自2009年來，標靶藥物用於治療晚期非小細胞肺癌的比率逐年提高，標靶藥物的出現雖為患者帶來希望，但長期的治療效果與潛在的風險，仍有待觀察。隨著其他疾病的發生，藥物種類的使用量也隨之增加，有些研究認為某些特定的疾病藥物會影響患者使用標靶藥物的效果，然而至今醫學上還有許多藥物尚未被證實會影響標靶藥物的療效，然而針對藥物之間的相互影響，醫學上大多以臨床或生物實驗為主，然而健保資料庫的出現為醫療上的數據分析帶來新的發展，期望能從大量的數據中獲取對醫療上有幫助的信息，而醫療上常見的分析手法有存活分析、多變量分析和貝式網路模型。 我們使用台灣健保資料庫去進行數據分析，健保資料庫具備數據量大且整合了病患長期的就醫資料等優勢，我們以標靶藥物為第一線藥物的肺癌患者為主要分析對象，整理患者在治療肺癌期間與其他藥物的使用情況，並透過建構貝式網路來尋找影響患者惡化的潛在原因，並探討病人在不同情況下使用不同種藥物所導致惡化的原因。 ;Since 2009, the ratio of target drugs for the treatment of advanced non-small cell lung cancer has increased year by year. Although the targeted drugs have brought hope to patients, the effect of long therapeutic procedure and potential risks remain to be observed. With the occurrence of other diseases, the use of drug types has also increased. Some studies believe that certain specific disease drugs will affect the effect of patients using target drugs. However, there are still many drugs in medicine that have not been confirmed to affect the efficacy of target drugs. And for the interaction of drugs, most of the medicines are mainly clinical or biological experiments. However, the rise of the health insurance database has brought new developments in medical data analysis. It is expected to obtain medically helpful information from a large amount of data. The common analytical methods in medical practice include survival analysis, multivariate analysis, and bayesian network models. We use the National Health Insurance Research Database for data analysis. The health insurance database has the advantages of numerous data and integration of long-term medical information of patients. We focus on the patients with lung cancer who use target drugs as first-line drugs. And summary the use of other drugs during the treatment of lung cancer. Through the construction of Bayesian network to find potential causes of deterioration. Discuss the reasons for the deterioration of patients using different drugs in different situations.
