Abstract: Objective To investigate the effect of Mailuoshutong pills on renal function indexes in the real world. Methods Based on the integrated data warehouse, the data on patients who took Mailuoshutong pills and those who did not between 2006 and 2016 was retrieved from the medical electronic database of fourteen tertiary hospitals in China. With the same gender and age of ±5 years as the matching conditions, patients in the Mailuoshutong pill group and control group were matched at the ratio of 1∶1. After matching, there were 2 395 patients in either group. The effect of medication on renal function indexes (Cr, BUN) was used as the outcome evaluation index. Propensity scoring was used to control the confounders in this study, and four statistical methods were used for safety analysis. Results Eleven confounders were weighted and balanced with the propensity scoring method, and the results obtained by the four statistical methods were all P>0.05. Abnormal changes in renal function indexes were as likely in the control group as in the Mailuoshutong pill group, and the difference between the two groups was not statistically significant. In the real world, the use of Mailuoshutong pills had no direct effect on renal function indexes. Conclusion Based on the data of this study, no significant effect of MailuoShutong pills on renal function indexes is found, but this drug should be used with caution.

Key words: Mailuoshutong pills, hospital information system, the real world, renal function index, propensity score, logistic regression