# Risk factor analysis and predictive model development for healthcare-associated infections post-coronary artery bypass grafting

## TL;DR

This study identifies risk factors for infections after heart surgery and creates a model to predict infection risk.

## Contribution

A high-accuracy predictive model for healthcare-associated infections after CABG is developed using clinical risk factors.

## Key findings

- Diabetes, blood transfusion, and drainage tube duration are significant risk factors for post-CABG infections.
- The predictive model achieved 90.5% sensitivity and 92.1% specificity with an area under the curve of 0.970.
- The model integrates multiple clinical variables to estimate infection risk accurately.

## Abstract

This study aimed to analyze the risk factors associated with healthcare-associated infections (HAIs) in individuals who underwent post-coronary artery bypass grafting (CABG) and to develop a predictive model for infection risk assessment.

Clinical data were retrospectively collected from patients who underwent CABG at our hospital between January 2019 and December 2023. Data sources included the hospital infection surveillance system, hospital information system, and a questionnaire for HAIs in patients after cardiac surgery. Patients were divided into an infection group and a non-infection group based on whether they developed HAIs during the postoperative hospitalization period. Logistic regression was used to identify independent risk factors and to develop a risk prediction model. The predictive performance of the model was assessed using receiver operating characteristic curve analysis.

Independent risk factors for HAIs post-CABG included diabetes (odds ratio [OR] = 1.467), preoperative white blood cell count (OR = 0.117), preoperative albumin levels (OR = −0.146), intraoperative blood transfusion (OR = 0.001), presence of an indwelling drainage tube (OR = 0.864), drainage volume (OR = 0.003), duration of ventilator use (OR = 0.656), and central venous catheterization time (OR = 0.103). The predictive model was established as: Ln (P/1−P) = −2.230 + 1.467 * diabetes + 0.117 * preoperative white blood cell count −0.146 * preoperative albumin + 0.001 * intraoperative blood transfusion + 0.864 * drainage tube indwelling + 0.003 * drainage volume + 0.656 * ventilator use time + 0.103 * central venous catheterization time. The Hosmer-Lemeshow test indicated a good model fit with observed values. Receiver operating characteristic curve analysis demonstrated that the model achieved an area under the curve of 0.970, with a sensitivity of 90.5% and a specificity of 92.1%.

The independent risk factors for HAIs after CABG were diabetes, body mass index, preoperative white blood cell count, intraoperative blood transfusion volume, duration of pericardial and mediastinal drainage tube placement, total drainage volume, duration of mechanical ventilation, and duration of central venous catheterization. The developed risk prediction model demonstrated high accuracy in estimating postoperative HAI risk.