Refining a Diagnostic Code List to Investigate Low-Acuity Utilization by Veterans

ABSTRACT

Objectives: Recent Veterans Affairs (VA) legislation expanding access to community (non-VA) emergency care use has important implications for budget and policy. Understanding drivers of community emergency department (ED) and urgent care (UC) utilization for low-acuity conditions is crucial for budgeting and improving care continuity, but current tools for identifying low-acuity ED visits have significant limitations. This study aimed to update and make publicly available a dichotomous diagnosis code list to identify low-acuity ED visits. We describe low-acuity community ED/UC visits and their users and assess patient-level characteristics associated with the use of community facilities.

Study Design: Retrospective cross-sectional study.

Methods: We evaluated all International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnoses with at least 100 ED visits across fiscal years 2018-2023. Severity probabilities were calculated and reviewed by 3 study personnel. Summary statistics were used to describe low-acuity community ED/UC visits. A logistic regression model was fitted to determine patient-level factors associated with using community settings for low-acuity care.

Results: For code list refinement, 4694 ICD-10 codes were reviewed, with 2439 (52.0%) rated as low acuity. Visits to community EDs and UCs for low-acuity conditions rose over the study period. Low-acuity diagnoses constituted 35.1% of community ED visits and 83.8% of community UC visits. Being younger, being female, and living farther from a VA ED were associated with greater odds of using community settings for low-acuity care.

Conclusions: Community ED and UC utilization for low-acuity conditions remains below rates in VA settings but is rising, resulting in increased non-VA spending. Targeted interventions may be useful in redirecting low-acuity community ED use to lower-cost settings.

Am J Manag Care. 2026;32(4):e118-e125. https://doi.org/10.37765/ajmc.2026.89924

Takeaway Points

In this article, the authors developed an updated, dichotomous code list for identifying low-acuity emergency department (ED) visits for population-level analyses, health system operations, and policy. When applied to national Veterans Affairs (VA) data, this code list predicted hospital admission and revealed growing use and predictors of non-VA services for low-acuity unscheduled care. Findings can facilitate VA policy and budgeting, identify gaps in VA service access, and inform targeted interventions.

• Low-acuity ED visits result in longer wait times, poorer patient experience, and higher cost for health systems.
• We update and share a diagnostic code list to identify low-acuity ED visits among veterans.
• When applying this list, we found growing use of non-VA services for low-acuity unscheduled care, likely related to recent policy changes.
• Using non-VA services for low-acuity care was associated with patients living farther from VA care sites and being younger and female.
• Diagnostic code lists are inherently limited and best used for population-level analyses to understand trends, not determining the appropriateness of seeking ED care.

Rising emergency department (ED) utilization and overcrowding are growing concerns across health systems worldwide. One driver is low-acuity ED visits, characterized by nonurgent medical concerns or care related to chronic conditions, which lead to poorer clinical outcomes, longer wait times, and increased costs.^1-4 To optimize ED services for those who will benefit most, many health systems, including the Veterans Affairs (VA) health care system, have sought to guide low-acuity visits to alternative care sites.

Efforts to address rising ED utilization within the VA have been shaped by recent legislative changes that expanded veterans’ access to non-VA care (ie, community care).⁵ These changes have redefined the VA’s role from health care provider to purchaser of health care services, with notable financial implications. Spending on VA community care rose from $7.9 billion in 2013 to $23.4 billion in 2022, approximately 25% of the VA’s budget.^6,7 National leaders have raised concerns about this rapid spending growth.

ED and unscheduled care account for more than 30% of VA community care spending.⁸ This has prompted VA policy efforts to divert low-acuity community ED visits to alternative settings.⁹ These include VA primary care, VA EDs and urgent care (UC) facilities, VA telehealth programs, and lower-cost community UC facilities, which became part of a new benefit program in June 2019. Attempts by other systems to reduce low-acuity ED visits to optimize health care savings and care coordination have yielded mixed results.^10,11

To inform these programs, VA leaders have requested a dichotomous categorization (low acuity vs emergent) using standard structured data to track population-level trends and inform targeted interventions. A dichotomous approach simplifies analysis and enables scalable estimates and predictive models. In contrast, current probabilistic approaches assign probabilities across multiple categories, with more ambiguous interpretation, especially for operations or policy. Existing approaches are also limited by difficult-to-access health record data, excluded visit types, or access fees.^12-14

