PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Emerg Med. Author manuscript; available in PMC 2017 May 1.
Published in final edited form as:
PMCID: PMC4851931
NIHMSID: NIHMS753384

Racial Differences in Pediatric Emergency Department Triage Scores

Abstract

Background

Racial disparities are frequently reported in emergency department (ED) care.

Objectives

To examine racial differences in triage scores of pediatric ED patients. We hypothesized that racial differences existed but could be explained after adjusting for socio-demographic and clinical factors.

Methods

We examined all visits to two urban, pediatric EDs between August 2009 and March 2010. Demographic and clinical data were electronically extracted from the medical record. We used logistic regression to analyze racial differences in triage scores, controlling for possible covariates.

Results

There were 54,505 ED visits during the study period, with 7216 (13.2%) resulting in hospital admission. White patients accounted for 36.4% of visits, African Americans 28.5%, Hispanics 18.0%, Asians 4.1%, and American Indians 1.8%. After adjusting for potential confounders, African American (adjusted odds ratio [aOR] 1.89, 95% confidence interval [CI] 1.69 – 2.12), Hispanic (aOR 1.77, CI 1.55 – 2.02), and American Indian (aOR 2.57, CI 1.80 – 3.66) patients received lower acuity triage scores than Whites. In 3 out of 4 subgroup analyses based on presenting complaints (breathing difficulty, abdominal pain, fever), African Americans and Hispanics had higher odds of receiving low acuity triage scores. No racial differences were detected for patients with presenting complaints of laceration/head injury/arm injury. However, among patients admitted to the hospital, African Americans (aOR 1.47, CI 1.13 – 1.90) and Hispanics (aOR 1.71, CI 1.22 – 2.39) received lower acuity triage scores than Whites.

Conclusion

After adjusting for available socio-demographic and clinical covariates, African American, Hispanic, and American Indian patients received lower acuity triage scores than Whites.

Keywords: triage, health disparities, race, pediatrics

Introduction

Racial disparities have been frequently reported in emergency department (ED) care (1-14). African Americans and Hispanics have been reported to experience 12-25% longer ED wait times to see a physician compared to Whites, to be about 40% less likely to receive opioid analgesia prescriptions at discharge from the ED after long bone fracture, to have 34% lower odds of receiving an opioid prescription during pain-related visits, and to have 24% lower odds of radiological testing during their ED visit (3-5,9,10). Additionally, the odds of pediatric African American patients leaving the ED before complete evaluation and treatment may be as much as 60% higher than the odds for White patients (6). These findings suggest a wide range of racial disparities in ED care.

The size of disparity depends on the accuracy of risk adjustment. If the risk adjusters, such as triage score, are differentially assigned by race, disparities in ED care might be larger than previously estimated. The ED triage score is often used to adjust analyses of disparities (1-3,5-6,8,11). While triage scores should estimate illness severity and anticipated resource utilization and are assumed to be assigned without systematic bias, they do contain a measure of subjectivity. Previous studies have reported an association between minority race and lower acuity ED triage scores for adults and in a single study of pediatric patients (12-14). However, studies of triage scores have often utilized national databases, which do not permit adjustment for socio-demographic determinants that influence ED utilization, such as income level and distance from the patient's residence to the ED (15).

Differences in triage scores might reflect patients' varying racial and cultural attitudes towards ED utilization and not represent a true disparity in care. This could give the appearance of bias, but actually reflect ED visits by minority populations secondary to poor access to primary care (16-18).

We wanted to determine if socio-demographic or clinical factors could explain racial differences in triage scores among pediatric ED patients. We hypothesized that racial differences in triage scores existed but could be accounted for by socio-demographic, clinical, or ED utilization factors. Our null hypothesis was that racial differences in triage scores did not exist.

Materials and Methods

Study Design and Sample

To study the relationship of race and triage scores in the ED, we used a cross-sectional design encompassing all visits to either of two pediatric EDs from August 1, 2009 to March 31, 2010. The August 1, 2009 start date was chosen because a new 5-level triage system, the Emergency Severity Index, version 4 (ESI), was introduced on July 1, 2009 (19-23). Both EDs SERVE primarily an urban, multicultural population. We excluded visits of patients who eloped, died, or had missing data (Figure 1). Due to the potential influence of factors such as lack of primary care access, poverty, and proximity to the hospital on ED utilization, we included a variable to represent distance from the patient's residence to the ED (15,24-26). Our clinical experience has been that patients who live close to the ED are more likely to visit the ED than a primary care provider for a variety of clinical complaints (25). We utilized inpatient admission as an independent marker for illness severity.

