Background and purpose: Blood-based biomarkers could enable early and cost-effective diagnostics for acute stroke patients in the prehospital setting to support early initiation of treatments. However, large prehospital sample sets required for biomarker discovery and validation are missing, and the feasibility of large-scale blood sampling by emergency medical services (EMS) has not been determined. We set out to establish extensive prehospital blood sampling of thrombolysis candidates in the catchment area of our comprehensive stroke center, with a 1.5 million population base.
Methods: EMS personnel were trained to collect prehospital blood samples using a cannula-adapter technique. Time delays, sample quality and performance bottlenecks were investigated between May 20, 2013 and May 19, 2014.
Results: Prehospital blood sampling and study recruitment were successfully performed in 430 thrombolysis candidates, of which 55.3% were admitted outside office hours. The median (interquartile range) emergency call to prehospital sample time was 33 minutes (25-41), and the median time from reported symptom onset or wake-up to prehospital sample was 53 minutes (38-85; n=394). Prehospital sampling was performed 31 minutes (25-42) earlier than admission blood sampling, and 37 minutes (30-47) earlier than admission neuroimaging. Quality control data from 25 participating EMS units indicated a 4-minute increase in median transport time (from arrival on-scene to hospital door) for study patients compared to patients of the preceding year. The hemolysis rate in serum and plasma samples was 6.5% and 9.3% for EMS samples, and 0.7% and 1.6% for admission samples collected with venipuncture.
Conclusions: Prehospital biomarker sampling is feasible in standard EMS units and provides a median timesaving of over 30 minutes to obtain first blood samples. Large biobanks of prehospital blood samples will facilitate development of ultra-acute stroke biomarkers.