Previous Approaches for Classifying ED Visits

The New York University ED Algorithm (NYU-EDA) is the most widely used tool for retrospectively assessing ED visits as urgent, preventable, or optimally treated in an ED. This publicly available tool uses probabilistic algorithms and categorizes common primary ED discharge diagnoses into 4 overlapping categories: (1) nonemergent; (2) emergent, primary care treatable; (3) emergent, ED care needed, preventable with timely ambulatory care; and (4) emergent, ED care needed, not preventable. Rather than assigning a diagnosis to a single category, the algorithm provides a set of probabilities across categories. It also classifies certain diagnoses into separate, uncategorized groups such as injuries, psychiatric conditions, alcohol/drug related, or unclassified.¹⁵

Since the original algorithm’s publication, there have been multiple updates. Ballard et al aggregated the NYU-EDA probabilities for nonemergent and emergent, primary care–treatable visits and compared this with the total probability of the categories for which emergent ED care is needed.¹⁶ Depending on which was greater, visits were classified as nonemergent, emergent, or intermediate if these probabilities were equal.¹⁶ This version maintained carve-outs for diagnoses related to injuries, psychiatric conditions, and substance use.¹⁶ Johnston et al provided a patch by applying the original probabilistic weights to new codes.¹⁷ The most recent algorithm update, led by researchers at Johns Hopkins University (NYU/JHU-EDA), assigned each primary discharge diagnosis to 1 of 11 categories.¹⁸ Updated assignments were based on the category with the highest probability, resolving cases of equal probability in favor of the highest level of urgency.¹⁸ Despite the advancements, limitations remain. First, this updated algorithm is not publicly available. Second, with 11 categories, it does not provide a simple dichotomous classification, making it more challenging to separate low-acuity ED visits or understand how they impact health care systems and resource allocation.

Given these considerations, this study aims to bridge the knowledge gap regarding low-acuity ED visits among veterans by updating and making available a low-acuity code list for identification. Additionally, using nationally representative data, this study seeks to (1) describe the frequency and characteristics of low-acuity community ED and UC visits and (2) assess individual characteristics associated with the use of community vs VA facilities for low-acuity care. In doing so, we endeavor to facilitate decisions and policies related to the complexities of providing and purchasing emergency care.

METHODS

Defining Low-Acuity ED Visits

Developing the VA low-acuity code list. To create our dichotomous low-acuity code list, we employed a systematic methodology. First, we created a list of International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnoses that were associated with at least 100 VA and/or community ED visits across fiscal years (FYs) 2018-2023.

For these diagnoses, we applied the method used by Ballard et al and others, aggregating the probabilistic weights from the NYU-EDA to determine the likelihood that a given diagnosis was emergent.^16,19 Diagnoses with at least 50% probability were classified as emergent; those with less than 50% were classified as low acuity. All classified codes were then reviewed by a 3-person expert panel of health services researchers—2 emergency medicine physicians (A.A.V., J.S.) and 1 paramedic (Jeffrey Rollman, MPH)—to confirm appropriateness of categorization. This panel reviewed all codes for validity using prespecified clinical criteria. Emergent conditions were those generally requiring medical attention within 12 hours and/or resources typically found in EDs. Low-acuity conditions were classified as generally not requiring medical attention within 12 hours or typically manageable in UC or primary care settings. This dichotomous classification system was designed to differentiate between cases that genuinely require immediate, comprehensive ED care and those that could potentially be managed elsewhere.

Next, for the remaining ICD-10-CM codes that were not previously classified with NYU-EDA assigned probabilities, including previously excluded diagnosis codes related to injury, drug, alcohol, and psychiatric reasons, the panel conducted a separate, independent review. Panel members were instructed to categorize these conditions as emergent or low-acuity independently using the criteria outlined previously (see eAppendix 1 for more detail and examples [eAppendices available at ajmc.com]). Discrepant codes were discussed regularly until unanimous (3/3) consensus was reached. The final code list (eAppendix 2) was independently reviewed and formally endorsed by the VA National Emergency Medicine Office.

