To develop and validate a simple scoring system for predicting the risk of infection after acute myocardial infarction (AMI) in patients undergoing percutaneous coronary intervention (PCI).

Continuous patients with ST-segment elevation myocardial infarction (STEMI) who received PCI in our hospital were divided into training set and validation set using a cohort approach. All patients received oral rosuvastatin calcium tablets 10-20 mg per day combined with evolocumab injection therapy 140 mg. The primary endpoint was post-AMI infection during hospitalization, while all-cause mortality and major adverse cardiovascular events (MACE) were considered secondary endpoints. The area under the curve (AUC) and the calibration between predicted and observed infection risks were calculated. A backward stepwise logistic regression model was selected to develop the risk score.

A total of 359 patients were included, with 217 in the training set and 142 in the validation set, and no patient dropout occurred during the study. A 24-point risk scoring system was developed, with infection risks ranging from 0.70% to 99.60% in patients with the lowest and highest scores, respectively. Seven variables including age, Killip class, insulin use, white blood cell count, serum albumin, diuretics and transfemoral approach were incorporated into the model. The model achieved equally high discrimination in both the development and validation cohorts (C-statistic: 0.85) and demonstrated good calibration in both datasets. In both development and validation cohorts, the incidence of post-myocardial infarction infection increased steadily across risk score groups. Additionally, this risk score exhibited good predictive ability for infection, in-hospital all-cause mortality.

This study established a simple bedside tool, a risk scoring system for assessing the risk of infection and other in-hospital outcomes in STEMI patients undergoing PCI. Clinicians can utilize this risk scoring system to evaluate infection risk and make evidence-based decisions accordingly.