Varhaisen varoituksen mittareiden toimivuus suomalaisessa ensihoidon potilasaineistossa

Background Early warning scores (EWSes) are an established tool for identifying clinical deterioration, with the capability of preventing excess mortality and cardiac arrest. Their effectiveness in a pre-hospital setting is less explored. We set out to compare the performance of four EWSes as well as dichotomous medical emergency team criteria for mortality, serious adverse events, and transportation to a tertiary referral center among heterogeneous pre-hospital patients after adjustment for multiple confounding factors. Methods The cohort consists of all 1,629 adults treated by Emergency Medical Services (EMS) and transported to the tertiary referral center emergency department during June 2015. EWSes were retrospectively calculated based on physiological observations documented at the scene. Five different scoring systems were compared against the outcome of interest: 48-hour mortality. A secondary endpoint was a 48-hour serious adverse event (SAE), defined here as a combined endpoint of 48-hour mortality, MET alert, Intensive Care Unit (ICU) admission or in-hospital cardiac arrest within 48 hours of hospital admittance. Results The National Early Warning Score (NEWS) prevailed as the best performing EWS, with the area under the receiver-operating characteristic (AUROC) of 0.81 (0.69–0.93) against 48-hour mortality, with other EWSes being less sensitive. The dichotomous criteria were insensitive to outcome. Patients with NEWS ≥7 had an elevated risk against 48-hour mortality of OR 10.9 (95 CI 3.66–26.67). Conclusions Vital dysfunctions are independent predictors of morbidity and mortality among pre-hospital patients. Elevated NEWS is associated with increased adverse outcomes. The NEWS is validated in the pre-hospital setting and can be expected to improve patient safety and outcomes