Reliability of brain CT evaluation by stroke neurologists in telemedicine

Abstract

Objective: To determine the reliability and therapeutic impact of standardized cerebral CT evaluation and quantification of early ischemic changes (EIC) with the Alberta Stroke Program Early CT Score (ASPECTS) by stroke neurologists in the Stroke Eastern Saxony Network (SOS-NET), which provides telemedical consultations for patients with acute ischemic stroke.

Methods: Two neuroradiologists re-evaluated all CT scans of consecutive SOS-NET patients in 2009 blinded to clinical information providing reference standard. We defined discrepant CT findings as all false-positive or false-negative EIC and brain pathology findings and ASPECTS deviations >1 point. We subsequently discussed the clinical impact of discrepant CT findings unblinded to clinical information. Weighted kappa (κ(w)) statistic was used to determine the interobserver agreement for ASPECTS.

Results: Of 582 patients, complete imaging data were available for 536 patients (351 cerebral ischemic events, 105 primary intracranial hemorrhages, and 80 stroke mimics). The neuroradiologists detected discrepant CT findings in 43 patients (8.0%) that were rated as clinically relevant in 9 patients (1.7%). Stroke neurologists recommended IV thrombolysis in 8 patients despite extensive EIC (ASPECTS ≤5). One of these patients had symptomatic intracranial hemorrhage. In 1 nonthrombolyzed patient, the stroke neurologist missed subdural hematoma. The interobserver agreement on ASPECTS between stroke neurologists and expert readers was substantial (κ(w) = 0.62; 95% confidence interval 0.54-0.71).

Conclusions: Clinically relevant misinterpretation of the CT scans was rare in our acute telestroke service. ASPECTS is a reliable tool to assess the extent of EIC by stroke neurologists in telemedicine in real time.