleed, pneumonia, pulmonary embolism, syncope, and urinary tract infection. Conditions herein describe a category of related diagnoses and were defined according to International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnosis codes, with the majority based on publicly available definitions (eAppendix Table 1 [eAppendix available at ajmc.com]).15,16 Conditions were determined to be homogeneous according to DRG, although admissions following ED visits for diabetes long-term complications notably had the largest range of DRGs (eAppendix Table 2).

After potential index ED events were retrospectively identified, 30-day nonoverlapping episodes of care were created in chronological order based on established methods.17 A patient could be in multiple ED-based episodes as long as the 30-day windows did not overlap. An episode of care contained all facility and professional claims for an individual during their index event and the 30-day postindex window (eAppendix Figure 1). Claims within each episode were categorized in a mutually exclusive manner as follows: index ED event (potentially inclusive of hospitalization occurring at the index ED event), postindex inpatient hospitalizations, professional services received during the index ED event and postindex period, and postacute care following the index ED event (inclusive of ED, skilled nursing facility, home health, long-term acute care hospital, inpatient or outpatient rehabilitation, and outpatient services).

The analytic cohort for this study included 30-day ED-based episodes of care for the identified 15 high-volume ED conditions among adult patients 18 years and older who were insured by BCBSM/BCN commercial or Medicare Advantage plans, Medicare FFS, or Medicaid. ED encounters were included between January 1, 2017, and December 31, 2022. Due to data availability at the time of analysis, ED-based episodes from Medicare FFS claims were through November 30, 2022. Index events were limited to ED visits at 105 nonfederal, nonpediatric MVC-participating hospitals in Michigan, inclusive of 83 general acute care hospitals and 22 critical access hospitals.

Thirty-Day ED Episode-of-Care Measures

Several measures to evaluate health care utilization were derived from the 30-day ED-based episodes of care. The primary measure was a risk- and price-standardized 30-day total episode-of care payments measure. This measure encompassed any insurance provider payments for facility and professional claims identified during the episode of care as outlined above. Other payment measures were created for facility index claim payments, postindex inpatient hospitalizations, professional services, and categories of postacute care services within episodes. For all payment measures, price standardization methodology was applied to ensure that calculated payments reflected the type and quantity of services rendered and not contractual pricing differences according to payer or site of care. Price standardization was achieved using established methods of assigning Medicare FFS paid amounts to equivalent services rendered across payers.18,19 For example, payments for facility inpatient claims were standardized based on claim DRG, and facility outpatient claims and professional claims were standardized using Current Procedural Terminology codes. More detailed information on price standardization methodology can be found in the MVC Data Guide in the eAppendix.

Risk-adjustment models incorporating covariates for patient age, sex, insurance provider, 6-month prior health care payments, and comorbidities were fitted to generate expected payment values for health care utilization outcomes within each episode. Model coefficients were selected using linear regression models with stepwise selection (P < .05), and Poisson regression was implemented for final models. Coefficients for patient age, sex, and binary indication of high or low price-standardized prior 6-month claim payments were always included. Comorbidities used in the models were grouped into 79 hierarchical condition categories (HCCs) (according to CMS-HCC Risk Adjustment Model version 22) identified by ICD-10-CM diagnosis codes on claims in the 6 months prior to each episode index event. Risk-adjusted payments were then calculated based on the ratio of observed to expected episode-level payments, multiplied by the overall sample mean payment.

To complement the payment measures, we also generated binary indicators for health care utilization measures for several components of care occurring within the 30-day episode of care, including inpatient hospitalization at time of index ED visit, postindex inpatient hospitalization, secondary ED visit without inpatient admission, home health care admission, skilled nursing facility care admission, and use of outpatient services.

Analysis

Univariate analysis described the frequency of ED episodes of care according to the principal diagnosis for the visit for all patients and stratified by age category (18-64 years and ≥ 65 years) and sex (male and female). Patient characteristics and characteristics of the index hospitals in the sample were also described. Patient characteristics included age, sex, race and ethnicity, year of admission, and comorbid conditions. Hospital characteristics were derived from the 2022 American Hospital Association Annual Survey and included bed size, teaching status, ownership type, rural or urban location, and critical access hospital status.

