Epstein Richard H, Dexter Franklin
Department of Anesthesiology, Perioperative Medicine and Pain Management, University of Miami, Miami, FL, USA.
Division of Management Consulting, Department of Anesthesia, University of Iowa, Iowa City, IA, USA.
J Am Med Inform Assoc. 2017 Jul 1;24(4):845-850. doi: 10.1093/jamia/ocw181.
Comorbidity adjustment is often performed during outcomes and health care resource utilization research. Our goal was to develop an efficient algorithm in structured query language (SQL) to determine the Elixhauser comorbidity index.
We wrote an SQL algorithm to calculate the Elixhauser comorbidities from Diagnosis Related Group and International Classification of Diseases (ICD) codes. Validation was by comparison to expected comorbidities from combinations of these codes and to the 2013 Nationwide Readmissions Database (NRD).
The SQL algorithm matched perfectly with expected comorbidities for all combinations of ICD-9 or ICD-10, and Diagnosis Related Groups. Of 13 585 859 evaluable NRD records, the algorithm matched 100% of the listed comorbidities. Processing time was ∼0.05 ms/record.
The SQL Elixhauser code was efficient and computationally identical to the SAS algorithm used for the NRD.
This algorithm may be useful where preprocessing of large datasets in a relational database environment and comorbidity determination is desired before statistical analysis. A validated SQL procedure to calculate Elixhauser comorbidities and the van Walraven index from ICD-9 or ICD-10 discharge diagnosis codes has been published.
在结果和医疗资源利用研究中，常进行共病调整。我们的目标是开发一种用结构化查询语言（SQL）编写的高效算法，以确定埃利克斯豪泽共病指数。
我们编写了一个SQL算法，根据诊断相关组和国际疾病分类（ICD）编码来计算埃利克斯豪泽共病情况。通过与这些编码组合的预期共病情况以及2013年全国再入院数据库（NRD）进行比较来验证。
对于ICD - 9或ICD - 10与诊断相关组的所有组合，SQL算法与预期共病情况完全匹配。在13585859条可评估的NRD记录中，该算法与列出的所有共病情况匹配率为100%。处理时间约为0.05毫秒/记录。
SQL埃利克斯豪泽代码效率高，在计算上与NRD使用的SAS算法相同。
在关系数据库环境中对大型数据集进行预处理并在统计分析前确定共病情况时，该算法可能有用。已发表了一个经过验证的SQL程序，用于根据ICD - 9或ICD - 10出院诊断代码计算埃利克斯豪泽共病情况和范瓦尔拉文指数。
