Hostname: page-component-586b7cd67f-t7fkt Total loading time: 0 Render date: 2024-11-26T06:03:36.381Z Has data issue: false hasContentIssue false

How Common SOFA and Ventilator Time Trial Criteria Would Have Performed During the COVID-19 Pandemic: An Observational Simulated Cohort Study

Published online by Cambridge University Press:  09 June 2022

B. Corbett Walsh*
Affiliation:
Division of Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, New York University Grossman School of Medicine, New York, NY, USA
Deepak Pradhan
Affiliation:
Division of Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, New York University Grossman School of Medicine, New York, NY, USA New York University Langone Health, New York, NY, USA Bellevue Hospital Center, NYC Health & Hospitals, New York, NY, USA
Vikramjit Mukherjee
Affiliation:
Division of Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, New York University Grossman School of Medicine, New York, NY, USA Bellevue Hospital Center, NYC Health & Hospitals, New York, NY, USA
Amit Uppal
Affiliation:
Division of Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, New York University Grossman School of Medicine, New York, NY, USA Bellevue Hospital Center, NYC Health & Hospitals, New York, NY, USA
Mark E. Nunnally
Affiliation:
New York University Langone Health, New York, NY, USA Department of Anesthesiology, Perioperative Care, and Pain Medicine, New York University Grossman School of Medicine, New York, NY, USA
Kenneth A. Berkowitz
Affiliation:
Division of Pulmonary, Critical Care, and Sleep Medicine, Department of Medicine, New York University Grossman School of Medicine, New York, NY, USA National Center for Ethics in Health Care, Veterans Health Administration, Washington, DC, USA Division of Medical Ethics, Department of Population Health, New York University Grossman School of Medicine, New York, NY, USA
*
Corresponding author: B. Corbett Walsh, Email: [email protected].
Rights & Permissions [Opens in a new window]

Abstract

Objectives:

To evaluate how key aspects of New York State Ventilator Allocation Guidelines (NYSVAG)—Sequential Organ Failure Assessment score criteria and ventilator time trials —might perform with respect to the frequency of ventilator reallocation and survival to hospital discharge in a simulated cohort of coronavirus disease (COVID-19) patients.

Methods:

Single center retrospective observational and simulation cohort study of 884 critically ill COVID-19 patients undergoing ventilator allocation per NYSVAG.

Results:

In total, 742 patients (83.9%) would have had their ventilator reallocated during the 11-day observation period, 280 (37.7%) of whom would have otherwise survived to hospital discharge if provided with a ventilator. Only 65 (18.1%) of the observed surviving patients would have survived by NYSVAG. Extending ventilator time trials from 2 to 5 days resulted in a 49.2% increase in simulated survival to discharge.

Conclusions:

In the setting of a protracted respiratory pandemic, implementation of NYSVAG or similar protocols could lead to a high degree of ventilator reallocation, including withdrawal from patients who might otherwise survive. Longer ventilator time trials might lead to improved survival for COVID-19 patients given their protracted respiratory failure. Further studies are needed to understand the survival of patients receiving reallocated ventilators to determine whether implementation of NYSVAG would improve overall survival.

Type
Original Research
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of Society for Disaster Medicine and Public Health, Inc.

By the time the World Health Organization (WHO) declared coronavirus disease (COVID-19) a pandemic on March 11, 2020, 113 countries had confirmed cases, with 80 955 cases reported in China and 696 in the United States. 1,2 By the end of that month, 5% of all worldwide cases 3 and 1 in 3 of US cases and deaths were in New York City. 4,5 This created a real concern that hospitals might suffer a shortfall of resources and need to ration ventilators. Reference Cha, Bernstein and Sellers6Reference Rosenthal and Goldstein9 Developed in 2007 and updated in 2015 to address such a scenario, the New York State Department of Health along with New York State Task Force on Life and the Law released the New York State Ventilator Allocation Guidelines (NYSVAG). 10 This document draws upon a community’s duty to care, resource stewardship, transparency, and distributive justice to save the most lives. Briefly, this unvalidated strategy heavily utilizes the Sequential Organ Failure Assessment (SOFA), a summative 0-24 score of organ dysfunction across neurologic, pulmonary, cardiovascular, hematologic, hepatobiliary, and renal domains, to prioritize (colored red for highest or yellow for intermediate) and potentially exclude patients from a ventilator (colored blue). Patients triaged and provided a ventilator receive 1 for a pre-specified duration of time. After the ventilator time trial, their SOFA scores are recalculated, and patients are re-triaged to maintain their ventilators or have them reallocated to other patients. NYSVAG are very similar to other available crisis standard of care guidelines, Reference White and Lo11Reference Piscitello, Kapania and Miller15 62.1% of which utilize a SOFA score. Moreover, and similar to NYSVAG, 55.6% exclude patients from a ventilator for a SOFA score > 11, and 65.5% use the same ventilator time trial intervals (2 days, 5 days, and every subsequent 2 days). Reference Fiest, Krewulak and Plotnikoff14,Reference Piscitello, Kapania and Miller15 With only 1 documented case of ventilator rationing by NYSVAG exclusionary critiera in New York City during the original surge, Reference Fins16 and with Alaska 17 and Idaho 18 having both activated regional crisis standards of care that explicitly permit ventilator rationing (Alaska uses verbatim SOFA cutoffs as NYSVAG and Idaho use a similar SOFA and time trial criteria as NYSVAG), it remains crucial to understand how these guidelines might perform.