Analysis of Low-Acuity Community Care Utilization

Design, data, and measures. Analysis was performed using VA outpatient encounter data from the VA Corporate Data Warehouse and community care claims data from the Office of Integrated Veteran Care. For descriptive analysis, the study sample included community ED and UC visits for veterans 18 years and older from October 1, 2017, to September 30, 2023. Community UC visits were included from the start of the community UC benefit program, June 6, 2019, to September 30, 2023. For regression analysis evaluating community vs VA utilization for low-acuity conditions, VA ED visits from October 1, 2017, to September 30, 2023, were included. Visits outside the 50 US states or missing a low-acuity diagnosis category were excluded. Community ED visit dates, primary ICD-10-CM diagnosis codes (organized into categories using Agency for Healthcare Research and Quality Clinical Classifications Software), total payments, and visit disposition (discharged, admitted, or died) were derived from administrative claims.²⁰

Data were extracted to describe age, race, ethnicity, sex, service-connected disability, VA priority group, housing status, rurality, US region, Elixhauser Comorbidity Index score, and Area Deprivation Index (ADI) percentile.^21-23 Access-to-care variables included driving time to VA primary care and differential distance between nearest VA and non-VA EDs. For more variable specification details, see eAppendix 3. This study was approved by the Stanford University Institutional Review Board (protocol No. 30780).

Statistical analyses. The count and proportion of low-acuity and emergent visits were calculated for community UCs and EDs. We compared proportions of low-acuity and emergent visits resulting in admission vs discharge from the ED. Standardized mean differences (SMDs) were used to compare population characteristics between low-acuity and emergent visits. The most frequent and most costly low-acuity conditions (determined by multiplying median cost by visit frequency) were identified.

To determine factors associated with community vs VA ED utilization for low-acuity conditions, a logistic regression model was fitted. This analysis was restricted to low-acuity visits, defined as visits with a primary ICD-10-CM diagnosis labeled “low acuity” and resulting in discharge. The primary outcome was visit location (community ED/UC vs VA ED). Predictor variable selection was guided by the Andersen health care utilization model, which describes 3 key dynamics: (1) predisposing factors, (2) individual and contextual enabling factors, and (3) need for care.²⁴ Predisposing factors included age, sex, and race/ethnicity. Enabling factors included driving time to VA primary care and differential distance between VA and non-VA EDs, rurality, housing stability, visit timing (weekend/holiday vs weekday), VA service connection, VA priority group, and ADI percentile. Need factors (clinical and perceived) included Elixhauser Comorbidity Index score and frequent ED utilizer status (defined as > 3 ED visits in any year of the study). Data were assessed for collinearity, distribution, cell size, missingness, and balance using an SMD cutoff of 0.5. All analyses were conducted using R 3.6.1 (R Foundation for Statistical Computing).²⁵ 

RESULTS

Using our refined methodology, we rated 4702 diagnosis codes. Of these, we classified 2439 (52.0%) codes as low-acuity and 2255 (48.0%) codes as emergent, and we recommended that 8 codes be excluded (eg, Z53.21: procedure and treatment not carried out due to patient leaving prior to being seen by health care provider). The full code list is provided in eAppendix 2.

Low-Acuity vs Emergent Community ED and UC visits

In total, 5,439,090 community ED and 1,681,448 community UC visits were made by 2,464,787 veterans from FYs 2018-2023. Encounter-level characteristics are described in Table 1 [part A and part B]. Most visits were by veterans who were male, White, and resided in urban areas. Modest differences between low-acuity and emergent community encounters were observed across several variables including age, sex, and medical comorbidities (Table 1).

Among the community ED visits, 35.1% were classified as low acuity, 61.5% as emergent, 0.1% were excluded diagnoses, and 3.4% remained unclassified. For community UC visits, 83.8% were low acuity, 15.2% were emergent, 0.01% were excluded from rating, and 1.0% were unclassified. The Figure shows monthly counts of low-acuity visits in community settings, including a sharp decline in visits from February 2020 to May 2020 (corresponding to the initial COVID-19 pandemic emergency declaration), followed by a rebound beginning in June 2020. A second, smaller dip occurred in December 2021, consistent with the Omicron variant surge. Despite these transient decreases, the overall trend in community ED and UC utilization was steady growth over the study period. Despite an overall rise in low-acuity utilization across community settings during the study period, the proportion of low-acuity visits remained stable. ED disposition analysis found that 7.1% and 33.1% of low-acuity and emergent ED visits resulted in admission, respectively. Death was rare in both groups, at 0.4% of emergent visits and 0.02% of low-acuity visits. The OR of requiring hospital admission for emergent visits vs low-acuity visits was 6.47 (95% CI, 6.44-6.57).

