我國大數據於職業病流行病學研究之現況挑戰

The Current Situation and Challenges of Applying Big Data to Epidemiological Research on Occupational Diseases

我國職業病發現率低估之原因，可能與職業傷病預防檢查尚未普及、職業病因果關係不易釐清、勞工作業環境危害物暴露資料庫及生物偵測資料庫不完整、勞工工作資訊不完整等因素有關。檢視我國職業病流行病學研究工具，多仰賴橫斷面的問卷調查，由於職業病具長期累積的特性，單一時段的調查無法進行疾病因果關係分析，因此本研究認為巨量數據分析可能為其中解套的方式之一。本研究羅列出職業流行病學上常使用來評估職業病與職業暴露關係之流行病學方法及每種研究設計方法之優點及其限制；另，本研究亦列出現行我國政府各部會已建構之巨量數據庫，盤點職業病暴露因素相關之變項，及各資料庫間之水平整合及串聯，以更精準的描繪出職業病與暴露因素之間的樣態。本研究提出幾點對於現行勞保資料庫面臨之挑戰及建議：1.勞保資料庫缺乏職業別欄位。2.勞保資料之準確性尚須加強。3.勞動部所公告之職業相關疾病與國際疾病分類診斷代碼無法互相對應。4.面對突發事件能快速組建應變小組且定期召開專家會議討論，高風險之職業相關疾病。5.應該加強勞工之健康識能。6.擴大健康檢查之族群，並建置離職後健康檢查之監測追蹤。

The reasons for the underestimation of the occupational disease detection rate in Taiwan may be due to the failure of occupational injury and disease prevention inspections, difficulty in clarifying the causal relationship of occupational diseases, incompleteness of labor working environment exposure database and biodetection database, and lack of labor and work information, and so forth. Reviewing the research tools of occupational epidemiology in Taiwan relies on cross-sectional questionnaire surveys which cannot be used for disease causality analysis due to the long-term accumulation of occupational diseases. Therefore, this study believes that the analysis of big database may be the only way to solve the problem. This study lists the epidemiological methods commonly used in occupational epidemiology to assess the relationship between occupational diseases and occupational exposure, and the advantages and limitations of each research design method; in addition, this study lists various constructed big database from many ministries and variables related to occupational diseases and exposures. Through integrating and connecting each database, it more accurately describes the relationship between occupational diseases and exposure factors. This research also puts forward several challenges and suggestions for the current labor insurance database: 1.The labor insurance database lacks occupation variable. 2.The accuracy of labor insurance data needs to be strengthened. 3.The occupational-related diseases announced by the Ministry of Labor and the codes of the International Disease Classification cannot correspond to each other. 4.It can quickly set up a response team to face emergencies and hold regular expert meetings and closely discuss potentially high-risk occupational-related diseases. 5.The health awareness of workers should be strengthened. 6.Expand the population of health checkups, and establish monitoring and tracking of post-employment health checkups.