评估和控制分解住院的方法初探

An initial study of identifying and controlling episode separations

目的模拟判别分解住院的过程，为建立评估分解住院的可操作方法提供依据。方法从北京市公共卫生信息中心采集2010年北京市全部出院病例的病案首页信息。以“北京版诊断相关组（BJ—DRGs）”为风险调整工具，分别在地区层面和医院层面计量各DRGs的两周再入院率。比对并筛选出分解住院的“可疑对象”。继而针对“可疑对象”进行病例数据细致检验，判断是否存在分解住院的情况。结果2010年北京市有13类疾病的两周再人院率超过10％。以11家三甲医院“痴呆及其他慢性大脑功能失调”两周再人院率比较为例。比较后发现，G医院为可疑对象。进行病例细致分析后，发现了G医院在该类病例分解住院的情况。结论借助DRGs作为风险调整工具，充分利用客观数据结合“标杆法“筛查可疑对象，有助于降低分解住院的监管成本。

Objective To imitate the process of identifying episode separations, which helps to improve the methodology of control episode separations. Methods Abstracts of inpatient record on all patients discharged from the hospitals in Beijing in 2010 were drawn from Beijing Public Health Information Center. Risk-adjusted by Diagnosis Related Groups-Beijing version（BJ-DRGs）, two-weeks readmission rate was calculated at the whole city and hospital levels respectively. Comparing the readmission rate between these two levels, ＂suspicious objects＂ of episode separations can be found. In turn,the episode separations can be identified by detail analysis of these suspicious objects. Results In 2010,there were 13 DRGs with readmission rate exceeding 10%. Take the cases of dementia and other chronic brain disorders for example, the readmission rate for this type of inpatients in hospital G was much higher than the average level of the whole city. After the detail analysis of hospital G, the episode separations were found. Conclusion It is an effective method to indentify episode separations that calculating readmission rate based on the DRG risk-adjustment and using city level readmission rate as bench-mark.