Abstract 12020: Symptom Clusters in Patients Ruled Out for Acute Coronary Syndrome Do Not Differ by Gender

Abstract

Introduction: Patients with potential acute coronary syndrome (ACS) present to the emergency department (ED) on the basis of their symptoms. Determining the likelihood that symptoms represent ACS and warrant cardiac evaluation remains challenging. Few studies have reported symptoms and symptoms clusters prospectively in patients ruled out for ACS, yet such profiles could assist in risk stratification.

Hypotheses: In a cohort of ED patients with symptoms suggestive of ACS who were subsequently ruled out for ACS, we hypothesized that: 1) symptoms would differ by gender; 2) subgroups of patients with similar symptom clusters could be identified.

Methods: A sample of 318 patients (148 women; 170 men) was recruited from 3 EDs in 3 cities. Symptom data were collected prospectively upon ED presentation using a valid 13 item Symptom Checklist. After initial stabilization, measures of functional status, clinical status, and demographics were collected by interview. Chi-square was used to compare symptoms by gender. Latent class analysis was used to identify patient groups (latent classes) with similar symptom profiles.

Results: Mean age was 57.6 (range 21-98). The most frequently reported symptom in both genders was chest discomfort (69.4%). Women were more likely than men to report chest pressure (69% vs. 54%, p = .008), lightheadedness (55% vs. 42%, p = .024) and upper back pain (37% vs. 24%, p = .017). Analysis of 13 symptoms resulted in a 3 class solution. Class 1 (n = 114) included 7 high probability symptoms: chest pressure, chest discomfort, shortness of breath, unusual fatigue, nausea, lightheadedness, chest pain. Class 2 (n = 102) included 3 high probability symptoms, all in the chest including pressure, discomfort, and pain. Class 3 (n = 102) included 3 moderate probability symptoms: shortness of breath, unusual fatigue, lightheadedness. Class 3 was significantly older: 63.6 years vs. 54 years (class 1) and 55.7 years (class 2) (p<.0167). Gender improved the fit of the model, but the classes did not differ by gender.

Conclusions: Patients who present to the ED with ACS-like symptoms and are ruled out can be classified into subgroups by symptom clusters that do not differ by gender. The role of symptom clusters as a risk stratification tool in ACS requires further study.