258 Can clinical experts agree on a diagnosis of TIA

Background Transient ischaemic attacks (TIAs) are sudden onset focal neurological deficits of vascular pathogenesis that resolve within 24 hours. TIAs remain a diagnostic challenge due to its clinical heterogeneity and lack of biomarkers. Previous studies on the diagnostic accuracy of TIAs by non-specialists have used TIA experts (neurologists or stroke physicians) as the gold standard. However, the inter-rater variability in TIA diagnoses among these experts is not firmly established. Here we conduct a meta-analysis of the inter-rater variability of TIA diagnosis between experts. Methods We performed a systematic review of studies from 1988 to present day using MEDLINE, EMBASE, and PubMed. Two reviewers independently screened for eligible studies and extracted inter-rater variability measurements using Cohen’s Kappa scores between expert clinicians. Results 14 original studies reporting on variability in TIA diagnosis for 15,907 patients were found. Meta-analysis revealed an overall agreement between experts of κ=0.73 (95%CI 0.63–0.84) Conclusions Overall agreement between experts is good for TIA diagnosis, but variation still exists for a sizable proportion of cases. Further research into reliability of the accepted gold standard as well as alternative clinical, imaging or blood biomarkers are required given the important clinical implications for making a TIA diagnosis.