Operations Research and Health Care
Operations Research and Health Care
复制标题
运筹学和医疗保健
DOI：
--
复制
发表时间：
2004
期刊：
影响因子：
--
通讯作者：
W. Pierskalla
中科院分区：
文献类型：
--
作者：
F. Hillier;M. Brandeau;F. Sainfort;W. Pierskalla
传染病、接触石棉等有毒物质、环境损害以及吸烟等生活方式风险都对公共卫生构成风险。必须在医疗干预或立法之前进行风险评估，这需要概率、统计和计算方法。本章的导言讨论了我们对公共卫生风险的看法是如何变化的，并确定了风险分析方法的一些趋势。风险评估的主要特点是基于概率的统计推断，使用疾病强度的点过程模型作为空间和时间位置的函数。条件似然法如考克斯偏似然法和配对逻辑回归法被广泛用于消除混杂变量。给出了使用这种条件似然性的两个例子。在第一，新的测试的传染病的病例特征的时空聚类推导和举例说明。在条件似然的第二个应用中，一些研究的风险志贺氏菌感染的学童在学校或玩伴。这一章的原始内容是两个新的测试时空聚类，并使用一个不寻常的条件似然的案例研究。
Risks to public health arise from infectious disease, exposure to toxic substances such as asbestos, environmental insults, and from lifestyle risks such as smoking. The risk assessment that must precede healthcare interventions or legislation requires probabilistic, statistical and computational methodologies. The introduction to this chapter discusses how our perception of the risks to public health is changing, and identifies some trends in the methodologies used for risk analysis. Risk assessment is largely characterised by likelihood-based statistical inference, using pointprocess models of disease intensity as a function of position in space and time. Conditional likelihoods such as Cox’s partial likelihood and matchedpairs logistic regression are widely used to eliminate confounding variables. Two examples of the use of such conditional likelihoods are given. In the first, new tests for the space-time clustering of cases characteristic of infectious disease are derived and exemplified. In a second application of conditional likelihood, some research on risks of Shigella infection to schoolchildren arising at school or from playmates is presented. The original content of this chapter is two new tests of space-time clustering, and a casestudy using an unusual conditional likelihood.