Abstract
Objective:
To identify disease processes in which patients experienced the greatest rates of emergency department (ED) revisit disparities by race, ethnicity, language, insurance, and Childhood Opportunity Index.
Methods:
We performed a retrospective, cross-sectional study of ED visits at a comprehensive pediatric health system encompassing three EDs. We included all pediatric (0 to 18 y) ED encounters that led to a discharge home (2018 to 2022). The primary outcome was ED revisit within 7 days of the index encounter. We performed multivariable logistic regression to assess the relationship between diagnoses, patient demographics, and 7-day ED revisit.
Results:
There were 1,008,651 total ED encounters included [53% (n = 529,960) were male]. Of all encounters, 52,176 (5.2%) had a 7-day ED revisit within the system. Patients who identified as black/African American (adjusted odds ratio [aOR]: 1.14, 95% CI: 1.11–1.17), multiracial (aOR: 1.14, 95% CI: 1.06–1.22), Hispanic (aOR: 1.21, 95% CI: 1.17–1.26), and those with public insurance (aOR: 1.19, 95% CI: 1.16–1.23) had greater odds of ED revisit. Disparities in ED revisits among Hispanic patients were concentrated in a few high-volume diagnoses, particularly within ear, nose, throat (ENT)/dental/mouth diseases (aOR: 1.11, 95% CI: 1.01–1.21) such as upper respiratory infections and infectious ear diseases. Conversely, disparities in ED revisits for black/African American patients were widespread across a broader range of diagnoses.
Conclusions:
Disease-focused interventions are needed to reduce disparities in ED return visits. Focusing on upstream determinants of health may better reduce the burden of ED revisits for minoritized populations at risk of ED revisits.
Keywords: health disparities, ED revisits, determinants of health
Approximately 3% to 5% of pediatric emergency department (ED) encounters lead to return visits within 72 hours to 7 days.1–3 Many ED return visits may be due to the progression of illness, while a portion may occur due to inadequate clinical care during the initial encounter, inadequate follow-up, or barriers to accessing outpatient care.4 ED return visits ultimately contribute to family frustration and present a financial and time burden to families and health care systems.2,4,5
Results from prior studies suggest there are greater ED return visit rates for children who are publicly insured,2,6 children whose families have preferred languages other than English,7–9 and for those who live in neighborhoods with lower Childhood Opportunity Indices (COIs), a composite indicator of neighborhood-level resources.10 Disparities in ED return visit rates also exist among minoritized racial or ethnic groups for specific disease processes such as asthma, mental health, fractures, pneumonia, migraines, and gastroenteritis.11–17 However, an assessment of ED return visit variability across disease processes is lacking. Understanding, if ED return visit disparities are greatest among certain disease processes, could inform more targeted interventions to reduce specific inequities.
To this end, our primary objective was to assess (1) factors associated with ED return visits and (2) variability in ED revisit rates across disease processes by patient race, ethnicity, preferred language, insurance, and COI.
METHODS
Study Design
We performed a retrospective, cross-sectional study of pediatric ED visits at a pediatric health system. We evaluated encounters over a 5-year period (January 2018 to December 2022). This study was approved by our hospital’s Institutional Review Board.
Study Setting
The study was performed at a large, comprehensive pediatric health system encompassing three EDs in the Atlanta, Georgia, metro region, and capturing all representative demographic groups in the area. The system has over 250,000 annual ED encounters and includes a tertiary care facility and the only level 1 pediatric trauma center in the state. It also involves a mixture of academic hospitals and a private hospital model.
Study Population
We included all ED encounters for pediatric patients aged 0 to 18 years that resulted in patient discharge to home. We excluded those resulting in transfer to another facility at the index ED visit, as we felt these indicated a different patient illness severity than those discharged at the index visit. Index visit was defined as an ED encounter without a preceding ED encounter in the previous 7 days. If a patient had more than one revisit within the 7 days after an index ED visit, we only considered the first ED revisit in this analysis.
