Point of care coagulometry in prehospital emergency care: an observational study

Abstract
Haemostatic impairment can have a crucial impact on the outcome of emergency patients, especially in cases of concomitant antithrombotic drug treatment. In this prospective observational study we used a point of care (POC) coagulometer in a prehospital physician-based emergency medical system in order to test its validity and potential value in the treatment of emergency patients. During a study period of 12 months, patients could be included if venous access was mandatory for further treatment. The POC device CoaguChek® was used to assess international normalized ratio (INR) after ambulance arrival at the scene. Results were compared with in-hospital central laboratory assessment of INR. The gain of time was analysed as well as the potential value of POC testing through a questionnaire completed by the responsible prehospital emergency physician. A total of 103 patients were included in this study. POC INR results were highly correlated with results of conventional assessment of INR (Bland-Altman-bias: 0.014). Using a cutoff value of INR >1.3, the device’s sensitivity to detect coagulopathy was 100 % with a specificity of 98.7 %. The median gain of time was 69 min. Treating emergency physicians considered the value of prehospital POC INR testing ‘high’ in 9 % and ‘medium’ in 21 % of all patients. In patients with tracer diagnosis ‘neurology’, the value of prehospital INR assessment was considered ‘high’ or ‘medium’ (63 %) significantly more often than in patients with non-neurological tracer diagnoses (24 %). Assessment of INR through a POC coagulometer is feasible in prehospital emergency care and provides valuable information on haemostatic parameters in patients. Questionnaire results suggest that POC INR testing may present a valuable technique in selected patients. Whether this information translates into an improved management of respective patients has to be evaluated in further studies.

This publication has 18 references indexed in Scilit: