An analysis of accuracy and reliability of emergency department ECG interpretations
Mustafa SEVER1, Özgür KARCIOğLU2, Özgür Aslan3, Fidan SEVER4, ismet PARLAK5, Murat Ersel6
1Harran Üniversitesi Tıp Fakültesi Hastanesi, Acil Tıp Anabilim Dalı, Şanlıurfa
2Dokuz Eylül Üniversitesi Tıp Fakültesi Hastanesi,acil Tıp Anabilim Dalı,izmir
3Dokuz Eylül Üniversitesi Tıp Fakültesi Hastanesi,acil Tıp Anabilim Dalı, İzmir
4Bornova Şifa Tıp Merkezi, İzmir
5Mersin Üniversitesi Tıp Fakültesi Hastanesi, Acil Tıp Anabilim Dalı, Mersin
6Ege Üniversitesi Tıp Fakültesi Hastanesi, Acil Tıp Anabilim Dalı, İzmir
Abstract
Objectives: To determine the accuracy and reliability of emergency physician (EP) interpretations of ECGs obtained in the emergency department (ED).
Materials and Methods: 300 consecutive adult patients of whom ECGs had been obtained in the ED were enrolled in the study. One patient was excluded due to misreplacement of ECG electrodes. Tracings were blindly analyzed and allocated into three categories (Category I: clinically insignificant abnormalities, Category II: potentially significant abnormalities and Category III: clinically significant ECG pathologies) by an EP and a second interpretation by a staff cardiologist as the gold standard. ECG interpretation concordance of EP and cardiologist was evaluated using kappa test.
Results: Discrepancies of interpretations of abnormal ECGs between EP and cardiologists were 42.2%. Ten (3.3%) of these were within Category III, 22 (7.3%) were in Category II and 95 (31.6%) in Category I. The most frequently missed findings were nonspecific ST-T segment elevation and atrial enlargment or hypertrophy. Total rate of discordance was 49.2% including misinterpretations of normal ECGs (n=148, p<0.05, CI: 95%).
Conclusion: Our data suggests presence of clinically significant discordance in interpretations of ED ECGs between EPs and cardiologists