Background Bedside international normalised ratio (INR) testing in emergency departments (ED) to assess for coagulopathies has not previously been examined.

Objectives To compare point-of-care INR (POCINR) testing in an ED with laboratory-based results (LABINR) using a prospective observational cohort study.

Methods Prospectively recruited patients requiring laboratory INR testing had simultaneous POC testing using a CoaguChek XS Plus in an urban tertiary hospital ED. Clinicians were blinded to the POC result and clinical decisions were based on the laboratory result. Result agreement was considered a priori to be a POCINR result within 15% of the laboratory result. Secondary outcomes included potentially incorrect management if POCINR had been used as the diagnostic test.

Results Two hundred and ninety-three patients were included. Agreement within a 15% range occurred in 245 patients (83.6% (95% CI 78.9% to 87.4%)). Following independent medical record review by the authors, no patients with POCINR agreement within a 30% range of LABINR would have had changes in management. Eleven patients had POCINR results which may have resulted in significant changes to management. Ten patients were incorrectly identified by the POCINR as having a clinically significant coagulopathy not confirmed by the LABINR result, which may have resulted in inappropriate management. One patient with snake venom-induced coagulopathy had a normal POCINR (1.4). No other patients with a normal POCINR had a clinically significant coagulopathy.

Conclusions POCINR testing can exclude clinically significant coagulopathy in the ED. LABINR is required to confirm non-normal INR results, particularly in the supratherapeutic range.