Variables and Data Sources
The primary outcome was ED revisit within 7 days of a prior ED visit. A 7-day window was used because results from prior studies suggest that it captures a more substantial share of ED return visits most likely to be related to the index visit than the traditionally used 48 and 72-hour windows or other > 7-day windows.1,3,18 Furthermore, there is evidence in readmission literature that a 0 to 7-day window captures follow-up hospital encounters with different markers of acute illness burden and may be more preventable than hospital encounters after 7 days.19,20 We utilized the electronic health record as the primary data source. Patient-level demographic variables included age, sex, race, ethnicity, preferred language, financial payor status, and zip-code level COI. Patient race and ethnicity were determined by the patient or family and categorized according to the Office of Management and Budget standards, which included a separate question for ethnicity and race where individuals can select multiple racial categories from the following: American Indian or Alaska Native, Asian, black or African American, Native Hawaiian or other Pacific Islander, and white.20 Patients were categorized as Multiracial if they selected more than one racial category. Registration recorded the preferred languages as determined by patient or family, and for our analysis, these were classified as English, Spanish, or other preferred language. Our rationale for this decision was based on the existence of in-person Spanish interpreter services and system-level Spanish discharge instructions, although we recognize that the grouping of non-English and non-Spanish languages may obscure other potential disparities. Financial payor status was categorized as public (eg, Medicaid), private, and uninsured. COI was derived from census tract data if available from patient’s reported home address as this provides a more granular level of COI data, otherwise it was obtained from patient zip codes which provides an aggregated estimate of several census tracts.21
We also collected clinical data on patients’ number of complex chronic conditions as described by Feudtner et al.22 We determined whether patients had an identifiable primary care provider through a review of the electronic health record in which documentation of the name and contact information for primary care providers is entered to facilitate follow-up. Encounter level variables included the Emergency Severity Index (ESI) score assigned by a triage nurse that classifies presenting patient acuity from 1 to 5, with 1 to 2 generally classified as high acuity and 3 to 5 as medium to low acuity.23 The primary encounter diagnoses, as selected by the clinical providers, were obtained according to International Classification of Diseases, Tenth Revision (ICD-10) classifications. The diagnoses were further categorized according to the Pediatric Emergency Care Applied Research Network (PECARN) Diagnosis Grouping System,24 a widely adopted clinical classification scheme for pediatric diagnoses seen in EDs that includes 21 clinically relevant major diagnosis groups and 77 subgroups based on body system or body region involved.
Statistical Analyses
We assessed the relationship between patient demographics (race, ethnicity, preferred language, insurance, and COI), primary diagnoses, and the binary outcome of 7-day ED revisit. These demographic variables were chosen based on previously cited literature suggesting their associations with ED revisit disparities.2,7,10,12,16 Continuous and categorical variables were summarized using means with SDs and counts with percentages, respectively. We tested the association between ED revisits and a priori-determined candidate variables by univariate logistic regression. Variables that had P < 0.05 in the univariate screen were included in a multivariable logistic regression model. We adjusted for patient race, ethnicity, preferred language, insurance, number of complex chronic conditions, and ESI score. We also conducted multivariable logistic regression moderator analyses by the PECARN major diagnosis groupings to assess for potential differences in ED revisit rates. For the moderator analysis, COI and ESI were analyzed as ordinal variables instead of nominal, given their hierarchical classifications. Lastly, we plotted the predictive values obtained from multivariable logistic regression models for race, ethnicity, language, and insurance moderated by the top PECARN major and subgroup diagnosis groupings. For all regression models, odd ratios (ORs), adjusted ORs, and 95% CI were calculated. All analyses were conducted using R Statistical Software (v4.2.1; R Core Team 2022).
RESULTS
There were 1,008,651 total ED encounters included in our analysis. Among these, 52,176 (5.2%) had an ED revisit within 7 days of the index ED visit. Of the total population, 53% (n = 529,960) were males, 59% (n = 541,016) self-identified as black/African American, 35% (n = 325,616) self-identified as white, and there were 89,668 encounters where race data was missing. 81% (n = 808,996) self-identified as non-Hispanic/Latino, and 88% (n = 888,545) reported English as their preferred language. Among the total population, 69% (n = 692,371) had public insurance. Patient age and ESI scores were similar between index ED encounters that had a 7-day ED revisit and those that did not have an ED return visit (Tables 1 and 2).
TABLE 1.
