Centre for Big Data Research in Health, University of New South Wales, Sydney, New South Wales, Australia.
MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, Glasgow, UK.
BMJ Open. 2019 Feb 22;9(2):e027639. doi: 10.1136/bmjopen-2018-027639.
Preventable hospitalisations are used internationally as a performance indicator for primary care, but the influence of other health system factors remains poorly understood. This study investigated between-hospital variation in rates of preventable hospitalisation.
Linked health survey and hospital admissions data for a cohort study of 266 826 people aged over 45 years in the state of New South Wales, Australia.
Between-hospital variation in preventable hospitalisation was quantified using cross-classified multiple-membership multilevel Poisson models, adjusted for personal sociodemographic, health and area-level contextual characteristics. Variation was also explored for two conditions unlikely to be influenced by discretionary admission practice: emergency admissions for acute myocardial infarction (AMI) and hip fracture.
We found significant between-hospital variation in adjusted rates of preventable hospitalisation, with hospitals varying on average 26% from the state mean. Patients served more by community and multipurpose facilities (smaller facilities primarily in rural areas) had higher rates of preventable hospitalisation. Community hospitals had the greatest between-hospital variation, and included the facilities with the highest rates of preventable hospitalisation. There was comparatively little between-hospital variation in rates of admission for AMI and hip fracture.
Geographic variation in preventable hospitalisation is determined in part by hospitals, reflecting different roles played by community and multipurpose facilities, compared with major and principal referral hospitals, within the community. Care should be taken when interpreting the indicator simply as a performance measure for primary care.
可预防的住院治疗被国际上用作初级保健的绩效指标，但其他卫生系统因素的影响仍知之甚少。本研究调查了可预防住院率的医院间差异。
这是一项针对澳大利亚新南威尔士州 266826 名 45 岁以上人群的队列研究，将健康调查和住院数据进行了关联。
使用交叉分类多成员多水平泊松模型，调整了个人社会人口统计学、健康和地区层面的背景特征，对可预防住院的医院间差异进行了量化。还对两种不太可能受到自由裁量性入院实践影响的情况（急性心肌梗死和髋部骨折的紧急入院）进行了差异探索。
我们发现，可预防住院的调整后率存在显著的医院间差异，医院的平均差异为 26%，与州平均水平相比。由社区和多用途设施服务的患者（主要在农村地区的小型设施）的可预防住院率较高。社区医院的医院间差异最大，其中包括可预防住院率最高的设施。急性心肌梗死和髋部骨折的入院率的医院间差异相对较小。
可预防住院的地理差异部分由医院决定，这反映了社区内社区和多用途设施与主要转诊医院相比所扮演的不同角色。在将该指标简单地解释为初级保健的绩效衡量标准时，应谨慎行事。