Figure 1
Flow diagram of study sample. Percentages are out of the original sample of 59 719 ED visits.

Subgroups were also analyzed based on the patient's presenting complaint. Subgroups included visits with presenting complaints of: 1) “breathing difficulty,” “wheezing,” “asthma,” or “cough” (n = 8594, 15.8% of visits), 2) “abdominal pain” or “stomach pain” (n = 1868, 3.4% of visits), 3) “fever” (n = 9516, 17.5% of visits), and 4) “laceration,” “head injury” (with or without loss of consciousness), or “arm injury” (n = 4170, 7.7% of visits). In addition, we separately analyzed a subgroup of patients who were later admitted to the hospital (n = 7216, 13.2% of visits). This study was approved by the hospitals' Institutional Review Board (#1003-026).

Outcome Measurements and Independent Variables

The primary outcome measure was triage level. Triage was performed by an ED nurse who documented the patient's chief complaint, obtained a short history, recorded vital signs, and performed a brief examination, as needed. The nurse then assigned a triage score ranging from level 1 (most acute) to level 5 (least acute) using the ESI system (19). We dichotomized the ESI levels into levels 1-3 vs. levels 4-5 for analysis.

All demographic, insurance, and clinical data were extracted from the electronic medical record. At registration, caregivers were asked to report their child's race, primary language, age, sex, and address. Registrars also recorded their mode of transportation to the ED, which we categorized as “private” (private vehicle, public transport, walked, or other) or “urgent” (ambulance, helicopter, plane, or police). Distance between the patient's residence and the ED was determined using ArcGIS software (Environmental Systems Research Institute, Inc., Redlands, CA) and calculated from the center of the patient's zip code to the ED at which the patient sought care. Patients were assigned the median income for the zip code in which they resided using Truven Health Analytics (Ann Arbor, MI) data. Median incomes were grouped into quartiles based on the median incomes of all zip codes in which ED patients from the 7-county Twin Cities metro area resided. Data on patients living farther from the EDs than the 7-county metro area were not included in this analysis, since those patients would likely have been referred. Insurance type was dichotomized as Medical Assistance/Self-Pay/Public insurance or Private/Other insurance.

We utilized an ED activity/overcrowding score to estimate ED busyness at the time of patient registration and to account for any impact this may have on triage scores (6). We have locally validated this score as predicting wait time to see a physician and elopement (6). The influence of ED activity/overcrowding was not linear, so we grouped scores into quartiles with the lowest quartile (least busy ED) used as the reference (6). To address the influence of frequent ED use on triage scores, we calculated the total number of ED visits for each patient during the study period.

Statistical Analysis

We used Chi-square tests to examine univariate associations. Potential covariates were selected based on a hypothesized association with triage score and a significant association with triage score in univariate analyses (p < 0.001). We developed a pair-wise correlation matrix to identify variables that were correlated. Covariates with a correlation coefficient ≥ 0.7 were not included in the same model. We considered a variance inflation factor (VIF) > 2.5 as excessive (27). The overall mean VIF was ≤ 1.40 and the highest VIF for a specific variable was ≤ 2.23. Median income and public insurance logically measured the same characteristic, so we chose not to include both variables in the final model. We kept public insurance in the model since we felt it was more informative than zip code median income. Statistical analyses were performed using Stata version 13.1 (StataCorp, College Station, TX). The final model utilized logistic regression (routine “logit”) and adjusted for race, insurance type, distance from patient residence to the ED, primary language, age, sex, ED activity/overcrowding, ED campus, previous ED visits, inpatient admission, and mode of transportation to the ED. We also adjusted the model for clustering of visits within each patient. Furthermore, we examined potential interactions of race with insurance type, primary language, and distance from the ED, and incorporated them into the final model if they were significant. We considered a p-value < 0.05 as significant.

Results

Demographics

This study included 54,505 ED visits and 38,549 patients (Figure 1). Our study sample was racially diverse with Whites constituting only 36.4% (19,845 / 54,505) of visits and 41.7% of patients (16,075/38,549). Demographic characteristics are shown in Table 1.