Objectives

It remains uncertain how such ventilator allocation guidelines would perform if implemented. This study simulates how specific SOFA cutoffs for resource prioritization and the ventilator time trials, used in NYSVAG and other crisis standards of care documents, might affect ventilator allocation and patient survival in a COVID-19 cohort during the pandemic. To address this research question, the relevant portions NYSVAG were simulated on every patient independently to one another and independent to the number of available ventilators, thereby aligning temporally discordant patient hospitalizations. This design can reveal important guideline inefficiencies that might potentiate mortality and remain obscure when incorporating a fixed ventilator supply with variable or low ventilator strain. Moreover, this analysis is helpful to evaluate how explicit triage criteria (SOFA cutoffs and ventilator time trials) would perform in an isolated manner that is independent to the resources available at a facility. Such an evaluation can determine whether a resource allocation strategy may be effective given the trajectory of patient survival or whether it might reallocate resources away from patients who would otherwise survive. Additionally, a retrospective analysis where each patient received a ventilator provides a unique control group to simulate resource allocation protocols because their survival with a ventilator is known. The results of this analysis remain germane to developing efficient crisis standard of care procedures and could provide reassurance of existing ventilator allocation guidelines or suggest the need for further revisions.

Methods

Adult patient charts were retrospectively reviewed from a single academic hospital system encompassing 1 tertiary care referral center and 2 community hospitals from 3/1/2020–7/1/2020. Patients were included if they were SARS-CoV-2 positive and were invasively mechanically ventilated. Components of the SOFA score were obtained for each patient daily as close to 10:00 am as possible. Missing components of the SOFA score were presumed normal. Reference Knaus, Wagner and Draper19,Reference Wunsch, Hill and Bosch20 If there was no reported SOFA score on the day of intubation, the SOFA score from the next day was used instead for the day of intubation (and for relevant comparisons). Every patient received a ventilator as medically indicated, and ventilator allocation or reallocation per NYSVAG only occurred in a virtual simulation.

NYSVAG were interpreted for this study in the following manner: Step 1, the application of exclusionary criteria, was deferred because the number of such patients was negligible (Supplement Figure 1). Reference Wunsch, Hill and Bosch20 Steps 2 and 3, the allocation and potential reallocation of ventilators respectively, were simulated on each patient as described in NYSVAG (Table 1; Supplement Figures 2, 3a-b). Patients categorized as Red received the highest priority for a ventilator, followed by those patients classified as Yellow. Patients classified as Blue were ineligible for a ventilator.

Table 1. New York State ventilator allocation guidelines steps 2 and 3 10

a Three gaps in NYSVAG that were addressed as follows: (1) Patients with a SOFA score of 7 on the day of intubation were categorized as Red; (2) individuals with a SOFA score greater than 7 on any assessment after 5 days of mechanical ventilation were ineligible for a ventilator; and (3) patients ventilated for 7 or more days who continued to show a SOFA score ≤ 7 and decreasing were placed in the Yellow category.

NYSVAG were applied to the cohort in the following simulation (see Supplemental Methods for additional details): Every patient’s SOFA score was calculated on the day of intubation, Day 2 of mechanical ventilation, Day 5, and every subsequent 2 days of mechanical ventilation. Any patient who was characterized as Red or Yellow would receive (or maintain) a ventilator; however, any patient characterized as Blue would have her or his ventilator withheld (or reallocated) on the day that patient was characterized as ineligible for a ventilator. This design was chosen because this analysis can evaluate a central component of the guidelines (SOFA score cutoffs and ventilator time trials) independent to the number of available ventilators, patient throughput, ventilator strain, or impact from randomization. Patients who had their ventilator withheld or reallocated were simulated to expire as modeled in other simulation studies withholding or withdrawing life-sustaining treatment. Reference Kanter21,Reference Bhavani, Luo, Miller, Sanchez-Pinto, Han, Mao, Sandıkçı, Peek, Coopersmith, Michelson and Parker22 Patients were also removed from the simulation if they expired, were extubated, or placed on extracorporeal membrane oxygenation (ECMO) between assessment days.

Outcome measurements include both observed (what occurred in the real world) and simulated (what occurred hypothetically) endpoints. Because every patient in our cohort received a ventilator in the real world, their performance with respect to survival to hospital discharge with continued ventilation is known (referenced as an observed outcome). The primary observed endpoints include death or survival to discharge. The primary simulated endpoints include whether the patient’s ventilator was simulated to be withheld or reallocated on the day of intubation, 2, 5, 7, 9, or 11 days post-intubation (Figure 1). Combined simulated-observed endpoints include whether the patient had her or his ventilator reallocated (simulated) but was observed to survive to hospital discharge (observed, how the patient would have performed if provided with a ventilator). Comorbidities were queried from the ICD-10 problem list.

Figure 1. Study Methodology. Patients were observed from time of intubation until observed endpoints of death or survival to discharge. Patient’s had SOFA reassessments on day 2, 5, 7, 9, and 11 to determine if their ventilator would have been reallocated in the simulation. Patients received a ventilator in the simulation if their SOFA score on day of intubation was less than 12. On the Day 2 Reassessment, patients would have had their ventilator reallocated if their interval SOFA score increased, did not change from an initial SOFA of 8-11, or was greater than 11. On the Day 5 Reassessment, patients would have had their ventilator reallocated if their SOFA score increased, was greater than 7, or was less than 7 and without a change in their previous SOFA score. On the Day 7 Reassessment and every 2 days thereafter, patients would have had their ventilator reallocated if their SOFA score did not improve or was greater than 7. A patient who did not have their ventilator reallocated in the simulation (No *) would only progress to the subsequent simulated ventilator reallocation assessment if they were not extubated, expired, or placed on ECMO prior to that assessment.

