Inter-Rater Reliability for Thrombolysis in Cerebral Infarction with TICI 2c Category

J Stroke Cerebrovasc Dis. 2017 May;26(5):992-994. doi: 10.1016/j.jstrokecerebrovasdis.2016.11.008. Epub 2016 Dec 2.

Abstract

Background and purpose: Thrombolysis in cerebral infarction (TICI) with 2b/3 (>50% of occluded territory/complete reperfusion) has been regarded as a successful angiographic outcome. To account for near-perfect angiographic results, the category TICI 2c (near-complete reperfusion) has been introduced. As the degree of inter-rater reliability for TICI with 2c category remains poorly studied, we strived to evaluate the agreement among stroke-treating specialists.

Methods: All consecutive patients, who underwent stent-retriever thrombectomy for acute ischemic stroke in the period between January 2014 and April 2016 at the Department of Neurointerventional Radiology, were analyzed. Digital subtraction angiography (DSA) images were interpreted using previously reported modified TICI score with TICI 2c (near-complete reperfusion). All DSA runs were scored independently by stroke-treating specialist, by consensus of neuroradiologist and stroke neurologist, and by consensus of neurointerventional fellow and attending. Reliability analysis was performed using Krippendorff's alpha (K-alpha).

Results: Sixty-one patients were included into analysis of inter-rater agreement. Mean age was 70 years (SD ± 12), 48% were women, and median admission National Institutes of Health Stroke Scale was 16 (IQR = 12-19). Median admission ASPECTS (Alberta Stroke Program Early CT Score) was 8 (IQR 7-9). Forty patients (65%) received intravenous thrombolysis. Agreement for complete modified TICI scale (compared with consensus of neurointerventional fellow and attending) was as follows: fair for stroke physician (K-alpha .36), moderate for neuroradiologist (K-alpha .48), and moderate for neurointerventional fellow (K-alpha .56). Agreement increased to almost perfect when evaluated by consensus of stroke neurologist and neuroradiologist (K-alpha .82).

Conclusion: Inter-rater agreement for modified TICI increased to almost perfect when scored by consensus of stroke-treating specialists.

Keywords: Inter-rater reliability; category 2c; mechanical thrombectomy; modified TICI; stroke.

Publication types

  • Comparative Study

MeSH terms

  • Aged
  • Aged, 80 and over
  • Angiography, Digital Subtraction*
  • Cerebral Angiography / methods*
  • Cerebral Infarction / diagnostic imaging*
  • Cerebral Infarction / physiopathology
  • Cerebral Infarction / therapy
  • Cerebrovascular Circulation*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Neurologists*
  • Observer Variation
  • Patient Care Team
  • Predictive Value of Tests
  • Radiologists*
  • Reproducibility of Results
  • Retrospective Studies
  • Thrombectomy
  • Thrombolytic Therapy