Table 1
Patient Demographics by Triage Levela

Univariate Analysis

Among all visits, African American, Hispanic, and American Indian racial categories were associated with lower acuity triage scores than Whites (Table 1). Speaking Spanish or Somali was associated with lower acuity scores than speaking English or Hmong. Patients ranging from 1-9 years old had a larger proportion of low acuity scores than other age groups. As distance from the patient's residence to the ED increased, the proportion of low acuity triage scores decreased and the proportion of White patients increased, reflecting the local demographics. We expected that increasing ED activity might be associated with lower acuity triage scores, but the opposite was observed. Increasing ED activity/overcrowding scores were associated with higher acuity triage scores (Table 1).

Logistic Regression Analysis of Race and Triage Level

Adjustment for potential confounders using logistic regression had little influence on the odds ratios of low acuity triage scores (ESI levels 4-5) for African American (adjusted odds ratio [aOR] 1.89, 95% confidence interval [CI] 1.69 – 2.12), Hispanic (aOR 1.77, CI 1.55 – 2.02), and American Indian (aOR 2.57, CI 1.80 – 3.66) patients compared to Whites. We identified significant interactions between race and insurance type, and race and distance from the ED. Triage score differences persisted between Whites and African Americans, Hispanics, and American Indians for patients with both public and private insurance; however, the differences were greater among those with private insurance (Table 2). Although racial differences were present for both distance categories, the odds of low acuity triage scores compared to Whites were greater among African American (aOR 2.06, CI 1.88 – 2.25) and Hispanic (aOR 2.08, CI 1.85 – 2.34) patients living farther (> 5 miles) from the ED (Table 2). On the other hand, American Indians (aOR 2.26, CI 1.82 – 2.80) had greater odds of low acuity triage scores than Whites among patients living closer (≤ 5 miles) to the ED (Table 2). Asian patients had significantly lower odds of low acuity triage scores than Whites among patients on public insurance and among those living ≤ 5 miles from the ED (Table 2).

Table 2
Logistic Regression Analysis

Subgroup Analyses

We were unable to directly control for illness severity, so we examined four sub-populations of patients with similar presenting complaints (see Methods) to attempt to control for this variable. At visits with presenting complaints of “breathing difficulty,” “abdominal pain,” or “fever,” triage score acuity was significantly lower in African American and Hispanic patients compared to Whites (Table 3). Among visits with presenting complaints of “laceration,” “head injury” (with or without loss of consciousness), or “arm injury,” there were no racial differences in triage scores (Table 3). There were too few American Indian patients to accurately study within each subgroup.

Table 3
Presenting Complaint and Inpatient Admission Subgroup Analyses

We then analyzed only patients who were admitted as inpatients. Among these patients, African Americans (aOR 1.47, CI 1.13 – 1.90) and Hispanics (aOR 1.71, CI 1.22 – 2.39) had significantly higher odds of low acuity triage scores than Whites (Table 3).

Discussion

This study found racial differences in pediatric ED triage scores for African American, Hispanic, and American Indian patients compared to White patients. Adjustment for insurance type, distance from the ED, primary language, age, sex, ED activity/overcrowding, previous ED visits, inpatient admission, and other clinical factors did not eliminate these differences. Subgroup analyses confirmed the results of our primary analysis in four of the five subgroups examined, including the subgroup of patients who were admitted as inpatients.

These findings are similar to those reported in a previous pediatric study and in several adult studies C of racial differences in triage scores (12-15). We were able to analyze interactions between race, insurance type, and distance from the ED, that to our knowledge have not been reported in previous triage studies. These interactions were significant, but did not eliminate the differences in triage scores between Whites and African Americans, Hispanics, and American Indians (Tables 2 and and33).

The triage score may affect wait time to see an emergency physician, which can increase the risk of elopement and is associated with other ED outcome measures (1,3,5-6,9,11-12). Studies of racial differences in treatment or outcome measures often use triage scores to adjust for level of acuity, so any potential bias in the triage score may result in underestimated racial disparities in the pediatric ED (1-3,5,6,8,11).

