Vincent Brenda M, Wiitala Wyndy L, Burns Jennifer A, Iwashyna Theodore J, Prescott Hallie C
VA Center for Clinical Management Research, Ann Arbor, MI, USA.
Department of Internal Medicine, University of Michigan, Ann Arbor, MI, USA.
Health Serv Outcomes Res Methodol. 2018 Sep;18(3):143-154. doi: 10.1007/s10742-018-0178-3. Epub 2018 Feb 19.
Hospital readmission is a key metric of hospital quality, such as for comparing Veterans Affairs (VA) hospitals to private sector hospitals. To calculate readmission rates, one must first identify individual hospitalizations. However, in the VA Corporate Data Warehouse (CDW), data are organized by "bedded stays," that is, any stay in a healthcare facility where a patient is provided a bed, not hospitalizations. Thus, CDW data poses several challenges to identifying hospitalizations including: (1) bedded stays include both non-acute inpatient stays (i.e. nursing home, mental health) and acute inpatient hospital care; (2) transfers between VA facilities appear as separate bedded stays; and (3) VA care may also be fragmented by non-VA care. Thus, we sought to develop a rigorous method to identify acute hospitalizations using the VA CDW. We examined all VA bedded stays with an admission date in 2009. Non-acute portions of a stay were dropped. VA to VA transfers were merged when consecutive discharge and admission dates were within one calendar day. Finally, hospitalizations that occurred in a non-VA facility were merged. The 30-day readmission rate was calculated at each step of the algorithm to demonstrate the impact. The total number of VA medical hospitalizations in 2009 with live discharges was 505,861. The 30-day readmission rate after adjusting for VA to VA transfers and incorporating non-VA care was 18.3% (95% confidence interval (CI): 18.2, 18.4%).
医院再入院率是衡量医院质量的关键指标，比如用于比较退伍军人事务部（VA）医院与私立医院。为了计算再入院率，首先必须确定个体住院情况。然而，在VA企业数据仓库（CDW）中，数据是按照“有床位的住院时间”来组织的，也就是说，是指患者在医疗机构中有床位的任何住院时间，而非住院情况。因此，CDW数据在确定住院情况方面带来了诸多挑战，包括：（1）有床位的住院时间既包括非急性住院患者的住院时间（即疗养院、心理健康方面），也包括急性住院患者的医院护理；（2）VA设施之间的转院表现为单独的有床位的住院时间；（3）VA护理也可能因非VA护理而变得分散。因此，我们试图开发一种严谨的方法，利用VA CDW来确定急性住院情况。我们检查了2009年所有有入院日期的VA有床位的住院时间。住院时间中的非急性部分被剔除。当连续出院和入院日期在一个日历日内时，VA之间的转院被合并。最后，在非VA设施中发生的住院情况被合并。在算法的每个步骤都计算了30天再入院率，以证明其影响。2009年有存活出院的VA医疗住院情况总数为505,861例。在对VA之间的转院进行调整并纳入非VA护理后，30天再入院率为18.3%（95%置信区间（CI）：18.2，18.4%）。
