Abstract: Objective To establish an active monitoring model for adverse reactions of cephalosporins in hospitalized children using global trigger tools (GTT) in order to provide data for the prevention and treatment of related adverse reactions. Methods A total of 102 pediatric medical records of allergic reactions to cephalosporins in a hospital between 2017 and 2018 were collected as the test group, while another 269 patients who used the drug during the same period but without any allergic reactions were randomly selected as the control group. Single-factor analysis and multi-factor binary logistic regression analysis were conducted to establish an active monitoring model of adverse drug reactions (ADR). Then, the model's calibration was tested with the Hosmer-Lemeshow (HL) method, and the predictive value of the monitoring model was tested by the receiver operating characteristic (ROC) curve. Results There was no apparent difference in age and gender between the 102 cases with ADR after using cephalosporins. Skin rash and hypotension were the most common clinical manifestations in these cases. The independent variable factors with P≤0.2 in the results of the univariate analysis were selected to establish an active monitoring model of ADR of cephalosporins in pediatric patients: Logit(p)=In$\frac{p}{-p}$1.822=1.822 (neurological disease)-3.43 (white blood cell count<3×10⁹/L)-4.162(with hormone drugs)-5.956 (excessive sedation/hypotension/drowsiness)-5.041 (skin rash). The goodness-of-fit test results showed that the model was highly accurate, specific and sensitive. Conclusion The active monitoring model of ADR induced by cephalosporins in pediatrics is of high predictive value. Active intervention in a child's primary disease, white blood cell count, drug use, blood pressure, consciousness, rash and other risk factors can help reduce the chance of adverse drug reactions.

Key words: pediatrics, cephalosporins, global trigger tools, active monitoring model, adverse drug reaction