Strengths

Our study was able to examine several factors that have not always been available in previous studies. Distance between patient residence and the ED, number of previous ED visits, and ED activity/overcrowding were all available to us and were associated with triage score. Controlling for these factors narrowed the differences between African American, Hispanic, and White patients' triage scores and often eliminated the differences between Asians and Whites. Our study sample was relatively large, with a diverse racial distribution of patients, which allowed us to analyze differences among subgroups categorized by presenting complaints. We identified that African American and Hispanic patients have over 50% higher odds of low acuity triage scores compared to Whites.

Limitations

There are several limitations to this study. Our data come from only two EDs in one metropolitan area. The sample was racially diverse, but it may not represent the racial distributions seen in other EDs nationally. There were so few American Indian patients that we could not obtain reliable estimates of triage score associations for all subgroups. We were limited in available potential covariates to those that were present and could be electronically extracted from the medical record. We did not have access to data on vital signs or past medical history, nor could we determine whether the patient was referred by a primary care practitioner, all of which may appropriately influence the triage score assignment (28).

Importantly, we were not able to directly account for the patient's illness severity. The very nature of the triage process requires subjective, contextual assessment and clinical instinct that may identify patient differences not captured in our data set or even in the medical record. Therefore, racial differences in triage scores may reflect accurate “real-time” assessment that appears biased in retrospect.

Conclusion

In summary, we have shown racial differences in triage score assignments that could not be explained by available socio-demographic, clinical, or ED utilization factors. Whatever their origin, these differences certainly warrant further investigation.

Summary

1. Why is this topic important?

A variety of Emergency Department (ED) patient care factors are influenced by the triage score, including wait time, ordering laboratory or radiology tests, and receiving pain medication. Triage scores should estimate illness severity and are assumed to be assigned without systematic bias, so any differences in triage scores while controlling for socio-demographic and clinical variables may indicate a lack of appropriate ED care.

2. What does this study attempt to show?

This study attempts to demonstrate that racial differences in pediatric ED triage scores exist and cannot be explained by available socio-demographic or clinical factors.

3. What are the key findings?

African American, Hispanic, and American Indian pediatric ED patients received lower acuity triage scores than White patients. African Americans and Hispanics also had lower acuity triage scores than Whites in subgroups of patients with the same presenting complaints, for all except laceration/head injury/arm injury. Racial differences in triage scores persisted for minority groups among patients who were later admitted to the hospital.

4. How is patient care impacted?

These findings will help to improve the care for all children by identifying gaps in knowledge and areas of concern in our treatment of evaluation of children in the ED setting.

