晚期診斷是改善肺癌預後的一大障礙，利用低劑量電腦斷層篩檢可以降低肺癌的死亡。然而，目前對於利用電腦斷層篩檢早期肺癌的成本效果仍沒有共識。因為晚期肺癌接受同一線不同藥物的整體存活往往沒有差異，選擇生活品質較好的治療愈形重要。為了優化肺癌照護，本研究有兩大目標：一、應用實際的生活品質效用值，估計在台灣運用電腦斷層篩檢早期肺癌的成本效果；二、比較晚期EGFR突變非小細胞肺癌三種第一線標靶藥物治療後的生活品質變化。
　　為估計電腦斷層篩檢的成本效果，我們從全國長達13年追蹤的世代估算各組織型態、各期別肺癌的生活品質調整後預期壽命、損失的生活品質調整後預期壽命（loss-of-QALE）與終身醫療花費。National Lung Screening Trial裡的期別轉變被拿來計算：電腦斷層篩檢挽救的loss-of-QALE、多花的終身醫療花費。篩檢陰性、偽陽性以及輻射誘發肺癌多花的費用也被包含在內，進而從公眾支付者的角度來計算成本效果增量比。另外，為了比較不同標靶藥物治療後的生活品質，我們2011至2016年間量測病人的生活品質分數。利用核心平滑的方式，畫出治療後隨時間變化的生活品質曲線。同時我們也重覆收錄會影響生活品質的變項，放入混和模式處理干擾。
　　電腦斷層篩檢增量的費用為每人22,769美金。將之除以挽救的loss-of-QALE（1.16生活品質調整後存活年），成本效果增量比為一個生活品質調整後存活年美金19,695元。拿Dutch-Belgian Lung Cancer Screening Trial篩檢診斷肺癌的期別分布來做敏感度分析，成本效果增量比會掉到一個生活品質調整後存活年美金10,953元。而不同標靶藥物治療後生活品質的研究，共有344病人參與，得到934筆重複測量的資料。我們發現使用afatinib較之gefitinib在生理、心理、社會範疇及10個子項生活品質分數較差，控制了日常體能狀態、exon 19 deletions、疾病惡化等干擾後亦是如此。這個差別似乎在治療的10個月後特別明顯。
　　總結來說，在台灣對於高危險的吸菸者實施低劑量電腦斷層篩檢符合成本效果。然而台灣女性只有約5 %吸菸，未來的研究應嘗試找出適合肺癌篩檢的非吸菸高危險族群。我們第一個提出病人使用afatinib相較gefitinib有著較差的生活品質，不過這還需要未來更多研究的支持。
Abstract:
　　Late diagnosis is a fundamental obstacle to improving the outcomes of lung cancer, screening with low-dose computed tomography (CT) can reduce lung-cancer mortality. However, there is still a lack of consensus on the cost-effectiveness of CT-screening to detect operable lung cancer. Because different palliative treatments for inoperable lung cancer might not show difference in overall survival, choosing a treatment with better quality of life (QoL) becomes important. To optimize lung cancer care, this study aims to: 1. Using the real-world utility values to estimate the cost-effectiveness of implementing CT screening for lung cancer in Taiwan; 2. Compare dynamic changes in QoL after three first-line therapies for EGFR mutation-positive advanced non-small cell lung cancer.
　　To evaluate the cost-effectiveness of CT screening, we estimated quality-adjusted life expectancy (QALE), loss-of-QALE, and lifetime healthcare expenditures per case of lung cancer stratified by pathology and stage from a nation-wide, 13-year follow-up cohort. Cumulative stage distributions for CT-screening and no-screening were assumed equal to those for CT-screening and radiography-screening in the National Lung Screening Trial to estimate the savings of loss-of-QALE and additional costs of lifetime healthcare expenditures after CT screening. Costs attributable to screen-negative subjects, false-positive cases and radiation-induced lung cancer were included to obtain the incremental cost-effectiveness ratio from the public payer’s perspective. To compare QoL after different tyrosine kinase inhibitors (TKIs), we assessed the QoL scores of patients from 2011 to 2016. QoL functions after initiation of treatment were estimated using a kernel smoothing method. Dynamic changes in major determinants were repeatedly assessed for constructing mixed models.
　　After dividing the incremental costs (US$22,769) by savings of loss-of-QALE (1.16 quality-adjusted life year (QALY)), the incremental cost-effectiveness ratio for CT screening was US$19,695 per QALY. This ratio would fall to US$10,953 per QALY if the stage distribution for CT-screening was the same as that of screen-detected cancers in the NELSON trial. A total of 344 patients with 934 repeated assessments were enrolled in the second study. There were significantly lower QoL scores for afatinib versus gefitinib in physical, psychological, social domains, and 10 facets. After controlling for performance status, exon 19 deletions, disease progression, and other confounders in the mixed models, these effects still existed. The differences seemed to appear 10 months after initiation of treatment.
　　In conclusion, low-dose CT screening for lung cancer among high-risk smokers would be cost-effective in Taiwan. As only about 5% of our women are smokers, future research is necessary to identify the high-risk groups among non-smokers and increase the coverage. The initial detection of lower scores of QoL for afatinib in comparison with gefitinib warrants more studies for corroboration.