Common and Costly Low-Acuity Community Visits

The most common low-acuity conditions seen in community EDs were musculoskeletal pain, superficial contusions, and skin infections, whereas the most common in community UCs were screening for infectious diseases, upper respiratory infections, and musculoskeletal pain. The low-acuity conditions with the highest median payments are detailed in Table 2.

Associations of Low-Acuity Visits in Community vs VA EDs

Individual associations with low-acuity community ED utilization among veterans are summarized in Table 3. Low-acuity visits accounted for 54.9% of visits to VA EDs. Among low-acuity encounters, the factor most strongly associated with utilizing community settings (vs VA) was differential distance between VA and non-VA EDs, with a dose-response effect. Having a VA ED 31 to 60 miles farther than a non-VA ED (vs being closer) had an adjusted OR of 15.26 (95% CI, 15.10-15.42). Having a VA ED more than 60 miles farther than a non-VA ED had an OR of 31.75 (95% CI, 31.51-32.01). Other factors included being female (OR, 1.39; 95% CI, 1.37-1.39), younger (aged 18-45 vs > 65 years: OR, 1.20; 95% CI, 1.20-1.21), and seeking care on a weekend/holiday (OR, 1.03; 95% CI, 1.03-1.04). Remaining in the VA for low-acuity care was associated with being a frequent ED utilizer (OR, 0.85; 95% CI, 0.83-0.86), being non-Hispanic Black (OR, 0.55; 95% CI, 0.54-0.56), and being from a highly deprived neighborhood (OR, 0.94; 95% CI, 0.94-0.95).

DISCUSSION

In this study, we developed the first public, comprehensive dichotomous code list to distinguish low-acuity from emergent ED visits, offering a valuable tool for evaluating acute care utilization. Our updates address literature gaps and provide an open-access resource for operations, research, and population health management. In our sample, we found a significant association between emergent visit categorization and hospitalization with an OR of 6.47 (95% CI, 6.44-6.57), similar to but slightly stronger than in previously published diagnosis-based algorithms such as Ballard et al (OR, 3.37; 95% CI, 3.31-3.44)¹⁶ and NYU-EDA (OR, 5.28; 95% CI, 4.93-5.66),²⁶ indicating strong performance at identifying conditions requiring the higher-level needs of an inpatient hospital admission. Applying this list allowed analysis of policy impacts and utilization trends and identified patient-level associations with low-acuity utilization, offering insights for reducing costs, enhancing care coordination, and guiding patients to appropriate in-network resources.

Applying our code list indicates that low-acuity community ED utilization has increased over time, likely reflecting expanded access to community services following the VA MISSION (Maintaining Internal Systems and Strengthening Integrated Outside Networks) Act of 2018. These findings are consistent with prior studies documenting substantial rises in overall community ED utilization.²⁷ The proportion of low-acuity visits in community EDs remained well below that of VA EDs (35% vs 55%) and stable, suggesting that veterans are not preferentially using community EDs for low-acuity care. Nevertheless, it reveals an important trend with implications for VA budget and policy.²⁸

Since introducing the community UC benefit in June 2019, low-acuity visits to community UCs have steadily increased, nearing the volume of community ED visits, building on findings from the first 9 months of the program.⁸ Expanding community UC access can facilitate care shifting from community EDs to community UCs, particularly for patients living farther from VA facilities. This shift could improve patient satisfaction and wait times while reducing non-VA spending, particularly for common low-acuity conditions such as musculoskeletal pain or urinary tract infection (UTI).²⁹ For example, UTI visits cost $125 at community UCs vs $682 at community EDs. Shifting such visits could reduce UTI-related community care spending by 81.7%. Although the number of UTI visits that could safely be shifted away from EDs is unknown, this example underscores the cost-saving potential of the UC benefit for low-acuity care.

