Related Experiment Videos

Interrater reliability of an etiologic classification of ischemic stroke

C J Johnson1, S J Kittner, R J McCarter

  • 1Department of Neurology, Johns Hopkins Bayview, Baltimore, MD 21224.

Stroke
|January 1, 1995
PubMed

Insights

Developing an etiology-specific ischemic stroke classification improves accuracy for research. This system demonstrated fair to good reliability, aiding epidemiological and pathophysiological studies of stroke causes.

Area of Science:

  • Neurology
  • Epidemiology
  • Medical Informatics

Background:

  • Accurate stroke cause identification is vital for research and clinical practice.
  • Existing ischemic stroke classifications show significant variability.
  • An etiology-specific classification system was developed to address this need.

Purpose of the Study:

  • To develop and implement an etiology-specific classification for ischemic stroke.
  • To evaluate the interrater reliability of this new classification system.

Main Methods:

  • Reviewed 160 ischemic stroke cases in young adults.
  • Paired neurologists assigned cases to prioritized etiological categories.
  • Interrater agreement was assessed using kappa statistics.

Main Results:

  • Kappa values indicated fair to good reliability for most stroke causes.
  • Lacunar stroke showed the greatest pair-to-pair variation (kappa = 0.31).
  • Strokes of undetermined or hematologic cause had borderline fair reliability.

Conclusions:

  • The utility of a stroke classification depends on its application.
  • An etiologic classification is valuable for stroke epidemiology and pathophysiology research.
  • Explicit criteria can achieve fair to good reliability for stroke etiology classification.
Abstract

Related Concept Videos