Demographic Characteristics
| Variable | Overall, N = 1,008,651* | Encounters With a Revisit, N = 52,176* | Encounters Without a Revisit, N = 956,475* |
|---|---|---|---|
|
| |||
| Age categories | |||
| 0–28 d | 16,605 (2) | 821 (2) | 15,784 (2) |
| 29 d-1 y | 138,080 (14) | 9920 (19) | 128,160 (13) |
| 1–4 y | 293,756 (29) | 17,614 (34) | 276,142 (29) |
| 5–9 y | 253,358 (25) | 11,266 (22) | 242,092 (25) |
| 10–14 y | 178,573 (18) | 7094 (14) | 171,479 (18) |
| 15–18 y | 128,279 (13) | 5461 (10) | 122,818 (13) |
| Sex | |||
| Male | 529,960 (53) | 27,660 (53) | 502,300 (53) |
| Female | 478,663 (47) | 24,516 (47) | 454,147 (47) |
| Missing | 28 | 0 | 28 |
| Race | |||
| Black or AA | 541,016 (59) | 28,651 (61) | 512,365 (59) |
| White | 325,616 (35) | 15,936 (34) | 309,680 (36) |
| 2 or more races | 26,722 (3) | 1379 (3) | 25,343 (3) |
| Asian | 18,641 (2) | 1036 (2) | 17,605 (2) |
| Native NH/PI | 3490 (0) | 171 (0) | 3319 (0) |
| AI/AN | 3435 (0) | 178 (0) | 3257 (0) |
| Other | 64 (0) | 2 (0) | 62 (0) |
| Missing | 89,667 | 4823 | 84,844 |
| Ethnicity | |||
| Non-Hispanic or Latino | 808,996 (81) | 41,106 (79) | 767,890 (81) |
| Hispanic or Latino | 195,674 (19) | 10,970 (21) | 184,704 (19) |
| Missing | 3981 | 100 | 3881 |
| Language | |||
| English preferred | 888,545 (88) | 45,971 (88) | 842,574 (88) |
| Spanish preferred (interpreter present) | 106,396 (11) | 5558 (11) | 100,838 (11) |
| Other (no interpreter) | 12,329 (1) | 647 (1) | 11,682 (1) |
| Missing | 1381 | 0 | 1381 |
| Insurance | |||
| Public | 692,371 (69) | 38,428 (74) | 653,943 (68) |
| Private | 263,152 (26) | 11,929 (23) | 251,223 (26) |
| Uninsured | 53,123 (5) | 1818 (3) | 51,305 (5) |
| Missing | 5 | 1 | 4 |
| Primary care provider | 846,385 (84) | 44,922 (86) | 801,463 (84) |
| No. complex chronic condition categories | |||
| 0 | 915,782 (91) | 45,674 (88) | 870,108 (91) |
| 1 | 40,096 (4) | 2384 (5) | 37,712 (4) |
| 2 | 14,796 (1) | 1023 (2) | 13,773 (1) |
| 3 | 8759 (1) | 665 (1) | 8094 (1) |
| 4 + | 29,218 (3) | 2430 (5) | 26,788 (3) |
| COI | |||
| Very low | 347,445 (35) | 18,498 (36) | 328,947 (35) |
| Low | 113,924 (11) | 5644 (11) | 108,280 (11) |
| Moderate | 165,808 (17) | 8560 (17) | 157,248 (17) |
| High | 183,144 (18) | 9520 (18) | 173,624 (18) |
| Very high | 184,267 (19) | 9497 (18) | 174,770 (19) |
| Missing | 14,063 | 457 | 13,606 |
n (%); mean (SD).
AA indicates African American; AI/AN, American Indian or Alaskan Native; COI, Childhood Opportunity Index; NH/PI, Native Hawaiian/Pacific Islander.
TABLE 2.
Diagnosis and ESI Scores
| Variable | Overall, N = 1,008,651* | Encounters With a Revisit, N = 52,176* | Encounters Without a Revisit, N = 956,475* |
|---|---|---|---|
|
| |||
| Diagnosis | |||
| ENT, dental, and mouth diseases | 182,167 (20) | 10,479 (22) | 171,688 (20) |
| Trauma | 129,499 (14) | 3618 (7) | 125,881 (14) |
| Gastrointestinal diseases | 124,914 (14) | 8046 (17) | 116,868 (13) |
| Systemic states | 118,981 (13) | 8859 (18) | 110,122 (13) |
| Respiratory diseases | 108,058 (12) | 5871 (12) | 102,187 (12) |
| Skin, dermatology,c and soft tissue diseases | 45,720 (5) | 2268 (5) | 43,452 (5) |
| Neurologic diseases | 31,643 (3) | 1828 (4) | 29,815 (3) |
| Musculoskeletal and connective tissue diseases | 23,335 (3) | 845 (2) | 22,490 (3) |
| Other | 20,654 (2) | 992 (2) | 19,662 (2) |
| No code entered | 16,697 (2) | 23 (0) | 16,674 (2) |
| Psychiatric and behavioral diseases and substance abuse | 16,294 (2) | 556 (1) | 15,738 (2) |
| Diseases of the eye | 15,866 (2) | 754 (2) | 15,112 (2) |
| Urinary tract diseases | 13,563 (1) | 807 (2) | 12,756 (1) |
| Allergic, immunologic, and rheumatologic diseases | 12,596 (1) | 706 (1) | 11,890 (1) |
