Association Between Availability of ECMO and Mortality in COVID-19 Patients Eligible for ECMO: A Natural Experiment

RATIONALE: The outcomes of patients with COVID-19 who are medically eligible to receive ECMO, but do not because of limited health system capacity, have not been reported. METHODS: We analyzed prospectively collected clinical data from consecutive patients with SARS-CoV-2 referred for ECMO to a single center between January 1, 2021 and August 31, 2021. Each patient underwent a systematic assessment of medical eligibility to receive ECMO followed by a separate assessment of the health system's resources to provide ECMO. When health system resources were available, the patient was transferred to an ECMO center. When health system resources were not available, the patient was not transferred to an ECMO center and did not receive ECMO. Patients were followed until the time of death or hospital discharge. Among medically eligible patients, we compared those for whom health system capacity permitted transfer to receive ECMO to those for whom health system capacity did not permit transfer to receive ECMO with regard to the primary outcome of all-cause in-hospital mortality using Cox proportional hazards regression analysis adjusting for age, acute kidney injury, and receipt of vasopressors. RESULTS: Among the 240 patients with COVID-19 referred for ECMO, 90 patients (37.5%) were determined to be medically eligible to receive ECMO and were included in this study. Median age was 40 years (IQR, 34-48). The health system capacity to provide ECMO was available for 35 patients (38.9%), of whom 32 received ECMO and 3 died or developed a contraindication to ECMO after transfer but prior to cannulation. Death before hospital discharge occurred in 15 of the 35 patients (42.9%) for whom health system capacity permitted transfer to receive ECMO compared with 49 of the 55 patients (89.1%) for whom health system capacity did not permit transfer to receive ECMO (adjusted hazard ratio 0.23;95% confidence interval, 0.12 to 0.43;P < 0.001) (Figure 1). CONCLUSIONS: In this cohort of adults with COVID-19, nearly 90% of patients who were eligible for ECMO but did not receive it due to limited health system resources died before hospital discharge, despite young age and limited comorbidities. Periods of resource limitation during which provision of ECMO is determined by resource availability and not patient characteristics may act as a natural experiment, and these results suggest that ECMO provides a significant mortality benefit in the treatment of COVID- 19.

Race, SpO2, and SaO2 in Critically Ill Adults

Rationale: For critically ill adults, oxygen saturation is continuously monitored using pulse oximetry (SpO2) as a surrogate for arterial oxygen saturation (SaO2). Skin pigmentation may affect accuracy of pulse oximetry by introducing error due to statistical bias (e.g., consistently lower SaO2 at a given SpO2), variance (e.g., wider range of SaO2 values at a given SpO2), or both. We evaluated relationships between patient race, SpO2, SaO2, and likelihood of hypoxemia or hyperoxemia. Methods: We analyzed prospectively collected data from all adults receiving mechanical ventilation in a single ICU from July 2018 to February 2021, excluding patients with Covid-19. Race was reported by patients or their surrogates in routine care;we included those with race documented as Black or White. Values for SpO2 were recorded from bedside monitors every 1 minute. Values for SaO2 from all arterial blood gas measurements were collected. First, we analyzed all pairs of SpO2 and SaO2 values obtained within 1 minute of each other in a generalized linear mixed-effects model with SaO2 as the dependent variable, a random effect for patient, and fixed effects for race, SpO2, and their interaction. Second, we compared the unadjusted incidence of hypoxemia (SaO2<88%) and hyperoxemia (PaO2>150 mmHg) by race, with SpO2 values 92-96% within 10 minutes of SaO2 measurement. Results:This analysis included 994 patients with 5,989 paired measurements of SpO2 and SaO2 collected within 10 minutes, of which 4,211 were within 1 minute. Of 5,989 SpO2-SaO2 pairs, 792 (13%) were from Black patients and 5,194 (87%) from White patients. In mixed-effects model analysis, the mean difference in SaO2 between White patients and Black patients was -0.1 (95% CI, -1.3 to 1.2) at SpO2 of 88%, 0.4 (-0.6 to 1.3) at SpO2 of 90%, 0.8 (0.0 to 1.6) at SpO2 of 92%, 1.1 (0.24 to 1.9) at SpO2 of 95%, and 1.0 (0.3 to 1.8) at SpO2 of 96% (Figure 1). Among patients with an SpO2 of 92-96%, Black patients were more likely than White patients to have hypoxemia (4.3% vs 1.4%) and hyperoxemia (5.1% vs 2.1%). Conclusions:At SpO2 values of 92-98%, the associated SaO2 value was 1 percentage point lower on average for Black patients compared to White patients, consistent with statistical bias in pulse oximetry. Additionally, greater variation in SaO2 values at a given SpO2 resulted in a higher incidence of both hypoxemia and hyperoxemia for Black patients with SpO2 values of 92-96%, compared to White patients.