We identified several factors that influence veterans’ decisions to use community settings for low-acuity conditions instead of VA facilities, including being younger, female, and seeking care on a weekend/holiday. The strongest association with community utilization was differential distance from the patient to VA and non-VA EDs, suggesting that community options may improve access for veterans living farther from VA facilities, especially because rural patients were more likely to rely on community settings for low-acuity care.³⁰ This reflects established patterns in nonveteran populations, where distance from primary care sites and younger age have been associated with greater ED reliance for low-acuity care, with ED utilization lowering modestly by facilitating primary care utilization.^1,31,32 The strength of this association suggests that more focus must be paid to the unscheduled care needs of veterans living more than 30 miles from VA facilities. Targeted education around VA telehealth programs may help retain low-acuity care within the VA.^33,34 Female veterans were more likely to seek community sites for low-acuity conditions, perhaps because they are often referred for gender-specific community services.^35,36 Because women represent the fastest-growing group of veterans, it is crucial to facilitate care for female veterans within the VA, especially for unscheduled low-acuity visits.

Factors associated with remaining within the VA for low-acuity care included living in highly deprived areas and experiencing homelessness. Veterans experiencing homelessness have higher ED utilization than housed veterans, partly for unmet psychosocial needs.²⁹ Veterans utilizing EDs for low-acuity psychosocial concerns are more likely to find these resources at VA EDs, which offer more targeted housing support and care coordination than community settings.^37,38 This is the first study to demonstrate an association between ADI percentile and low-acuity ED utilization among veterans, although similar patterns have been observed in other populations.³⁹ Previously demonstrated barriers to community emergency care access, such as a lack of information about policy changes, eligibility requirements, and concerns about medical bills, are likely more pronounced among veterans with unstable housing or those from high-deprivation areas.⁴⁰ Additional outreach may be necessary to ensure these patients are aware of their eligibility for community resources.

Limitations and Strengths

Although 2 decades of evidence have supported the NYU-EDA, it is essential to recognize the limitations of relying on discharge diagnoses for acuity characterization. Studies have shown associations between emergent visits and factors such as likelihood of death and hospitalization.^16,18,26 However, it is imperative to heed the warnings of researchers and emergency clinicians against using visit classifications as the sole basis for interventions aimed at reducing “unnecessary” visits or denying payment.^41,42 No classification tool based on diagnosis codes fully captures the clinical complexity or appropriateness of a given ED visit. We found a small proportion (~7%) of visits classified as low acuity that resulted in hospital admission, underscoring this limitation. Our intent was not to assess individual clinical appropriateness but to provide population-level insights into unscheduled care trends, aligned with the tool’s intended purpose. Several factors contribute to ED utilization, including symptom severity, comorbidity, and access to care, which are not captured by our algorithm.^43-45 Code-based classifications rely on the coding specificity, which can vary between providers.⁴⁶ Also, individual patient circumstances may warrant ED utilization despite a low-acuity discharge diagnosis; for instance, patients with significant abdominal pain may appropriately seek ED evaluation, only to have a UTI identified as the cause. Therefore, although ED visit classifications are tools for understanding population health care patterns and conducting retrospective evaluations as performed here, they should not be conflated with the medical needs of individual patients.

Conversely, our study has several strengths, including creating an updated, comprehensive, open-access code list for evaluating low-acuity emergency care and providing the first national analysis of low-acuity unscheduled care by veterans in community settings. We demonstrate how this tool can be used to understand population-level utilization patterns, identify patients for targeted outreach, and optimize resources. Our large sample size contributes to statistically significant associations for all variables examined; we therefore focus discussion on larger effect sizes with highest clinical and policy relevance. However, there are important limitations to consider. This analysis did not include VA UC visits, another possible site for low-acuity care. VA UCs represent a small share of overall visits compared with VA EDs, accounting for only 4.8% of visits in FY 2022, so their exclusion is unlikely to substantially affect findings. Finally, disposition data for UC visits were assumed to result in discharge. Literature suggests that only 2% to 3% of UC visits are transferred for additional care, supporting this assumption.⁴⁷

CONCLUSIONS

As the VA considers policy changes that could reshape veterans’ care-seeking patterns, the ability to accurately capture, track, and understand drivers of low-acuity acute care utilization will be crucial. Our novel code list and comprehensive analysis of low-acuity community utilization can strengthen the VA’s understanding of patient preferences and care needs. By leveraging these insights, the VA can work toward ensuring that all veterans have equitable and affordable access to appropriate care. 

