Clinical prediction models for multidrug-resistant organism colonisation or infection in critically ill patients: a systematic review protocol

Machine learning-based antibiotic resistance prediction models: An updated systematic review and meta-analysis

BACKGROUND

The widespread use of antibiotics has led to a gradual adaptation of bacteria to these drugs, diminishing the effectiveness of treatments.

OBJECTIVE

To comprehensively assess the research progress of antibiotic resistance prediction models based on machine learning (ML) algorithms, providing the latest quantitative analysis and methodological evaluation.

METHODS

Relevant literature was systematically retrieved from databases, including PubMed, Embase and the Cochrane Library, from inception up to December 2023. Studies meeting predefined criteria were selected for inclusion. The prediction model risk of bias assessment tool was employed for methodological quality assessment, and a random-effects model was utilised for meta-analysis.

RESULTS

The systematic review included a total of 22 studies with a combined sample size of 43,628; 10 studies were ultimately included in the meta-analysis. Commonly used ML algorithms included random forest, decision trees and neural networks. Frequently utilised predictive variables encompassed demographics, drug use history and underlying diseases. The overall sensitivity was 0.57 (95% CI: 0.42-0.70; p< 0.001; I2= 99.7%), the specificity was 0.95 (95% CI: 0.79-0.99; p< 0.001; I2 = 99.9%), the positive likelihood ratio was 10.7 (95% CI: 2.9-39.5), the negative likelihood ratio was 0.46 (95% CI: 0.34-0.61), the diagnostic odds ratio was 23 (95% CI: 7-81) and the area under the receiver operating characteristic curve was 0.78 (95% CI: 0.74-0.81; p< 0.001), indicating a good discriminative ability of ML models for antibiotic resistance. However, methodological assessment and funnel plots suggested a high risk of bias and publication bias in the included studies.

CONCLUSION

This meta-analysis provides a current and comprehensive evaluation of ML models for predicting antibiotic resistance, emphasising their potential application in clinical practice. Nevertheless, stringent research design and reporting are warranted to enhance the quality and credibility of future studies. Future research should focus on methodological innovation and incorporate more high-quality studies to further advance this field.

Population pharmacokinetics and individualized dosing of tigecycline for critically ill patients: a prospective study with intensive sampling

Background: Due to the heterogeneity of critically ill patients, the pharmacokinetics of tigecycline are unclear, and the optimal dosing strategy is controversial. Methods: A single-center prospective clinical study that included critically ill patients who received tigecycline was performed. Blood samples were intensively sampled (eight samples each), and plasma drug concentrations were determined. A population pharmacokinetic (PPK) model was developed and evaluated by goodness-of-fit plots, bootstrap analysis and visual predictive checks. Monte Carlo simulation was conducted to optimize the dosage regimen. Results: Overall, 751 observations from 98 patients were included. The final PPK model was a two-compartment model incorporating covariates of creatinine clearance on clearance (CL), body weight on both central and peripheral volumes of distribution (V1 and V2), γ-glutamyl transferase and total bilirubin on intercompartment clearance (Q), and albumin on V2. The typical values of CL, Q, V1 and V2 were 3.09 L/h, 39.7 L/h, 32.1 L and 113 L, respectively. A dosage regimen of 50 mg/12 h was suitable for complicated intra-abdominal infections, but 100 mg/12 h was needed for community-acquired pneumonia, skin and skin structure infections and infections caused by less-susceptive bacteria. Conclusion: The Tigecycline PPK model was successfully developed and validated. Individualized dosing of tigecycline could be beneficial for critically ill patients.

Development and Validation of a Nomogram Prediction Model for Multidrug-Resistant Organisms Infection in a Neurosurgical Intensive Care Unit

Objective

To develop a predictive model for assessing the risk of multidrug-resistant organisms (MDROs) infection and validate its effectiveness.

We conducted a study on a total of 2516 patients admitted to the neurosurgery intensive care unit (NICU) of a Grade-III hospital in Nantong City, Jiangsu Province, China, between January 2014 and February 2022. Patients meeting the inclusion criteria were selected using convenience sampling. The patients were randomly divided into modeling and validation groups in a 7:3 ratio. To address the category imbalance, we employed the Synthetic Minority Over-sampling Technique (SMOTE) to adjust the MDROs infection ratio from 203:1558 to 812:609 in the training set. Univariate analysis and logistic regression analysis were performed to identify risk factors associated with MDROs infection in the NICU. A risk prediction model was developed, and a nomogram was created. Receiver operating characteristic (ROC) analysis was used to assess the predictive performance of the model.

Patients and Methods

Results

Logistic regression analysis revealed that sex, hospitalization time, febrile time, invasive operations, postoperative prophylactic use of antibiotics, mechanical ventilator time, central venous catheter indwelling time, urethral catheter indwelling time, ALB, PLT, WBC, and L% were independent predictors of MDROs infection in the NICU. The area under the ROC curve for the training set and validation set were 0.880 (95% CI: 0.857-0.904) and 0.831 (95% CI: 0.786-0.876), respectively. The model's prediction curve closely matched the ideal curve, indicating excellent predictive performance.

Conclusion

The prediction model developed in this study demonstrates good accuracy in assessing the risk of MDROs infection. It serves as a valuable tool for neurosurgical intensive care practitioners, providing an objective means to effectively evaluate and target the risk of MDROs infection.

Preventing Multidrug-Resistant Bacterial Transmission in the Intensive Care Unit with a Comprehensive Approach: A Policymaking Manual

Patients referred to intensive care units (ICU) commonly contract infections caused by multidrug-resistant (MDR) bacteria, which are typically linked to complications and high mortality. There are numerous independent factors that are associated with the transmission of these pathogens in the ICU. Preventive multilevel measures that target these factors are of great importance in order to break the chain of transmission. In this review, we aim to provide essential guidance for the development of robust prevention strategies, ultimately ensuring the safety and well-being of patients and healthcare workers in the ICU. We discuss the role of ICU personnel in cross-contamination, existing preventative measures, novel technologies, and strategies employed, along with antimicrobial surveillance and stewardship (AMSS) programs, to construct effective and thoroughly described policy recommendations. By adopting a multifaceted approach that combines targeted interventions with broader preventive strategies, healthcare facilities can create a more coherent line of defense against the spread of MDR pathogens. These recommendations are evidence-based, practical, and aligned with the needs and realities of the ICU setting. In conclusion, this comprehensive review offers a blueprint for mitigating the risk of MDR bacterial transmission in the ICU, advocating for an evidence-based, multifaceted approach.