OBJECTIVE To explore the potential risk factors for neonatal hospital-acquired sepsis, construct a scientific and effective prediction model, and evaluate its efficacy.

METHODS A total of 74 neonates with hospital-acquired sepsis admitted to the neonatal ward of Tianjin First Central Hospital from Jan. 1, 2019 to Dec. 31, 2023 were selected as the infection group. Meanwhile, 148 neonates admitted to the same neonatal ward during the same period but without hospital-acquired infections were randomly selected as the control group at a ratio of 1∶2. Multivariate logistic regression analysis was conducted to identify the risk factors for neonatal hospital-acquired sepsis, and a prediction model was established. The model′s efficacy was evaluated with the area under the receiver operating characteristic curve (AUC) and calibration curve.

RESULTS The multivariate logistic regression model revealed that birth weight odds ratio (OR)=0.998, 95%CI: 0.997 to 0.999, use of a central venous catheter (OR=10.740, 95%CI: 1.491 to 219.946), gestational age (OR=0.705, 95%CI: 0.504 to 0.955), length of hospital stay (OR=1.037, 95%CI: 1.007 to 1.071) and amniotic fluid volume (OR=0.994, 95%CI: 0.988 to 0.998) were risk factors for neonatal hospital-acquired sepsis (P < 0.05). A nomogram prediction model was established based on these independent risk factors, with AUC values of 0.943 and 0.947 for the training and validation sets, respectively. The Hosmer-Lemeshow test indicated good accuracy in model fitting (training set: χ²=4.522, P=0.340, validation set: χ²=5.279, P=0.260).

CONCLUSIONS Birth weight, use of a central venous catheter, gestational age, length of hospital stay and amniotic fluid volume are independent predictors of neonatal hospital-acquired sepsis. The nomogram model constructed based on these factors demonstrates relatively high predictive efficacy, aiding in the early assessment and identification of high-risk potential neonates, thereby reducing the risk of neonatal hospital-acquired sepsis.