Acknowledgments

The authors thank Mr Jeffrey Rollman for his help in classifying ICD-10-CM diagnosis codes for this project and Ms Tracy Urech for her work on project coordination and manuscript feedback.

Author Affiliations: Center for Innovation to Implementation, Veterans Affairs (VA) Palo Alto Health Care System (AR, SMA, AAV), Menlo Park, CA; Department of Health Policy, School of Medicine, Stanford University (AR), Stanford, CA; Center of Innovation to Accelerate Discovery and Practice Transformation, Durham VA Health Care System (JS), Durham, NC; Department of Emergency Medicine, Duke University School of Medicine (JS), Durham, NC; Quantitative Sciences Unit (DB) and Division of Primary Care and Population Health (SMA), Stanford University Department of Medicine, Stanford, CA; Department of Emergency Medicine, University of California, San Francisco (AAV), San Francisco, CA; Department of Emergency Medicine, Stanford University (AAV), Stanford, CA.

Source of Funding: Funding was provided by the US Department of Veterans Affairs Health Services and Research Development Service and the Office of Integrated Veteran Care.

Author Disclosures: Dr Asch attended a June 2025 meeting sponsored by Novo Nordisk, which supported his travel and time. Dr Vashi receives funding from the VA Office of Research and Development, the VA Office of Rural Health and the VA Office of Connected Care. 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 (AR, JS, SMA, AAV); analysis and interpretation of data (AR, JS, DB, SMA, AAV); drafting of the manuscript (AR, JS, SMA, AAV); critical revision of the manuscript for important intellectual content (DB); statistical analysis (AR, JS, DB, SMA, AAV); obtaining funding (AAV); administrative, technical, or logistic support (AR, JS, SMA, AAV); and supervision (AAV).

Address Correspondence to: Anu Ramachandran, MD, MPH, Stanford University, 1140 Rhode Island St, San Francisco, CA 94107. Email: anu12@stanford.edu.

REFERENCES

1. Weinick RM, Burns RM, Mehrotra A. Many emergency department visits could be managed at urgent care centers and retail clinics. Health Aff (Millwood). 2010;29(9):1630-1636. doi:10.1377/hlthaff.2009.0748

2. Luo D, Bayati M, Plambeck EL, Aratow M. Low-acuity patients delay high-acuity patients in the emergency department. SSRN. January 5, 2018. Revised October 28, 2025. doi:10.2139/ssrn.3095039

3. Uscher-Pines L, Pines J, Kellermann A, Gillen E, Mehrotra A. Emergency department visits for nonurgent conditions: systematic literature review. Am J Manag Care. 2013;19(1):47-59.

4. Schull MJ, Kiss A, Szalai JP. The effect of low-complexity patients on emergency department waiting times. Ann Emerg Med. 2007;49(3):257-264.e1. doi:10.1016/j.annemergmed.2006.06.027

5. VA MISSION Act of 2018, S 2372, 115th Cong, 2nd Sess (2018). Accessed September 12, 2024. https://www.congress.gov/bill/115th-congress/senate-bill/2372/text

6. The Veterans Community Care Program: background and early effects. Congressional Budget Office. October 2021. Accessed August 4, 2024. https://www.cbo.gov/publication/57583

7. Panangala SV, Sussman JS. Department of Veterans Affairs FY2023 appropriations. Congressional Research Service. February 14, 2023. Accessed March 12, 2026. https://www.congress.gov/crs-product/R47423

8. Vashi AA, Urech T, Wu S, et al. Community urgent care use following implementation of the Veterans Affairs Maintaining Internal Systems and Strengthening Integrated Outside Networks Act. Med Care. 2021;59(suppl 3):S314-S321. doi:10.1097/MLR.0000000000001549

9. Greenstone CL, Peppiatt J, Cunningham K, et al. Standardizing care coordination within the Department of Veterans Affairs. J Gen Intern Med. 2019;34(suppl 1):4-6. doi:10.1007/s11606-019-04997-6