Missing data were amended in 2 subsequent simulations. First, all Glasgow Coma Scale (GCS) SOFA subscores were normalized (GCS-nl). Second, missing pulmonary SOFA subscores were supplemented by the modified SOFA pulmonary category (mSOFA-FiO2). As oxygen saturation was not gathered and FiO2 was, pulmonary mSOFA scores were computed with a presumptive oxygen saturation of 85% or better. A fourth simulation delayed the Day 2 reassessment to Day 5, the Day 5 reassessment to Day 7, and the subsequent 2-day reassessments to start on Day 9 (MVD 2to5).

Categorical data were analyzed with the chi-square test using the 95% confidence interval (CI). Continuous variables were compared with a 2-tailed t-test using a 5% significance level.

This study was approved by the NYU Grossman School of Medicine Institutional Review Board (study #i20-01653). No treatment, ventilator allocation, or reallocation decisions were made for any patient. Patients were exempt from informed consent.

Results

Observed Results

COVID-19 patients, 884 in total, were intubated between 3/1/2020 and 7/1/2020 and included in the simulation. The average age of the cohort was 63 years, predominately male (70.0%), with varied racial ethnicities (Table 2). Comorbidity information was only available for 50% the cohort (Supplement Table 1). The median (IQR) duration of mechanical ventilation for the cohort was 10.2 (3.74, 23.73) days, with 41.0% survival to discharge.

Table 2. Patient cohort demographics

Simulation Results

NYSVAG were applied to the observed patient SOFA scores in a simulation shown in Figure 2 and Table 3. Table 4 depicts how simulated patients, on each day of reassessment, were triaged as ineligible to receive or maintain a ventilator (Blue) and how many were observed to survive to discharge.

Figure 2. Sankey diagram of a limited NYSVAG simulation on a cohort of COVID-19 patients. Each color represents the number of patients triaged to that color on the day of triage (Day 0 or potential intubation, Day 2, 5, 7, 9, and 11). Patients triaged as Red or Yellow received or maintained their ventilator until the next Reassessment Day while those triaged as Blue had their ventilator withheld or reallocated and were simulated to expire. Flow between each day of assessment describes the number of patients who moved between triage categories. The paths from left to the far right bars describe the course of patients who satisfied the combined simulated-observed endpoint of patients who had their ventilator reallocated (simulated) but were observed to survive to hospital discharge (observed, how the patient would have performed if provided a ventilator in the simulation). Patients exited the simulation because they were extubated, expired, or placed onto ECMO.

Table 3. Numerical representation of a simulated application of NYSVAG to a cohort of critically ill COVID-19 patients

Numerical representation of Figure 2 depicting the simulated application of NYSVAG on a cohort of critically ill COVID-19 patients.

a Survival to hospital discharge is an observed outcome.

b Characterized as ineligible for a ventilator (Blue) and simulated for ventilator removal/withholding is a simulated outcome.

c Combined simulated-observed outcome; those who were classified as ineligible for a ventilator (Blue) in the simulation but were also observed to survive to hospital discharge when they were provided a ventilator in the real world.

Table 4. Etiology of patients triaged as Blue and ineligible for a ventilator during simulation

Etiology of patients classified as Blue and triaged as ineligible for a ventilator on the specified reassessment day of mechanical ventilation.

a See Figure 1 and Table 1 regarding specific criteria triaging patients as Blue or ineligible to receive or maintain a ventilator for the specific day reassessed.

b P value comparing SOFA > 11 of the same day.

c P value comparing SOFA 8-11 of the same day.

Step 2: Assessment at intubation

In total, 884 patients underwent Step 2 of NYSVAG—prioritization and allocation of a ventilator. Average SOFA on the day of triage for mechanical ventilation was 7.4 ± 3.3; 478 (54.1%) individuals qualified as Red, the highest priority for a ventilator, 285 (32.2%) as Yellow or intermediate priority for a ventilator, and 121 (13.7%) as Blue or ineligible for a ventilator. Of the 121 individuals triaged as Blue and not provided with a ventilator in the simulation, 31 (25.6%) were observed to survive to discharge. Of the 763 patients who were initially prioritized to receive a ventilator (Red or Yellow), 249 (32.6%) would later be triaged in the simulation to have their ventilator reallocated (Blue) but would have otherwise been observed to survive to discharge (observed outcome).

Step 3: Day two reassessment

From the day of intubation to 2 days post-intubation, 41 individuals had expired, 35 had been extubated, and 10 had been placed onto ECMO. As such, 678 patients were reassessed on post-intubation Day 2. Of this group, 223 (32.9%) were characterized as Red and 50 (7.4%) as Yellow. Per NYSVAG, for those being reassessed after 2 days of mechanical ventilation, 404 individuals (59.7%) would have had their ventilator reallocated, 136 (33.7%) of whom were observed to survive to discharge. Among these 404 individuals, both the 90 individuals (22.3% of all reassessed as Blue on this reassessment, 47.8% observed survival) who had a worsening SOFA score less than 7 and the 204 individuals (50.5%, 35.3% survival) who had a worsening SOFA score of 8-11 were more statistically likely to be discharged alive if provided a ventilator than those 110 individuals (27.2%, 19.1% survival) who had a SOFA score greater than 11 (P < 0.001 and P = 0.003, respectively). Individuals characterized as Blue with a worsening SOFA score less than 7 were also more likely to survive to hospital discharge with a ventilator than those with a SOFA score of 8-11 (P = 0.04).