| Not categorized | 10,981 (1) | 433 (1) | 10,548 (1) |
| Hematologic diseases | 9637 (1) | 905 (2) | 8732 (1) |
| Genital and reproductive diseases | 9564 (1) | 407 (1) | 9157 (1) |
| Endocrine, metabolic, and nutritional diseases | 9132 (1) | 200 (0) | 8932 (1) |
| Fluid and electrolyte disorders | 8114 (1) | 344 (1) | 7770 (1) |
| Toxicologic emergencies (including environment) | 7087 (1) | 152 (0) | 6935 (1) |
| Child abuse | 3254 (0) | 74 (0) | 3180 (0) |
| Circulatory and cardiovascular diseases | 3126 (0) | 150 (0) | 2976 (0) |
| Neoplastic diseases (cancer, not benign neoplasms) | 1008 (0) | 36 (0) | 972 (0) |
| Missing | 86,761 | 3823 | 82,938 |
| ESI score | |||
| Critical | 2243 (0) | 6 (0) | 2237 (0) |
| High risk | 224,188 (22) | 10,908 (21) | 213,280 (22) |
| 2+ resources | 359,239 (36) | 19,277 (37) | 339,962 (36) |
| 1 resource | 356,081 (35) | 19,175 (37) | 336,906 (35) |
| Fast track | 62,922 (6) | 2795 (5) | 60,127 (6) |
| Missing | 3978 | 15 | 3963 |
n (%)
ENT indicates ear, nose, throat; ESI, Emergency Severity Index.
Overall Factors Associated With Emergency Department Return Visits
In the multivariable analyses including all diagnoses, there were greater adjusted odds of ED revisit for patients who identified as black/African American (aOR: 1.14, 95% CI: 1.11–1.17), multiracial (aOR: 1.14, 95% CI 1.06–1.22) compared with white, and for Hispanic (aOR: 1.21, 95% CI: 1.17–1.26) compared with non-Hispanic (Fig. 1, Supplemental Table, Supplemental Digital Content 1, http://links.lww.com/PEC/B415). There were also greater odds of ED revisits for those who had public insurance (aOR: 1.19, 95% CI: 1.16–1.23) compared with private insurance, and those with at least one complex chronic condition (aOR: 1.21, 95% CI: 1.16–1.27), and those living in higher COI level neighborhoods (aOR: 1.09, 95% CI: 1.07–1.12; Fig. 1). There were lower odds of ED revisits for patients whose preferred language was Spanish (aOR: 0.91, 95% CI: 0.86–0.96) compared with English and those who were uninsured (aOR: 0.83, 95% CI: 0.79–0.88) compared with private insurance.
FIGURE 1.
Forest plot of multivariable logistic regression for encounters with 7-day ED revisit. AA indicates African American; AI/AN, American Indian/Alaskan Native; COI, Childhood Opportunity Index; ED, emergency department; ESI, Emergency Severity Index; NH/PI, Native Hawaiian/Pacific Islander; OR, odd ratio.
Disparities in Emergency Department Return Visit Rates by Diagnostic Groupings
The most common diagnostic group during the index visit that had ED revisits were ear, nose, throat (ENT)/dental/mouth diseases (22%; eg, infectious ear disorders and upper respiratory illnesses), followed by systemic states (18%; eg, fever and viral illnesses), gastrointestinal (GI) diseases (17%; eg, gastroenteritis and abdominal pain), respiratory diseases (12%; eg, asthma and pneumonia), and trauma (7%; Table 2). Black/African American patients had higher adjusted odds of revisits for gastrointestinal diseases (aOR: 1.12, 95% CI: 1.05–1.2) and systemic states (aOR: 1.09, 95% CI: 1.03–1.17) than white patients (Fig. 2). Hispanic patients had greater adjusted odds of revisits for ENT/dental/mouth diseases (aOR: 1.11, 95% CI: 1.01–1.21), systemic states (aOR: 1.26, 95% CI: 1.16–1.37), respiratory diseases (aOR: 1.27, 95% CI 1.14–1.41) than non-Hispanic patients. Patients who had public insurance had greater adjusted odds of revisits for all of the most common disease categories compared with those with private insurance (ENT/dental mouth diseases, aOR: 1.14, 95% CI: 1.07–1.22; systemic states, aOR: 1.18, 95% CI: 1.11–1.25; respiratory diseases, aOR: 1.10, 95% CI: 1.03–1.19; GI diseases, aOR: 1.22, 95% CI: 1.14–1.3).
FIGURE 2.