Anti-gm-csf monoclonal antibody gimsilumab improved ventilator-free survival, decreased hospitalization length, and prevented nt-probnp rise in invasively ventilated patients with hyperinflammatory covid19 pneumonia: A subgroup analysis from the breathe trial suggests neurohormonal role for gm-csf inhibition

Background: Granulocyte-macrophage colony-stimulating factor (GM-CSF), a myeloid cell growth factor and pro-inflammatory cytokine, may drive the overactive host immune response in COVID-19. We conducted a clinical trial assessing the anti-GM-CSF monoclonal antibody gimsilumab for hyperinflammatory COVID-19 pneumonia (BREATHE). Here, we report a pre-specified subgroup analysis demonstrating a signal of benefit in patients invasively ventilated at baseline. Methods: BREATHE (NCT04351243) was a double-blind, randomized, placebo-controlled trial at 21 US locations. Patients were randomized 1:1 to receive two doses of IV gimsilumab or placebo one week apart. The study included hospitalized COVID-19 patients with hyperinflammation (CRP ≥50 mg/L or ferritin ≥1,000 ng/mL) and pre-ARDS lung injury or ARDS. The primary endpoint was all cause mortality at day 43, and key secondary outcomes assessed ventilator use and hospitalization length. Results: 225 patients were randomized and dosed. 41 patients were invasively ventilated at baseline. Steroid use and baseline characteristics were generally balanced across study arms in this subgroup. Ventilated patients treated with gimsilumab demonstrated improvements over placebo on the primary and key secondary endpoints (Table 1). Contrasting the placebo group, gimsilumabtreated patients did not experience a sharp rise in NT-proBNP, a marker of heart failure, through day 43 (Figure 1). Conclusions: GM-CSF inhibition may be therapeutic in ventilated COVID-19 patients through a neurohormonal mechanism. More studies are needed to assess the role of GM-CSF in COVID-19-associated cardiomyopathy, volume status, and ARDS.

Effect of COVID-19 pandemic on the use of video laryngoscopy during tracheal intubation of critically ill adults

Rationale: Significant variability exists in the use of direct versus video laryngoscopy for the intubation of critically ill adults. Concerns regarding risks of infection to healthcare providers have led many experts to recommend the use of video laryngoscopy for intubation of patients with COVID-19. This recommendation is based on the belief that using video laryngoscopy allows more physical distance between the operator and the patient, potentially decreasing the risk to providers. Information on the effect of the COVID-19 pandemic on intubation practices is currently limited. Methods: We conducted a survey of intubation practices in 21 emergency departments and intensive care units participating in either of two randomized trials (clinicaltrials.gov identifiers: NCT03928925 and NCT03787732) within the Pragmatic Critical Care Research Group. At each site, the site investigator estimated the prevalence of various airway practices among 3 groups of patients: 1) patients intubated prior to the COVID-19 pandemic, 2) patients intubated with known or suspected COVID-19, and 3) patients intubated during the pandemic without known or suspected COVID-19. The site investigator provided information on use of personal protective equipment, preintubation fluid bolus administration, preoxygenation, sedative choice, paralytic choice, approach to oxygenation from induction to intubation, approach to laryngoscopy, bougie use, primary operator, and ancillary staff present. The primary outcome was the proportion of intubations at a study site that was estimated to be performed using video laryngoscopy. Results: We received responses from 19 of 21 sites (90%). Investigators reported that video laryngoscopy was used in a median of 65% of intubations [IQR: 50-76%] at their site prior to the pandemic compared with a median of 100% of intubations [IQR: 76-100%] for patients with known or suspected COVID-19 (p= 0.0002). Prior to the pandemic only 2 sites (10.5%) reported using exclusively video laryngoscopy, compared to 10 sites (52.6%) that reported using exclusively video laryngoscopy for COVID-19 patients. For patients without known or suspected COVID-19, use of video versus direct laryngoscopy did not differ significantly between patients intubated before the pandemic and patients intubated during the pandemic. Conclusions: Among patients intubated at 19 emergency departments and intensive care units participating in two ongoing clinical trials, we found that the perceived use of video laryngoscopy was greater for patients with known or suspected COVID-19 compared to patients intubated prior to the pandemic, whereas perceived use of video laryngoscopy for patients without COVID-19 during the pandemic was similar to use prior to the pandemic.

