Article Text

Download PDFPDF
E-146 Clinical decision making for thrombectomy using CT versus CTP – an interobserver study
  1. P Nazari1,
  2. P Golnari1,
  3. S Ansari2,
  4. M Hurley2,
  5. A Shaibani2,
  6. M Potts1,
  7. B Jahromi1
  1. 1Neurosurgery, Northwestern University, Chicago, IL
  2. 2Radiology, Northwestern University, Chicago, IL


Introduction Accurate and early determination of ischemic penumbra versus infarcted core is crucial in selecting patients who are likely to benefit from thrombectomy in acute ischemic stroke (AIS). CT perfusion (CTP) measures regional cerebral blood flow, blood volume, and transit time, and its secondarily derived estimates of core versus penumbra have been used to help select patients in numerous recent randomized trials of thrombectomy for AIS with great success. However, CTP may overestimate core infarct in early time periods, and some have advocated only using ‘clinical mismatch’ and CT (to rule out an already-established large core infarct) when deciding whether to proceed with thrombectomy. The purpose of this study was to evaluate clinical decision making for thrombectomy based on CT versus CTP by multiple observers; to assess how CTP affected clinical decision making and to investigate the inter-rater agreement in imaging interpretation.

Materials and methods Imaging sets including non-contrast CT (NCCT) and CTP [RAPID (iSchemaView Inc, Menlo Park, CA)] from patients admitted with the diagnosis of large vessel occlusion (MCA and ICA) stroke at a single institution were retrospectively reviewed from 12/15/2016 to 9/15/2017 by five neurovascular interventionists independently. In order to isolate the contribution of imaging to decision-making, all observers were blinded to the demographics and clinical symptoms of patients and were told that imaging were performed between 0 to 6 hours after the stroke. Each interventionist was asked whether they would intervene in AIS first based upon NCCT, and then after viewing CTP images for the same patient. Inter-rater agreement was calculated using Fleiss’ kappa statistic for multiple observers.

Results Datasets of thirty eight consecutive patients (female/male: 14/24; mean age ±SD: 63.24±17.42; side of stroke left/right: 21/17; Thrombectomy no/yes: 19/19) were reviewed. Inter-rater agreement was moderate (κ=0.43) for thrombectomy based on NCCT. Our interventionists had fair agreement (κ=0.33) for thrombectomy based on CTP and had poor agreement (κ=0.17) on change of decision by assessing CTP after NCCT. On average, CTP lead to change of decision for thrombectomy in 13.16% (25) of the cases (table 1 demonstrates values for each interventionist).

Abstract E-146 Table 1

Change in decision regarding thrombectomy after viewing CTP

Conclusion Our study demonstrates heterogeneity in decision-making when only using neuroimaging to determine whether to proceed with thrombectomy within the 0–6 hour time-window. This was true whether CT or CTP was the final imaging modality. Overall, in 13.16% of cases CTP caused the change of decision for thrombectomy. Further work is needed to determine if similar heterogeneity exists within the 6–24 hour time window and/or at other institutions.

Disclosures P. Nazari: None. P. Golnari: None. S. Ansari: None. M. Hurley: None. A. Shaibani: None. M. Potts: None. B. Jahromi: None.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.