Step 3: Day five reassessment

Of the 678 patients from Day 2, 24 individuals were extubated, 16 expired, 7 were placed onto ECMO, and 405 were extubated due to the ventilator reallocation that occurred because of the Day 2 reassessment. As such, 226 patients were present to be reassessed after 5 days of mechanical ventilation. At this time, 25 individuals (11.1%) were triaged as Red and 39 (17.3%) as Yellow; 162 patients (71.7%) were triaged as Blue and were simulated to have their ventilator reallocated, 75 (46.3%) of whom were observed to survive to discharge. Among these 162 individuals, both the 88 individuals (54.3% of all reassessed as Blue, 55.7% observed survival) who had a worsening SOFA score less than 7 and the 61 individuals (37.7%, 41.0% survival) who had a SOFA score of 8-11 were more statistically likely to be discharged alive if provided with a ventilator than those 13 individuals (8.0%, 7.7% survival) who had a SOFA score greater than 11 (P = 0.001 and P = 0.02, respectively).

Step 3: Day seven reassessment

Of the 226 patients from Day 5, 5 individuals were extubated, 3 expired, and 162 were extubated due to the ventilator reallocation on the Day 5 Reassessment. As such, 56 patients were present to be reassessed after 7 days of mechanical ventilation. At this time, 7 individuals were prioritized for a ventilator. 49 patients (87.5%) were characterized as Blue and simulated to have their ventilator reallocated, 34 (69.4%) of whom were observed to survive to discharge. Amongst these 49 individuals, the 42 individuals (85.7% of all reassessed as Blue, 78.6% observed survival) who had a worsening SOFA score less than 7 were more statistically likely to be discharged alive if provided with a ventilator than and the 7 individuals (14.3%, 14.3% survival) who had a SOFA score of 8-11 (P < 0.001).

After 11 days of simulation, all 884 patients would have either been extubated and survived to hospital discharge, expired, had their ventilator reallocated, or placed onto ECMO; 742 patients would have had their ventilator reallocated (83.9% of the original cohort), 280 (37.7%) of whom were observed to survive to discharge if provided with a ventilator. Of the total 360 patients observed to survive to discharge, only 65 (18.1%) would have survived to discharge without having their ventilator reallocated in the simulation. There was no statistically significant ventilator reallocation by age, gender, race, or ethnicity at any day of assessment, among all who had their ventilator reallocated, or among those who were simulated to survive.

Missing data and additional analyses

Our cohort required 1852 patient-SOFA-days (a SOFA score for each day a ventilated patient was assessed). The majority of missing information was on the day of intubation, with 100 patients having no SOFA score on the day of intubation (Supplement Table 2). Additional modeling with adjusted or supplemented SOFA scores, as well as deferring NYSVAG Day 2 reassessment to Day 5, is shown in Figure 3. We failed to find a significant difference in simulated surviving individuals when normalizing the GCS or supplementing pulmonary SOFA subscores (72 GCS-nl versus 65; see Figure 3). A delay in the Day 2 reassessment to Day 5 (MVD 2to5) resulted in a 49.2% increase in simulated survival to discharge within NYSVAG criteria.

Figure 3. Variations on the limited NYSVAG simulation on a cohort of critically ill COVID-19 patients. Solid represents total remaining patients. OG, original applied NYSVAG. GCS-nl, normalizing all GCS SOFA subscores. mSOFA-FiO2, missing pulmonary SOFA subscores are supplemented with mSOFA pulmonary subscore presuming an oxygen saturation greater than 85%. MVD 2to5, moving NYSVAG OG Day 2 reassessment to Day 5. Dashed lines represent the total number of simulated patients who were extubated & survived to hospital discharge (combined endpoint).

Discussion

Our findings simulate the performance of common triage criteria (SOFA cutoffs and ventilator time trials) on a large cohort of critically ill COVID-19 patients. Ventilator reallocation was simulated to occur for 83.9% (742/884) of the cohort, 37.7% (280/742) of whom would have survived to discharge if provided with a ventilator. Additionally, only 18% (65/362) of all individuals who were observed to survive to discharge were also simulated to survive to discharge within the simulated NYSVAG criteria. This raises a significant concern that under this ventilator allocation protocol, ventilators might be taken away from patients who would otherwise survive. This concern would be assuaged if the subsequent patients who received the reallocated ventilators would have fared much better, thereby improving overall group survival. This would satisfy the NYSVAG utilitarian goal of doing the most good for the most people, commonly interpreted as saving the most lives. 10

Others have articulated that any redistribution of lifesaving resources must be fair, transparent, and promote equity among vulnerable and marginalized populations. Reference Ashana, Anesi, Liu, Escobar, Chesley, Eneanya, Weissman, Miller, Harhay and Halpern23Reference Cleveland Manchanda, Couillard and Sivashanker26 It is therefore encouraging that any ventilator distribution per NYSVAG was not statistically significant among any age, gender, ethnic, or racial groups.

