Predicting Futility in Severely Injured Patients: Using Arrival Lab Values and Physiology to Support Evidence-Based Resource Stewardship

Abstract

BACKGROUND: The recent pandemic exposed a largely unrecognized threat to medical resources, including daily available blood products. Some of the most severely injured patients who arrive in extremis consume tremendous resources yet succumb shortly after arrival. We sought to identify cut points available early in the patient’s resuscitation that predicted 100% mortality. STUDY DESIGN: Cut points were developed from a previously collected data set of all level 1 trauma patients admitted January 2010 to December 2016. Objective values available on or shortly after arrival were evaluated. Once generated, we then validated these variables against (1) a prospective data set November 2017 to October 2021 of severely injured patients and (2) a multicenter, randomized trial of hemorrhagic shock patients. Analyses were conducted using STATA 17.0 (College Station, TX), generating positive predictive value (PPV), negative predictive value, sensitivity, and specificity. RESULTS: The development data set consisted of 9,509 patients (17% mortality), with 2,137 (24%) and 680 (24%) in the two validation data sets. Several combinations of arrival vitals and labs had 100% PPV. Patients undergoing CPR in the field or on arrival (with subsequent return of spontaneous circulation) required lower fibrinolysis LY-30 (30%) than those with systolic blood pressures of ≤50 (30 to 50%), ≤70 (80 to 90%), and ≤90 mmHg (90%). Using a combination of these validated variables, the Suspension of Transfusions and Other Procedures (STOP) criteria were developed, with each element predicting 100% mortality, allowing physicians to cease further resuscitative efforts. CONCLUSIONS: The use of evidence-based STOP criteria provides cut points of futility to help guide early decisions for discontinuing aggressive treatment of severely injured patients arriving in extremis.