dcyphr | Racial Disparity in Emergency Department Triage


Previous studies have shown that black people have lower triage acuity scores (TAS), which determine how urgent a patient’s condition is in emergency departments (EDs), than white people. However, these studies did not determine whether this was due to bias of triagers or proper triage according to the severity of patients’ conditions. In this study, the researchers wanted to determine if black people are assigned lower TAS even after adjusting for demographic and clinical differences. They also wanted to understand if lower TAS caused longer wait times in EDs. The researchers analyzed data from an ED electronic medical record system, including demographic and clinical information, TAS, and ED waiting times. Triage scores were assigned from 1-5, with 1 being the most urgent. TAS and wait times were compared by race after adjusting for age, gender, insurance status, time of day, day of week, comorbidities, and abnormal vital signs. The researchers found that black people had an average TAS of 2.97, compared to 2.81 for white people, which was a significantly lower acuity rating. Black people also had longer wait times compared to white people. The results of this study show that racial bias may influence the triage process.


The researchers wanted to understand if black people were assigned lower TAS even after adjusting for demographic factors, and also if they experienced longer wait times in the ED.


Triage is the most important factor that determines how quickly patients are seen by a physician in EDs. Several studies have shown poor outcomes for patients who deteriorate while waiting for care in EDs. The triage process is under scrutiny in light of stresses on limited resources and disparities in medical care. Triage systems try to use objective data to assess the severity and urgency of patients’ medical conditions.

However, the effectiveness of triage is dependent on people who are doing the triage itself. All triage systems allow a lot of subjectivity in assessing patients. The current 5-point scoring system allows upgrading of scores so that patients with more urgent conditions are seen faster. Although this seems like a reasonable system, the scoring system might allow bias to play a factor in triaging patients.

Previous studies have reported lower TAS and longer wait times for black people, but no studies have adjusted for clinical variables such as vital signs and co-morbidities, which could justify lower TAS and longer wait times. In this study, the researchers sought to determine if racial disparities exist in triage at an urban Level 1 emergency department. They wanted to understand if TAS was influenced by race even after adjusting for several demographic and clinical factors. Wait time as assessed as a secondary outcome to understand if TAS affected patient care.


The data of 87,685 patients was initially collected. After excluding patients without one of eight defined chief complaints (see methods), had missing data, had a race of other, or a wait time of 0 minutes, 19,726 visits were left, which included 6,456 white patients and 13,270 black patients. 4,210 matched pairs were generated for the remaining visits. For these pairs, the average TAS for black patients was 2.97, compared with 2.81 for white patients. Black patients were twice as likely to have lower TAS than white patients.

A significant number of black patients were triaged 2-3 points lower below their white matches. 81 black patients were assigned a TAS of 4 while their matched white patients had a TAS of 2. The opposite occurrence only happened 13 times. Significant differences in TAS were seen in all chief complaints except for syncope and altered mental status.

Overall, black patients had a longer ED wait time than their white matches, with an 11-minute increase. Wait times were longer regardless of co-morbidities.


Accurate triage is challenging in EDs due to significant subjectivity in current practices. This subjectivity may cause bias to play a role in triage. The results of this study supports the hypothesis that racial bias may influence triage, even after adjusting for other variables.

Wait time for black patients was overall 11 minutes longer than white patients. It is difficult to determine how clinically important this is. However, for chest pain and shortness of breath, the difference in wait times was 21 minutes and 23 minutes respectively. These complaints can be serious and the differences in these wait times are in a range that could be clinically significant.


Study Design and Setting

Data was collected from a large, urban, academic teaching hospital that saw 87,000+ ED visits annually. At this hospital, triage was performed by nurses using the Emergency Severity Index (ESI), a 5-point system that accurately predicts urgency of patients’ medical conditions. Of the nurses, 85%% were female and 90%% were white. Demographic information of patients was also collected.


Patients who presented in the ED with chest pain, abdominal pain, weakness and dizziness, syncope, shortness of breath, altered mental status, back pain, and headache were included. Patients without demographic data were also excluded.

Methods of Measurement

Age was a continuous independent variable and gender was categorized as male or female. Black and white patients were organized into cohorts, and were matched 1:1 for each of the symptoms. They were matched by day of the week and within 4-hour time blocks.

Data Collection and Processing

Data was collected from the hospital’s ED electronic medical record system for all patient visits during the study period. Age, race, gender, chief complaint, insurance status, day and time of service, TAS, ED process times and whether or not the patient was discharged was collected. Co-morbidities were also identified, which was also used to match patients for analysis. Triage vital signs were collected and classified using the MEWS score, which predicts the need for intensive medical services. Triage acuity was assigned using the ESI 5-point system, with 1 being the most urgent and 5 being the least urgent.

Primary Data Analysis

Mean ESI was analyzed after matching black and white patients to adjust for demographic and clinical variables. Cases were matched by chief complaint, abnormal vital signs, co-morbidites, gender, discharge/admission, insurance status, day of the week, age, and time of ED visit. Mean wait times were also compared between matched patients. Finally, the researchers examined differences in ESI and wait times for each chief complaint. All analysis was done in SAS/STAT.


This study demonstrated higher TAS and longer wait times for black patients than white patients. Recognition of racial bias in other aspects of medical practice supports these findings.