Predictors of mortality at 28-days in infection associated acute kidney injury in cirrhosis

Predictors of Short-term Mortality in Patients of Cirrhosis of Liver Presenting as Acute Kidney Injury: An In-hospital Prospective Observational Study

Background

Acute kidney injury (AKI) is known to be associated with increased short-term mortality among cirrhotic patients. On this background, we designed this study to evaluate various causes of AKI among admitted patients with cirrhosis of liver and predictors of 90-day mortality.

Methods

One hundred and two consecutive adult patients with cirrhosis of liver with AKI hospitalized between November 2016 and March 2018 were enrolled in this prospective study. Their detailed clinical profile, including biochemical parameters, the etiology of AKI, and their clinical outcome of survival or mortality at 90-days, were recorded.

Results

The most common causes of AKI were infections, followed by hypovolemia, seen in 55.88% and 31.37% of the patients, respectively. Hepatorenal syndrome (HRS) was seen in 10.78%, while parenchymal renal disease was the least common (1.9%). The in-hospital mortality rate was 28.4%, while 90-day mortality was 39.21%. The HRS group had a high 90-day mortality rate of 54.54%. ROC analysis of various biochemical parameters revealed that serum creatinine (sCr), Model for End-Stage Liver Disease (MELD), International Normalized Ratio (INR), and Neutrophil-Lymphocyte ratio (NLR), followed by Child Turcotte Pugh (CTP), had high area under the curves of 0.785, 0.773, 0.747, 0.740, and 0.718, respectively, for the prediction of 90-day mortality.

Conclusion

Infection is the commonest cause of AKI in cirrhosis; however, mortality in patients with HRS-AKI is higher than that in those with infection-related AKI. Serum creatinine at admission, INR, NLR, and CTP scores predict short-term mortality among patients with AKI in cirrhosis. Further, large prospective studies are needed to confirm these findings.

MACHINE LEARNING MODELS FOR PREDICTING ACUTE KIDNEY INJURY IN PATIENTS WITH SEPSIS-ASSOCIATED ACUTE RESPIRATORY DISTRESS SYNDROME

ABSTRACT

Background: Acute kidney injury (AKI) is a prevalent and serious complication among patients with sepsis-associated acute respiratory distress syndrome (ARDS). Prompt and accurate prediction of AKI has an important role in timely intervention, ultimately improving the patients' survival rate. This study aimed to establish machine learning models to predict AKI via thorough analysis of data derived from electronic medical records. Method: The data of eligible patients were retrospectively collected from the Medical Information Mart for Intensive Care III database from 2001 to 2012. The primary outcome was the development of AKI within 48 hours after intensive care unit admission. Four different machine learning models were established based on logistic regression, support vector machine, random forest, and extreme gradient boosting (XGBoost). The performance of all predictive models was evaluated using the area under receiver operating characteristic curve, precision-recall curve, confusion matrix, and calibration plot. Moreover, the discrimination ability of the machine learning models was compared with that of the Sequential Organ Failure Assessment (SOFA) model. Results; Among 1,085 sepsis-associated ARDS patients included in this research, 375 patients (34.6%) developed AKI within 48 hours after intensive care unit admission. Twelve predictive variables were selected and further used to establish the machine learning models. The XGBoost model yielded the most accurate predictions with the highest area under receiver operating characteristic curve (0.86) and accuracy (0.81). In addition, a novel shiny application based on the XGBoost model was established to predict the probability of developing AKI among patients with sepsis-associated ARDS. Conclusions: Machine learning models could be used for predicting AKI in patients with sepsis-associated ARDS. Accordingly, a user-friendly shiny application based on the XGBoost model with reliable predictive performance was released online to predict the probability of developing AKI among patients with sepsis-associated ARDS.