From the 2019 HVPAA National Conference
Dr. Henry Han (University of Michigan), Dr. Raymond Yeow (University of Michigan), Dr. James Henderson (University of Michigan), Ms. Katie Grzyb (University of Michigan), Mr. Matthew Johnson (University of Michigan), Ms. Emily Dibble (University of Michigan), Dr. Ashwin Gupta (University of Michigan), Dr. Steven Kronick (University of Michigan), Dr. David Somand (University of Michigan), Dr. Donald Giacherio (University of Michigan), Dr. Scott Flanders (University of Michigan), Dr. Geoffrey Barnes (University of Michigan)
Background
High sensitivity troponin T (hsTnT), approved by the FDA in 2017, has an increased sensitivity for cardiac injury or ischemia. Previous studies in patients at low risk for ACS have found an association between positive troponin I/T testing results and increased admission rates.
Objective
We sought to understand the impact of implementing hsTnT on final diagnoses and rates of admission or discharge in a non-chest pain population presenting to the emergency department.
Methods
We conducted a retrospective cohort analysis at a large, academic, tertiary care center. Our cohort included ED patients without a chief complaint of “chest pain” or “shortness of breath” who had troponin testing ordered prior to and after implementation of hsTnT testing (February 2018). Patients included in analysis received either traditional troponin I (January 2016- December 2017) or hsTnT (March 2018- February 2019). We identified the most common presenting chief complaints for which troponin testing was performed, most common final diagnoses (by ICD10), and seasonally-adjusted rates of admission and discharge in the pre- and post- implementation periods. Additionally, we calculated the relative risks of key final diagnoses in the pre- vs post-hsTnT periods.
Results
In total, 21,946 (pre-) and 10,008 (post-) patients were included. Baseline characteristics were similar between groups. Overall, a composite “other” group of chief complaints (those appearing <200 times) was the most common reason for ED patients to receive troponin testing (14.2%) after chest pain and shortness of breath. A composite “other” group was the most common final discharge diagnosis (25.5%), followed by pain in throat and chest (18.5%) and atrial fibrillation/flutter (3.9%). Acute MI was noted to be the 25th most common diagnosis among patients without a chief complaint of chest pain (0.9%). Implementation of hsTnT was not associated with changes in rates of admission in patients presenting with any of the most common chief complaints (Table 1). In pre- and post-hsTnT analysis, there was no significant difference in rates of acute myocardial infarction (MI) as the final discharge diagnosis (Table 2). However, final discharge diagnoses of atrial fibrillation/flutter, abnormalities of heart beat, abnormalities of function studies, hypertensive crisis, pulmonary embolism, and hypotension were more common (Table 2). Association between the most common chief complaints and a final diagnosis of acute MI among patients presenting without a chief complaint of chest pain or shortness of breath did not change following implementation of hsTnT testing (Table 3).
Conclusions
Implementing hsTnT testing in a non-“chest pain” ED population was associated with similar rates of admission, no significant change in acute MI diagnoses but an increase in other non-acute MI diagnoses. There was no statistically significant association between non-“chest pain” chief complaints and increased acute MI diagnosis after hsTnT implementation.
Clinical implications
Hospitals considering a switch to the hsTnT test for patients in the emergency department will likely see changes in their final discharge diagnoses of non-acute MI. It is important for hospitals to consider how a switch to hsTnT will impact both patients presenting to the emergency department with chest pain and non-chest pain chief complaints.
Figures