Our results contrast with that reported by Wunsch et al., a similar retrospective simulation of applying NYSVAG to 40,439 intubated critically ill patients from 2014-2015. Reference Wunsch, Hill and Bosch20 Our cohort of COVID-19 patients was sicker than their mixed medical/surgical population (initial COVID-19 SOFA of 7.4 ± 3.3 compared to 4.5 ± 3.7), which may account for the larger proportion triaged as ineligible for a ventilator (Blue, 13.7% vs 8.9%). On the second or fifth reassessment day, only 4.7% or 28.3% of their cohort was classified as ineligible for a ventilator (Blue), respectively. This contrasts concerningly to the 59.7% and 71.7% simulated in our cohort of COVID-19 patients. Moreover, their median (IQR) duration of mechanical ventilation was 1.94 (0.7, 5.1) days, whereas our COVID-19 patient cohort required a median of 10.2 (3.74, 23.73) days. COVID-19 patients do not recover on a similar trajectory to other critical respiratory illnesses, including influenza, and allocation guidelines may need to extend their reassessment intervals. Determination of the appropriate length of time for a trial of mechanical ventilation can be difficult since the physiologic trajectory of the disease—even with optimal treatment—may be unknown or variable. Too short a trial of ventilation and patients may have an inadequate length of time to demonstrate a clinical benefit. Short of objective evidence that a patient is failing therapy, premature reassessment that results in the removal of resources that might otherwise later benefit the patient may promote rapid cycling between patients irrespective of prognosis and worsen overall mortality. Too long between reassessments might lead to disproportionate resource utilization without optimizing overall survival.

Our results may signal that the recommended length of time for a ventilator time trial was too short to demonstrate improvement for COVID-19 patients. This is supported by the observation that ventilator reallocation would have occurred for over half of the simulated remaining patients on or after 2 days of mechanical ventilation, of which over one third were observed to survive to discharge (see Table 3). It is also corroborated by the 49.2% increase in patients who would survive to discharge if the second day of assessment of ventilator allocation was delayed to Day 5 (see Figure 3). This may suggest that the appropriate length of a time trial should be extended for COVID-19 patients compared to other typical causes of respiratory failure.

Potential Limitations

This study has several potential limitations. Although this simulation is not based on a fixed supply of ventilators and includes an unrealistic dynamic number of available ventilators, it provides an important analysis that best represents the true performance of commonly used SOFA cutoffs and reassessment intervals with respect to patient mortality. This design is crucial because guideline underperformance might be obscured under circumstances of low ventilator strain where the majority of patients are provided with a ventilator regardless of their eligibility. A true understanding of which patients are classified as ineligible for lifesaving resources and their chances of survival remains paramount to understanding how guidelines might perform with increasing resource strain. This cohort study also facilitates a unique opportunity to understand how many patients would have had their ventilator withheld (simulated) but would have otherwise survived if a ventilator was provided (because they were provided with a ventilator in the real world). Our simulation demonstrates the important concern that, according to the NYSVAG criteria studied, many patients who may have had their ventilators withdrawn would have subsequently survived. This concern would be mitigated, and NYSVAG indeed validated, if subsequent patients receiving reallocated ventilators had a higher rate of survival. We acknowledge that our study design cannot explicitly assess the mortality of subsequent patients receiving reallocated ventilators. However, our results suggest that patients ventilated for 5 or more days and selected to have their ventilator reallocated may have a higher chance of surviving to discharge with a ventilator (46.3-75.0%) than subsequent patients receiving their reallocated ventilator (43.4%) (see Table 3). This conjecture, if substantiated, may have significant implications as to how new patients should be triaged in the setting of existing ventilated patients. The survival of patients receiving reallocated ventilators must be explicitly studied before changes to existing protocols are proposed and adopted.

A simulation including a fixed number of available ventilators and patient throughput incorporating ventilator strain will produce different results and allow for the explicit study of patients receiving reallocated ventilators. In such a scenario, however, it remains concerning that the majority of patients classified as Blue on or after 5 days of mechanical ventilation had a SOFA score of less than 7. Randomization may not maximize patient survival if it occurs among all Blue individuals and does not prioritize subcategories that are more likely to survive (see Table 4). A triage officer or committee may provide a more sensitive selection process; however, NYSVAG only permits randomization within a priority color. 10 (p. 68) Further subcategorization of all patients classified as ineligible for a ventilator should be considered in resource allocation protocols. Subsequent studies incorporating a fixed number of available ventilators, real patient throughput, and the comparison of alternative triage strategies such as first-come first-served should be simulated. Bhavani et al. performed a limited simulation; however, they do not incorporate patient throughput or dynamic real-world surge strain. Reference Bhavani, Luo, Miller, Sanchez-Pinto, Han, Mao, Sandıkçı, Peek, Coopersmith, Michelson and Parker22

We believe our results are broadly generalizable. While NYSVAG was designed for an influenza pandemic, the NYSVAG Q&A approve its application for a broader respiratory crisis. 27 Although we analyzed limited aspects of NYSVAG, these elements are similar to several other resource allocation policies. Reference White and Lo11Reference Piscitello, Kapania and Miller15 We believe our sizeable cohort is generalizable to other COVID-19 patients as our SOFA score and PaO2/FiO2 at intubation Reference Cummings, Baldwin and Abrams28Reference Grasselli, Zangrillo and Zanella32 and duration of mechanical ventilation Reference Cummings, Baldwin and Abrams28,Reference Ziehr, Alladina and Petri29 are consistent with other reports. Our high mortality is consistent with other local reports during the study period Reference Petrilli, Jones and Yang33Reference Santos, Levy and Gong35 and likely related to the unprecedented surge and associated ICU strain that would be consistent in communities activating crisis standards of care. Reference Toth, Tatem and Hosseinipour34 . Our NYSVAG simulation results are additionally similar to a 200-patient mixed COVID-19 cohort from a comparable local institution. Reference Santos, Levy and Gong35 While all ventilated patients would participate in the same resource allocation policy, non-COVID-19 patients were excluded from this simulation because they were marginally present during our institutions’ surge.

