BACKGROUND: Venovenous extracorporeal membrane oxygenation (VV-ECMO) is a therapy for patients with refractory respiratory failure. The decision to decannulate someone from extracorporeal membrane oxygenation (ECMO) often involves weaning trials and clinical intuition. To date, there are limited prognostication metrics to guide clinical decision-making to determine which patients will be successfully weaned and decannulated.
OBJECTIVE: This study aims to assist clinicians with the decision to decannulate a patient from ECMO, using Continuous Evaluation of VV-ECMO Outcomes (CEVVO), a deep learning-based model for predicting success of decannulation in patients supported on VV-ECMO. The running metric may be applied daily to categorize patients into high-risk and low-risk groups. Using these data, providers may consider initiating a weaning trial based on their expertise and CEVVO.
METHODS: Data were collected from 118 patients supported with VV-ECMO at the Columbia University Irving Medical Center. Using a long short-term memory-based network, CEVVO is the first model capable of integrating discrete clinical information with continuous data collected from an ECMO device. A total of 12 sets of 5-fold cross validations were conducted to assess the performance, which was measured using the area under the receiver operating characteristic curve (AUROC) and average precision (AP). To translate the predicted values into a clinically useful metric, the model results were calibrated and stratified into risk groups, ranging from 0 (high risk) to 3 (low risk). To further investigate the performance edge of CEVVO, 2 synthetic data sets were generated using Gaussian process regression. The first data set preserved the long-term dependency of the patient data set, whereas the second did not.
RESULTS: CEVVO demonstrated consistently superior classification performance compared with contemporary models (P<.001 and P=.04 compared with the next highest AUROC and AP). Although the model\'s patient-by-patient predictive power may be too low to be integrated into a clinical setting (AUROC 95% CI 0.6822-0.7055; AP 95% CI 0.8515-0.8682), the patient risk classification system displayed greater potential. When measured at 72 hours, the high-risk group had a successful decannulation rate of 58% (7/12), whereas the low-risk group had a successful decannulation rate of 92% (11/12; P=.04). When measured at 96 hours, the high- and low-risk groups had a successful decannulation rate of 54% (6/11) and 100% (9/9), respectively (P=.01). We hypothesized that the improved performance of CEVVO was owing to its ability to efficiently capture transient temporal patterns. Indeed, CEVVO exhibited improved performance on synthetic data with inherent temporal dependencies (P<.001) compared with logistic regression and a dense neural network.
CONCLUSIONS: The ability to interpret and integrate large data sets is paramount for creating accurate models capable of assisting clinicians in risk stratifying patients supported on VV-ECMO. Our framework may guide future incorporation of CEVVO into more comprehensive intensive care monitoring systems.

BACKGROUND: Early diagnosis of acute brain injury (ABI) is critical for patients on veno-arterial extracorporeal membrane oxygenation (V-A ECMO) to guide anticoagulation strategy; however, neurological assessment in ECMO is often limited by patient sedation.
METHODS: In this pilot study of adults from June 2018 to May 2019, plasma samples of glial fibrillary acidic protein (GFAP), neurofilament light chain (NFL), and tubulin associated unit (Tau) were collected daily after V-A ECMO cannulation and measured using a multiplex platform. Primary outcomes were occurrence of ABI, assessed clinically, and neurologic outcome, assessed by modified Rankin Scale (mRS).
RESULTS: Of 20 consented patients (median age = 48.5°years; 55% female), 8 (40%) had ABI and 15 (75%) had unfavorable neurologic outcome at discharge. 10 (50%) patients were centrally cannulated. Median duration on ECMO was 4.5°days (IQR: 2.5-9.5). Peak GFAP, NFL, and Tau levels were higher in patients with ABI vs. without (AUC = 0.77; 0.85; 0.57, respectively) and in patients with unfavorable vs. favorable neurologic outcomes (AUC = 0.64; 0.59; 0.73, respectively). GFAP elevated first, NFL elevated to the highest degree, and Tau showed limited change regardless of ABI.
CONCLUSIONS: Further studies are warranted to determine how plasma biomarkers may facilitate early detection of ABIs in V-A ECMO to assist timely clinical decision-making.