Multivariable logistic regressions of revisits by common major diagnostic groupings. AA indicates African American; AI/AN, American Indian/Alaskan Native; COI, Childhood Opportunity Index; NH/PI, Native Hawaiian/Pacific Islander; OR, odd ratio.
Table 3 shows the further breakdown of the most common subgroup diagnoses within the top 5 major diagnostic groupings of the PECARN Diagnostic Grouping System. In the moderator analyses of these common subgroup diagnoses, there was a higher probability of ED revisits for Hispanic patients for upper respiratory illnesses, infectious ear disorders, infectious respiratory disease, asthma, fever, viral illness, and acute systemic states (Fig. 3). For black/African American patients, ED return visits for most high-volume subgroup diagnoses were not more common than other race groups (Fig. 3). Patients on public insurance had a higher probability of revisits for lacerations/amputations, fractures/dislocations, brain and skull trauma, and for the remaining subgroup diagnoses in ENT and GI disorders (Supplemental Figure, Supplemental Digital Content 2, http://links.lww.com/PEC/B416).
TABLE 3.
Top 3 Subgroup Diagnosis for Each Overall Diagnosis Grouping by Revisit
| Diagnosis Grouping | Encounters With a Revisit | Encounters Without a Revisit | ||
|---|---|---|---|---|
|
|
|
|||
| Subgroup Diagnosis | n (%) | Subgroup Diagnosis | n (%) | |
|
| ||||
| ENT, dental, and mouth diseases | n = 10,479 | — | n = 171,688 | — |
| Infectious nose and sinus disorders, including URI | 4695 (44.8) | Infectious nose and sinus disorders, including URI | 65,988 (38.43) | |
| Infectious ear disorders | 3075 (29.34) | Infectious ear disorders | 51,574 (30.04) | |
| Infectious mouth and throat disorders | 1448 (13.82) | Infectious mouth and throat disorders | 27,109 (15.79) | |
| Gastrointestinal diseases | n = 8046 | — | n = 116,868 | — |
| GE | 2303 (28.62) | GE | 32,376 (27.7) | |
| Vomiting | 1922 (23.89) | Other gastrointestinal diseases | 27,598 (23.61) | |
| Other gastrointestinal diseases | 1665 (20.69) | Vomiting | 26,114 (22.34) | |
| Respiratory diseases | n = 5871 | — | n = 102,187 | — |
| Infectious respiratory diseases | 3251 (55.37) | Infectious respiratory diseases | 43,546 (42.61) | |
| Asthma | 1762 (30.01) | Asthma | 33,012 (32.31) | |
| Bronchospasm and wheezing | 494 (8.41) | Other respiratory diseases | 17,569 (17.19) | |
| Systemic states | n = 8859 | — | n = 110,122 | — |
| Fever | 5350 (60.39) | Fever | 55,010 (49.95) | |
| Viral illnesses | 2650 (29.91) | Viral illnesses | 39,374 (35.75) | |
| Acute systemic states | 799 (9.02) | Acute systemic states | 14,656 (13.31) | |
| Trauma | n = 3618 | — | n = 125,881 | — |
| Lacerations, amputations, and uninfected foreign bodies (external) | 932 (25.76) | Lacerations, amputations, and uninfected foreign bodies (external) | 29,795 (23.67) | |
| Fractures and dislocations (extremities) | 689 (19.04) | Fractures and dislocations (extremities) | 27,020 (21.46) | |
| Brain and skull trauma | 652 (18.02) | Brain and skull trauma | 19,264 (15.3) | |
ENT indicates ear, nose, throat; GE, gastroenteritis; URI, upper respiratory infection.
FIGURE 3.
Moderator analyses of most common major diagnostic groupings and top diagnoses by race and ethnicity. AA indicates African American; AI/AN, American Indian/Alaskan Native; BSP/WZ, bronchospasm/wheezing; BST, bone and skull trauma; Dis, diseases; ENT, ear, nose, throat; Fract/Disloc (Ext), fracture/dislocations of extremity; GE, gastroenteritis; Hist/Lat, Hispanic/Latino; Inf, infectious; lac/amp/uninf FB, laceration/amputations/uninfected foreign bodies; NH/PI, Native Hawaiian/Pacific Islander; Non-Hist/Lat, non-Hispanic/Latino; Resp, respiratory; SS, systematic states; URI, upper respiratory infection.
