Analysis effects of shuxuening injection on patient's liver function based on electronic medical data.
- Author:
Teng-Teng HAO
;
Yan-Ming XIE
;
Wei YANG
;
Yong-Yan WANG
;
Lin LI
;
Yan ZHUANG
- Publication Type:Journal Article
- MeSH:
Adolescent;
Adult;
Aged;
Aged, 80 and over;
Drugs, Chinese Herbal;
administration & dosage;
adverse effects;
Electronic Health Records;
Female;
Humans;
Injections;
Liver;
drug effects;
metabolism;
Logistic Models;
Male;
Middle Aged;
Young Adult
- From:
China Journal of Chinese Materia Medica
2014;39(18):3609-3615
- CountryChina
- Language:Chinese
-
Abstract:
To assess whether the Shuxuening injection influences liver function, this paper analyzes the hospital information system (HIS) data from 20 national hospitals. The patients were divided two groups: patients (5 353 cases) using Shuxuening injection acted as the exposed group, then the patients (4 863 cases) not using Shuxuening injection acted as the unexposed group. Take the abnormal changes of alanine aminotransferase and aspartate aminotransferase indexes as the result. Control confounding factors by stratification analysis, classic logistic regression analysis, before logistic regression based on propensity score weighting method and logistic regression based on generalized boosted models (GBM) of propensity score weighted combination of covariate adjustment. Then, the study analyzes the effects of Shuxuening injection on liver function by observing the changes of the indexes. Analysis of the impact of parenterally administered Shuxuening on ALT shows that, before logistic regression based on propensity score, then the regression is -0.232 (P value is 0.017). However, after logistic regression based on propensity score method the regression became -0.150 (P value is 0.370). Whereas, analysis of the impact of parenterally administered Shuxuening on AST shows that, before logistic regression based on propensity score, then the regression is -0.034 (P value is 0.767). However, after logistic regression based on propensity score method the regression became -0.091(P value is 0.538). Thus, propensity score method can control a large number of confounding factors, making the results more close to the clinic. Hence, based on existing electronic medical data, this study didn't find that parenterally administered Shuexuening influences ALT and AST.