Patients treated with ECMO are at great risk of nosocomial infections, and around 10% of isolates are gram-positive pathogens. Linezolid (LZD) is effective in the treatment of these infections but appropriate dosing is challenging. The aim was to evaluate the occurrence of thrombocytopenia during ECMO when treated with LZD. An LZD trough concentration of 8 mg/L was set as the cutoff value for thrombocytopenia occurrence among critically ill patients who received parenteral LZD therapy at a dose of 600 mg every 8 h during ECMO. Eleven patients were included in this prospective observational study. Median LZD trough concentrations were 7.85 (interquartile range (IQR), 1.95-11) mg/L. Thrombocytopenia was found in 81.8% of patients. Based on the median LZD trough concentrations cutoff value, patients were divided into two groups, 1.95 (IQR, 0.91-3.6) and 10.3 (IQR, 9.7-11.7) mg/L, respectively. Median platelet values differed significantly between groups on admission, ECMO day 0, ECMO day 1, and LZD sampling day [194 and 152.5, (p < 0.05)], [113 and 214, (p < 0.05)], [76 and 147.5, (p < 0.01)], and [26 and 96.5, (p < 0.01)], respectively. Duration of LZD therapy was similar between the groups. Significant platelet reduction was observed in both groups, emphasizing the need for closer monitoring to prevent LZD-associated thrombocytopenia.

Patients undergoing extracorporeal membrane oxygenation (ECMO) often require therapy with anti-infective drugs. The pharmacokinetics of these drugs may be altered during ECMO treatment due to pathophysiological changes in the drug metabolism of the critically ill and/or the ECMO therapy itself. This study investigates the latter aspect for commonly used anti-infective drugs in an ex vivo setting. A fully functional ECMO device circulated an albumin-electrolyte solution through the ECMO tubes and oxygenator. The antibiotic agents cefazolin, cefuroxim, cefepime, cefiderocol, linezolid and daptomycin and the antifungal agent anidulafungin were added. Blood samples were taken over a period of four hours and drug concentrations were measured via high-pressure liquid chromatography (HPLC) with UV detection. Subsequently, the study analyzed the time course of anti-infective concentrations. The results showed no significant changes in the concentration of any tested anti-infectives throughout the study period. This ex vivo study demonstrates that the ECMO device itself has no impact on the concentration of commonly used anti-infectives. These findings suggest that ECMO therapy does not contribute to alterations in the concentrations of anti-infective medications in severely ill patients.

BACKGROUND: Patients with severe respiratory failure due to COVID-19 who are not under mechanical ventilation may develop severe hypoxemia when complicated with spontaneous pneumomediastinum (PM). These patients may be harmed by invasive ventilation. Alternatively, veno-venous (V-V) extracorporeal membrane oxygenation (ECMO) may be applied. We report on the efficacy of V-V ECMO and invasive ventilation as initial advanced respiratory support in patients with COVID-19 and acute respiratory failure due to spontaneous PM.
METHODS: This was a retrospective cohort study performed between March 2020 and January 2022. Enrolled patients had COVID-19 and acute respiratory failure due to spontaneous PM and were not invasively ventilated. Patients were treated in the intensive care unit (ICU) with invasive ventilation (invasive ventilation group) or V-V ECMO support (V-V ECMO group) as the main therapeutic option. The primary outcomes were mortality and ICU discharge at 90 days after ICU admission.
RESULTS: Twenty-two patients were included in this study (invasive ventilation group: 13 [59%]; V-V ECMO group: 9 [41%]). The V-V ECMO strategy was significantly associated with lower mortality (hazard ratio [HR] 0.33 [95% CI 0.12-0.97], p = 0.04). Five (38%) patients in the V-V ECMO group were intubated and eight (89%) patients in the invasive ventilation group required V-V ECMO support within 30 days from ICU admission. Three (33%) patients in the V-V ECMO group were discharged from ICU within 90 days compared to one (8%) patient in the invasive ventilation group (HR 4.71 [95% CI 0.48-45.3], p = 0.18).
CONCLUSIONS: Preliminary data suggest that V-V ECMO without invasive ventilation may improve survival in COVID-19-related acute respiratory failure due to spontaneous PM. The study\'s retrospective design and limited sample size underscore the necessity for additional investigation and warrant caution.

