Abstract
BACKGROUND: Candidacy for venovenous extracorporeal membrane oxygenation is dictated by ECMO to Rescue Lung Injury in Severe ARDS (EOLIA) criteria. We evaluated the effect of modifying candidacy on the basis of escalating demand and limited resources. METHODS: We retrospectively reviewed adult patients diagnosed with COVID-19-related severe acute respiratory distress syndrome who failed to respond to conventional ventilation and required extracorporeal support at our institution. Candidacy was restricted with a published probability model because of supply-demand mismatch and high mortality observed after the first surge. Age <55 years, mechanical ventilation days <3, and minimal comorbidities were prioritized. Primary outcomes included time to decannulation, extubation, tracheostomy, discharge, and death. Hospital and intensive care unit length of stay and hospitalization costs were evaluated. Predictors included cannulation strategy, before and after criteria implementation, use of cytoreductive techniques, timing of tracheostomy, and body mass index. Propensity score matching, multistate Cox proportional hazards models, and generalized linear models were used. RESULTS: Our sample comprised 105 patients, 26 from before criteria implementation ("before" phase) and 79 after ("after" phase). Propensity score results indicated no significant differences in death (P = .152) and costs (P = .063) between the groups. Patients who received cytoreductive therapy had lower total costs (P = .033). Those who underwent single-site cannulation had higher probability of decannulation (P = .009), discharge (P < .001), tracheostomy (P < .001), and extubation alive (P < .001) and lower risk of death (P = .017). CONCLUSIONS: Modifying candidacy by objective criteria with the use of adjunctive therapies may improve outcomes and lower costs during periods of supply-demand mismatch.