Frequency, characteristics, and outcome of patients hospitalized with acute coronary syndromes with undetermined electrocardiographic patterns Academic Article uri icon


  • dial infarction. The case report form also included details regarding the demographic, clinical, and ECG characteristics of the patient, diagnostic and treatment modalities, in-hospital complications, and discharge status. Patient follow-up was available at 30 days for 90% of the patients. Patients with ACS were divided into 2 groups based on their initial ECG pattern as interpreted by the attending admitting physician: undetermined ECG or determined ECG patterns (including ST elevation or non-ST elevation). We compared the clinical charac- teristics, clinical course while in-hospital, treatment, and the in-hospital and 30-day outcomes between the 2 groups. The chi-square and Student'st tests were used to determine the significance of differences be- tween variables and means, respectively. Unadjusted and multivariate adjusted odds ratios (with 95% con- fidence intervals (CI)) of in-hospital and 30-day mor- tality were calculated using logistic regression analy- sis. ECG patterns (undetermined vs determined), age, gender, prior myocardial infarction, Killip class, dia- betes, and hypertension were included in the multi- variate model. Continuous variables are presented as mean SD. SAS software (SAS Institute, Cary, North Carolina) was used for statistical analysis. Of the 10,484 patients enrolled in the Euro Heart Survey of ACS, 686 patients (6.5%) were admitted with an undetermined ECG pattern; the remaining 9,798 patients (93.5%) had determined ECG patterns. The rhythm on the initial electrocardiogram among the undetermined group was atrial fibrillation in 86 patients (12.5%), paced rhythm in 60 (8.7%), ventric- ular or supraventricular tachycardia in 14 (2%), ad- vanced atrioventricular block in 11 (1.6%), junctional rhythm in 7 (1%), and sinus rhythm in 467 (68.1%). One hundred fifty-five patients (22.6%) had left bun- dle branch block of uncertain duration, and 81 (11.8%) presumably had new left bundle branch block. In 41 patients (6%), the initial ECG description was missing in the case report, although it was avail- able to the physician who categorized it as being an undetermined ECG pattern. Fifty-one percent of the patients with an undeter- mined ECG pattern were hospitalized in coronary care units, 25.4% in cardiology departments, and 20.4% in internal medicine departments, compared with 63%, 21.9%, and 13.4%, respectively, in the determined ECG pattern group (p 0.001 for comparison be- tween the 2 groups). Thus, patients with undetermined ECG patterns were less likely to be hospitalized in coronary care units and more likely to be hospitalized in internal medicine wards compared with the deter-

publication date

  • January 1, 2003