Takeuchi, Tomonori

Abstract Background: Thrombocytopenia commonly occurs in critically ill patients with acute kidney injury (AKI) receiving continuous renal replacement therapy (CRRT) and is associated with poor outcomes. While previous studies have examined static platelet counts and relative declines, the prognostic significance of platelet trajectory patterns before and after CRRT initiation remains unclear. This study investigates the association between platelet count trajectories and in-hospital mortality in critically ill adult patients with AKI receiving CRRT. Methods: This study is a single-center retrospective cohort study utilizing electronic health record (EHR) data from critically ill adults (≥18 years) with AKI who received CRRT at the University of Alabama at Birmingham between January 2012 and December 2023. Platelet trajectories were assessed using mixed-effects models incorporating all platelet measurements within seven days before and after CRRT initiation. The slopes of platelet count change were categorized as descending (≤ -10 × 10³/μL/day), stable (-10 to 10), or ascending (≥ 10), creating a total of nine trajectory patterns by combining the three pre-CRRT and three during-CRRT categories. The primary outcome was in-hospital mortality, which was analyzed in relation to these trajectory patterns using logistic regression models adjusted for demographics, comorbidities, Sequential Organ Failure Assessment (SOFA) score, medications, and use of other organ support devices as covariates. Results: The study included 2,226 critically ill adults with AKI on CRRT, encompassing 8,569 patient-days before and 11,699 patient-days after CRRT initiation for platelet slope calculation. The median age was 59 years (IQR 48–69) and the median SOFA score at CRRT initiation was 12 (9–14). The in-hospital mortality rate was 57.8%. Before CRRT initiation, the most common trajectory was descending (46.2%), while after CRRT initiation, the stable pattern was most prevalent (58.3%). All patterns with a descending platelet trajectory during-CRRT were significantly associated with higher in-hospital mortality (aOR 1.92 [95% CI: 1.16–3.21] for ascending-descending; 1.49 [1.07–2.08] for stable-descending; 1.52 [1.04–2.23] for descending-descending). In contrast, all patterns with stable or increasing platelet counts during-CRRT tended to have lower mortality. Conclusions: A descending platelet trajectory after CRRT initiation was independently associated with increased in-hospital mortality.

Key Points Pooled discrimination metrics were acceptable (area under the receiver operating characteristic curve >0.70) for all AKI risk classification categories in both internal and external validation. Better performance was observed in most recently published studies and those with a low or unclear risk of bias. Significant heterogeneity in patient populations, definitions, clinical predictors, and methods limit implementation in real-world clinical scenarios. Background Artificial intelligence through machine learning models seems to provide accurate and precise AKI risk classification in some clinical settings, but their performance and implementation in real-world settings has not been established. Methods PubMed, Excerpta Medica (EMBASE) database, Web of Science, and Scopus were searched until August 2023. Articles reporting on externally validated models for prediction of AKI onset, AKI severity, and post-AKI complications in hospitalized adult and pediatric patients were searched using text words related to AKI, artificial intelligence, and machine learning. Two independent reviewers screened article titles, abstracts, and full texts. Areas under the receiver operating characteristic curves (AUCs) were used to compare model discrimination and pooled using a random-effects model. Results Of the 4816 articles initially identified and screened, 95 were included, representing 3.8 million admissions. The Kidney Disease Improving Global Outcomes (KDIGO)-AKI criteria were most frequently used to define AKI (72%). We identified 302 models, with the most common being logistic regression (37%), neural networks (10%), random forest (9%), and eXtreme gradient boosting (9%). The most frequently reported predictors of hospitalized incident AKI were age, sex, diabetes, serum creatinine, and hemoglobin. The pooled AUCs for AKI onset were 0.82 (95% confidence interval, 0.80 to 0.84) and 0.78 (95% confidence interval, 0.76 to 0.80) for internal and external validation, respectively. Pooled AUCs across multiple clinical settings, AKI severities, and post-AKI complications ranged from 0.78 to 0.87 for internal validation and 0.73 to 0.84 for external validation. Although data were limited, results in the pediatric population aligned with those observed in adults. Between-study heterogeneity was high for all outcomes (I 2 >90%), and most studies presented high risk of bias (86%) according to the Prediction Model Risk of Bias Assessment Tool. Conclusions Most externally validated models performed well in predicting AKI onset, AKI severity, and post-AKI complications in hospitalized adult and pediatric populations. However, heterogeneity in clinical settings, study populations, and predictors limits their generalizability and implementation at the bedside.

Key Points In patients with cirrhosis and AKI requiring renal replacement therapy (RRT), hepatorenal syndrome-AKI was not associated with an increased 90-day mortality when compared with other AKI etiologies. Etiology of AKI may not be a critical factor regarding decisions to trial RRT in acutely ill patients with cirrhosis and AKI. Although elevated, mortality rates in this study are comparable with those reported in general hospitalized patients with AKI requiring RRT. Background While AKI requiring renal replacement therapy (AKI-RRT) is associated with increased mortality in heterogeneous inpatient populations, the epidemiology of AKI-RRT in hospitalized patients with cirrhosis is not fully known. Herein, we evaluated the association of etiology of AKI with mortality in hospitalized patients with cirrhosis and AKI-RRT in a multicentric contemporary cohort. Methods This is a multicenter retrospective cohort study using data from the HRS-HARMONY consortium, which included 11 US hospital network systems. Consecutive adult patients admitted in 2019 with cirrhosis and AKI-RRT were included. The primary outcome was 90-day mortality, and the main independent variable was AKI etiology, classified as hepatorenal syndrome (HRS-AKI) versus other (non–HRS-AKI). AKI etiology was determined by at least two independent adjudicators. We performed Fine and Gray subdistribution hazard analyses adjusting for relevant clinical variables. Results Of 2063 hospitalized patients with cirrhosis and AKI, 374 (18.1%) had AKI-RRT. Among them, 65 (17.4%) had HRS-AKI and 309 (82.6%) had non–HRS-AKI, which included acute tubular necrosis in most cases (62.6%). Continuous renal replacement therapy was used as the initial modality in 264 (71%) of patients, while intermittent hemodialysis was used in 108 (29%). The HRS-AKI (versus non–HRS-AKI) group received more vasoconstrictors for HRS management (81.5% versus 67.9%), whereas the non–HRS-AKI group received more mechanical ventilation (64.3% versus 50.8%) and more continuous renal replacement therapy (versus intermittent hemodialysis) as the initial RRT modality (73.9% versus 56.9%). In the adjusted model, HRS-AKI (versus non–HRS-AKI) was not independently associated with increased 90-day mortality (subdistribution hazard ratio, 1.36; 95% confidence interval, 0.95 to 1.94). Conclusions In this multicenter contemporary cohort of hospitalized adult patients with cirrhosis and AKI-RRT, HRS-AKI was not independently associated with an increased risk of 90-day mortality when compared with other AKI etiologies. The etiology of AKI appears less relevant than previously considered when evaluating the prognosis of hospitalized adult patients with cirrhosis and AKI requiring RRT.