空間階層模型在偵測台灣疾病群聚的應用

Application of Hierarchical Models for Detection of Spatially Clustered Diseases in Taiwan

張美齡(Mei-Ling Chang) ； 林培生(Pei-Sheng Lin) ； 鄒小蕙(Hsiao-Hui Tsou)

《中國統計學報》 46卷1期 (2008/03) Pp. 22-35

https://doi.org/10.29973/JCSA.200803.0003

摘要

在本文中，我們使用階層的卜瓦松廣義線性模型作爲空間群聚偵則的模型，同時利用米徹普勒氏－哈士提(Metropolis-Hasting)演算法對模型中的隨機因素來抽取樣本。在給定不同群聚選取的先驗分配下，我們使用後驗分配和貝氏因子去偵測群聚的可能位置。我們同時將此方法應用在台灣2005年12-64歲心臟病、痛風和精神病的資料上，以檢測出是否存在空閒群聚的現象。最後，我們也對貝氏因子和後驗分配在群聚檢測上的效能做簡單比較。

關鍵字

空間群聚偵測 ； 階層模型 ； 貝氏因子

並列摘要

In this paper, a hierarchical generalized linear model with a Markov Chain Monte Carlo method is used to detect clustering effects for the disease rates. We first follow the criteria by Waller et al. (1997) and Gangnon and Clayton (2003) to assign prior distributions for the random effects in the model. With the Metropolis-Hasting algorithm, values of the random effects are simulated according to different selection methods of potential clusters. We then compute the Bayes factor and posterior distribution to decide the possible clusters for the diseases. This method is applied to analyze the data from National Health Research Institute. A simple comparison between the Bayes factor and posterior distribution for clustering detection is also made.

並列關鍵字

Detection of spatial cluster ； hierarchial model ； Bayes factor

參考文獻

Besag, J.,Green, P.,Higdond, D.,Mengersenk, K.(1995).Bayesian computation and stochastic systems.Statistical Science.10,3-41.

Best, N.G.,Arnold, R.A.,Thomas, A.,Waller, L.A.,Conlon, E.M.(1999).Bayesian models for spatially correlated disease and exposure data.Bayesian Statistics.6,131-156.

Denison, D.,Holmes, C.(2001).Bayesian partitioning for estimating disease risk.Biometrics.57,143-149.

Ferreira, J.,Denison, D.,Holmes, C.,Lawson, A.,Dension, D. (eds)(2002).Spatial Cluster Modelling.London:Chapman & Hall.

Gangnon, R.E.,Clayton, M.K.(2000).Bayesian detection and modeling of spatial disease clustering.Statistics in Medicine.22,3213-3228.