Further analysis of ED-based episode-of-care measures focused on 6 of the highest-volume conditions that were identified as initial priorities for collaborative-wide ED quality improvement: abdominal pain, cellulitis, chest pain (nonspecific), COPD, CHF, and urinary tract infection. Risk-adjusted, price-standardized episode payment components and health care utilization rates were described for these 6 conditions as outlined above. For analysis of within-condition variation, episodes were placed into evenly sized quartiles by condition based on rank-ordered risk-adjusted, price-standardized 30-day total episode payments. Average episode payment components were compared between episodes in the lowest and highest quartiles.

RESULTS

A total of 2,657,818 ED-based episodes of care at 105 Michigan hospitals were identified in the claims data. Across all ED-based episodes for the 15 index conditions, 61.3% were for female patients and the mean (SD) age of ED utilizers was 54.1 (21.2) years, with the largest group of patients (23.9%) being aged between 18 and 35 years (Table 1). Approximately one-quarter (24.3%) of patients identified as Black, 65.7% as White, and the remaining patients as other race categories or unknown. Most episodes were from patients insured by Michigan Medicaid (39.3%), followed by Medicare FFS (30.6%), BCBSM/BCN commercial plans (18.9%), and BCBSM/BCN Medicare Advantage plans (11.2%). Approximately 5.3% of ED episodes began at critical access hospitals. The yearly number of ED-based episodes identified was approximately 500,000 in 2017, 2018, and 2019, decreasing to below 400,000 annually for the years 2020-2022.

Among the 15 defined conditions, the most common condition identified by primary diagnosis at the ED index event was nonspecific chest pain, comprising 553,756 (20.8%) of ED-based episodes among this cohort, followed closely by 550,124 (20.7%) episodes for abdominal pain (Table 1 and eAppendix Figure 2). eAppendix Figure 3 and eAppendix Table 3 show the distribution of ED episodes by age and sex categories. Certain conditions were more prevalent among younger patients, including asthma, diabetes short-term complications, and abdominal pain, for which more than 80% of patients were younger than 65 years.

The univariate distribution of total episode-of-care payments across all 15 conditions is displayed in Figure 1. Mean total episode payments and the 4 main payment components comprising the total episode payment varied widely across the 6 conditions selected for analysis. Abdominal pain and nonspecific chest pain episodes had the lowest mean (SD) total price-standardized 30-day episode payments ($3262 [$5533] and $3495 [$6106], respectively), whereas CHF episodes had the highest mean aggregate payment ($17,941 [$15,552]) (Table 2). For all 6 conditions, most payments across the episode of care were attributed to the index event, which comprised 32.7% of the total episode payments for abdominal pain episodes and up to 51.5% for CHF episodes.

Substantial variation was seen in health care utilization measures across conditions, particularly in the rate of same-day inpatient admissions, ranging from 1.6% for abdominal pain to 80.9% for CHF (Table 3). Other types of care with wide intercondition variation in 30-day utilization were home health care (ranging from approximately 2.0% for chest pain and abdominal pain episodes to 25.0% for CHF episodes) and skilled nursing facility care (ranging from approximately 1% for abdominal pain and chest pain to 15.4% for CHF). Secondary ED visits without inpatient admission were most prevalent following index ED events for abdominal pain (22.9%).

Comparison of episodes in the lowest and highest quartiles of total episode payments for each condition provides insight on the drivers of variation in episode payments. For abdominal pain ED episodes, episodes in the lowest and highest quartiles differed little on index event payments (15.5% of the total difference) but differed greatly on postindex inpatient hospitalizations (34.9%) and postacute care (29.8%). In contrast, episodes for COPD differed substantially on facility index event payments (43.5% of the total difference), with small variation in postacute care (14.4%) (Figure 2).

DISCUSSION

The objective of this study was to describe the development of ED-based episodes of care based on multipayer administrative claims data in Michigan to support ED-based collaborative quality improvement efforts. With more than 2.6 million episodes of care for 15 common medical conditions, the ED-based episode-of-care data provide a novel opportunity to better understand the quality and value of ED-based care in Michigan. Analysis of these data highlighted substantial variation in 30-day risk- and price-standardized episode-of-care payments and in rates of acute and postacute care utilization across and within conditions, presenting opportunities for quality and value improvement of ED care more broadly.

