台灣不同危險因子族群愛滋感染的地理空間變異

Abstract

背景： 如何有效阻止愛滋傳染的擴散是當前愛滋預防的首要問題，台灣三個主要的愛滋感染高危險族群分別是男性間不安全性行為者、海洛英靜脈注射者以及異性間不安全性行為者。當我們面對愛滋疫情和欲建立相關政策的時候，首先必須了解這些高危險族群的傳播網絡與分布區域是否重疊? 因為愛滋相關政策必須依照不同目標族群的傳染模式、需求和分布地點來設計，才能達到有效控制和預防的目的。雖然在愛滋病毒基因型的分析上、以及從生活型態與危險行為模式的角度去探討，不同族群基本上具有相異的社交網路與生活範疇；然而限於資料的取得困難和現實的障礙，我們無法明確地辦定是否為完全相異的族群；同時，台灣目前仍然沒有研究是針對愛滋感染的空間分布來進行探討與分析，無法掌握愛滋感染高危險族群的地理分布情況。如果不同族群具有不同的地理分布，在資源的分配上就必須依照目標族群所分布的地區來規劃適合的政策。 目標： 分別從空間分布和感染愛滋相關之社會經濟因子的角度，來分析是否不同危險因子的愛滋感染族群分屬於不同的愛滋疫情(HIV epidemic)。 方法： 我們同時擷取疾病管制局法定傳染病資料庫與台大醫院的愛滋感染者資料，分別以兩個具有不同地理範圍及型態的資料來分析。疾管局的資料以鄉鎮市為空間單位，分析性傳染途徑愛滋感染族群與毒癮愛滋感染族群在台灣的地理分布;而臺大醫院的資料是依據愛滋感染者的居住地點(點資料)，探討男性間性行為者、海洛英靜脈注射者和異性間性行為者三個族群在台灣北部四個縣市(台北縣市、桃園縣、基隆市)的地理空間分布。檢驗不同危險因子的愛滋感染族群之間是否具有空間上的依存性或相關性，並藉由空間聚集的偵測來比較各族群在地理分佈上之異同。 另外，我們從整體的角度分析社會相關因子(各鄉鎮市的都市化程度、性活躍人數、經濟和教育程度等)與愛滋感染分佈之間的關係，是否不同危險因子愛滋感染族群具有相異的社會經濟決定因子。同時也試圖探討愛滋感染者居住地點與男同性間性行為者容易發生危險性行為之場域兩者之間的空間分布，從到特殊公共場域的距離來推論是否在空間上為相異的愛滋感染族群。 結果： 無論是空間上的相依性或是聚集發生的地點，不同危險因子的愛滋感染族群具有相異的地理空間分布，而且相同的社會相關因子對不同的愛滋感染族群具有不同的相關性。男性間性行為者與異性間性行為者主要分布在都市化程度高、社經地位較高的區域；而毒癮愛滋者則主要分布於都市化程度較低的次都會區。在市中心屬於交通發達的地方，相較於靜脈注射藥癮愛滋或異性戀愛滋感染者，男同性間性行為之愛滋感染者的分佈範圍雖然較靠近特殊公共場域之分布，但兩者之間的分佈並無明顯相關，顯示男同性間性行為族群並不會為了遷就到場域的可近性而選擇居住的地點，主要還是分布於住宅區。 結論： 我們對已知的愛滋感染者進行探討，發現不同危險因子的愛滋感染族群具有相異的地理空間分布與社會決定因子，在地理空間上分屬於不同的愛滋疫情，因此推論不同的愛滋感染高危險族群具有相異的地理分布，所以在政策的實施上就必須依照目標族群的分布地點來進行規劃，同時必須再進一步研究背後的成因。研究結果也顯示，從巨觀社會環境因子切入，可以找出不同愛滋感染族群好發的鄉鎮市分布，對於愛滋的傳染，除了個人的因素之外，社會環境也扮演了一個重要的因子，因此對於整個政策的規劃必須同時從個人與整體環境下手。對於防疫而言，針對各高危險族群易發生愛滋病毒傳染的地點(特殊場域)是最有效的介入方式，然而這個部分的資訊比較不完整，仍需要更多的探討研究，讓整體愛滋預防政策更趨於實際與完善。

Background: The three main high-risk groups for HIV infection in Taiwan are men who have sex with men (MSM), heroin injection drug users (IDU), and persons who have unprotected heterosexual sex. Formulating the appropriate control strategy for HIV epidemics among these high-risk groups requires the knowledge whether the transmission networks and locations overlap or not in the first place. To maximize the effectiveness, HIV preventive programs need to be delivered according to the locations of specific populations. Genotyping analysis and qualitative researches suggested that different high-risk groups have distinct social transmission networks and social contexts generally. However, the extent of overlapping has not been known. Moreover, there is no study which explores and analyses the spatial distribution of HIV infection. If high-risk groups have different geographic distributions, we must allocate resources according to the locations of target groups and establish appropriate policy.

Objectives: The main purpose of this research is to examine whether HIV-infected groups with different risk factors could be distinguished as different HIV epidemics from the aspects of geo-spatial as well as socioeconomic factors. Methods: We examine the geo-spatial distributions of different HIV-infected group by constructing two scales of data to analysis: explore the spatial distributions of sexually transmitted HIV-infected group and IDU-related HIV-infected group in Taiwan by township level, and the spatial relationship within the MSM group, IDU group and heterosexuality group (HIV-infected patients of National Taiwan university Hospital) in Northern Taiwan by point data. We would like to inspect whether there was spatial association or dependency between different HIV-infected groups and compare their geographic distribution by spatial pattern analysis and spatial cluster detection method. Second, we would like to perform a population-based study to examine the relationship between distribution of HIV infection and socioeconomic factors (degree of urbanization, sexually active population, level of economy and education at township level and see if HIV-infected groups with different risk factors have the different socioeconomic determinants. Moreover, we try to explore the spatial relationship between HIV-infected patients’ residential locations and gay places where MSM easily take high-risk sexual behavior. Through this relationship we could investigate whether there are different HIV-infected groups from the spatial aspect by comparing the average distances to gay places. Results: (1) Spatial distributions of HIV-infected groups with different risk factors were not identical and the sexually transmitted HIV epidemic (MSM and Heterosexuality) was more clustered than IDU-related HIV epidemic. The former was more likely to reside in urban places, but the latter was more concentrated in the suburban area. Although the spatial distributions were not the same, there were locations where clusters overlapped. (2) HIV epidemics with different risk factors had different socioeconomic determinants. Sexually transmitted HIV cases were more likely to occur in townships with high average household income, high level of education, and high degree of urbanization; while IDU-related cases were more likely to occur in townships with lower income, middle-level of education and middle degree of urbanization. (3) There was significant shorter distance to gay places for MSM cases in comparison with IDU or heterosexual cases. However, spatial dependency was not found between MSM and gay places, which may be explained by the observation that MSM resided in residential district but gay places were located in business district generally. Conclusions: HIV epidemics among groups with different risk factors were not similar in geo-spatial distribution and socioeconomic determinants. We identified the locations of clusters for HIV-infected groups and the possible place where high-risk groups would concentrate. It is necessary to develop suitable program and strategies for specific high-risk groups at the appropriate locations. Key words: HIV infection, spatial analysis