Chaudhary, Sanjay

Backround: Atrial fibrillation (AF) is a common complication following veno-venous extracorporeal membrane oxygenation (VV-ECMO), leading to further complications from anticoagulation or stroke. The Artificial Intelligence AF Prediction Dashboard developed at Mayo Clinic generates two predictive outputs: (1) a continuous probability score and (2) a binary threshold-based signal to predict the likelihood of AF at an individual level after 12-lead electrocardiogram (ECG) is obtained. Using both continuous risk (%) and binary (yes/no) likelihood, we present a novel method to understand AF risk after VV-ECMO cannulation based on pre-ECMO 12-lead ECG. Methods: We reviewed a total of 147 patients on ECMO between 2018 and 2024, from which 85 VV-ECMO patients with complete data were included. AI Dashboard outputs were quantified as binary (1/0) or percent likelihood for AF. R was utilized for statistical analysis. Results: 85 patients met our inclusion criteria from which AF after cannulation occurred in 31 (36%). 17 patients had a history of AF prior to VV-ECMO initiation. The mean age of our cohort was 49 years with a mean BMI of 29.3. Baseline AI Dashboard continuous assessment identified 39 patients (46%) compared to the binary assessment identifying 31 patients (36%). Logistic regression in the binary assessment demonstrated an odds ratio of 1.08 (95% CI 1.04 – 1.12) with a p value of <0.001. The binary AF Prediction yielded a 58% sensitivity, 76% specificity, 58% PPV, and 76% NPV, with an accuracy of 67%. The continuous AF Probability score achieved 72% balanced accuracy with sensitivity 74%, specificity 70%, PPV 59% and NPV 83%, indicating fair overall performance and suggesting good utility for ruling out AF in low-risk patients. Although both outputs provided moderate predictive accuracy, the continuous score reported before ECMO cannulation demonstrated stronger overall performance with fewer false negatives and higher rule-out value. Conclusions: We demonstrate the novel use of an AI predictive model in the VV-ECMO population for the prediction of AF risk with both continuous and binary risk cutoffs from baseline ECG. This data highlights the need for further development in ECMO-specific AI models to help mitigate risk after the onset of atrial arrhythmias. As AI-driven tools become increasingly integrated into care, their performance must be continuously validated and recalibrated for specific populations to ensure their safe and effective use.

The evidence on indications, outcomes, and complications with the use of extracorporeal membrane oxygenation (ECMO) in the setting of interstitial lung disease (ILD) is limited in the existing literature. We performed a systematic review and meta-analysis for the use of ECMO in the setting of ILD to study the prognostic factors associated with in-hospital mortality. Eighteen unique studies with a total of 1,356 patients on ECMO for ILD were identified out of which 76.5% were on ECMO as a bridge to transplant (BTT) and the rest as a bridge to recovery (BTR). The overall in-hospital mortality was 45.76%, with 71.3% and 37.8% for BTR and BTT, respectively. Among the various prognostic factors, mortality was lower with younger age (mean difference = 3.15, 95% confidence interval [CI] = 0.82–5.49), use of awake veno-arterial (VA)-ECMO compared to veno-venous (VV)-ECMO (unadjusted odds ratio [OR] = 0.22, 95% CI = 0.13–0.37) in the overall cohort. In the setting of BTT, the use of VA-ECMO had a decreased hazard ratio (HR) compared to VV-ECMO (adjusted HR = 0.34, 95% CI = 0.15–0.81, p = 0.015). The findings of our meta-analysis are critical but are derived from retrospective studies with small sample sizes and thus are of low to very low-GRADE certainty.

Abstract Background Early diagnosis, prognostication, and treatment initiation are the cornerstones of sepsis management. There remains an immense interest in exploring the diagnostic and prognostic roles of neutrophil-lymphocyte ratio (NLR) in the sepsis population. Methods We performed a retrospective multi-center observational study including patients admitted to different Mayo Clinic sites with a diagnosis of sepsis between October 2018 and August 2022. Patients were excluded if they were aged &lt; 18 years, lacked research authorization, or had missing Neutrophil or Lymphocyte count at admission. Data were collected from an existing sepsis database and Mayo Data Explorer (MDE). Categorical variables were summarized as percentages and continuous variables were summarized as medians. Chi-square test for significance, independent sample t-test, and multivariate Cox Proportional Hazard models were performed on IBM SPSS Statistics v28.0. Results Our study cohort consisted of 13968 patients, of which the majority were males (55.9%), Caucasian (91.6%), and non-Hispanic or Latino (94.3%). The median age of the cohort was 71 (60, 80) years, and the median BMI was 27.7 (23.4, 33.3) kg/m2. Among all sepsis patients, 6.0% had a positive blood culture and the most common organisms were E. coli (26.5%), Klebsiella sp. (15.6%), Streptococcus sp. (12.8%) and Staphylococcus aureus (9.9%). 13131 patients (94.1%) survived the hospital stay and 837 (5.9%) did not. There was no difference in mortality based on blood culture positivity status (6.0% vs 6.6%, p=0.41). However, the median NLR was higher among patients with a positive blood culture (16.4 vs. 12.1, p&lt; 0.001) and among non-survivors (14.3 vs 12.2, p&lt; 0.001). NLR did not significantly vary based on the type of organism growing in the blood culture. A multivariate model revealed NLR as an independent predictor of mortality (HR 1.003, 95% CI: 1.000-1.006, p=0.029) after adjusting for comorbidities, baseline clinical and laboratory variables. (Tables 1 & 2) Conclusion Our retrospective multicentric study showed that the Neutrophil-Lymphocyte ratio at admission was higher among bacteremic sepsis patients and non-survivors. Further prospective studies are needed to explore its diagnostic and prognostic utility. Disclosures All Authors: No reported disclosures