臺灣全國子宮頸癌篩檢計畫效益評估

Abstract

子宮頸癌是世界排名第二常見的癌症，其臨床可偵測期漫長，並有操作簡易且不具侵襲性的檢測方法，使得該疾病適於利用篩檢預防侵襲癌的發生。已開發國家所施行的組織性篩檢政策已有效減少子宮頸癌的發生率及死亡率，伺機性篩檢亦有部份效果，但不如前者。臺灣的全國性篩檢政策開始於民國84年，針對三十歲以上婦女提供每年一次的抹片篩檢服務。篩檢效益在臺灣可利用累積至今的篩檢資料進行評估，且可以世代研究設計進行個人資料的分析。 本研究利用數個國家型資料庫進行篩檢效益評估，包含子宮頸癌篩檢登記資料庫、子宮頸暨陰道切片登記資料庫、戶籍登記資料庫、死亡登記資料庫、癌症登記資料庫、以及重大傷病登記資料庫，這些資料庫在本研究的使用經過行政院衛生署國民健康局同意。長期趨勢分析僅使用年齡別及年代別發生數、死亡數、及人口數進行計算，並套以布瓦松迴歸模式（Poisson regression）進行年齡別及年代別因子的估計。篩檢參與、抹片異常後續追蹤及子宮頸癌發生的相關性則利用Cox氏比例危險模式（Cox’s proportional hazards model）進行估計，累積發生率則以Nelson-Aalen估計法計算。 臺灣子宮頸癌三年篩檢率在民國八十四年至民國九十六年間，自小於10%提升到51%，雖然參與率仍低，但已使得子宮頸癌發生率降低48%，死亡率降低49%。在民國84年至民國88年五年間，參加過一年、二年、三年以上篩檢的婦女，相較於該期間未參加篩檢者，可降低於民國90年至民國95年間的子宮頸侵襲癌發生率，分別達41%、55%、以及66%；且可分別降低該期間死於子宮頸癌的風險達61%、72%、以及79%。較近期篩檢正常的婦女罹患子宮頸癌的危險性較較遠期篩檢正常的婦女低，然而一旦發現癌前病變，未每年做抹片檢查的婦女有較高罹患子宮頸癌的機會，在近期或較早期間發現癌前病變的罹癌風險沒有差異。本研究亦發現，子宮頸癌抹片不僅可有效降低子宮頸鱗狀細胞癌的發生率，亦可降低子宮頸腺細胞癌的發生率，然而後者降低幅度不及前者。若婦女在抹片中發現第一次的非典型鱗狀細胞、非典型腺細胞、或低度鱗狀上皮細胞病變，過去二年前曾做過抹片，子宮頸癌發生率較過去未做過抹片者減少20%；若在過去二年內曾做過抹片，則可減少35%子宮頸原位癌及60%子宮頸侵襲癌的發生。婦女第一次發現上述細胞異常之後，有參與抹片或切片追蹤者，可降低12%至48%高度鱗狀上皮細胞病變的發生率、降低41%至49%子宮頸原位癌的發生率、以及降低49%至77%子宮頸侵襲癌的發生率。非典型腺細胞相較於非典型鱗狀細胞以及低度鱗狀上皮細胞病變有更重要的臨床意義。 本研究利用長期趨勢分析以及世代研究設計證實透過全國性篩檢計畫可成功降低子宮頸癌對婦女的威脅，建議要持續積極邀請近三年未做抹片的婦女接受篩檢，以有效提高篩檢率。即便婦女在篩檢中僅發現低度病變，仍然需要監測其後續追蹤情形，建議將非典型腺細胞納入篩檢計畫的陽性追蹤管理系統。

Cervical cancer is the second leading female cancer in the world. The long period of pre-cancerous detectable stage and the availability of a valid simple test with its non-invasive property make cervical cancer ideal for screening. Organized screening program in developed countries has been well documented to effectively reduce the incidence and mortality of cervical cancer, while opportunistic screening may also reduce cervical cancer morbidity and mortality with a lesser effect. Nation-wide screening program in Taiwan has been launched since 1995 to reimburse annual Pap smear to women aged more than 30 years old. The efficacy of the screening program in Taiwan was comprehensively investigated in this study using the long-term continuous data registered in the cervical cancer screening registry. The analysis of efficacy was assessed at population and individual levels, respectively. This study utilized several national registries to evaluate the cervical cancer screening efficacy, including cervical cancer screening registry, cervical and vaginal biopsy registry, household registry, death certificate registry, cancer registry, and catastrophic illness registry with the permission by the Bureau of Health Promotion, Department of Health, Executive Yuan, Taiwan. The secular trend of age- and calendar year-specific incidence and mortality of cervical cancer was assessed by Poisson regression analysis including calendar years and age groups in the models. Two cohort studies were conducted to assess the efficacy of cervical cancer screening at the individual level using computerized linkage to obtain information of screening history, the first abnormality detected by screening, follow-up attendance of Pap smear and biopsy examinations, date of cervical cancer diagnosis, and date at death. The associations with cervical cancer risk were assessed for screening and follow-up attendance by Cox’s proportional hazards models, and the cumulative incidences and mortality of cervical cancer were estimated by Nelson-Aalen method. The triennial screening participation in Taiwan increased from less than 10% to 51% from 1995 to 2007. Despite of the screening participation rate was not high, there was a 48% and 49% reduction in invasive cancer incidence and mortality, respectively. Compared with women who never attended cervical cancer screening from 1995 to 1999, women who attended cervical cancer screening once, twice, and three or more times had a 37%, 52%, and 60% reduction in invasive cervical cancer incidence, respectively; and a 58%, 70%, and 76% reduction in cervical cancer mortality, respectively, from 2001 to 2006. The risks of developing invasive cancer among women attended screening in recent years found to be normal were less than those not attended recently. However, if abnormal lesions have been found, women were not followed annually have higher risk, and there was no difference to detect abnormal lesion in the early or later period. The Pap smear screening also significantly reduced the incidence and mortality of invasive adenocarcinoma with a smaller reduction. Women first detected with atypical squamous cells, atypical glandular cells, or low-grade squamous intraepithelial lesions had at least 20% reduction in cervical cancer risk if they attended screening at two or more years before the detection, and at least 35% and 60% reduction of carcinoma in situ and invasive cervical cancer if they attended screening within 2 years before the detection compared with women who never attended screening. The attendance of follow-up examination after the first detection of cytological abnormality significantly reduced 12% to 48% of HSIL, 41% to 49% of carcinoma in situ, and 49% to 77% of invasive cervical cancer. In comparison with atypical squamous cells and low-grade squamous intraepithelial lesions, atypical glandular cells had the highest risk of developing clinically significant lesion. The success of reducing disease burden of cervical cancer through the implementation of a nationwide screening program was verified by the decreasing secular trend of cervical cancer incidence and mortality in women population, and by the reduction in risk of developing end-stage cervical diseases through the cohort study on 4.7 million women in Taiwan. Continuing to actively invite women not attended screening in recent three years was recommended. Women detected to be affected with low-grade abnormalities by screening need to be monitored carefully. It is suggested that women with atypical glandular cells should be included in the intensive positivity follow-up system in the screening program.