Statistical Models of Bio-surveillance for Monitoring Disease Outbreak

2001年的炭疽恐怖袭击、严重急性呼吸综合征(SARS)的爆发以及对大流行性流感的恐惧促使许多公共卫生部门在美国开发疾病爆发的早期检测系统。及早发现疾病暴发使公共卫生官员能够尽早实施疾病控制和预防措施。特别是考虑到生物恐怖主义的威胁，最近人们对开发能够检测疾病空间模式变化的生物监测系统感兴趣。在这项研究中，我们开发了四种可以在监控系统中实现的统计模型/方法。我们首先提出了一种形状灵活的空间扫描统计量，可以在每个区域的相对较小的邻域内检测到不规则形状的簇。将该空间扫描统计量与基于Mo…的Kulldorff圆形空间扫描统计量的性能进行了比较通过考虑几种圆形和非圆形热点集群模型进行了更多的nte-Carlo模拟。实验结果表明，该空间扫描统计量具有良好的通常功率，并且能够比圆形热点簇更准确地检测出非圆形热点簇。其次，我们提出了簇检测测试的一种扩展功率，它包含了通常的功率作为特例。此外，我们还定义了扩展能力的分布，这对于评估和比较几种簇检测测试具有重要的意义。两个测试--Kulldorff的圆形空间扫描统计量(1995,1997)和Tango和Takahashi(2005)提出的灵活的空间扫描统计量--验证了所提出的扩展功率及其分布。第三，在弹性空间扫描统计量的基础上，提出了一种形状灵活的空时扫描统计量。特别是，我们开发了一种前瞻性的时空扫描统计方法，用于早期检测疾病爆发。利用Kulldorff等人提供的基准数据，将所提出的空时扫描统计量与柱面扫描统计量的性能进行了比较。为了比较空时扫描统计量的性能，对纯空间二元功率分布进行了扩展，提出了一种空时功率分布。使用美国马萨诸塞州的每日症状监测数据来说明所提出的检验统计量，并比较了弹性时空扫描统计量和柱面时空扫描统计量的性能。最后，提出了一类基于病例控制点数据的二次型卫生事件空间聚集性检测方法。它包括库齐克和爱德华兹的检验统计量(1990)。尽管他们使用了检验统计量的渐近正态性质，但我们表明，对于中等大的样本量，这样的近似通常是很差的。相反，我们建议将中心卡方分布作为测试统计量的渐近分布的更好近似。此外，为了在聚类尺度上估计未知参数的最优值，并通过改变参数值来适应多次测试，我们提出了参数的检验统计量的轮廓p值的最小值作为综合检验统计量。较少

The anthrax terrorist attacks in 2001, the severe acute respiratory syndrome (SARS) outbreak, and a fear of pandemic flu have motivated many public health departments to develop early disease outbreak detection systems in the United States. Early detection of disease outbreaks enables public health officials to implement disease control and prevention measures at the earliest possible time. Particularly in light of the perceived threat of bioterrorism, there has been a spate of recent interest in the development of biosurveillance systems that can detect changes in spatial patterns of disease. In this research, we developed four statistical models/methods that can be implemented in the surveillance system. We first proposed a flexibly shaped spatial scan statistic that can detect irregular shaped clusters within relatively small neighborhoods of each region. The performance of the proposed spatial scan statistic is compared to that of Kulldorff's circular spatial scan statistic with Mo … More nte Carlo simulation by considering several circular and noncircular hot-spot cluster models. The proposed spatial scan statistic is shown to have good usual powers plus the ability to detect the noncircular hot-spot clusters more accurately than the circular one. We second we proposed an extended power of the cluster detection tests, which includes the usual power as a special case. Further, we define the profile of the extended power, which can be expected to play an important role in the evaluation and comparison of several cluster detection tests. The proposed extended power and its profile are demonstrated by two tests -- Kulldorff's circular spatial scan statistic (1995,1997) and a flexible spatial scan statistic proposed by Tango and Takahashi(2005). Third, we propose a flexibly shaped space-time scan statistic based on the flexible spatial scan statistic. In particular, we develop a prospective space-time scan statistic for early detection of disease outbreaks. The performance of the proposed space-time scan statistic is compared with that of the cylindrical scan statistic, using the benchmark data provided by Kulldorff et al. In order to compare the performance of space-time scan statistics we propose a space-time power distribution by extending the purely spatial bivariate power distribution. Daily syndromic surveillance data in Massachusetts, USA, are used to illustrate the proposed test statistic and also to compare the performance of the flexible space-time scan statistic with that of the cylindrical one. Finally, a class of tests with quadratic forms for detecting spatial clustering of health events based on case-control point data was proposed. It includes Cuzick and Edwards' test statistic(1990). Although they used the property of asymptotic normality of the test statistic, we showed that such an approximation is generally poor for moderately large sample sizes. Instead, we suggested a central chi-square distribution as a better approximation to the asymptotic distribution of the test statistic. Furthermore, not only to estimate the optimal value of the unknown parameter on the scale of cluster but also to adjust for multiple testing due to repeating the procedure by changing the parameter value, we proposed the minimum of the profile p-value of the test statistic for the parameter as an integrated test statistic. Less