10. Kirkland SW, Soleimani A, Rowe BH, Newton AS. A systematic review examining the impact of redirecting low-acuity patients seeking emergency department care: is the juice worth the squeeze? Emerg Med J. 2019;36(2):97-106. doi:10.1136/emermed-2017-207045

11. Early S, Gordon S. Privatization warning. The American Prospect. April 11, 2024. Accessed September 14, 2024. https://prospect.org/health/2024-04-11-privatization-warning-veterans-affairs/

12. Chen AT, Muralidharan M, Friedman AB. Algorithms identifying low-acuity emergency department visits: a review and validation study. Health Serv Res. 2022;57(4):979-989. doi:10.1111/1475-6773.14011

13. Slagman A, Fischer-Rosinský A, Legg D, Schmieder K, Möckel M. Identification of low-acuity attendances in routine clinical information documented in German emergency departments. BMC Emerg Med. 2023;23(1):64. doi:10.1186/s12873-023-00838-2

14. Mistry RD, Brousseau DC, Alessandrini EA. Urgency classification methods for emergency department visits: do they measure up? Pediatr Emerg Care. 2008;24(12):870-874. doi:10.1097/PEC.0b013e31818fa79d

15. Feldman J. The NYU classification system for ED visits: WSHA technical concerns. Washington State Hospital Association. August 2010. October 12, 2024. https://www.nispcnet.org/files/169/NYU_Classification_System_for_EDVisits.pdf

16. Ballard DW, Price M, Fung V, et al. Validation of an algorithm for categorizing the severity of hospital emergency department visits. Med Care. 2010;48(1):58-63. doi:10.1097/MLR.0b013e3181bd49ad

17. Johnston KJ, Allen L, Melanson TA, Pitts SR. A “patch” to the NYU emergency department visit algorithm. Health Serv Res. 2017;52(4):1264-1276. doi:10.1111/1475-6773.12638

18. Lemke KW, Pham K, Ravert DM, Weiner JP. A revised classification algorithm for assessing emergency department visit severity of populations. Am J Manag Care. 2020;26(3):119-125. doi:10.37765/ajmc.2020.42636

19. Yu J, Hammond G, Waken RJ, Fox D, Joynt Maddox KE. Changes in non-COVID-19 emergency department visits by acuity and insurance status during the COVID-19 pandemic. Health Aff (Millwood). 2021;40(6):896-903. doi:10.1377/hlthaff.2020.02464

20. Clinical Classifications Software (CCS) for ICD-9-CM. Healthcare Cost and Utilization Project. Updated September 10, 2025. Accessed September 15, 2025. https://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp

21. Kind AJH, Buckingham WR. Making neighborhood-disadvantage metrics accessible — the Neighborhood Atlas. N Engl J Med. 2018;378(26):2456-2458. doi:10.1056/NEJMp1802313

22. Kind AJH, Jencks S, Brock J, et al. Neighborhood socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study. Ann Intern Med. 2014;161(11):765-774. doi:10.7326/M13-2946

23. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. doi:10.1097/00005650-199801000-00004

24. Andersen R, Newman JF. Societal and individual determinants of medical care utilization in the United States. Milbank Q. 2005;83(4). doi:10.1111/j.1468-0009.2005.00428.x

25. Wickham H, Averick M, Bryan J, et al. Welcome to the tidyverse. J Open Source Softw. 2019;4(43):1686. doi:10.21105/joss.01686

26. Gandhi SO, Sabik L. Emergency department visit classification using the NYU algorithm. Am J Manag Care. 2014;20(4):315-320.

27. Vashi AA, Urech T, Wu S, Tran LD. Community emergency care use by veterans in an era of expanding choice. JAMA Netw Open. 2024;7(3):e241626. doi:10.1001/jamanetworkopen.2024.1626

28. Axon RN, Gebregziabher M, Everett CJ, Heidenreich P, Hunt KJ. Dual health care system use is associated with higher rates of hospitalization and hospital readmission among veterans with heart failure. Am Heart J. 2016;174:157-163. doi:10.1016/j.ahj.2015.09.023

29. Ekwall A, Gerdtz M, Manias E. The influence of patient acuity on satisfaction with emergency care: perspectives of family, friends and carers. J Clin Nurs. 2008;17(6):800-809. doi:10.1111/j.1365-2702.2007.02052.x