DISCUSSION
In a comprehensive pediatric health system, we identified significant disparities in the rates of ED return visits for minoritized populations. Black/African American, multiracial, and Hispanic patients had significantly higher odds of ED revisits compared with their white and non-Hispanic counterparts when diagnoses were included in aggregate, consistent with prior research demonstrating ED care and outcomes disparities for minoritized patients.25 We also found higher odds of ED revisits for patients on public insurance and with greater numbers of complex chronic conditions which has been shown previously.2 However, our study on a system-wide variability in ED revisit disparities extends these prior findings and provides a new approach to assess disparities across several diagnostic groupings and subgroupings which helps underscore the need for targeted interventions to address these inequities.
The greater ED revisit probabilities for Hispanic patients with upper respiratory illnesses, infectious ear disorders, infectious respiratory disease, asthma, and viral illness suggest certain disease processes where targeted culturally centered disease-specific interventions could be particularly beneficial. One might consider assessing the differences in cultural beliefs and perceptions of these diagnoses and designing interventions addressing these beliefs in verbal and written discharge instructions.26,27 There did not appear to be higher probability of revisits for black/African American patients with most higher-volume diagnoses, which may reflect the fact that the system-level disparities observed for black/African American patients are derived more from disparities spread broadly throughout all diagnoses. This could inform future targets for reducing ED revisit racial disparities by focusing on upstream determinants of health and impacts of structural racism instead of specific disease targets for black/African American patients.28,29
Patients with public insurance had significantly higher odds of ED revisits across all the top major diagnostic categories. This finding aligns with existing literature that highlights the challenges faced by publicly insured patients, such as limited access to primary care providers, longer wait times for specialty care, and increased likelihood of utilizing EDs.30,31 The greater odds of revisit for trauma-related conditions, such as lacerations, amputations, fractures, dislocations, and brain and skull trauma, among publicly insured patients, suggest that these patients might experience higher exposure to environmental risks or may lack access to timely and effective primary care follow-ups for trauma-related illnesses.
Our study suggests that Spanish-speaking patients and those living in lower COI neighborhoods had lower odds of revisits across the most common major diagnosis groups in our institution, which is in contrast to prior literature.7,10 This counterintuitive finding could be attributed to a variety of factors, including potential underutilization of repeated health care services due to geographic access barriers, language barriers or local community mistrust of the health care system.32 Further research is needed to understand the underlying causes of this regional variability seen in our institution and to develop strategies to ensure that these populations are receiving appropriate care.
Other studies have shown a variety of hospital-based methods to address disparities in pediatric EDs, ranging from implementing real-time data dashboards with quality metrics visible by all staff members, to holding workshops on implicit biases of clinical teams, to incorporating a focus on equity into higher-level hospital operations.33–35 Disease-specific interventions such as those targeting asthma encounters have also shown promise in reducing ED revisits.36,37 In addition, the use of case management services to facilitate follow-up care across settings has also shown promise in reducing frequent ED revisits.38,39 Furthermore, certain states like North Carolina have found success in addressing upstream determinants of health through partnerships between the department of health and various community organizations that focus on nonmedical drivers of health.40 Future efforts to reduce ED revisit disparities may benefit from a blend of disease-specific interventions while also devoting resources to tackling upstream determinants and utilizing a mechanism of real-time feedback for ongoing initiatives.
Our findings are subject to several limitations. There are a variety of reasons families return to the ED, ranging from illness progression to inadequate initial care or barriers to accessing outpatient care follow-up. However, this study did not assess reasons for ED revisits as these were not available given the retrospective nature of our study. The scope of this study also precluded an examination of differences between patients who were admitted versus discharged on the subsequent ED visit. In addition, we do not know if patients returned to other EDs outside of the health system studied, though we feel these potential outside revisits would likely reflect a minority of cases given the magnitude of pediatric care seen within the studied health system in Atlanta, GA. Our study was performed in a single health care system and therefore the results may not be generalizable to other care settings, however the approach taken to explore variability in disparities by disease process still may provide widespread utility. For a subset of families, our COI variable was derived from patient zip codes, which do not give as granular data as census tract data, and which may limit the conclusions from the COI analyses. In addition, the third most commonly reported race after black/African American and white was “missing,” and it is unclear if there existed any bias in which encounters had a “missing” race category, which, if present, could limit the assessment of variability by diseases. Furthermore, due to low numbers of non-English and non-Spanish preferred languages, we combined this group into an “other” category and, therefore, could not analyze for disparities in any of these less common language domains.
CONCLUSIONS
This study highlights significant disparities in the odds of pediatric ED revisits by race, ethnicity, and insurance status, and how they vary by disease process. For certain variables, the disparities were most prevalent within specific disease groups, while for others, the disparities were spread across a much broader swath of disease processes. Addressing these inequities will require targeted, disease-specific interventions and comprehensive strategies that encompass clinical, community, and policy-level interventions to target more upstream determinants of ED return visits. Future research should explore the mechanisms driving these disparities and evaluate the effectiveness of interventions designed to reduce them.
