Background
Emergency department triage is designed to quickly identify which patients need immediate attention and which can safely wait. In busy emergency departments, however, triage is not always perfectly accurate. Some patients who are actually quite sick may be classified as lower priority, a problem known as undertriage. Others with relatively low urgency may be assigned a higher priority than necessary, known as overtriage. Both situations can affect how quickly patients receive care and, potentially, their outcomes.
This study examined whether triage mismatch in emergency departments was associated with delays in care, longer stays in the emergency department, admission to intensive care, and short-term mortality. The investigators also explored whether patient history and recent health care use differed among patients who were undertriaged or overtriaged compared with those whose triage status better matched their actual care needs.
Why Triage Accuracy Matters
Triage in the emergency department is often based on the Emergency Severity Index, or ESI, a five-level system used in many hospitals. The ESI helps staff estimate how urgently a patient needs care and how many resources they are likely to require. Ideally, a patient’s assigned ESI level should match the intensity of evaluation and treatment they ultimately need.
When the assignment is too low, a seriously ill patient may wait longer than appropriate. When it is too high, less urgent patients may receive faster attention than needed, which can crowd the system and delay care for others. Because emergency departments face constant pressure from high patient volume and limited staffing, even small triage errors may have system-wide effects.
Study Design
This was a retrospective cohort study of emergency department encounters from 2016 through 2020 across 21 emergency departments. The study included more than 5.3 million adult visits, making it one of the larger real-world assessments of triage accuracy and downstream outcomes.
The researchers used operational measures of triage accuracy by comparing the assigned ESI level with the patient’s downstream intensity of care and resource use. Patients whose triage level did not match their later care needs were categorized as undertriaged high-acuity or overtriaged low-acuity. Patients whose triage level aligned with their downstream use of resources were classified as true low-, mid-, or high-acuity.
The main outcome was delay in care. Secondary outcomes included emergency department length of stay, intensive care unit admission, and short-term mortality.
Who Was Included
Among the 5,315,081 adult emergency department encounters, the average age was 51.7 years. Women accounted for 2,962,827 visits, or 56% of the cohort. The racial and ethnic distribution included 590,566 Asian patients (11.1%), 800,966 Black patients (15.1%), 2,336,012 non-Hispanic White patients (44.0%), 1,137,444 Hispanic patients (21.4%), and 450,093 patients identified as other, unknown, or multiracial (8.5%).
The study also found that patients who were undertriaged and later proved to have high acuity tended to have higher burdens of chronic illness, use more high-risk medications, and have more recent health care encounters than patients classified as true mid- or high-acuity. This suggests that elements of recent medical history may provide useful clues to triage nurses and clinicians if they are available at the point of triage.
Key Findings
In adjusted analyses, both undertriage and overtriage were associated with small delays in care. The effect sizes were modest, but consistent.
Undertriaged high-acuity patients experienced an average delay of 8 minutes in care compared with true high-acuity patients. Overtriaged low-acuity patients experienced an average delay of 3 minutes in care and had a total emergency department length of stay that was 42 minutes longer than true low-acuity patients.
These findings are important because they show that triage mismatch does not only affect the patient who is misclassified; it may also influence the flow of care throughout the emergency department. Overtriage can contribute to crowding and inefficiency, while undertriage can threaten patient safety if seriously ill patients do not receive prompt evaluation and treatment.
Although the study focused on measured delays and disposition outcomes, it did not show large differences in short-term mortality attributable solely to triage mismatch. That said, even small delays can matter for time-sensitive conditions such as sepsis, stroke, major trauma, or severe respiratory distress, where rapid recognition and intervention are critical.
Interpretation of the Results
The overall message of the study is that triage is generally effective, but not perfect. When mismatches occur, they are linked to measurable delays in emergency care. The impact appears larger for overtriage on emergency department throughput, with a meaningful increase in total length of stay, than for short delays in direct care timing alone.
The finding that undertriaged high-acuity patients had more comorbidities, higher-risk medication exposure, and greater recent health care use points to a practical opportunity. Triage systems may improve if they incorporate more patient history data, such as recent hospitalizations, chronic disease burden, medication lists, and recent emergency visits. These data may help identify patients who look stable initially but have a higher risk of deterioration.
At the same time, triage systems must balance sensitivity and specificity. If too many patients are escalated, overtriage can worsen crowding and delay care for everyone. If triage thresholds are too strict, undertriage can miss patients who need urgent treatment.
Clinical and Operational Implications
For emergency clinicians and hospital leaders, this study reinforces several practical points.
First, triage should remain a high priority in emergency department operations. The first minutes of an encounter often determine how quickly a patient is placed in the right care pathway.
Second, triage accuracy can likely be improved by better use of existing clinical data. In many health systems, electronic health records contain useful information about prior diagnoses, medications, recent visits, and previous admissions. Integrating those data into triage workflows may help staff identify patients at risk of being undertriaged.
Third, hospitals should monitor both undertriage and overtriage. Quality improvement efforts often focus on missed high-acuity patients, but excessive overtriage can also be harmful by increasing crowding and reducing overall efficiency.
Finally, the findings support the value of continuous triage training, especially in settings with high patient volume or complex case mixes. Nurse education, decision support tools, and standardized reassessment protocols may help reduce mismatch.
What Patients Should Know
For patients and families, the triage process may seem quick or impersonal, but it is meant to ensure that the sickest people are seen first. If your symptoms worsen while waiting, it is important to tell staff immediately. Symptoms such as chest pain, difficulty breathing, confusion, sudden weakness, severe pain, or fainting should never be ignored.
Patients with chronic illness, multiple medications, or recent hospital visits may benefit from clearly sharing this information during triage, since it can help clinicians understand risk more accurately.
Study Limitations
As with all retrospective studies, this one has limitations. Because it used observational data, it can show associations but cannot prove that triage mismatch directly caused the observed delays or outcomes. Other factors, such as staffing levels, crowding, or differences among hospitals, may also have influenced care timing.
The classification of undertriage and overtriage was based on downstream resource use and intensity of care, which is practical for research but not identical to every clinical judgment made at the bedside. Also, the study included adult emergency department encounters only, so the findings may not fully apply to children or specialized emergency settings.
Even so, the large sample size and inclusion of multiple emergency departments strengthen the reliability and real-world relevance of the findings.
Conclusion
This large multicenter study found that emergency department mistriage was associated with small but measurable delays in care. Undertriaged high-acuity patients waited longer than appropriately triaged high-acuity patients, and overtriaged low-acuity patients experienced longer emergency department stays.
The results suggest that improving triage accuracy remains an important emergency medicine goal. Using patient history data, recent health care use, and risk factors more effectively may help clinicians identify high-risk patients earlier and reduce both undertriage and overtriage. In a busy emergency department, even modest improvements in triage can support safer, more efficient care for everyone.