Dhiman, Radha K

INTRODUCTION: Nosocomial infections (NIs) in cirrhosis are associated with high mortality but could be preventable. Logistic regression (LR) models have failed to identify high-risk patients. We aimed to develop machine learning (ML) models to predict NI. METHODS: The CLEARED consortium consists of prospectively enrolled cirrhosis inpatients from >120 centers. Using day-of-admission clinical data, 3 ML approaches (random forest [RF], extreme gradient boosting, and neural networks [NNs]) were used to predict NI. Data were split 80:20 for training and testing stratified by the outcome. Models were compared using area under the receiver operating characteristic curve (AUC). RESULTS: In total, 8,263 patients (55.90 ± 13.34 years; 64.1% men) from 127 centers in 37 countries were included. NI developed in 869 (10.5%), a median of 6 (4–11) days of postadmission. Major NIs were respiratory (29.6%) and urinary tract infection (15.7%), and spontaneous bacterial peritonitis (13.5%). NIs occurred more frequently in patients from low/low-middle income countries and those with severe liver disease, alcohol etiology, and admission infections. NIs were associated with inpatient mortality (31.9% vs 8.1%, P < 0.001) and liver transplantation (4.8% vs 1.9%, P < 0.001). Although the RF model (AUC 0.69) showed good calibration (Brier score 0.09), outperforming extreme gradient boosting, neural network, and LR models (AUC 0.66 for all; LR comparison P = 0.043), no model achieved AUC ≥0.80 for clinical utility. At 10% predicted probability threshold, the RF model demonstrated only 75.4% sensitivity, 52.9% specificity, and 15.9% positive predictive value (PPV). DISCUSSION: NIs cannot be accurately predicted from day-of-admission data using ML models, even in a large, prospective, global cirrhosis cohort. Every hospitalized patient with cirrhosis should receive protocolized infection control measures.

SummaryBackground and AimsMachine learning (ML) can identify the hidden patterns without hypothesis in heterogeneous diseases like acute‐on‐chronic live failure (ACLF). We employed ML to describe and predict yet unknown clusters in ACLF.MethodsClinical data of 1568 patients with ACLF from a tertiary care centre (2015–2023) were subjected to distance‐, density‐ and model‐based clustering algorithms. Final model was selected on best cluster separation, viz. Silhouette width and Dunn's index (for distance‐ or density‐based algorithms) and minimum BIC (for model‐based algorithms). Cluster assignments, patient trajectories and survival were analysed through inferential statistics. Supervised ML models were trained in 70% data that predicted clusters in remaining 30% data followed by an temporal validation.ResultsThe cohort was male‐predominant (87%), aged 44.3 years, with alcohol‐associated hepatitis (62.9%) and survival of 50.5%. Due to poor performance of distance‐ and density‐based algorithms and better explainability, the latent class model (LCM) was selected for exploration. LCM revealed four clusters with distinct trajectories, reversibility and survival (independent of MELD, CLIF‐C ACLF and AARC scores). Cluster1 had patients with none/one organ failure and highest reversibility. Cluster2 had females with viral hepatitis and two organ failures. More‐than‐one acute precipitant, severity, infections, organ failures and irreversibility escalated from clusters 1 to 4. Circulatory and renal failures critically influenced cluster assignments. Incorporating clusters to CLIF‐C ACLF, infection and ACLF definition improved the discriminative accuracy of CLIF‐C‐ACLF by 11%. Extreme gradient boost and decision trees could predict clusters with AUCs of 0.989 (0.979–0.995) and 0.875 (0.865–0.890). MELD, CLIF‐C‐OF, haemoglobin, lactate, CLIF‐C‐ACLF and ALT were critical variables for cluster prediction. Clusters with distinct survival were documented in a temporal validation cohort.ConclusionsML for the first time could identify clusters with distinct phenotypes, trajectories and outcomes in ACLF. Stratification into clusters can address heterogeneity, guide prognosis, recruitment in trials, resource allocation and liver transplant discussions in ACLF.