曾春典2006-07-252018-06-292006-07-252018-06-292003-07-31http://ntur.lib.ntu.edu.tw//handle/246246/4713膽固醇被認為是引起心臟血管疾病， 尤其是缺血性心臟病重要危險因子。不過 利用單一脂質指標是否適用於心血管疾病 篩檢，而且定出最佳的脂質篩檢指標則仍 未有定論。由於診斷指標不但視個人危險 因子而定，且受到資料是來自臨床病例或 社區無症狀早期個案所影響，因此是否這 些指標適合社區大規模脂質篩檢，進而減 低心臟血管疾病仍值得探討。從預防醫學 角度而言，利用血脂質做為心臟血管疾病 篩檢指標， 通常有兩種策略， 高危險 (High-risk group) 策略及大規模族群 (Population wide)策略，然而台灣地區缺乏 社區脂質篩檢經驗，因此到底是使用高危 險策略或大規模族群策略是值得研究之問 題。此外，欲進行成本效益分析，也需要 了解脂質篩檢策略、篩檢指標、以及治療 之效能。 目的：本研究目的是利用一個社區性複合 式篩檢計畫進行脂質篩檢降低心臟血管危 險性流行病學評估及成本效益分析，其特 殊研究目的包括： (1) 估計以社區為主脂質篩檢所得到不同 脂質指標異常之盛行率及與心臟血管 疾病（如冠心症或缺血性心臟病）之關 係【（如總膽固醇，低密度脂蛋白等）】。 (2)在調整了危險因子（如性別，抽煙，家 族病史）進行不同脂質成份與心臟血管 疾病之關係。 (3)利用上述（2）及不同篩檢指標如膽固 醇，低密度脂蛋白（LDL）及TC/HDL 比值配合調整相關危險因子評估不同組 合下其單一指標敏感度及精確度，並進 行ROC 曲線分析尋找最佳組合。 (4)配合貝氏分析發展多重指標下（如總膽 固醇加上低密度脂蛋白（LDL），TC/HDL 及其他危險因子），脂質篩檢其一系列 逐次加入不同指標及危險因子之敏感度 及精確度之組合，並與目前健保給付所 依據脂質分類所得到敏感度及精確度相 比較。Although cholesterol has been regarded as the most influence risk factor for cardiovascular disease, in particular ischemic heart disease, whether a single lipid index is appropriate for screening for cardiovascular disease and whether the optimal cutoff point can be obtain are uncertain. This may depend on the presence of risk factors and on whether cases come from clinical series or community subjects with asymptomatic syndrome. In addition, inconsistent results were also found among large clinical trials that found the effectiveness of anti-cholesterol drugs also depends on the presence risk factors in individuals. Preventive strategies for cardiovascular disease have two approaches, high-risk group approach, and population-wide strategy. However, which strategy is appropriate is in dilemma. The above findings imply that cost-effectiveness for lipid screening should encompass lipid screening index, the efficacy 3 of treatment, and screening strategies. Objectives: The purposes are to (1) estimate abnormality of lipid and their association with cardiovascular disease; (2) elucidate the association between lipids and cardiovascular disease after adjustment for relevant factors; (3) to identify the optimal cut-off point given a combination of LDL, TC/HDL using Receiver Operating Characteristics curve; (4) to develop a multiple index using Bayesian sequential method to obtain the optimal sensitivity and specificity and to compare the result with those paid by health insurance. Results: Taking 240 mg/dl as the cut-off point of cholesterol the sensitivity and specificity were 11% and 94%. The corresponding figures were 55% and 42% for HDL≦35 mg/dl and 4% and 98% for LDL≧ 160mg/dl. Using Bayesian Analysis, given 240mg/dl of cholesterol, the odd ratios were increased from 0.09 to 11.42 with the corresponding range of HDL from 80mg/dlto 10mg/dl. Conclusion: The optional cut-off point of the combination of lipid profiles were estimated using Bayesian Approach.application/pdf34138 bytesapplication/pdfzh-TW國立臺灣大學公共衛生學院預防醫學研究所脂質篩檢社區性研究ROC 曲線成本效益分析lipid screeningcommunity-based studyROC curvecost-effectiveness analysis[SDGs]SDG3reporthttp://ntur.lib.ntu.edu.tw/bitstream/246246/4713/1/912320B002172.pdf