Objectives: The Emergency Severity Index (ESI) is a prospectively validated, five-level emergency department (ED) triage system designed to match triage acuity to both patient acuity and appropriate resource allocation. The study hypothesis was that, in practice, there exists an inappropriate bias toward triaging patients with abdominal pain to a higher ESI level based solely upon their mode of arrival to the ED.
Methods: The authors performed a retrospective case-control study of patients presenting with abdominal pain. Patients were matched on sex, age (± 5 years), and date of arrival. Cases were those patients triaged to a Level 2, and controls were those triaged as Level 3. Conditional multiple variable logistic regression was used to evaluate the effect of the following variables on the odds of being triaged as Level 2: mode of arrival, systolic blood pressure (<90 mm Hg; normal, >140 mm Hg), heart rate, severe pain score (≥ 8 of 10), fever, race, history of cancer, and previous abdominal surgery. Age was also included in the regression modeling to confirm that matching was adequate. One-hundred cases and 100 controls were necessary to provide adequate sample size. A backward modeling technique was used, requiring a p < 0.05 for retention.
Results: Of the 200 subjects, 52 arrived by emergency medical services (EMS) and 148 walked in. After matching for sex, age, and date of arrival, and after adjusting for heart rate, cancer diagnosis, and severe pain, the odds ratio (OR) for being triaged ESI Level 2 was 7.19 (95% confidence interval [CI] = 2.75 to 18.8, p < 0.0001) for EMS patients compared to walk-in patients. The admission rate for Level 2 patients was not different from that of Level 3 patients (49% vs. 35% of Level 3 patients, p = 0.06), but EMS patients were more likely to be admitted, regardless of ESI level assignment (65% vs. 34%, p < 0.001).
Conclusions: After adjusting for covariates, EMS patients with abdominal pain were more likely to be triaged to a higher acuity level. Triage level was not associated with admission, but patients arriving by EMS were more likely to be admitted. This may indicate that the effect of EMS arrival on triage level assignment is actually appropriate. Further research is necessary to validate whether mode of arrival should be incorporated in the initial ESI triage acuity assignment.
© 2011 by the Society for Academic Emergency Medicine.