Sources of subjective variability in the assessment of left ventricular regional wall motion from contrast ventriculograms

Am J Cardiol. 1987 Jul 1;60(1):153-7. doi: 10.1016/0002-9149(87)91003-4.

Abstract

Intraobserver and interobserver variability in the subjective interpretation of angiographic regional wall motion was quantitated in 135 contrast left ventriculograms showing a wide range of normal and abnormal cardiac function, and the effects of rating-scale complexity and myocardial regional differences were examined. Three experienced observers separately graded endocardial motion on a 6-point scale (0 = normal to 5 = dyskinesia) in each of 5 regions. Scores were also tabulated on a 3-point scale and as normal or abnormal. Average intraobserver agreement (2 evaluations per observer) was 68% using the 6-point scale, 86% using the 3-point scale and 90% for normal/abnormal. Interobserver agreement (3 observers) was 47% for exact agreement using the 6-point scale, 75% using the 3-point scale and 80% for normal/abnormal. Interobserver agreement was 84% within a range of 1 wall motion grade on the 6-point scale. Subjectively normal wall motion was most frequent in the 2 basal segments and least frequent in the apical and anterolateral segments. Disagreements were most frequent in the latter 2 segments, but when normal segments were excluded, these interregional differences disappeared. Thus, intraobserver and interobserver agreement is higher than previously reported, and may exceed 80%. Variability depends on whether regional wall motion is normal, but is also affected by the complexity of the rating scale. The distinctions implied by a 6-point subjective rating scale are probably not reliable, but variability is greatly reduced by use of a simplified scoring system.

MeSH terms

  • Angiocardiography*
  • Heart Diseases / diagnosis
  • Heart Diseases / diagnostic imaging
  • Heart Diseases / physiopathology*
  • Humans
  • Myocardial Contraction*