Factors related to antibiotic-associated diarrhea in patients in the intensive care unit receiving antifungals: a single-center retrospective study

Interpretable machine learning models for predicting the incidence of antibiotic- associated diarrhea in elderly ICU patients

BACKGROUND

Antibiotic-associated diarrhea (AAD) can prolong hospitalization, increase medical costs, and even lead to higher mortality rates. Therefore, it is essential to predict the incidence of AAD in elderly intensive care unit(ICU) patients. The objective of this study was to create a prediction model that is both interpretable and generalizable for predicting the incidence of AAD in elderly ICU patients.

METHODS

We retrospectively analyzed data from the First Medical Center of the People's Liberation Army General Hospital (PLAGH) in China. We utilized the machine learning model Extreme Gradient Boosting (XGBoost) and Shapley's additive interpretation method to predict the incidence of AAD in elderly ICU patients in an interpretable manner.

RESULTS

A total of 848 adult ICU patients were eligible for this study. The XGBoost model predicted the incidence of AAD with an area under the receiver operating characteristic curve (ROC) of 0.917, sensitivity of 0.889, specificity of 0.806, accuracy of 0.870, and an F1 score of 0.780. The XGBoost model outperformed the other models, including logistic regression, support vector machine (AUC = 0.809), K-nearest neighbor algorithm (AUC = 0.872), and plain Bayes (AUC = 0.774).

CONCLUSIONS

While the XGBoost model may not excel in absolute performance, it demonstrates superior predictive capabilities compared to other models in forecasting the incidence of AAD in elderly ICU patients categorized based on their characteristics.

Antibiotics and antibiotic-associated diarrhea: a real-world disproportionality study of the FDA adverse event reporting system from 2004 to 2022

BACKGROUND

Our study aimed to assess the risk signals of antibiotic-associated diarrhea (AAD) caused by various antibiotics using real-world data and provide references for safe clinical applications.

METHODS

We analyzed data extracted from the FDA Adverse Event Reporting System (FAERS) database, covering the period from the first quarter of 2004 to the third quarter of 2022. We computed the reporting odds ratio (ROR) for each antibiotic or antibiotic class to compare the signal difference. Furthermore, we also examined the differences in the onset times and outcomes of AAD caused by various antibiotics.

RESULTS

A total of 5,397 reports met the inclusion requirements. Almost all antibiotics, except tobramycin and minocycline (ROR 0.98; 95%CI: 0.64-1.51 and 0.42; 95%CI: 0.16-1.11, respectively), showed a significant correlation with AAD. The analysis of the correlation between different classes of antibiotics and AAD revealed that lincomycins (ROR 29.19; 95%CI: 27.06-31.50), third-generation cephalosporins (ROR 15.96; 95%CI: 14.58-17.47), and first/second generation cephalosporins (ROR 15.29; 95%CI: 13.74-17.01) ranked the top three. The ROR values for antibiotics from the same class of antibiotics also varied greatly, with the ROR values for third-generation cephalosporins ranging from 9.97 to 58.59. There were also differences in ROR values between β-lactamase inhibitors and their corresponding β-lactamase drugs, such as amoxicillin-clavulanate (ROR = 13.31; 95%CI: 12.09-14.65) and amoxicillin (ROR = 6.50; 95%CI: 5.69-7.44). 91.35% of antibiotics have an onset time of less than four weeks.

CONCLUSIONS

There is a significant correlation between almost all antibiotics and AAD, particularly lincomycins and β-lactam antibiotics, as well as a different correlation within the same class. These findings offer valuable evidence for selecting antibiotics appropriately.