OBJECTIVE: When out-of-hospital cardiac arrest (OHCA) becomes refractory, extracorporeal cardiopulmonary resuscitation (ECPR) is a potential option to restore circulation and improve the patient\'s outcome. However, ECPR requires specific materials and highly skilled personnel, and it is unclear whether increased survival and health-related quality of life (HRQOL) justify these costs.
RESULTS: This cost-effectiveness study was part of the INCEPTION study, a multi-centre, pragmatic randomized trial comparing hospital-based ECPR to conventional CPR (CCPR) in patients with refractory OHCA in 10 cardiosurgical centres in the Netherlands. We analysed healthcare costs in the first year and measured HRQOL using the EQ-5D-5L at 1, 3, 6, and 12 months. Incremental cost-effectiveness ratios (ICERs), cost-effectiveness planes, and acceptability curves were calculated. Sensitivity analyses were performed for per-protocol and as-treated subgroups as well as imputed productivity loss in deceased patients. In total, 132 patients were enrolled: 62 in the CCPR and 70 in the ECPR group. The difference in mean costs after 1 year was €5109 (95% confidence interval -7264 to 15 764). Mean quality-adjusted life year (QALY) after 1 year was 0.15 in the ECPR group and 0.11 in the CCPR group, resulting in an ICER of €121 643 per additional QALY gained. The acceptability curve shows that at a willingness-to-pay threshold of €80.000, the probability of ECPR being cost-effective compared with CCPR is 36%. Sensitivity analysis showed increasing ICER in the per-protocol and as-treated groups and lower probabilities of acceptance.
CONCLUSIONS: Hospital-based ECPR in refractory OHCA has a low probability of being cost-effective in a trial-based economic evaluation.

Monitoring for thrombosis and hemolysis is crucial for patients under extracorporeal or mechanical circulatory support, but it can be costly. We investigated correlations between hemolysis index (HI) and plasma-free hemoglobin (PFH) levels on one hand, and between the HI and plasma lactate dehydrogenase (LDH) levels on the other, in critically ill patients with and without extracorporeal or mechanical circulatory support. Additionally, we calculated the cost reductions if monitoring through HI were to replace monitoring through PFH or plasma LDH. In a single-center study, HI was compared with PFH and plasma LDH levels in blood samples taken for routine purposes in critically ill patients with and without extracorporeal or mechanical circulatory support. A cost analysis, restricted to direct costs associated with each measurement, was made for an average 10-bed ICU. This study included 147 patients: 56 patients with extracorporeal or mechanical circulatory support (450 measurements) and 91 patients without extracorporeal or mechanical circulatory support (562 measurements). The HI correlated well with PFH levels (r = 0.96; p < 0.01) and poorly with plasma LDH levels (r = 0.07; p < 0.01) in patients with extracorporeal or mechanical circulatory support. Similarly, HI correlated well with PFH levels (r = 0.97; p < 0.01) and poorly with plasma LDH levels (r = -0.04; p = 0.39) in patients without extracorporeal or mechanical circulatory support. ROC analyses demonstrated a strong performance of HI, with the curve indicating excellent discrimination in the whole cohort (area under the ROC of 0.969) as well as in patients under ECMO or mechanical circulatory support (area under the ROC of 0.988). Although the negative predictive value of HI for predicting PFH levels > 10 mg/dL was high, its positive predictive value was found to be poor at various cutoffs. A simple cost analysis showed substantial cost reduction if HI were to replace PFH or plasma LDH for hemolysis monitoring. In conclusion, in this cohort of critically ill patients with and without extracorporeal or mechanical circulatory support, HI correlated well with PFH levels, but poorly with plasma LDH levels. Given the high correlation and substantial cost reductions, a strategy utilizing HI may be preferable for monitoring for hemolysis compared to monitoring strategies based on PFH or plasma LDH. The PPV of HI, however, is unacceptably low to be used as a diagnostic test.

