Photo Credit: ngraphix1
The following is a summary of “Factors and machine learning models for predicting successful discontinuation of continuous renal replacement therapy in critically ill patients with acute kidney injury: a retrospective cohort study based on MIMIC-IV database,” published in the November 2024 issue of Nephrology by Sheng et al.
The factors predicting successful discontinuation of CRRT in patients with AKI remain unclear. Identifying these factors could enhance decision-making in critical care.
Researchers conducted a retrospective study to identify factors for successful CRRT discontinuation in patients with AKI and developed predictive models for this outcome.
They conducted a retrospective study on adult patients with AKI who received CRRT from the Medical Information Mart for Intensive Care (MIMIC-IV) database. Successful CRRT discontinuation was defined as no CRRT requirement within 72 hours after cessation. They analyzed predictive factors and applied machine learning algorithms, including logistic regression (LR), decision tree (DT), random forest (RF), XGBoost, and K-nearest neighbor (KNN), to develop predictive models.
The results showed that 599 patients were included, with 475 (79.3%) successfully discontinuing CRRT. Key risk factors for successful discontinuation included urine output, non-renal SOFA score, bicarbonate, systolic blood pressure, and blood urea nitrogen. The KNN model had the highest AUC (0.870), followed by RF (0.847), XGBoost (0.830), LR (0.739), and DT (0.691). The ensemble models (RF and XGBoost) showed superior performance.
Investigators identified factors for successful CRRT discontinuation and developed predictive models for future use.
Source: bmcnephrol.biomedcentral.com/articles/10.1186/s12882-024-03844-z
