Do Gender and Race Make a Difference in Acute Coronary Syndrome Pretest Probabilities in the Emergency Department?

Abstract

Objectives: The objective was to test for significant differences in subjective and objective pretest probabilities for acute coronary syndrome (ACS) in a large cohort of chest pain patients stratified by race or gender. Secondarily we wanted to test for any differences in rates of ACS, rates of 90-day returns, cost, and chest radiation exposure after these stratifications.

Methods: This is a secondary analysis of a prospective outcomes study of ED patients with chest pain and shortness of breath. We performed two separate analyses. The data set was divided by gender for analysis 1 while the analysis 2 stratification was made by race (nonwhite vs. white). For each analysis, groups were compared on several variables: provider visual analog scales (VAS) for likelihood of ACS, PREtest Consult ACS probabilities, rates of ACS, total radiation exposure to the chest, total costs at 30 days, and 90-day recidivism (ED, overnight observations, and inpatient admissions).

Results: A total of 844 patients were studied. Gender information was present on all 844 subjects, while complete race/ethnicity information was available on 783 (93%) subjects. For the first analysis, female patients made up 57% (478/844) of the population and their mean provider VAS scores for ACS were significantly lower (p = 0.000) at 14% (95% confidence interval [CI] = 13% to 16%) than that of males at 22% (95% CI = 19% to 24%). This was consistent with the objective pretest ACS probabilities subsequently calculated via the validated online tool, PREtest Consult, which were also significantly lower (p = 0.000) at 2.7% (95% CI = 2.4% to 3.1%) for females versus 6.6% (95% CI = 5.9% to 7.3%) for males. However, comparing females to males, there was no significant difference in diagnosis of ACS (3.6% vs. 1.6%), mean chest radiation doses (5.0 mSv vs. 4.9 mSv), total costs at 30 days ($3,451.24 vs. $3,847.68), or return to the ED within 90 days (26% each). For analysis 2 by race, nonwhite patients also comprised 57% (444/783) of individuals. Similar to the gender analysis, mean provider VAS scores for ACS were found to be significantly lower (p = 0.000) at 15% (95% CI = 13% to 16%) for nonwhite versus 20% (95% CI = 18% to 23%) for white subjects. Concordantly, objective pretest ACS probabilities were also significantly lower (p = 0.000) at 3.4% (95% CI = 2.9% to 3.9%) for nonwhite versus 5.3% (95% CI = 4.7% to 5.9%) for white subjects. There were no significant differences in outcomes in nonwhite versus white subjects when compared on diagnosis of ACS (3.2% vs 2.4%), mean chest radiation dose (4.6 mSv vs. 5.0 mSv), cost ($3,156.02 vs. $2,885.18), or 90-day ED returns (28% vs. 23%).

Conclusions: Despite consistently estimating the risk for ACS to be lower for both females and minorities concordantly with calculated objective pretest assessments, there does not appear to have been any significant decrease in subsequent evaluation of these perceived lower-risk groups when radiation exposure and costs are taken into account. Further studies on the impact of pretest assessments on gender and racial disparities in ED chest pain evaluation are needed.