An assessment of ST-segment measurement variability between two core electrocardiogram laboratories

J Electrocardiol. 2014 Jan-Feb;47(1):38-44. doi: 10.1016/j.jelectrocard.2013.10.005. Epub 2013 Oct 18.

Abstract

Objectives: We evaluated inter-reader agreement of the ST-segment between two electrocardiogram (ECG) core laboratories.

Background: Accurate measurement of the ST-segment is key to diagnosis and management of acute coronary syndromes (ACS). Clinical trials also rely on adherence to the pre-specified ECG eligibility criteria.

Methods: 150 patients (100 ST-segment elevation (STE)-ACS, 50 non-STE-ACS) were selected. An experienced ECG reader from each laboratory measured ST-segment deviation on the baseline ECGs (nearest 0.1mm).

Results: ∑ST-segment deviation showed excellent inter-reader agreement (R=0.965, intraclass correlation coefficient (ICC) 0.949, 95% CI (0.930-0.963)). Similar agreement was observed when ∑ST-segment elevation (∑STE) and ∑ST-segment depression (∑STD) were assessed separately. Better agreement was evident in STE-ACS cohort (ICC (95% CI): 0.968 (0.953-0.978, 0.969 (0.954-0.979), 0.931 (0.899-0.953)) compared to NSTE-ACS patients (ICC (95% CI): 0.860 (0.768-0.917), 0.816 (0.699-0.890), 0.753 (0.605-0.851) across measurement of ∑ST-segment deviation, ∑STE, and ∑STD.

Conclusions: We demonstrated excellent agreement on ST-segment measurements between two experienced readers from two ECG core laboratories.

Keywords: Acute coronary syndromes; Electrocardiogram; Reliability.

Publication types

  • Comparative Study
  • Multicenter Study

MeSH terms

  • Aged
  • Alberta
  • Electrocardiography / methods*
  • Electrocardiography / statistics & numerical data*
  • Female
  • Humans
  • Laboratories, Hospital / statistics & numerical data*
  • Male
  • Middle Aged
  • Missouri
  • Myocardial Infarction / diagnosis*
  • Observer Variation*
  • Reproducibility of Results
  • Sensitivity and Specificity