There remain specific limitations related to the general dependence on SOFA scores for reallocation policies. While an initial report suggested that SOFA scores could predict mortality among COVID-19 patients, Reference Zhou, Yu and Du36 subsequent studies have not been as supportive. Reference Fiest, Krewulak and Plotnikoff14,Reference Raschke, Agarwal and Rangan37,Reference Shahpori, Stelfox and Doig38 Despite such shortcomings, using an objective score to prioritize resource allocation with the intent of saving the most lives achieves impartiality, preserves fairness toward patients, promotes transparency with the community, maintains reproducibility between providers, eliminates the encroachment of bias or prejudices by decision-makers, and may reduce provider distress about ventilator allocation decisions. Such scores may require further augmentation to strive for equity and not worsen existing health disparities, Reference Ashana, Anesi, Liu, Escobar, Chesley, Eneanya, Weissman, Miller, Harhay and Halpern23,Reference White and Lo39 or to better reflect community values. Reference Biddison, Gwon and Schoch-Spana40 While many critically ill COVID-19 patients are at risk of multiple organ sequelae, some remain “single organ” dysfunction with respiratory failure, whereupon SOFA scores may not be the most accurate reflection of the severity of their illness.

Dependence on SOFA scores can be additionally problematic as COVID-19 patients often require prone positioning, which can complicate the interpretation of the oxygenation component, or heavy sedation or paralytics, which can interfere with the neurological assessment. In our simulation, GCS scores were normalized when they were not documented, which was overwhelmingly common and may be related to patients’ clinical contraindications for interruption of sedation or paralytics versus incomplete documentation. The frequency of missing data in Supplemental Table 2 underscores this point and may only be partially mitigated by using the mSOFA score. Our normalization of missing information underestimates a patient’s true SOFA score and therefore may underestimate the frequency of ventilator reallocation; 11.3% of our cohort was intubated without any SOFA information, possibly suggesting that first responders may need to make triage decisions with incomplete information.

Conclusion

States across the nation continue to battle against surges in the COVID-19 pandemic. Federal, state, and local governments should continue to study and develop their disaster preparedness plans. Our results suggest that if NYS Ventilator Allocation Guidelines were implemented, patients might have their ventilator reallocated when they might otherwise survive. Further studies should determine whether subsequent patients receiving reallocated ventilators would have a higher survival than the original cohort before changes to the guidelines are enacted.

Supplementary Material

To view supplementary material for this article, please visit https://doi.org/10.1017/dmp.2022.154

Acknowledgments

We are grateful to Kelly Paluch for her statistical contributions.

Author Contributions

All authors contributed to all aspects except MEN, who was not involved in the study design.

The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States Government.

Funding Statement

This research is unfunded and approved by the NYU Grossman School of Medicine Office of Science and Research Institutional Review Board i20-01653. This project was presented as an abstract during the American Thoracic Society 2021.

Conflict(s) of Interest

The authors have no conflicts of interest.