Acknowledgments

The authors thank Jeffrey Barney for providing demographic and geographic information and Jill Dreyfus for assistance with statistical analysis. Research reported in this publication was supported by the National Institute on Minority Health and Health Disparities of the National Institutes of Health under Award Number U54MD008164. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Disclosure statement: The authors declare no conflicts of interest.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Park CY, Lee MA, Epstein AJ. Variation in emergency department wait times for children by race/ethnicity and payment source. Health Serv Res. 2009;44:2022–2039. [PMC free article] [PubMed]
2. Mills AM, Shofer FS, Boulis AK, et al. Racial disparity in analgesic treatment for ED patients with abdominal or back pain. Am J Emerg Med. 2011;29:752–756. [PubMed]
3. Haywood C, Jr, Tanabe P, Naik R, et al. The impact of race and disease on sickle cell patient wait times in the emergency department. Am J Emerg Med. 2013;31:651–656. [PMC free article] [PubMed]
4. Pletcher MJ, Kertesz SG, Kohn MA, et al. Trends in opioid prescribing by race/ethnicity for patients seeking care in US emergency departments. JAMA. 2008;299:70–78. [PubMed]
5. Payne NR, Puumala SE. Racial disparities in ordering laboratory and radiology tests for pediatric patients in the emergency department. Pediatr Emerg Care. 2013;29:598–606. [PubMed]
6. Harrison B, Finkelstein M, Puumala S, et al. The complex association of race and leaving the pediatric emergency department without being seen by a physician. Pediatr Emerg Care. 2012;28:1136–1145. [PubMed]
7. Agency for Healthcare Research and Quality. [Accessed October 1, 2013];National Healthcare Quality and Disparities Reports. 2011 http://www.ahrq.gov/research/findings/nhqrdr/nhqrdr11/8_respiratorydiseases/T8_3_2_1_2.html.
8. Quazi S, Eberhart M, Jacoby J, et al. Are racial disparities in ED analgesia improving? Evidence from a national database. Am J Emerg Med. 2008;26:462–464. [PubMed]
9. James CA, Bourgeois FT, Shannon MW. Association of race/ethnicity with emergency department wait times. Pediatrics. 2005;115:e310–e315. [PubMed]
10. Ortega HW, Vander Velden H, Lin CW, et al. Race, ethnicity, and analgesia provision at discharge among children with long-bone fractures requiring emergency care. Pediatr Emerg Care. 2013;29:492–497. [PubMed]
11. Goodacre S, Webster A. Who waits longest in the emergency department and who leaves without being seen? Emerg Med J. 2005;22:93–96. [PMC free article] [PubMed]
12. Schrader CD, Lewis LM. Racial disparity in emergency department triage. J Emerg Med. 2013;44:511–518. [PubMed]
13. Lopez L, Wilper AP, Cervantes MC, et al. Racial and sex differences in emergency department triage assessment and test ordering for chest pain, 1997-2006. Acad Emerg Med. 2010;17:801–808. [PubMed]
14. Hambrook JT, Kimball TR, Khoury P, et al. Disparities exist in the emergency department evaluation of pediatric chest pain. Congenit Heart Dis. 2010;5:285–291. [PubMed]
15. Hong R, Baumann BM, Boudreaux ED. The emergency department for routine healthcare: race/ethnicity, socioeconomic status, and perceptual factors. J Emerg Med. 2007;32:149–158. [PubMed]
16. Tang N, Stein J, Hsia RY, et al. Trends and characteristics of US emergency department visits, 1997-2007. JAMA. 2010;304:664–670. [PMC free article] [PubMed]
17. Lowe RA, McConnell KJ, Fu R, et al. Changes in access to primary care for Medicaid beneficiaries and the uninsured: the emergency department perspective. Am J Emerg Med. 2006;24:33–37. [PubMed]
18. Becker DJ, Blackburn J, Morrisey MA, et al. Co-payments and the use of emergency department services in the children's health insurance program. Med Care Res Rev. 2013;70:514–530. [PubMed]
19. Gilboy N, Tanabe P, Travers D, et al. Emergency Severity Index (ESI): A Triage Tool for Emergency Department Care, Version 4 Implementation Handbook, 2012 Edition. Rockville, MD: Agency for Healthcare Research and Quality; 2011.
20. Wuerz RC, Milne LW, Eitel DR, et al. Reliability and validity of a new five-level triage instrument. Acad Emerg Med. 2000;7:236–242. [PubMed]
21. Tanabe P, Gimbel R, Yarnold PR, et al. Reliability and validity of scores on The Emergency Severity Index version 3. Acad Emerg Med. 2004;11:59–65. [PubMed]
22. Durani Y, Brecher D, Walmsley D, et al. The Emergency Severity Index version 4: reliability in pediatric patients. Pediatr Emerg Care. 2009;25:751–753. [PubMed]
23. Green NA, Durani Y, Brecher D, et al. Emergency Severity Index version 4: a valid and reliable tool in pediatric emergency department triage. Pediatr Emerg Care. 2012;28:753–757. [PubMed]
24. Hargraves JL, Hadley J. The contribution of insurance coverage and community resources to reducing racial/ethnic disparities in access to care. Health Serv Res. 2003;38:809–829. [PMC free article] [PubMed]
25. Willems S, Peersman W, De Maeyar P, et al. The impact of neighborhood deprivation on patients' unscheduled out-of-hours healthcare seeking behavior: a cross-sectional study. BMC Fam Pract. 2013;14:136. doi: 10.1186/1471-2296-14-136. [PMC free article] [PubMed] [Cross Ref]
26. Lorch SA, Silber JH, Even-Shoshan O, et al. Use of prolonged travel to improve pediatric risk-adjustment models. Health Serv Res. 2009;44:519–541. [PMC free article] [PubMed]
27. Hoffman JP. Generalized Linear Models, a Practical Approach. New York, NY: Pearson Education, Inc; 2004.
28. Rinderknecht AS, Ho M, Matykiewicz P, et al. Referral to the emergency department by a primary care provider predicts severity of illness. Pediatrics. 2010;126:917–924. [PubMed]