Laure Wynants

Objectives: To systematically review the risk of bias and applicability of published prediction models for risk of central line-associated bloodstream infection (CLA-BSI) in hospitalized patients. Introduction: CLA-BSIs are the most common source of hospital-acquired infections (HAIs), and are always associated with higher morbidity, longer length of stay and increased financial burdens. As a priority target for prevention, tools that were developed to predict the risk of CLA-BSI for individuals may help improve the infection control in hospitals. In this systematic review, we evaluated the current risk prediction models for CLA-BSI and discussed the practical problems for implementing the models. Inclusion criteria: All inpatients (no age limit) with at least a central line in place during their hospitalization anytime. Methods: Four key databases including PubMed (MEDLINE), Embase (Embase.com), Web of Science Core Collection (SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, BKCI-S, BKCI-SSH, ESCI, CCR-EXPANDED, IC) and Scopus were used to conduct the search. The search was conducted on the 10th of July 2023. We included studies that describe the development or validation of a prediction model for CLABSI and have at least two predictor variables to build multivariable predictive models. Articles that do not report original research (i.e. reviews), or that are not full papers (i.e. letters, notes, and conference abstracts), or qualitative studies were excluded. References were imported and deduplicated using EndNote (Clarivate Analytics) and Rayyan (Qatar Computing Research Institute). Titles and abstracts were initially screened for exclusion by at least two authors. As interreviewer agreement was considered reliable, the remaining title-abstract screening was done by one author and irrelevant articles were excluded. Then, full text of the potentially relevant articles were screened independently by two authors. Discrepancies were resolved through discussion with a third author. We also used forward and backward snowballing from the final articles that will be included within this review.