以Poisson Regression Model分析職場中可能曝露游離輻射員工罹患癌症之風險

Cancer risk analysis of possible Ionizing Radiation Exposure workers in enterprises with Poisson Regression Model.

謝明忠(Ming-Chung Hsieh)

指導教授 ： 張寶樹

高雄醫學大學/健康科學院/職業安全衛生研究所/碩士(2013年)

摘要

研究背景：
在我國，從事輻射相關工作人數眾多，依照民國 100年1月21日修正發布「勞工健康保護規則」第十四條之規定，從事第二條之特別危害健康作業，應建立健康管理資料，並依規定分級實施健康管理，這些勞工在工作過程中，曾經因工作需要而可能曝露於游離輻射環境中，這群勞工相關癌症風險為本研究關注焦點。
材料與方法：
本研究蒐集台灣從事游離輻射工作者，自1992年至2010年，19年間共34,404人年，罹患癌症共84名員工資料，包括工作時間和工作地點，依工作地點將員工分組為北部與南部，考慮輻射曝露誘發癌症之潛伏期，設定曝露時間未滿5年和工作史未滿10年的排除條件，由於女性案例僅有3例，本研究將女性排除，最終得到62位男性員工納入Poisson Regression Model分析罹癌風險，並將年度分組，包括不分組、5年和10年，探討不同年度間的罹癌風險。
研究結果：
本研究分析結果顯示，年度不分組部分，北部罹癌機率比南部高了0.14倍，沒有達到統計上顯著差異；年度依5年分成4組，北部罹癌機率比南部高了0.14倍，統計上未達到顯著差異；年度依10年分為2組，北部罹癌機率比南部高了0.14倍，統計上未達到顯著差異。
結論：
Poisson Regression Model分析罹癌風險，可以發現年度分組部分，各年代組風險都沒有顯著差異；10年分組部分，1992-1999年比2000-2010年高了0.12倍，未達到顯著差異，這個趨勢類似日本原子彈爆炸後倖存者餘命研究，在年輕時曝露風險較高。

關鍵字

游離輻射 ； 勞工健康保護規則 ； Poisson Regression Model

並列摘要

Background:
In Taiwan, there are many kinds of Ionizing Radiation-related workplaces, including Nuclear Power Plants, Medical Careers, Nuclear Engineering, Radiation Detection and etc. Hundreds of thousands employees are engaged in radiation-related works. According to Article 14 of Labor Health Protection Rules (LHPRs), amendment on January 21, 2011, enterprises with labors engage in operations with special health hazards (e.g. ionizing radiation) from Article 2 of LHPRs, must establish health data manager system, and implement the classifications of health manger system. We would like to analyze cancer risk of those workers who might be possible exposed to ionizing radiation in their past working history with Poisson Regression Model.
Materials and Methods:
Our objects are cancer cases in enterprises and with radiation exposure in Taiwan from year 1992 to 2010. Totally, we collected 84 cancer cases, and 34,404 person years. The collected data of 84 cancer employees included gender, exposure time and working areas. We grouped these workers into northern and southern areas by their working areas. Considering the latency period of radiation induced cancers, we set up criteria to exclude those workers whose exposure time is less than 5 years and working time is less than 10 years. We also excluded 3 female cases, because the case number is too few. Finally, 62 male workers’ data are suitable to input Poisson Regression Model. For the 62 workers, we grouped them into 1 year, 5 years and 10 years groups, to explore the cancer risk between different eras.
Results:
Our study results show that in groups of 1 year, cancer risk of northern working area is higher 0.14 time than southern, which is not statistically significant. In groups of 5 years, cancer risk of northern working area is 0.14 time than southern, and is not statistically significant. In groups of 10 years, cancer risk of northern working area is 0.14 time than southern, which is not statistically significant.
Conclusions:
In the Poisson Regression Model cancer risk analysis shows that in groups of 1 year, the cancer risk is not statistically significant. In groups of 10 years, cancer risk of the group 1992~1999 is 0.12 time higher than group 2000~2010, this trend is similar to the atomic bomb survivors’ Life Span Study (LSS) cohort in Japan, that is, cancer risk is higher when exposed at younger ages.

並列關鍵字

Ionizing radiation ； Labor Health Protection Rules ； Poisson Regression Model.

參考文獻

[1] 中華民國. 勞工安全衛生法. 2002.

[2] 中華民國. 勞工健康保護規則. 2011.

[3] United Nations Scientific Committee on the Effects of Atomic Radiation. (Accessed at http://www.unscear.org/.)

[4] 張寶樹. 游離輻射防護導論. 初版二刷(修訂版) edition. 台灣台北市: 合記圖書出版社, 2011.

[5] 中華民國. 游離輻射防護安全標準. 2005.