Construction of medication error prediction model for intravenous medication dispensing center in a hospital from 2023 to 2024
10.12206/j.issn.2097-2024.202512047
- VernacularTitle:2023年～2024年某医院静脉用药调配中心用药错误预测模型构建研究
- Author:
Yanqing ZHANG
1
;
Rutao WANG
2
;
Meina ZHAO
1
;
Dan SUN
1
;
Min ZHAO
1
;
Xiaoyan ZHOU
1
Author Information
1. Department of Pharmacy, the First Affiliated Hospital of Air Force Medical University, Xi’an 710032, China.
2. a. Department of Pharmacy, b.Department of Cardiology, The First Affiliated Hospital of Air Force Medical University, Xi 'an 710032, China.
- Publication Type:Pharmacyadministration
- Keywords:
static matching center;
medication errors;
influencing factors;
predictive model
- From:
Journal of Pharmaceutical Practice and Service
2026;44(10):529-533
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the influencing factors of medication errors （ME） in the intravenous admixture service center of a tertiary hospital in Xi’an and construct a prediction model. Methods A retrospective cohort study design was adopted, with 200 cases of ME recorded in the hospital’s intravenous dispensing center from January 1, 2023 to December 31, 2024, as the ME group. 400 cases of routine medication records without ME were randomly selected as the ME group. Baseline data from two groups were collected and compared, the influencing factors of ME in the hospital’s static fitting center were analyzed, and a decision tree mode was constructed. Results There was a statistically significant difference （P<0.05） in incomplete medical order information, high-risk drugs, preparation time, and pharmacist training frequency between the ME group and the non ME group. However, there was no statistically significant difference （P>0.05） in other data between the groups; Through logistics regression analysis, it was found that incomplete medical order information, high-risk drugs, preparation time, and pharmacist training frequency were all influencing factors for ME occurrence in hospital intravenous dispensing centers （OR>1, P<0.05）; The constructed decision tree model selected a total of four explanatory variables, among which the configuration time was the important root node variable. The AUC of the decision tree model was 0.808 （95% CI: 0.774−0.842）, the sensitivity was 0.840, the specificity was 0.635, and the Youden index was 0.475. Conclusion The occurrence of ME in the hospital's static dispensing center from 2023 to 2024 was related to incomplete medical order information, high-risk drugs, preparation time, and pharmacist training frequency. The decision tree model constructed based on the above indicators had certain predictive value for ME in the static dispensing center.
