ABSTRACT
OBJECTIVE: To model performance of sequential organ failure assessment (SOFA) score-based ventilator allocation guidelines during the COVID-19 pandemic. METHODS: A retrospective cohort study design was used. Study sites included three New York City hospitals in a single academic medical center. We included a random sample (205) of adult patients intubated (1002) from 3/25/20-4/29/20. Protocol criteria adapted from New York State's 2015 guidelines were applied to determine which patients would have had mechanical ventilation withheld or withdrawn. RESULTS: 117 (57%) patients would have been identified for ventilator withdrawal or withholding, based on the triage guidelines. Of those 117 patients, 28 (24%) survived hospitalization. Overall, 65 (32%) patients survived to discharge. CONCLUSIONS: Triage protocols aim to maximize survival by redirecting ventilators to those most likely to survive. Over half of this sample would have been identified as candidates for ventilator exclusion. Clinical judgment would therefore still be needed in ventilator reallocation, re-introducing bias and moral distress. These data suggest limited utility for SOFA score-based ventilator rationing. This raises the question of whether there is sufficient ethical justification to impose a life-ending decision based on a SOFA scoring method on some patients to offer potential benefit to a modest number of others.