Correlation Analysis Method Study of Chronic Disease and Multidisease Coexistence in Longitudinal Physical Examination Queue
Objective To explore effect of correlation analysis method for comorbidity of multiple diseases,to provide reference for the health management of population prone to chronic diseases.Methods The paper chose individual physical examination data based on consecutive years collected by the Information Department of the Western Theater Command Hospital of PLA from 2006 to 2016,and established longitudinal queue of physical examination data.Based on the modeling of longitudinal queue physical examination data,the paper constructed data-driven correlation analysis method for chronic comorbidities based on a mixed effect model,"hidden"incidence characteristics of individual chronic comorbiditiesandincidence risks,and analyzed relationship between non-alcoholic fatty liver disease(NAFLD)and comorbid hyperuricemia(HUA),and further elucidated applicability and advantages of the method in practical applications.ResultsThe model considering comorbidities risk had significantly lower AIC and BIC values than general mixed effect model,and the risk assessment was more accurate and credible.Conclusion Comorbidity analysis method can describe relationship between chronic comorbidities for chronic comorbidities based on the mixed effect model combined with the potential risk of chronic diseasesa ccurately and practically,which is worthy of promotion and application.
Disease correlation analysisLongitudinal examination queueRandom effect modelComorbidity of multiple diseases
NETL
NSTL
万方数据