Recently, patients with out-of-hospital cardiac arrest (OHCA) refractory to conventional resuscitation have started undergoing extracorporeal cardiopulmonary resuscitation (ECPR). However, the mortality rate of these patients remains high. This study aimed to clarify whether a center ECPR volume was associated with the survival rates of adult patients with OHCA resuscitated using ECPR. This was a secondary analysis of a retrospective multicenter registry study, the SAVE-J II study, involving 36 participating institutions in Japan. Centers were divided into three groups according to the tertiles of the annual average number of patients undergoing ECPR: high-volume (≥ 21 sessions per year), medium-volume (11-20 sessions per year), or low-volume (< 11 sessions per year). The primary outcome was survival rate at the time of discharge. Patient characteristics and outcomes were compared among the three groups. Moreover, a multivariable-adjusted logistic regression model was applied to study the impact of center ECPR volume. A total of 1740 patients were included in this study. The center ECPR volume was strongly associated with survival rate at the time of discharge; furthermore, survival rate was best in high-volume compared with medium- and low-volume centers (33.4%, 24.1%, and 26.8%, respectively; P = 0.001). After adjusting for patient characteristics, undergoing ECPR at high-volume centers was associated with an increased likelihood of survival compared to middle- (adjusted odds ratio 0.657; P = 0.003) and low-volume centers (adjusted odds ratio 0.983; P = 0.006). The annual number of ECPR sessions was associated with favorable survival rates and lower complication rates of the ECPR procedure.Clinical trial registration: https://center6.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000041577 (unique identifier: UMIN000036490).

Background: Extracorporeal membrane oxygenation (ECMO) is frequently used during lung transplantation. Unfractionated heparin (UFH) is mainly used as part of ECMO support for anticoagulation. One of the most common perioperative complications is bleeding, which high-dose UFH can aggravate. Methods: We retrospectively analyzed (n = 141) patients who underwent lung transplantation between 2020 and 2022. All subjects (n = 109) underwent central cannulated VA ECMO with successful intraoperative ECMO weaning. Patients on ECMO bridge, postoperative ECMO, heart-lung transplants and transplants without ECMO were excluded. The dose of UFH for the entire surgical procedure, blood loss and consumption of blood derivatives intraoperatively and 48 h after ICU admission were recorded. Surgical revision for postoperative bleeding were analyzed. Thrombotic complications, mortality and long-term survival were evaluated. Results: Lower doses of UFH administered for intraoperative ECMO anticoagulation contribute to a reduction in intraoperative blood derivates consumption and blood loss with no thrombotic complications related to the patient or the ECMO circuit. Lower doses of UFH may lead to a decreased incidence of surgical revision for hemothorax. Conclusion: Lower doses of UFH as part of intraoperative ECMO anticoagulation might reduce the incidence of complications and lead to better postoperative outcomes.

UNASSIGNED: Our study aimed to compare the outcomes of COVID-19 patients who met a low-risk inclusion criteria for veno-venous extra corporeal membrane oxygenation (VV ECMO) with those who did not meet criteria due to higher risk but were subsequently cannulated.
UNASSIGNED: This was a retrospective observational cohort study that included adult patients who were placed on VV ECMO for COVID-19 related acute respiratory distress syndrome (ARDS) at a tertiary care academic medical center. The primary outcome was the association between the low-risk criteria and mortality. The patients met the criteria if they met EOLIA severe ARDS criteria, no absolute contraindications (age > 60 years, BMI > 55 kg/m2, mechanical ventilation (MV) duration >7 days, irreversible neurologic damage, chronic lung disease, active malignancy, or advanced multiorgan dysfunction), and had three or less relative contraindications (age > 50 years, BMI > 45 kg/m2, comorbidities, MV duration > 4 days, acute kidney injury, receiving vasopressors, hospital LOS > 14 days, or COVID-19 diagnosis > 4 weeks).
UNASSIGNED: Sixty-five patients were included from March 2020 through March 2022. Patients were stratified into low-risk or high-risk categories. The median Sequential Organ Failure Assessment score was 7 and the median PaO2/FiO2 ratio was 44 at the time of ECMO cannulation. The in-hospital mortality was 47.8% in the low-risk group and 69.0% in the high-risk group (p = 0.096).
UNASSIGNED: There was not a statistically significant difference in survival between low-risk patients and high-risk patients; however, there was a trend toward higher survival in the lower-risk group.