Use of positive pressure ventilation during tracheal intubation before and during the COVID-19 pandemic

Rationale: Positive pressure ventilation (PPV) during tracheal intubation of critically ill adults reduces the risk of peri-procedural hypoxemia. Patients with COVID-19 are at increased risk of hypoxemia during tracheal intubation, but expert guidelines have recommended against the use of PPV for patients with suspected or known COVID-19 based on a concern that PPV may increase aerosolization and the risk of infection of healthcare providers. Evidence informing the impact of these recommendations on the airway management practices for patients with COVID-19 is currently limited. Methods: We conducted a survey of perceived intubation practices in 21 emergency departments (EDs) and intensive care units (ICUs) participating in either of two randomized trials (clinicaltrials.gov identifiers: NCT03928925 and NCT03787732) within the Pragmatic Critical Care Research Group between 9-3-2020 and 10-21-2020. At each site, the site investigator provided information on perceived airway management practices among patients intubated prior to the COVID-19 pandemic and during the pandemic. The primary outcome for this analysis was the use of any PPV, either bag-mask ventilation (BMV) or non-invasive ventilation (NIV), between induction and laryngoscopy. Results: We received responses from 19 of 21 sites (90%). Use of BMV was reported in a median of 35% of intubations [IQR: 0-67%] prior to the pandemic compared with a median of 0% of intubations [IQR: 0-20%] performed among patients with known or suspected COVID-19 (p=0.004). Investigators reported using NIV in a median of 10% of intubations [IQR: 0-25%] prior to the pandemic, compared to 10% [IQR 0-30%] of intubations among patients with known or suspected COVID-19 (p=0.20). Receipt of any positive pressure was reported in 50% [IQR: 0-85%] of intubations prior to the pandemic, compared to 30% [IQR: 0-55%] of intubations among patients with known or suspected COVID-19 (p=0.06) (Figure 1). Conclusions: Based on reported practices at 19 EDs and ICUs participating in either of two ongoing clinical trials, we found that the perceived use of BMV between induction and laryngoscopy decreased in patients with known or suspected COVID-19 while the perceived use of NIV did not change. Despite recommendations to avoid its use, perceived use of PPV remains common among patients with suspected or known COVID-19. This reflects the tension between balancing risk to providers and patients for a condition that places both groups at high risk of peri-procedural complications.

Maximum amplitude measurement on thromboelastography is inversely associated with development of venous thromboembolism in critically ill patients with COVID-19 pneumonia

RATIONALE: Coronavirus disease 2019 (COVID-19) is associated with high rates of venous thromboembolism despite prophylactic anticoagulation. Thromboelastography (TEG) has been suggested as a tool for identifying patients at increased risk of VTE. This study aims to determine if serial TEG evaluation predicts the development of VTE in patients with COVID-19 and respiratory failure admitted to the intensive care unit (ICU). METHODS: We conducted a prospective cohort study of patients admitted to the ICU with COVID-19 and respiratory failure without a prior diagnosis of acute VTE. Patients underwent TEG on days 1, 4, and 6 of ICU admission. The number of hypercoagulable days, defined as elevated TEG maximum amplitude, were recorded. The outcome was time from ICU admission to VTE during the hospital admission. Patients were censored at death and discharge. A Cox proportional hazards model was fit for time to first VTE controlling for patient age, sex, and exposure to antiplatelet agents. RESULTS: A total of 56 patients were enrolled in the study (11 females/45 males);41.1% had a hypercoagulable TEG (MA above the upper limit of normal) on ICU day 1, 55.4% on day 4, and 48.2% on day 6. The primary outcome of VTE occurred in 17 patients (30.4%). The median time from ICU admission to VTE was 7 days (IQR 3-12). Of the 14 patients with hypercoagulable MA on all 3 measured days, 1 patient (7.1%) developed a VTE compared to 15.4% of those with hypercoagulable MA for 2 days (N=13), 30.8% of those hypercoagulable for 1 day (N=13), and 62.5% of patients not hypercoagulable any of the 3 days (N=16). Multivariable Cox-proportional hazards model found that increasing number of hypercoagulable days by MA was associated with a decreased risk of VTE (OR 0.48, CI 0.27-0.85, p=0.01). A Cox proportional hazards regression curve for the number of hypercoagulable days by MA holding all other covariates at their medians is shown in Figure 1. CONCLUSION: In this study, more hypercoagulable days on TEG MA measurement was counterintuitively associated with a decreased risk of VTE in patients with COVID-19 respiratory failure. The normal response to inflammation and infection is upregulation of pro-coagulant acute phase reactants suggesting that increased TEG MA may be physiologic. TEG MA measures below the upper limit of normal are associated with an increased risk of VTE and may mark the development of a consumptive coagulopathy.