Quality evaluation of adverse drug reaction reports based on weighted TOPSIS-RSR model
10.3760/cma.j.cn114015-20240719-00606
- VernacularTitle:基于加权TOPSIS-RSR模型的药物不良反应报告质量的评价
- Author:
Liang WU
1
;
Jingbao CHEN
;
Xiaoxiao CHEN
;
Shanyue JIANG
;
Yun SHEN
Author Information
1. 安徽省六安市中医院药学部，六安 237006
- Publication Type:Journal Article
- Keywords:
Adverse drug reaction reporting systems;
Health care quality, access, and evaluation;
Technique for order preference by similarity to ideal solution;
Ran
- From:
Adverse Drug Reactions Journal
2025;27(4):218-224
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To understand the quality of adverse drug reaction (ADR) reports in Lu′an Hospital of Traditional Chinese Medicine (our hospital) and its change trend in recent years, and explore the methods of objectively evaluating the quality of ADR reports.Methods:According to the 20 evaluation indicators of the ADR report quality evaluation scoring table in the Appendixes 5 of Provisions for Adverse Drug Reaction Reporting and monitoring, the ADR reports submitted to the National Center for ADR Monitoring from 2013 to 2022 by our hospital were evaluated. The weighted technique for order preference by similarity to ideal solution (TOPSIS) combined with rank-sum ratio (RSR) model was used to rank the quality of ADR reports into the following 5 grades: excellent, good, medium, qualified and unqualified, according to the weight of each evaluation index. The quality grading results were tested to determine the rationality of grading. Results:A total of 3 947 ADR reports were included in the analysis, including 1 361 new/serious ADR reports (34.5%), and the average score of quality evaluation index was 87.9. After 2016, the number of ADR reports and the proportion of reports with scores ≥ 80 increased significantly. Among the 20 evaluation indicators, 10 had a high pass rate, 7 had a medium or upper pass rate, and 3 had a low pass rate. The TOPSIS-RSR model was used to classify the quality of ADR reports. The overall proportions of excellent, good, moderate, qualified, and unqualified reports were 4.7% (186/3 947), 23.0% (908/3 947), 45.3% (1 787/3 947), 23.4% (925/3 947), and 3.6% (141/3 947), respectively. The homogeneity of variance test showed that each grade met the homogeneity of variance, and the analysis of variance results showed that the differences between every 2 grades were statistically significant ( P<0.001), indicating that the quality grading was reasonable. Conclusions:After 2016, the quantity and quality of ADR reports in our hospital have significant improvement, but there are still some evaluation indicators with low pass rate. Using the weighted TOPSIS-RSR model to grade the quality of ADR reports can more objectively reflect the quality of ADR reports.