The episode-of-care framework built from claims data provides rich insights into patient health care journeys beyond any given institution’s electronic health record data. By incorporating records of patients’ index ED events along with all care documented on claims for the following 30 days, the full picture of patients’ utilization and costs can be seen and analyzed in respect to quality, timeliness, and variation across the state. ED episodes-of-care data reveal which ED visit conditions involve the most resources at the initial visit and over the next 30 days; our work to date identifies CHF as an ED condition that should remain at the forefront of ED quality initiatives due to its high rates of readmission and postacute care. The episodes also shed light on the use of ED care over time and suggest a substantial decrease in ED use post 2020 compared with pre-2020 that should be considered when assessing trends over time, as decreased frequency may be associated with greater acuity or severity of ED visits following the COVID-19 pandemic.

Findings highlighted several opportunities for standardization of care within specific ED-based conditions. For conditions such as COPD, for which the largest contribution to within-episode variation was attributed to differences in care received during the index event, attention should be paid to standardizing care pathways for patients while in the ED, including hospitalization practices. In contrast, patients visiting the ED for abdominal pain varied most on inpatient hospitalizations in the 1 to 30 days following their index ED visit, which may suggest opportunities for more robust outpatient care management and coordination upon ED discharge. For all conditions, more standardized care pathways would reduce health care costs and improve care received by the millions of patients visiting US EDs each year.

One important application of this work is to assess hospital variation in treatment and decision-making within the ED, along with the associated impact on post-ED outcomes. Studies have shown large interinstitutional variation in hospital admission rates following ED visits for low-mortality conditions such as chest pain,20 suggesting opportunities for cross-institution collaboration and better consensus on ideal treatment pathways for patients presenting to a given ED. Since 2023, individualized reports containing findings from ED episodes have been distributed to MVC hospital collaborative members to support quality improvement work. A sample hospital-based report containing information on ED-based episode-of-care payments and outcomes can be found in the eAppendix. Reports give providers an opportunity to understand and benchmark ED-based care within their institutions and to facilitate multi-institutional quality improvement efforts to improve ED care in Michigan. Future studies should further investigate drivers of variation between hospitals as a next step in identifying areas of quality improvement opportunity.

Possibilities for future application of this data framework include opportunities to understand ED-based care across vulnerable populations and geographic regions such as differences in the ED continuum of care by patient race and ethnicity or rurality.21,22 The data in this ED-based episodes framework may also shed more light on patients with barriers preventing use of primary care outside the ED who may be more likely to seek similar care in an ED setting,2 including older adults.3 Furthermore, this framework may be used to assess the critical role played by observation units at hospitals, efficiently diverting patients away from inpatient beds or ED space that can be used by higher-need patients.23

Future Directions

Moving forward, the statewide MVC and MEDIC collaboratives are working to incorporate additional conditions of interest into the ED episodes, including conditions relevant to behavioral health and substance use–related ED visits. The 2 collaboratives also continue to expand statewide and tailored reporting and analyses using ED-based episode metrics to support ongoing quality improvement initiatives across Michigan. A goal of this work is for the ED episodes framework to be used as a model to support similar quality improvement work in other states or countries.

Additionally, there is work to be done evaluating ED episodes of care among children. Study findings suggest that the association of postindex care with index ED visits may differ between children and adults.24 Furthermore, a focus on younger adults may be warranted because the demographic of those aged 20 to 44 years accounts for a large proportion of spending on ED care in the US.25

Limitations

There are limitations to consider with this study. First, although claims data present a rich opportunity to examine the breadth of care received regardless of time or location, claims only encompass care billed to an insurance payer. Second, although a large proportion of Michigan residents (approximately 84%) are insured by the payers represented in this study, patients who are uninsured or who have commercial insurance providers other than BCBSM/BCN are not included. Third, claims data are generated for purposes of payment and do not contain potentially important clinical information that may help predict differences in 30-day ED episode-of-care spending or health care utilization. Future work may provide opportunities to link these data with robust clinical data from the MEDIC data registry.

