Alexander J A, Anderson J G, Lewis B L
Med Care. 1985 Jul;23(7):913-32. doi: 10.1097/00005650-198507000-00007.
Inconsistency and lack of utility in multihospital systems research is often attributable to misspecification of the unit of analysis and lack of generalizability of research findings. The typical, system-level focus takes an "organizations are all alike" approach that masks potentially important variation among constituent hospitals in systems. This article develops and validates a multivariate, empirical typology of hospitals affiliated with multihospital systems. The objective of the typology is to identify prevalent types of system hospitals that share similar organizational and operational characteristics. Three hundred and sixty-six hospitals affiliated with secular, not-for-profit multihospital systems were subjected to cluster analysis within and across affiliation categories (owned, leased, and contract managed). Findings indicate that the population of hospitals in not-for-profit systems is organizationally heterogeneous but that distinct groupings of hospitals, based largely on size and affiliation status, occur within this population. Homogeneous groupings of hospitals were tested for differences in governance and performance characteristics. Although these clusters differ only slightly in their governance characteristics, they do manifest several differences in performance characteristics. Implications and applications of these findings are discussed.
多医院系统研究中的不一致性和实用性不足,往往归因于分析单位的错误设定以及研究结果缺乏普遍性。典型的系统层面研究采用“所有组织都相似”的方法,掩盖了系统内各成员医院之间潜在的重要差异。本文开发并验证了一种针对多医院系统附属医院的多变量实证类型学。该类型学的目的是识别具有相似组织和运营特征的系统医院的普遍类型。对隶属于世俗非营利性多医院系统的366家医院,在附属类别(自有、租赁和合同管理)内部及之间进行了聚类分析。研究结果表明,非营利性系统中的医院群体在组织上是异质的,但在这一群体中,主要基于规模和附属地位出现了不同的医院分组。对医院的同类分组进行了治理和绩效特征差异测试。虽然这些聚类在治理特征上仅略有不同,但在绩效特征上确实表现出一些差异。讨论了这些发现的影响和应用。