How would new york ventilator reallocation policies perform during a COVID-19 surge? An observational cohort study

COVID-19 has created significant strain on the supply of healthcare resources and, during the spring surge in New York City, many hospitals prepared resource allocation policies should the demand for ventilators exceed supply. In such circumstances, resources should remain allocated to patients most likely to survive. Understanding how these guidelines perform is an important consideration in disaster planning. Numerous allocation guidelines exist however nearly all utilize a Sequential Organ Failure Assessment (SOFA) score. We sought to evaluate the performance of ventilator reallocation by applying the New York State Ventilator Allocation Guidelines (NYVAG) to a large cohort of COVID-19 patients. Our retrospective cohort included 895 intubated COVID-19 patients admitted to an academic system in New York City. SOFA scores were calculated for every day of mechanical ventilation. Per NYVAG, patients would have their ventilator reallocated at 48 hours if their interval SOFA score increased, did not change from an initial SOFA of 8-11, or was greater than 11. At 120 hours it would be reallocated if their SOFA score worsened or was greater than 7. At 168 hours and every subsequent 48 hours it would be reallocated if their SOFA score worsened. Ventilator reallocation was simulated and no reallocation was made for any patient. The average SOFA (n=895) at the time of intubation was 7.1 ± 3.6. At the 48-hour reassessment (average SOFA 8.2 ± 3.6, n=759), 436 (57%) patients would have their ventilator reallocated, 145 (33%) of whom would later survive to discharge. At the 120-hour reassessment (average SOFA 7.8 ± 3.6, n=264) 173 (66%) of the 264 remaining simulated ventilated patients would have their ventilators reallocated, 83 (48%) of whom would later survive to discharge. At the 168-hour reassessment (average SOFA 7.8 ± 3.6, n=80) 66 (83%) of the 80 simulated remaining ventilated patients would have their ventilators reallocated. Overall, 685 patients (77%) of the starting cohort would have had their ventilator reallocated at some time during the first 168 hours of mechanical ventilation, 268 (40%) of whom survived to discharge. Our simulated study found that the application of NYVAG to the COVID-19 surge at one academic system would have resulted in a significant portion of ventilated patients having had their ventilators reallocated. This may be deeply concerning as a significant portion of patients ultimately survived to discharge. These results call for further confirmatory studies and have implications for optimal resource allocation strategies during pandemics.