Data mining and preventive measures for 239 cases of falls in hospitalized adults / 中华护理杂志

ABSTRACT

Objective To find out valuable association patterns hidden in the events of falls among hospitalized patients,and to provide scientific references for prevention of falls.Methods According to basic principles of data mining,totally 7 170 records of 239 cases of falls in a tertiary hospital in Zhejiang Province from 2011 to 2016 were collected.Apriori algorithm was conducted to mine association patterns,and chi-square test was used to test effectiveness for strong association patterns.Results Through the condition setting,245 association patterns were obtained,and 94 patterns were selected by chi-square test.Finally,18 strong patterns were obtained via analysis with professional knowledge.Conclusion Through data mining of falls in hospitalized patients,figuring out strong association patterns for factors of falls,can help to discover system's weaknesses,and to provide scientific and accurate references for further development of targeted preventive measures.