Factors associated with erroneous emergency department noncardiac chest pain classifications in men and women

Objective. To analyze the frequency and clinical characteristics associated with erroneous initial classifications of noncardiac chest pain (NCP) in men and women. Methods. We analyzed all case records in which chest pain was initially classified as noncardiac in origin according to clinical signs and electrocardiograms evaluated in our emergency department between 2008 and 2017. We considered the initial evaluation of NCP to be in error if the final diagnosis was acute coronary syndrome. A risk model for an erroneous initial classification of NCP was developed based on multivariable analysis of our patient data. We also used multivariable analysis to explore associations between 10 clinical signs of chest pain and an erroneous initial NCP classification. The data for men and women were analyzed separately. Results. NCP was the initial classification for 8093 women; their median (interquartile range) age was 54 (38-73) years. The classification was in error for 72 women (0.9%). Odds ratios (ORs) showed that patient risk factors associated with an erroneous NCP classification in the women in our series were obesity (OR, 0.40; 95% CI, 0.17-0.97) and cocaine consumption (OR, 5.18; 95% CI, 1.16-23.2). Clinical risk factors associated with erroneous NCP classification in women were recent physical exertion (OR, 2.01; 95% CI, 1.21-3.33), radiation exposure (OR, 2.05; 95% CI, 1.23-3.41), and vegetative symptoms (OR, 1.86; 95% CI, 1.02-3.41). For 9979 men with a median age of 47 (33.64) years, NCP was the initial classification; in 83 of the men (0.8%) the classification was erroneous. Patient factors associated with erroneous NCP classification in men were age over 40 years (OR, 1.74; 95% CI, 1.04-2.91) and hypertension (OR, 0.45; 95% CI, 0.24-0.84). No clinical signs of chest pain in men were associated with error. Conclusions. More clinical characteristics are associated with an erroneous classification of NCP in women. Our findings underline the need to assess the possibility of acute coronary syndrome differently in women, in whom the signs have usually been considered to be atypical.