Supplementary Material
Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.pec-online.com.
Disclosure:
C.A.R. was supported by the National Institutes of Health (K23HL173694). The funder played no part in this study’s design and execution, data collection, management, analysis, or interpretation, nor in the preparation, review, or approval of this manuscript.
Footnotes
The remaining authors declare no conflict of interest.
REFERENCES
- 1.Michelson KA, Lyons TW, Bachur RG, et al. Timing and location of emergency department revisits. Pediatrics. 2018;141:e20174087. [DOI] [PubMed] [Google Scholar]
- 2.Akenroye AT, Thurm CW, Neuman MI, et al. Prevalence and predictors of return visits to pediatric emergency departments: return Visits to Pediatric EDs. J Hosp Med. 2014;9:779–787. [DOI] [PubMed] [Google Scholar]
- 3.Rising KL, Victor TW, Hollander JE, et al. Patient returns to the emergency department: the time-to-return curve. Acad Emerg Med. 2014;21:864–871. [DOI] [PubMed] [Google Scholar]
- 4.Ali AB, Place R, Howell J, et al. Early pediatric emergency department return visits: a prospective patient-centric assessment. Clin Pediatr (Phila). 2012;51:651–658. [DOI] [PubMed] [Google Scholar]
- 5.Augustine EM, Bekker TD, Goel A, et al. A Survey of return visits to the pediatric emergency department: the caretakersʼ perspective. Pediatr Emerg Care. 2018;34:27–32. [DOI] [PubMed] [Google Scholar]
- 6.Pittsenbarger Z, Thurm C, Neuman M, et al. Hospital-level factors associated with pediatric emergency department return visits. J Hosp Med. 2017;12:536–543. [DOI] [PubMed] [Google Scholar]
- 7.Portillo EN, Stack AM, Monuteaux MC, et al. Association of limited English proficiency and increased pediatric emergency department revisits. Acad Emerg Med. 2021;28:1001–1011. [DOI] [PubMed] [Google Scholar]
- 8.Smith JA, Fletcher A, Mirea L, et al. Pediatric emergency department return visits within 72 hours: caregivers’ motives and analysis of ethnic and primary language disparities. Pediatr Emer Care. 2022;38:e833–e838. [DOI] [PubMed] [Google Scholar]
- 9.Greenky D, Levine A, Gillespie SE, et al. A reversed trend: care for limited English proficiency patients in the pediatric emergency department. Emerg Med Int. 2019;2019:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bettenhausen JL, Noelke C, Ressler RW, et al. The Association of the Childhood Opportunity Index on pediatric readmissions and emergency department revisits. Academic Pediatrics. 2022;22:614–621. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Parikh K, Berry J, Hall M, et al. Racial and ethnic differences in pediatric readmissions for common chronic conditions. J Pediatr. 2017;186:158–164.e1. [DOI] [PubMed] [Google Scholar]
- 12.Johnson LH, Beck AF, Kahn RS, et al. Characteristics of pediatric emergency revisits after an asthma-related hospitalization. Ann Emerg Med. 2017;70:277–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Beck AF, Simmons JM, Huang B, et al. Geomedicine: area-based socioeconomic measures for assessing risk of hospital reutilization among children admitted for asthma. Am J Public Health. 2012;102:2308–2314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Aratani Y, Addy S. Disparities in repeat visits to emergency departments among transition-age youths with mental health needs. PS (Wash DC). 2014;65:685–688. [DOI] [PubMed] [Google Scholar]
- 15.Dy CJ, Lyman S, Do HT, et al. Socioeconomic factors are associated with frequency of repeat emergency department visits for pediatric closed fractures. J Pediatr Orthop. 2014;34:548–551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Freedman SB, Thull-Freedman JD, Rumantir M, et al. Emergency department revisits in children with gastroenteritis. J Pediatr Gastroenterol Nutr. 2013;57:612–618. [DOI] [PubMed] [Google Scholar]
- 17.de Vos-Kerkhof E, Geurts DHF, Steyerberg EW, et al. Characteristics of revisits of children at risk for serious infections in pediatric emergency care. Eur J Pediatr. 2018;177:617–624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ramgopal S, Varma S, Victor TW, et al. Pediatric return visits to the emergency department: the time to return curve. Pediatr Emer Care. 2022;38:e1454–e1461. [DOI] [PubMed] [Google Scholar]
