有關近視眼研究的統計方法

Abstract

在科技進步的現代社會，近視已是日趨嚴重的公衛問題，不僅盛行率增加，發生年齡亦下降，其中又以華人社會最為嚴重。同時眼科醫學亦著力於近視的預防與治療。在近視眼的研究中，統計分析是不可或缺的角色。本論文整理了有助於近視研究的統計方法，內容包含：最常見的橫斷面研究，可找出近視的風險因子；利用分析長期追蹤資料，探討近視度數演進的變化。此外眼科測度為“成對”但彼此相關的資料，存活分析可以用來探討眼睛發生重要“事件” 所需的時間長度。我們亦回顧了台灣、新加坡及其他地區所發表有關於近視眼研究的醫學文獻，並整理這些論文使用的統計方法。最後我們以統計的角度，建議了未來研究可採用的方法。

Nowadays myopia has become a global critical health problem, especially in Chinese community. Statistical analysis plays an important role in myopia research. In the thesis, we investigate related statistical methods, including cross-sectional studies for identifying possible risk factors for myopia and longitudinal studies for studying the evolution of myopia. Eye measurements are “paired” observations so that we review methods for analyzing such data. Furthermore survival analysis can also be applied to study critical event times such as the onset of myopia. We also review medical literature on myopia, most of which are recent empirical studies in Taiwan and Singapore. At the end of the thesis, we provide some suggestions for future research from the viewpoint of statistical analysis.