Data mining and management inspiration of comprehensive hospital accreditation data based on association rules / 中华医院管理杂志

ABSTRACT

Objective:To conduct data mining on hospital accreditation results for inspirations and clues in hospital management.Methods:The Apriori function contained in the arules package of R software was used to extract the association rules. This practice aimed to analyze association rules of the accreditation results of 56 tertiary hospitals which were made based on tertiary hospital accreditation standard(2011)and Detailed Rules from 2017 to 2019; to explore the correlation between the clauses, and analyze the inspirations for hospital management.Results:A total of 6 566 138 and 247 rules were generated for all clauses and core clauses, receptively. The top 10 rules sorted by lift were selected as strong association rules. Among them, the minimum lift of all clauses and core clauses was 1.41 and 1.53, respectively. There was a strong correlation between the establishment of a quality and safety management team in the hospital and the development & implementation of relevant regulations by the hospital. There were strong correlations among emergency service procedure and regulations, patient rights to know, responsibility system by the first one to receive a complaint, medical safety(adverse)event reporting, development & implementation of antimicrobial management system, as well as training for prevention of multi-drug resistance infection measures, and multi-drug resistance infection hospital infection control system.Conclusions:This study suggested that hospital management should highlight the correlation between regulations development and staffing, that between indicator systems and monitoring feedback systems, and that between indicators in different sectors in the medical process. These correlations can be used as management clues and inspirations for hospital management.