- 19.Graham KL, Auerbach AD, Schnipper JL, et al. Preventability of early vs. late hospital readmissions in a national cohort of general medicine patients. Ann Intern Med. 2018;168:766–774. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Graham KL, Wilker EH, Howell MD, et al. Differences between early and late readmissions among medical patients, a cohort study. Ann Intern Med. 2015;162:741–749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Noelke C, McArdle N, Baek M, et al. Child Opportunity Index 2.0 Technical Documentation Published online 2020. diversitydatakids.org/researchlibrary/research-brief/how-we-built-it. [Google Scholar]
- 22.Feudtner C, Feinstein JA, Zhong W, et al. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr. 2014;14:199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Gilboy N, Tanabe P, Travers DA, et al. Emergency Severity Index, Version 4: Implementation Handbook. AHRQ Publication No. 05–0046-2. Rockville, MD: Agency for Healthcare Research and Quality. May 2005. [Google Scholar]
- 24.Fant C, Marin JR, Ramgopal S, et al. Updated diagnosis grouping system for pediatric emergency department visits. Pediatr Emer Care. 2023;39:299–303. [DOI] [PubMed] [Google Scholar]
- 25.Zhang X, Carabello M, Hill T, et al. Racial and ethnic disparities in emergency department care and health outcomes among children in the United States. Front Pediatr. 2019;7:525. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Acorda DE, DesOrmeaux CN, Rozmus CL, et al. Hispanic parental beliefs and practices in the management of common childhood illnesses: a review of the literature. J Transcult Nurs. 2020;31:502–518. [DOI] [PubMed] [Google Scholar]
- 27.Acorda DE, Engebretson J, DesOrmeaux C, et al. Exploring Latino perspectives in childhood fever: beliefs, practices, and needs. J Transcult Nurs. 2022;33:695–703. [DOI] [PubMed] [Google Scholar]
- 28.Johnson TJ. Racial bias and its impact on children and adolescents. Pediatr Clin North Am. 2020;67:425–436. [DOI] [PubMed] [Google Scholar]
- 29.Kharbanda AB. Racial inequities in pediatric emergency care. JAMA Network Open. 2021;4:e2034019. [DOI] [PubMed] [Google Scholar]
- 30.Bisgaier J, Rhodes KV. Auditing access to specialty care for children with public insurance. N Engl J Med. 2011;364:2324–2333. [DOI] [PubMed] [Google Scholar]
- 31.Kenney GM, Coyer C. National Findings on Access to Health Care and Service Use for Children Enrolled in Medicaid or CHIP. Urban Institute. 2012. Retrieved from: https://www.macpac.gov/wp-content/uploads/2012/03/National-Findings-on-Access-to-Health-Care-and-Service-Use-for-Children-Enrolled-in-Medicaid-or-CHIP.pdf. [Google Scholar]
- 32.Jaiswal J Whose responsibility is it to dismantle medical mistrust? Future directions for researchers and health care providers. Behav Med. 2019;45:188–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hartford EA, Thomas AA, Kerwin O, et al. Toward improving patient equity in a pediatric emergency department: a framework for implementation. Ann Emerg Med. 2023;81:385–392. [DOI] [PubMed] [Google Scholar]
- 34.Tsuchida RE, Haggins AN, Perry M, et al. Developing an electronic health record-derived health equity dashboard to improve learner access to data and metrics. AEM Educ Train. 2021;5(suppl 1):S116–S120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Yi S, Burke C, Reilly A, et al. Designing and developing a digital equity dashboard for the emergency department. J Am Coll Emerg Physicians Open. 2023;4:e12997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Gill I, Shah A, Lee EK, et al. Community interventions for childhood asthma ED visits and hospitalizations: a systematic review. Pediatrics. 2022;150:e2021054825. [DOI] [PubMed] [Google Scholar]
- 37.Allen ED, Montgomery T, Ayres G, et al. Quality improvement-driven reduction in countywide Medicaid acute asthma health care utilization. Acad Pediatr. 2019;19:216–226. [DOI] [PubMed] [Google Scholar]
- 38.Malebranche M, Grazioli VS, Kasztura M, et al. Case management for frequent emergency department users: no longer a question of if but when, where and how. CJEM. 2021;23:12–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Raven MC, Kushel M, Ko MJ, et al. The effectiveness of emergency department visit reduction programs: a systematic review. Ann Emerg Med. 2016;68:467–483.e15. [DOI] [PubMed] [Google Scholar]
- 40.Wortman Z, Tilson EC, Cohen MK. Buying health for North Carolinians: addressing nonmedical drivers of health at scale. Health Aff (Millwood). 2020;39:649–654. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.