References

WHO Director-General’s Opening Remarks at the Media Briefing on COVID-19. World Health Organization. https://www.who.int/dg/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19---11-march-2020. Published March 11, 2020. Accessed March 23, 2020.Google Scholar
Coronavirus Disease 2019 (COVID-19) Situation Report—51. World Health Organization. https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200311-sitrep-51-covid-19.pdf?sfvrsn=1ba62e57_10. Published March 11, 2020. Accessed January 13, 2021.Google Scholar
Jesse McKinley New York City Region is now an epicenter of the coronavirus pandemic. New York Times. https://www.nytimes.com/2020/03/22/nyregion/Coronavirus-new-York-epicenter.html. Published March 22, 2020. Accessed January 13, 2021.Google Scholar
Coronavirus. Worldometer. https://www.worldometers.info/coronavirus/. Published date unknown. Accessed July 22, 2020.Google Scholar
COVID-19: Data. Department NYCH. https://www1.nyc.gov/site/doh/covid/covid-19-data.page#download. Published date unknown. Accessed July 22, 2020.Google Scholar
Cha, A, Bernstein, L, Sellers, FSH, Harris S. Faced with a crush of patients, besieged NYC hospitals struggle with life-or-death decisions. The Washington Post. https://www.washingtonpost.com/health/2020/03/31/new-york-city-hospitals-coronavirus/. Published March 31, 2020. Accessed April 6, 2020.Google Scholar
Coronavirus: New York Forced to Redistribute Ventilators. BBC. https://www.bbc.com/news/world-us-canada-52148534. Published April 3, 2020. Accessed January 13, 2021.Google Scholar
Gutmann Koch, V, Han, SA. COVID in NYC: what New York did, and should have done. Am J Bioeth. 2020;20(7):153-155. doi: 10.1080/15265161.2020.1777350 CrossRefGoogle Scholar
Rosenthal, BM, Goldstein, J. N.Y. may need 18,000 ventilators very soon. It is far short of that. New York Times. https://www.nytimes.com/2020/03/17/nyregion/ny-coronavirus-ventilators.html?action=click&module=RelatedLinks&pgtype=Article. Published March 26, 2020. Accessed January 14, 2021.Google Scholar
Ventilator Allocation Guidelines. New York State Task Force on Life and the Law & New York State Department of Health. https://www.health.ny.gov/regulations/task_force/reports_publications/docs/ventilator_guidelines.pdf. Published November 2015. Accessed January 13, 2021.Google Scholar
White, DB, Lo, B. A framework for rationing ventilators and critical care beds during the COVID-19 pandemic. JAMA. 2020;323(18):1773-1774. doi: 10.1001/jama.2020.5046 CrossRefGoogle ScholarPubMed
Emanuel, EJ, Persad, G, Upshur, R, et al. Fair allocation of scarce medical resources in the time of COVID-19. N Engl J Med. 2020;382(21):2049-2055. doi: 10.1056/NEJMsb2005114 CrossRefGoogle ScholarPubMed
Antommaria, AHM, Gibb, TS, McGuire, AL, et al. Ventilator triage policies during the COVID-19 pandemic at U.S. hospitals associated with members of the Association of Bioethics Program Directors. Ann Intern Med. 2020;173(3):188-194 doi: 10.7326/M20-1738 CrossRefGoogle ScholarPubMed
Fiest, KM, Krewulak, KD, Plotnikoff, KM, et al. Allocation of intensive care resources during an infectious disease outbreak: a rapid review to inform practice. BMC Med. 2020;18(1):404. doi: 10.1186/s12916-020-01871-9 CrossRefGoogle ScholarPubMed
Piscitello, GM, Kapania, EM, Miller, WD, et al. Variation in Ventilator Allocation Guidelines by US state during the coronavirus disease 2019 pandemic: a systematic review. JAMA Netw Open. 2020;3(6):e2012606. doi: 10.1001/jamanetworkopen.2020.12606 CrossRefGoogle ScholarPubMed
Fins, JJ. Sunshine is the best disinfectant, especially during a pandemic. New York State Bar Association. https://nysba.org/sunshine-is-the-best-disinfectant-especially-during-a-pandemic/#_ftn30. Published June 1, 2020. Accessed April 3, 2022.Google Scholar
Alaska Activates the State’s Crisis Standards of Care for Multiple Health Care Facilities. Department of Health and Social Services SoA. https://health.alaska.gov/news/Documents/press/2021/DHSS_PressRelease_StateCSC_20211002.pdf. Published October 2, 2021. Accessed July 28, 2022.Google Scholar
Knaus, WA, Wagner, DP, Draper, EA, et al. The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults. Chest. 1991;100(6):1619-1636. doi: 10.1378/chest.100.6.1619 CrossRefGoogle ScholarPubMed
Wunsch, H, Hill, AD, Bosch, N, et al. Comparison of 2 triage scoring guidelines for allocation of mechanical ventilators. JAMA Netw Open. 2020;3(12):e2029250. doi: 10.1001/jamanetworkopen.2020.29250 CrossRefGoogle ScholarPubMed
Kanter, RK. Would triage predictors perform better than first-come, first-served in pandemic ventilator allocation? Chest. 2015;147(1):102-108. doi: 10.1378/chest.14-0564 CrossRefGoogle ScholarPubMed
Bhavani, SV, Luo, Y, Miller, WD, Sanchez-Pinto, LN, Han, X, Mao, C, Sandıkçı, B, Peek, ME, Coopersmith, CM, Michelson, KN, Parker, WF. Simulation of Ventilator Allocation in Critically Ill Patients with COVID-19. Am J Respir Crit Care Med. 2021;204(10):1224-1227. doi: 10.1164/rccm.202106-1453LE.CrossRefGoogle ScholarPubMed
Ashana, DC, Anesi, GL, Liu, VX, Escobar, GJ, Chesley, C, Eneanya, ND, Weissman, GE, Miller, WD, Harhay, MO, Halpern, SD. Equitably Allocating Resources during Crises: Racial Differences in Mortality Prediction Models. Am J Respir Crit Care Med. 2021;204(2):178-186. doi: 10.1164/rccm.202012-4383OC.CrossRefGoogle ScholarPubMed
Gershengorn, HB, Holt, GE, Rezk, A, et al. Assessment of disparities associated with a crisis standards of care resource allocation algorithm for patients in 2 US hospitals during the COVID-19 pandemic. JAMA Netw Open. 2021;4(3):e214149. doi: 10.1001/jamanetworkopen.2021.4149 CrossRefGoogle ScholarPubMed
Cleveland Manchanda, EC, Sanky, C, Appel, JM. Crisis standards of care in the USA: a systematic review and implications for equity amidst COVID-19. J Racial Ethn Health Disparities. 2020:1-13. doi: 10.1007/s40615-020-00840-5 Google ScholarPubMed
Cleveland Manchanda, E, Couillard, C, Sivashanker, K. Inequity in crisis standards of care. N Engl J Med. 2020;383(4):e16. doi: 10.1056/NEJMp2011359 CrossRefGoogle ScholarPubMed
Frequently Asked Questions: New York State Task Force on Life and the Law and the New York State Department of Health’s Ventilator Allocation Guidelines. https://www.health.ny.gov/regulations/task_force/reports_publications/docs/ventilator_guidelines_faqs.pdf. Published date unknown. Accessed July 28, 2022.Google Scholar
Cummings, MJ, Baldwin, MR, Abrams, D, et al. Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: a prospective cohort study. Lancet. 2020;395(10239):1763-1770. doi: 10.1016/s0140-6736(20)31189-2 CrossRefGoogle ScholarPubMed
Ziehr, DR, Alladina, J, Petri, CR, et al. Respiratory pathophysiology of mechanically ventilated patients with COVID-19: a cohort study. Am J Respir Crit Care Med. 2020;201(12):1560-1564. doi: 10.1164/rccm.202004-1163LE CrossRefGoogle ScholarPubMed
Ling, L, So, C, Shum, HP, et al. Critically ill patients with COVID-19 in Hong Kong: a multicentre retrospective observational cohort study. Crit Care Resusc. 2020;22(2):119-125.Google Scholar
Yang, X, Yu, Y, Xu, J, et al. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med. 2020;8(5):475-481. doi: 10.1016/s2213-2600(20)30079-5 CrossRefGoogle Scholar
Grasselli, G, Zangrillo, A, Zanella, A, et al. Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy Region, Italy. JAMA. 2020;323(16):1574-1581. doi: 10.1001/jama.2020.5394 CrossRefGoogle ScholarPubMed
Petrilli, CM, Jones, SA, Yang, J, et al. Factors associated with hospital admission and critical illness among 5279 people with coronavirus disease 2019 in New York City: prospective cohort study. BMJ. 2020;369:m1966. doi: 10.1136/bmj.m1966 CrossRefGoogle ScholarPubMed
Toth, AT, Tatem, KS, Hosseinipour, N, et al. Surge and mortality in ICUs in New York City’s public healthcare system. Crit Care Med. 2021;49:1439-1450. doi: 10.1097/CCM.0000000000004972 CrossRefGoogle ScholarPubMed
Santos, HT, Levy, BM, Gong, MN, et al. How useful are Sequential Organ Failure Assessment (SOFA) score-based ventilator triage guidelines during the COVID-19 pandemic? TP11. TP011 USE OF HEALTH SERVICES AND MEDICAL EDUCATION RESEARCH TO EVALUATE THE EFFECT OF COVID-19:A1494-A94.Google Scholar
Zhou, F, Yu, T, Du, R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020;395(10229):1054-1062. doi: 10.1016/s0140-6736(20)30566-3 CrossRefGoogle ScholarPubMed
Raschke, RA, Agarwal, S, Rangan, P, et al. Discriminant accuracy of the SOFA score for determining the probable mortality of patients with COVID-19 pneumonia requiring mechanical ventilation. JAMA. 2021;325:1469-1470. doi: 10.1001/jama.2021.1545 CrossRefGoogle ScholarPubMed
Shahpori, R, Stelfox, HT, Doig, CJ, et al. Sequential Organ Failure Assessment in H1N1 pandemic planning. Crit Care Med. 2011;39(4):827-832. doi: 10.1097/CCM.0b013e318206d548 CrossRefGoogle ScholarPubMed
White, DB, Lo, B. Mitigating inequities and saving lives with ICU triage during the COVID-19 pandemic. Am J Respir Crit Care Med. 2021;203(3):287-295. doi: 10.1164/rccm.202010-3809CP CrossRefGoogle ScholarPubMed
Biddison, ELD, Gwon, HS, Schoch-Spana, M, et al. Scarce resource allocation during disasters: a mixed-method community engagement study. Chest. 2018;153(1):187-195. doi: 10.1016/j.chest.2017.08.001 CrossRefGoogle ScholarPubMed
Figure 0

