老年婦女子宮頸癌變與抹片篩檢史之相關研究：配對病例對照研究

研究背景：根據世界衛生組織2012年最新統計指出，子宮頸癌的發生在全球女性前五好發癌症中居於第四位，雖然各國實施子宮頸抹片篩檢計劃的始點不同，但皆已使篩檢率大幅提升，子宮頸癌發生率及死亡率明顯降低。我國於民國84年起，提供30歲以上的婦女每年1次免費子宮頸抹片篩檢，推動至今18年，已使國內30歲以上每3年1次的篩檢率大幅上升，子宮頸發生率及死亡率也已明顯下降，顯示子宮頸抹片篩檢之重要性。但我國70歲以上婦女的篩檢率低於其他年齡層，再加上疾病自然史之演變以及以致從65歲以上開始子宮頸癌發生率與死亡率隨著年齡的增加而上升，從而使過去篩檢率低者至老年時期罹患侵襲癌之比率較高。由於台灣地區已邁入高齡化社會，現今國人女性的平均壽命為83歲，70歲以上婦女罹患子宮頸侵襲癌的發生率又高於原位癌許多，故本研究欲探討70歲以上婦女罹患子宮頸癌與其篩檢史之相關性，以了解老年婦女子宮頸抹片篩檢情形及其影響因素，提供有關當局子宮頸癌預防保健篩檢政策之參考依據。研究目的：本研究之目的在探討70歲以上女性有無罹患子宮頸癌與子宮頸抹片篩檢史之相關性。研究方法：本研究使用「癌症登記資料檔」、「子宮頸癌篩檢登記資料檔」、「子宮頸癌防治監測中心統整之人口資料檔」及「死亡登記資料檔」進行個案選取，並採配對式病例對照研究設計(Matched case-control study)，配對年齡±2歲及居住地區。使用條件式邏輯斯迴歸模式探討子宮頸原位癌及侵襲癌之發生與篩檢史之相關性，估算勝算比及95%信賴區間。研究結果：在原位癌發生與否的分析當中納入3605個分析對象，其中病例組有721人，對照組有2884人；在侵襲癌發生與否的分析當中納入4555個分析對象，其中病例組有911人、對照組有3644人。經年齡及居住地區配對後，70歲以上原位癌之病例組及對照組在年齡、居住地區及都市化程度之分佈上無差異，侵襲癌配對結果亦同。在原位癌發生與否的分析當中，過去曾經接受過子宮頸抹片篩檢之婦女罹患子宮頸原位癌之相對危險性是過去未曾接受過抹片篩檢之婦女的1.40倍(95% C.I. = 1.15-1.71)，隨著篩檢間隔距離越短，婦女發生子宮頸原位癌之機會越高，也就是原位癌之罹癌相對危險性隨篩檢間隔距離呈上升之趨勢。而侵襲癌發生與否的分析當中，過去曾經接受過子宮頸抹片篩檢之婦女罹患子宮頸侵襲癌之相對危險性是過去未曾接受過抹片篩檢之婦女的0.39倍(95% C.I. = 0.36-0.46)，隨著篩檢間隔距離越短，婦女發生子宮頸侵襲癌之機會越低，也就是侵襲癌之罹癌相對危險性隨篩檢間隔距離呈下降之趨勢。結論：本研究結果指出，發病前距離最後一次接受抹片之間隔顯著地與子宮頸原位癌與侵襲癌之發生有關。此結果顯示提高老年婦女子宮頸抹片之篩檢率之重要性。

關鍵字

子宮頸癌；子宮頸抹片篩檢；篩檢間隔

並列摘要

OBJECTIVES: This study aims to examine the association between incidence of cervical cancer and screening history among women aged 70 and over in Taiwan. METHODS: The data sources were the secondary data files in Taiwan. This study uses matched case-control study to investigate the relations between cervical cancer and screening history. All new cases of cervical cancer were identified from 2006-2010 cancer registry as the study cases, and the control cases were matched by age and residential areas. The screening intervals was defined as the number of days before the date of cancer diagnosis. Odds ratio and their 95% confidence intervals of screening history were estimated to predict the risk of cervical cancer in the conditional logistic regression models. RESULTS: In predicting the risk of cervical carcinoma in situ, women who had been screened before had higher risks than women who never had screened (OR=1.40, 95% C.I.=1.15-1.71). And the shorter the screening intervals the higher the risk of cervical carcinoma in situ. In predicting the risk of invasive cervical cancer, women who had been screened before had higher risks than women who never had screened before (OR=0.39, 95% C.I.=0.36-0.46). And the longer the screening intervals the higher the risk of invasive cervical cancer. CONCLUSIONS: Screening intervals was showed to have an effect on the incidence of cervical cancer among women aged 70 and over. Measures aiming to improve the screening coverage of older women are warranted to enhance early detection of cervical cancer.