Chuang Ying-Chih, Chuang Ting-Wu, Chao Hsing Jasming, Tseng Kuo-Chien, Chuang Kun-Yang
School of Public Health, Taipei Medical University, Taipei City.
Geospat Health. 2018 Nov 9;13(2). doi: 10.4081/gh.2018.700.
Geographical variations in teen birth rates (TBR) still persist despite controlling for contextual factors. The aim of this research was to identify spatial patterns of TBR in Taiwan and to examine spatial relationships among different contextual factors. Using townships as the unit of analysis (N=359), this research used social and demographic variables for the years 1995, 2000, 2005 and 2010 and conducted spatial modelling of TBR. Geographical maps of TBR were presented, and Local Indicator of Spatial Autocorrelations was used to identify TBR clustering. Comparisons were made between ordinary least-squares models and spatial lag models, in which township-level TBRs were regressed on other township-level contextual characteristics. Our study found that townships with a high TBR were mostly in eastern, central and some southern regions of Taiwan, while townships with a low TBR were in the vicinity of metropolitan areas. The significant spatial lag indicated that townships would have a higher expected prevalence rate if adjacent townships have had higher rates. Results also indicated that the percentage of aborigines and the percentage of college-educated people were consistently associated with TBR over the years. Interventions aimed at reducing TBR in Taiwan should consider the presence of spatial correlations and should incorporate neighbouring townships.
尽管对背景因素进行了控制，但青少年生育率（TBR）的地理差异仍然存在。本研究的目的是确定台湾青少年生育率的空间模式，并研究不同背景因素之间的空间关系。本研究以乡镇为分析单位（N = 359），使用了1995年、2000年、2005年和2010年的社会和人口变量，并对青少年生育率进行了空间建模。展示了青少年生育率的地理地图，并使用空间自相关局部指标来识别青少年生育率的聚类情况。对普通最小二乘法模型和空间滞后模型进行了比较，其中乡镇级青少年生育率与其他乡镇级背景特征进行了回归分析。我们的研究发现，青少年生育率高的乡镇大多位于台湾东部、中部和一些南部地区，而青少年生育率低的乡镇则位于大都市区附近。显著的空间滞后表明，如果相邻乡镇的青少年生育率较高，那么这些乡镇的预期患病率也会更高。结果还表明，多年来原住民的比例和受过大学教育的人的比例一直与青少年生育率相关。台湾旨在降低青少年生育率的干预措施应考虑空间相关性的存在，并应纳入相邻乡镇。