30. Ohl ME, Carrell M, Thurman A, et al. Availability of healthcare providers for rural veterans eligible for purchased care under the Veterans Choice Act. BMC Health Serv Res. 2018;18(1):315. doi:10.1186/s12913-018-3108-8

31. Fishman J, McLafferty S, Galanter W. Does spatial access to primary care affect emergency department utilization for nonemergent conditions? Health Serv Res. 2018;53(1):489-508. doi:10.1111/1475-6773.12617

32. Bradley CJ, Neumark D, Walker LS. The effect of primary care visits on health care utilization: findings from a randomized controlled trial. National Bureau of Economic Research working paper 24100. December 2017. Revised July 2018. Accessed December 4, 2024. https://www.nber.org/papers/w24100

33. Lum HD, Nearing K, Pimentel CB, Levy CR, Hung WW. Anywhere to anywhere: use of telehealth to increase health care access for older, rural veterans. Public Policy Aging Rep. 2020;30(1):12-18. doi:10.1093/ppar/prz030

34. Wray CM, Junge M, Keyhani S, Smith JE. Assessment of a multi-center tele-urgent care program to decrease emergency department referral rates in the Veterans Health Administration. J Telemed Telecare. 2023;29(10):749-754. doi:10.1177/1357633X211024843

35. Marshall V, Stryczek KC, Haverhals L, et al. The focus they deserve: improving women veterans’ health care access. Womens Health Issues. 2021;31(4):399-407. doi:10.1016/j.whi.2020.12.011

36. Dyer KE, Potter SJ, Hamilton AB, et al. Gender differences in veterans’ perceptions of harassment on Veterans Health Administration grounds. Womens Health Issues. 2019;29(suppl 1):S83-S93. doi:10.1016/j.whi.2019.04.016

37. Tsai J, Rosenheck RA. Risk factors for ED use among homeless veterans. Am J Emerg Med. 2013;31(5):855-858. doi:10.1016/j.ajem.2013.02.046

38. Tsai J, ed. Homelessness Among U.S. Veterans: Critical Perspectives. Oxford University Press; 2018.

39. Carlson LC, Kim J, Samuels-Kalow ME, et al. Comparing neighborhood-based indices of socioeconomic risk factors and potentially preventable emergency department utilization. Am J Emerg Med. 2021;44:213-219. doi:10.1016/j.ajem.2020.03.035

40. Nevedal AL, Wong EP, Urech TH, Peppiatt JL, Sorie MR, Vashi AA. Veterans’ experiences with accessing community emergency care. Mil Med. 2023;188(1-2):e58-e64. doi:10.1093/milmed/usab196

41. Raven MC, Lowe RA, Maselli J, Hsia RY. Comparison of presenting complaint vs discharge diagnosis for identifying “nonemergency” emergency department visits. JAMA. 2013;309(11):1145-1153. doi:10.1001/jama.2013.1948

42. Kellermann AL, Weinick RM. Emergency departments, Medicaid costs, and access to primary care — understanding the link. N Engl J Med. 2012;366(23):2141-2143. doi:10.1056/NEJMp1203247

43. Billings J, Parikh N, Mijanovich T. Emergency department use in New York City: a substitute for primary care? Issue Brief (Commonw Fund). 2000;(433):1-5.

44. O’Brien GM, Shapiro MJ, Woolard RW, O’Sullivan PS, Stein MD. “Inappropriate” emergency department use: a comparison of three methodologies for identification. Acad Emerg Med. 1996;3(3):252-257. doi:10.1111/j.1553-2712.1996.tb03429.x

45. Young GP, Lowe RA. Adverse outcomes of managed care gatekeeping. Acad Emerg Med. 1997;4(12):1129-1136. doi:10.1111/j.1553-2712.1997.tb03695.x

46. Peng M, Eastwood C, Boxill A, et al. Coding reliability and agreement of International Classification of Disease, 10th revision (ICD-10) codes in emergency department data. Int J Popul Data Sci. 2018;3(1):445. doi:10.23889/ijpds.v3i1.445

47. Urgent care data. Urgent Care Association of America. Accessed December 4, 2024. https://urgentcareassociation.org/about/urgent-care-data/