CONCLUSIONS

Episodes of care initialized by ED visits and created from multipayer claims data contain robust information on health care utilization measures following ED visits. This episode framework can be used to gain insight into opportunities for improving patients’ care continuum following visits to the ED.

Author Affiliations: Michigan Value Collaborative (KPH, CAP, MPT), Ann Arbor, MI; Michigan Emergency Department Improvement Collaborative (KEK), Ann Arbor, MI; Department of Emergency Medicine (KEK, AJ), Department of Learning Health Sciences (KEK), and Center for Healthcare Outcomes and Policy (MPT), University of Michigan Medical School, Ann Arbor, MI; US Department of Veterans Affairs (VA) Center for Clinical Management Research, VA Ann Arbor Healthcare System (KEK, AJ), Ann Arbor, MI; Clinician Scholars Program, Institute for Healthcare Policy and Innovation, University of Michigan/VA Ann Arbor Healthcare System (AJ), Ann Arbor, MI; Department of Cardiac Surgery, Michigan Medicine (MPT), Ann Arbor, MI.

Source of Funding:Support for the Michigan Emergency Department Improvement Collaborative and Michigan Value Collaborative is provided by Blue Cross Blue Shield of Michigan (BCBSM) as part of the BCBSM Value Partnerships program; however, the opinions, beliefs, and viewpoints expressed by the authors do not necessarily reflect those of BCBSM or any of its employees.

Author Disclosures: Dr Kocher reports grants received from Blue Cross Blue Shield of Michigan and Blue Care Network to support the Michigan Emergency Department Improvement Collaborative, a quality improvement network. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (KPH, KEK, AJ, CAP, MPT); acquisition of data (CAP); analysis and interpretation of data (KPH, KEK, AJ, MPT); drafting of the manuscript (KPH, MPT); critical revision of the manuscript for important intellectual content (KPH, KEK, AJ, CAP, MPT); statistical analysis (KPH); obtaining funding (MPT); administrative, technical, or logistic support (KPH, CAP, MPT); and supervision (CAP).

Address Correspondence to: Michael P. Thompson, PhD, Department of Cardiac Surgery, Michigan Medicine, 2800 Plymouth Rd, NCRC Building 16, Room 138E, Ann Arbor, MI 48109. Email: mthomps@med.umich.edu.

REFERENCES

1. Sharp AL, Cobb EM, Dresden SM, et al. Understanding the value of emergency care: a framework incorporating stakeholder perspectives. J Emerg Med. 2014;47(3):333-342. doi:10.1016/j.jemermed.2014.04.017

2. Rust G, Ye J, Baltrus P, Daniels E, Adesunloye B, Fryer GE. Practical barriers to timely primary care access: impact on adult use of emergency department services. Arch Intern Med. 2008;168(15):1705-1710. doi:10.1001/archinte.168.15.1705

3. Ionescu-Ittu R, McCusker J, Ciampi A, et al. Continuity of primary care and emergency department utilization among elderly people. CMAJ. 2007;177(11):1362-1368. doi:10.1503/cmaj.061615

4. Galarraga JE, Pines JM. Costs of ED episodes of care in the United States. Am J Emerg Med. 2016;34(3):357-365. doi:10.1016/j.ajem.2015.06.001

5. Hooker EA, Mallow PJ, Oglesby MM. Characteristics and trends of emergency department visits in the United States (2010-2014). J Emerg Med. 2019;56(3):344-351. doi:10.1016/j.jemermed.2018.12.025

6. Lane BH, Mallow PJ, Hooker MB, Hooker E. Trends in United States emergency department visits and associated charges from 2010 to 2016. Am J Emerg Med. 2020;38(8):1576-1581. doi:10.1016/j.ajem.2019.158423

7. Schull MJ, Guttmann A, Leaver CA, et al. Prioritizing performance measurement for emergency department care: consensus on evidence-based quality of care indicators. CJEM. 2011;13(5):300-309. doi:10.2310/8000.2011.110334

8. Ellimoottil C, Syrjamaki JD, Voit B, Guduguntla V, Miller DC, Dupree JM. Validation of a claims-based algorithm to characterize episodes of care. Am J Manag Care. 2017;23(11):e382-e386.

