Lau L M, Wright S O, Garlick-Longhurst T J, Graybill C S, Warner H R
University of Utah Health Sciences Center, Salt Lake City 84132, USA.
Clin Perform Qual Health Care. 1996 Jan-Mar;4(1):10-3.
Hospital characteristics vary greatly across a geographic area such as a state. Hospital peer groups internally exhibit similar characteristics and can be used as a basis for the analysis of data, the dissemination of information, and the adoption of continuous quality improvement project results. This paper reflects the efforts made toward the identification of hospital peer groups within the state of Michigan. Hospital characteristics data for fiscal year 1992 were obtained from the American Hospital Association's Annual Survey of Hospitals and the Health Care Financing Administration's MEDPRO database. Thirteen peer group clusters have been identified, reviewed, and commented on by the state's hospital association and have met general approval by hospital administrators across the state. The established peer groups are being used to identify the differences in patterns of care among hospitals in the state. The peer groups also are being used for the feedback of comparable data and the identification of hospitals for participation in continuous quality improvement projects. The next research objective is to experiment with other clustering techniques and other inpatient populations. The consistency of the peer groupings across all clustering techniques and across both Medicare and total inpatient populations will be studied.
在一个州这样的地理区域内，医院的特征差异很大。医院同行群体内部具有相似的特征，可作为数据分析、信息传播以及采用持续质量改进项目成果的基础。本文反映了在密歇根州内识别医院同行群体所做的努力。1992财年的医院特征数据来自美国医院协会的年度医院调查和医疗保健财务管理局的MEDPRO数据库。该州的医院协会已经识别、审查并评论了13个同行群体集群，并且得到了全州医院管理人员的普遍认可。已建立的同行群体正用于识别该州各医院护理模式的差异。这些同行群体还被用于可比数据的反馈以及确定参与持续质量改进项目的医院。下一个研究目标是试验其他聚类技术和其他住院患者群体。将研究所有聚类技术以及医疗保险和全部住院患者群体中同行分组的一致性。