Table 1. New York State ventilator allocation guidelines steps 2 and 310

Figure 1

Figure 1. Study Methodology. Patients were observed from time of intubation until observed endpoints of death or survival to discharge. Patient’s had SOFA reassessments on day 2, 5, 7, 9, and 11 to determine if their ventilator would have been reallocated in the simulation. Patients received a ventilator in the simulation if their SOFA score on day of intubation was less than 12. On the Day 2 Reassessment, patients would have had their ventilator reallocated if their interval SOFA score increased, did not change from an initial SOFA of 8-11, or was greater than 11. On the Day 5 Reassessment, patients would have had their ventilator reallocated if their SOFA score increased, was greater than 7, or was less than 7 and without a change in their previous SOFA score. On the Day 7 Reassessment and every 2 days thereafter, patients would have had their ventilator reallocated if their SOFA score did not improve or was greater than 7. A patient who did not have their ventilator reallocated in the simulation (No *) would only progress to the subsequent simulated ventilator reallocation assessment if they were not extubated, expired, or placed on ECMO prior to that assessment.

Figure 2

Table 2. Patient cohort demographics

Figure 3

Figure 2. Sankey diagram of a limited NYSVAG simulation on a cohort of COVID-19 patients. Each color represents the number of patients triaged to that color on the day of triage (Day 0 or potential intubation, Day 2, 5, 7, 9, and 11). Patients triaged as Red or Yellow received or maintained their ventilator until the next Reassessment Day while those triaged as Blue had their ventilator withheld or reallocated and were simulated to expire. Flow between each day of assessment describes the number of patients who moved between triage categories. The paths from left to the far right bars describe the course of patients who satisfied the combined simulated-observed endpoint of patients who had their ventilator reallocated (simulated) but were observed to survive to hospital discharge (observed, how the patient would have performed if provided a ventilator in the simulation). Patients exited the simulation because they were extubated, expired, or placed onto ECMO.

Figure 4

Table 3. Numerical representation of a simulated application of NYSVAG to a cohort of critically ill COVID-19 patients

Figure 5

Table 4. Etiology of patients triaged as Blue and ineligible for a ventilator during simulation

Figure 6

Figure 3. Variations on the limited NYSVAG simulation on a cohort of critically ill COVID-19 patients. Solid represents total remaining patients. OG, original applied NYSVAG. GCS-nl, normalizing all GCS SOFA subscores. mSOFA-FiO2, missing pulmonary SOFA subscores are supplemented with mSOFA pulmonary subscore presuming an oxygen saturation greater than 85%. MVD 2to5, moving NYSVAG OG Day 2 reassessment to Day 5. Dashed lines represent the total number of simulated patients who were extubated & survived to hospital discharge (combined endpoint).

Supplementary material: File

Walsh et al. supplementary material

Walsh et al. supplementary material

Download Walsh et al. supplementary material(File)
File 855.4 KB