Evaluation of ICD data quality utilizing DQA model
10.3969/j.issn.1671-332X.2025.05.012
- VernacularTitle:基于DQA模型的ICD数据质量评估研究
- Author:
Meijun DENG
1
;
Ying ZHANG
1
;
Guangming TAN
1
Author Information
1. 广东省妇幼保健院 广东 广州 511400
- Publication Type:Journal Article
- Keywords:
International classification of diseases;
Data quality assessment;
Medical record front pages
- From:
Modern Hospital
2025;25(5):701-704,708
- CountryChina
- Language:Chinese
-
Abstract:
Objective To systematically evaluate the ICD coding data on the front page of inpatient medical records filed out by clinicians,and to analyze the quality of ICD codes reasonably,objectively and comprehensively by employing the 3×3 Da-ta Quality Assessment(3×3 DQA)model.Methods A total of 1 410 medical records from a tertiary women and children's hospital between June 2021 to July 2024 were randomly selected.Based on the 3×3 DQA model,the data quality of ICD codes for diagnosis and surgical procedures on the front pages filed out by clinicians was assessed.Results Among 1,410 medical re-cords reviewed,778 had defects,resulting in a total defect rate of 55.18％,including a completeness defect rate of 32.98％and an accuracy defect rate of 30.57％.Totally,a total of 9 302 ICD codes were reviewed,among which 1 254 had defects,with a total defect rate of 13.48％.The defect rates for the principal diagnosis,other diagnosis,principal operation,and other operation were 22.20％,17.72％,4.53％and 4.92％,respectively.For diagnostic ICD codes,the completeness defect rate was relative-ly high,while for surgical ICD codes,the accuracy defect rate was higher.In terms of specialties,there were statistically signifi-cant differences in the overall defect rates among gynecology,obstetrics,and pediatrics,with pediatrics and gynecology having defect rates higher than the overall average.Conclusion The quality of ICD code data on the front page of medical records filed out by clinicians is suboptimal,and there are differences in data quality across different specialties and ICD categories.There-fore,it is still necessary to conduct research on the supervision and improvement measures for ICD data quality in hospital medical record management.
