Curtis Jeffrey R, Cheng Hong, Delzell Elizabeth, Fram David, Kilgore Meredith, Saag Kenneth, Yun Huifeng, Dumouchel William
Center for Education and Research on Therapeutics of Musculoskeletal Disorders, University of Alabama at Birmingham, Alabama 35294, USA.
Med Care. 2008 Sep;46(9):969-75. doi: 10.1097/MLR.0b013e318179253b.
Bayesian data mining methods have been used to evaluate drug safety signals from adverse event reporting systems and allow for evaluation of multiple endpoints that are not prespecified. Their adaptation for use with longitudinal data such as administrative claims has not been previously evaluated or validated.
In this pilot study, we evaluated the feasibility of adapting data mining methods using the empirical Bayes Multi-item Gamma Poisson Shrinkage (MGPS) algorithm to longitudinal administrative claims data. The Medicare Current Beneficiary Survey was used to identify a cohort of Medicare enrollees who were exposed to cyclooxygenase selective (coxib) or nonselective nonsteroidal anti-inflammatory drugs (NS-NSAIDs) from 1999 to 2003. Empirical Bayes MGPS algorithm was used to simultaneously evaluate 259 outcomes associated with current use of coxibs versus NS-NSAIDs while adjusting for key covariates and multiple comparisons. For comparison, a parallel analysis used traditional epidemiologic methods to evaluate the relationship between coxib versus NS-NSAID use and acute myocardial infarction, with the goal of establishing the concurrent validity of the data mining approach.
Among 9431 Medicare beneficiaries using NSAIDs and considering all 259 possible outcomes, empirical Bayes MGPS identified an association between current celecoxib use and acute myocardial infarction (Empirical Bayes Geometric Mean ratio 1.91) but not other outcomes. Rofecoxib use was associated with acute cerebrovascular events (Empirical Bayes Geometric Mean ratio 1.85) and several other diagnoses that likely represented indications for the drug. Results from the analyses using traditional epidemiologic methods were similar and indicated that the data mining results were valid.
Bayesian data mining methods seem useful to evaluate drug safety using administrative data. Further work will be needed to extend these findings to different types of drug exposures and to other claims databases.
贝叶斯数据挖掘方法已被用于评估不良事件报告系统中的药物安全信号，并允许对未预先指定的多个终点进行评估。此前尚未对其用于行政索赔等纵向数据的适用性进行评估或验证。
在这项试点研究中，我们评估了使用经验贝叶斯多项目伽马泊松收缩（MGPS）算法将数据挖掘方法应用于纵向行政索赔数据的可行性。利用医疗保险当前受益人调查来确定1999年至2003年期间接触环氧化酶选择性（coxib）或非选择性非甾体抗炎药（NS-NSAIDs）的医疗保险参保人群。经验贝叶斯MGPS算法用于同时评估与当前使用coxib与NS-NSAIDs相关的259种结局，同时对关键协变量和多重比较进行调整。作为比较，一项平行分析使用传统流行病学方法评估coxib与NS-NSAIDs使用与急性心肌梗死之间的关系，目的是确定数据挖掘方法的同时效度。
在9431名使用非甾体抗炎药的医疗保险受益人中，考虑所有259种可能的结局，经验贝叶斯MGPS确定当前使用塞来昔布与急性心肌梗死之间存在关联（经验贝叶斯几何平均比1.91），但与其他结局无关。使用罗非昔布与急性脑血管事件（经验贝叶斯几何平均比1.85）以及其他几种可能代表该药物适应证的诊断相关。使用传统流行病学方法的分析结果相似，表明数据挖掘结果是有效的。
贝叶斯数据挖掘方法似乎有助于使用行政数据评估药物安全性。需要进一步开展工作，将这些发现扩展到不同类型的药物暴露以及其他索赔数据库。