9. Schulman KA, Yabroff KR, Kong J, et al. A claims data approach to defining an episode of care. Health Serv Res. 1999;34(2):603-621.

10. Wingert TD, Kralewski JE, Lindquist TJ, Knutson DJ. Constructing episodes of care from encounter and claims data: some methodological issues. Inquiry. 1995;32(4):430-443.

11. Froimson MI, Rana A, White RE Jr, et al. Bundled Payments for Care Improvement Initiative: the next evolution of payment formulations: AAHKS Bundled Payment Task Force. J Arthroplasty. 2013;28(suppl 8):157-165. doi:10.1016/j.arth.2013.07.012

12. Wiler JL, Beck D, Asplin BR, et al. Episodes of care: is emergency medicine ready? Ann Emerg Med. 2012;59(5):351-357. doi:10.1016/j.annemergmed.2011.08.020

13. Burke LG, Burke RC, Epstein SK, Orav EJ, Jha AK. Trends in costs of care for Medicare beneficiaries treated in the emergency department from 2011 to 2016. JAMA Netw Open. 2020;3(8):e208229. doi:10.1001/jamanetworkopen.2020.8229

14. Kocher KE, Arora R, Bassin BS, et al. Baseline performance of real-world clinical practice within a statewide emergency medicine quality network: the Michigan Emergency Department Improvement Collaborative (MEDIC). Ann Emerg Med. 2020;75(2):192-205. doi:10.1016/j.annemergmed.2019.04.033

15. Healthcare Cost and Utilization Project. Clinical Classifications Software Refined (CCSR). Agency for Healthcare Research and Quality. December 9, 2022. Accessed March 7, 2024. https://hcup-us.ahrq.gov/toolssoftware/ccsr/ccs_refined.jsp

16. Prevention Quality Indicators in Inpatient Settings measures. Agency for Healthcare Research and Quality. Accessed March 7, 2024. https://qualityindicators.ahrq.gov/measures/pqi_resources

17. Ellimoottil C, Ryan AM, Hou H, Dupree JM, Hallstrom B, Miller DC. Implications of the definition of an episode of care used in the Comprehensive Care for Joint Replacement Model. JAMA Surg. 2017;152(1):49-54. doi:10.1001/jamasurg.2016.3098

18. Gottlieb DJ, Zhou W, Song Y, Andrews KG, Skinner JS, Sutherland JM. Prices don’t drive regional Medicare spending variations. Health Aff (Millwood). 2010;29(3):537-543. doi:10.1377/hlthaff.2009.0609

19. Thompson MP, Cain-Nielsen AH, Yost Karslake ML, et al. Hospital performance in a statewide commercial insurer episode-based incentive program. Am J Manag Care. 2023;29(8):e250-e256. doi:10.37765/ajmc.2023.89412

20. Sabbatini AK, Nallamothu BK, Kocher KE. Reducing variation in hospital admissions from the emergency department for low-mortality conditions may produce savings. Health Aff (Millwood). 2014;33(9):1655-1663. doi:10.1377/hlthaff.2013.1318

21. Zhang X, Carabello M, Hill T, Bell SA, Stephenson R, Mahajan P. Trends of racial/ethnic differences in emergency department care outcomes among adults in the United States from 2005 to 2016. Front Med (Lausanne). 2020;7:300. doi:10.3389/fmed.2020.00300

22. Greenwood-Ericksen MB, Kocher K. Trends in emergency department use by rural and urban populations in the United States. JAMA Netw Open. 2019;2(4):e191919. doi:10.1001/jamanetworkopen.2019.1919

23. Baugh CW, Venkatesh AK, Bohan JS. Emergency department observation units: a clinical and financial benefit for hospitals. Health Care Manage Rev. 2011;36(1):28-37. doi:10.1097/HMR.0b013e3181f3c035

24. Sills MR, Macy ML, Kocher KE, Sabbatini AK. Return visit admissions may not indicate quality of emergency department care for children. Acad Emerg Med. 2018;25(3):283-292. doi:10.1111/acem.13324

25. Scott KW, Liu A, Chen C, et al. Healthcare spending in U.S. emergency departments by health condition, 2006-2016. PLoS One. 2021;16(10):e0258182. doi:10.1371/journal.pone.0258182