基于医院信息系统数据的舒血宁注射液对肾功能影响的临床实效研究

药物的安全性问题一直是临床医生和患者共同关注的重要问题,肾脏是药物代泄的重要器官。该研究基于全国20家大型三甲医院信息系统(hospital information system,HIS)数据,以肾功能常规检测的血肌酐(Scr)和血尿素氮(BUN)作为结局指标,采用generalized boosted models(GBM)倾向性评分法对混杂因素进行平衡,将使用舒血宁注射液和未使用者肾功能变化进行对比,探讨临床使用舒血宁注射液是否对肾功能有影响。结果显示,经典logistic回归分析显示使用舒血宁注射液对Scr可能有影响(P为0.035),但使用倾向性评分后则无显著影响(P为0.513,0.590);倾向性评分前后结果均显示使用舒血宁注射液对BUN可能有影响(P均<0.05)。由于该研究为回顾性研究,虽然采用倾向性评分法对已知的混杂变量已控制,仍可能存在潜在偏倚,因此进一步研究需重点监测其对BUN是否真正存在影响,为临床安全用药提供依据。

Analysis of clinical effects of Shuxuening injectionon renal function based on hospital information system data

The safety of drugs has been the important issue that has caught the mutual attention of clinicians and patients. Kidney is one of the important organs in drug metabolism. The study is based on hospital information system(HIS) data from 20 nationwide general hospitals, indicated by Scr and BUN. Generalized boosted models(GBM) based on propensity score method were used to balance confounding factors. The study contrasts changes on renal function between used Shuxuening injectionpatients and the unused to explore whether Shuxuening injection affects renal function in clinical. The results showed that Shuxuening injection is likely to affect Scr in logistic regression (P value is 0.035);after the logistic regression based on propensity score method(P value became 0.513,0.590),no differences were found between the two groups; before and after the logistic regression based on propensity score method(P value was less than 0.05), Shuxuening injection possibly affects BUN. Nevertheless, this study is a retrospective analysis. Logistic regression based on propensity score method can control the known confounding factors very well,while there may be the unknown ones. Therefore, the further exploration is required to monitor whether Shuxuening injection affects really BUN,which will provide clinical guidance for Shuxuening injection safe application.