ABSTRACT
INTRODUCTION: Previous studies suggest that delayed initiation of extracorporeal membrane oxygenation (ECMO) is associated with higher patient mortality. Hence, we hypothesized that prolonged invasive mechanical ventilation (IMV) prior to ECMO was associated with higher mortality in patients with COVID-19. METHOD(S): The COVID-19 Critical Care Consortium, a prospective international multicenter registry, was queried for all patients with COVID-19 infection who received IMV and ECMO. Patients who were intubated prior to transfer to a study site were excluded. The primary variable was number of days on IMV prior to ECMO initiation and study endpoint was death or discharge from the study site. Cox proportional hazards model for the time between ECMO initiation and death was built using covariates including age, gender, selected comorbidities, and time intervals from ICU admission to IMV and IMV to ECMO initiation. RESULT(S): Between 1/1/2020 and 6/6/2022, A total of 593 patients from 107 study sites and 25 countries were included in the analysis. In this cohort, the median age was 50 (Interquartile range [IQR]: 40-58) years. Obesity and hypertension were prevalent among 220 (38.4%) and 223 (38.8%) of the patients, respectively. Twenty-four (4.2%) patients had chronic pulmonary disease. Prior to ECMO initiation, patients spent a median of 3.68 (IQR: 1.36-8.07) days in the ICU and a median of 2.49 (IQR: 0.88-5.65) days on IMV. Overall mortality was 47.2% with 3.9% patients' status not finalized or unknown. According to the final survival model, the number of days on IMV prior to ECMO initiation was not associated with mortality. The hazard ratios for 0, 3, 7, and 14 days of pre-ECMO IMV were 0.94 (95% confidence interval [CI]: 0.83 to 1.07), 1.02 (95% CI: 0.97 to 1.08), 1.09 (95% CI: 0.92 to 1.3) and 1.09 (95% CI: 0.83 to 1.42), respectively. Other noticeable contributory factors in the model included age and gender. CONCLUSION(S): Among patients with COVID-19 who received ECMO, the length of pre-ECMO IMV was not associated with hospital mortality. Further studies evaluating the ventilator settings before and after ECMO initiation are needed.
ABSTRACT
SESSION TITLE: Insights into the Care of Patients with Respiratory Failure SESSION TYPE: Original Investigations PRESENTED ON: October 18-21, 2020 PURPOSE: Lung ultrasound is a quick, non-invasive, and widely available tool used to assess for a multitude of disease processes. Recently, a quick point assessment of B-lines (B-line score) has been shown to be an accurate marker of extra-vascular lung water. In mechanically ventilated patients, static lung compliance is calculated as Cstat = VT/(Pplat-PEEP), with VT meaning tidal volume, Pplat plateau pressure, and PEEP positive end expiratory pressure. A limitation of this technique is that patients need to be passive on the ventilator to accurately measure PPlat, something which is difficult given efforts to reduce sedation. Other techniques, such as esophageal manometry, have been used to estimate transpulmonary pressures, but is limited by being invasive and subject to error. Our study sought to determine if the B-line score correlates with lung compliance in mechanically ventilated patients. We hypothesized that the B-line score would inversely correlate with static lung compliance. METHODS: This was a prospective observational study performed in the medical intensive care units of NYU Langone Health and Bellevue Hospital Center. Inclusion criteria included all adult patients requiring mechanical ventilation. Exclusion criteria included reasons to have reduced respiratory system compliance from an extrapulmonary etiology: BMI > 35, abdominal hypertension, significant pleural effusions, or pneumothorax. Lung ultrasound was performed at 4 points over the anterior chest. Each image was independently scored by two intensivists. The number of B-lines seen in each window were counted and added together to create the B-line score. Static lung compliance was determined using the formula above;inspiratory hold maneuvers were used to obtain plateau pressures. RESULTS: A total of 99 observations were performed. The mean B-line score was 4.73 +/- 0.60 with a range score from 0 to 25.5. Ultrasound B-line score inversely correlated with static lung compliance (Spearman's r = -0.60, p<0.001), net fluid balance in the 24 hours prior to the scan (Spearman's r = -0.26, p=0.02), and P/F ratio (Spearman's r = -0.37, p<0.001). CONCLUSIONS: Ultrasound B-line score correlates with for static lung compliance in mechanically ventilated patients. CLINICAL IMPLICATIONS: Our study demonstrates that an assessment of B-lines on lung ultrasound is an effective marker of important clinical characteristics in mechanically ventilated patients. In patients where assessments of these factors are limited, use of this quick and simple B-line score may be clinically useful. This is especially true in the recent COVID-19 pandemic, where imaging tests and monitoring may be limited given isolation precautions. Whether serial B-line scores in the same patient correlate with clinical improvement in the above-mentioned markers is yet to be determined. DISCLOSURES: No relevant relationships by Andrew Lehr, source=Web Response No relevant relationships by Vikramjit Mukherjee, source=Web Response, value=Consulting fee Removed 04/27/2020 by Vikramjit Mukherjee, source=Web Response No relevant relationships by Deepak Pradhan, source=Web Response No relevant relationships by Bishoy Zakhary, source=Web Response