Wong Timothy, Brovman Ethan Y, Rao Nikhilesh, Tsai Mitchell H, Urman Richard D
Department of Anesthesiology, University of Vermont College of Medicine, Burlington, VT, USA.
Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women's Hospital, Boston, MA, USA.
J Clin Med Res. 2020 Jan;12(1):18-25. doi: 10.14740/jocmr4029. Epub 2020 Jan 6.
Over the past several decades, diabetes mellitus has contributed to a significant disease burden in the general population. Evidence suggests that patients with a coexisting diabetes diagnosis consume more hospital resources, and have higher readmission rates compared to those who do not. Against the backdrop of bundled-payment programs, healthcare systems cannot underestimate the importance of monitoring patient health information at the population level.
Using the data from the Centers for Medicare and Medicaid Services (CMS) administrative claims database, we created a dashboard prototype to enable hospitals to examine the impact of diabetes on their all-cause readmission rates and financial implications if diabetes was present at the index hospitalization. The technical design involved loading the relevant 10th revision of International Classification of Diseases (ICD-10) codes provided by the medical team and flagging diabetes patients at the claim. These patients were then tagged for readmissions within the same database. The odds ratios were determined based on data from two groups: those with diabetes at index hospitalization which include type 1 only, type 2 only, and type 1 and type 2 diabetes, plus those without diabetes at index hospitalization.
The dashboard presents summary data of diabetes readmissions quality metrics at a national level. Users can visualize summary data of each state and compare odds ratios for readmissions as well as raw hospitalization data at their facility. Dashboard users can also view data classified by a diagnosis-related group (DRG) system. In addition to a "national" data view, for users who inquire about data specific to demographic regions, the DRG view can be further stratified at the state level or county level. At the DRG level, users can view data about the cost per readmissions for all index hospitalization with and without diabetes.
The dashboard prototype offers users a virtual interface which displays visual data for quick interpretation, monitors changes at a population level, and enables administrators to benchmark facility data against local and national trends. This is an important step in using data analytics to drive population level decision making to ultimately improve medical systems.
在过去几十年中，糖尿病给普通人群带来了沉重的疾病负担。有证据表明，与未患糖尿病的患者相比，同时患有糖尿病的患者消耗更多的医院资源，再入院率也更高。在捆绑支付计划的背景下，医疗系统不能低估在人群层面监测患者健康信息的重要性。
利用医疗保险和医疗补助服务中心（CMS）行政索赔数据库的数据，我们创建了一个仪表板原型，以使医院能够检查糖尿病对其全因再入院率的影响，以及如果在首次住院时患有糖尿病所产生的财务影响。技术设计包括加载医疗团队提供的相关国际疾病分类第十版（ICD-10）代码，并在索赔中标记糖尿病患者。然后在同一数据库中对这些患者的再入院情况进行标记。根据两组数据确定比值比：首次住院时患有糖尿病的患者，包括仅1型糖尿病、仅2型糖尿病以及1型和2型糖尿病患者，以及首次住院时未患糖尿病的患者。
该仪表板展示了全国范围内糖尿病再入院质量指标的汇总数据。用户可以直观看到每个州的汇总数据，并比较再入院的比值比以及其所在医疗机构的原始住院数据。仪表板用户还可以查看按诊断相关分组（DRG）系统分类的数据。除了“全国”数据视图外，对于询问特定人口区域数据的用户，DRG视图可以在州一级或县一级进一步分层。在DRG层面，用户可以查看所有有或无糖尿病的首次住院的每次再入院成本数据。
仪表板原型为用户提供了一个虚拟界面，该界面显示可视化数据以便快速解读，在人群层面监测变化，并使管理人员能够根据当地和全国趋势对医疗机构数据进行基准对比。这是利用数据分析推动人群层面决策以最终改善医疗系统的重要一步。
