| Abstract: | 在風險評估研究中, 通常是針對單一化學物質下去進行評估, 但在我們生活環境中潛藏著許多汙染物質, 像是空氣汙染、水污染和食物汙染, 這些汙染源主要來自於各種化學物質的過量使用或是處置不當而導致。人們每天生活在這種環境下自然會同時接觸到許多種不同的毒物或是化學物質, 因此累積風險評估(Cumulative Risk Assessment)
成為近年來廣泛探討的課題。
劑量反應模式(Dose-response Model) 通常使用於量測有毒物質劑量所產生不良反應的機率。Chen et al.(2001) 將廣泛應用於二分反應變項(dichotomous quantalresponse) 的標準邏輯斯模式, 結合劑量加法(dose addition) 的概念, 提出了一個新的劑量模式以估計混合物聯合暴露(combined exposures risk) 時所產生的累積風險,然而Chen et al. 的模式並未討論其累積風險之信賴區間估計, 且該模式在某些情況下會產生資料配適不良, 因此本研究提出了一個新的劑量反應模式, 並針對在二分反應變項下, 利用拔靴法計算累積風險之信賴區間。
另一方面, 在風險評估的研究中, 劑量反應模型可以用來對資料進行配適, 並加以預測在某特定劑量下反應的機率。大多數情況, 反應比例為100p% 時所需劑量是值得我們去探討的。此時的劑量定義為100p% 的有效劑量(effective dose) , 並註記為ED100p
。Chen et al.等人雖已提出一估計累積風險的反應劑量模式, 但是卻未特別針對主題有所探討。因此本篇研究亦提出一兩階段的方法(two-stage approach) 用以進行累積風險評估下其有效劑量之估計。
Risk assessment is usually conducted to evaluate the potential risks from exposure to a single toxic agent. However, people are frequently exposed to multiple chemicals simultaneously or in sequence by different routes from different sources, including, for instance, contaminants in food, drinking water, air, or the environment. Risk assessment for a single researchers to address such question. As a resut, cumulative risk assessment has been received notable interest of researchers in recent years.
Chen et al.(2001) employed the standard logistic model and combine the idea of doseaddition to introduce derived the dose-response function of a chemical mixture for quzntal response. Nevertheless, their dose-response model may not provide a good fit for a given set in some cases. For example, the procedure fails to accyrtly estimate the cumulative risk for combined exposures to multiple chemicals when the dose-response curve are not symmetric or the data have heavy tails. To this end, we proposed an extended doseresponse model for dichotomous response. The approach aim to estimate cumulative risk of the multiple chemicals with common mechanism of toxicity. Confidence interval of the estimated cumulative risk derived from the proposed model was also.
Dose-response model fitted to quantal data can be used not only to predict the probability of response but also to estimate the dose with 100p% response rate, usually called p% effective dose of a toxic chemical and denoted as ED100p. In this study, we propose the two-stage approach to find the mixture effective dose and the optimal individual dose associated with the mixture effective dose. |
