| 摘要: | 背景：在台灣，子宮頸癌是公共衛生的重大議題之一。台灣2009年的癌症登記報告，子宮頸癌發生率(年齡標準化發生率每十萬人口9.9人)為婦女癌症的第二位，死亡率(年齡標準化死亡率每十萬人口4.2人)也名列婦女十大癌症死因的第七位。雖比東南亞、非洲及中南美洲等開發中或未開發國家低，但相較歐美等已開發國家仍屬偏高，主要原因為國內自民國1995年才全面推動抹片，相較歐美國家推動起程較晚，因此國內篩檢率和子宮頸癌發生與死亡率尚需一段時間推動方能達到歐美水準。
目的：高篩檢順從率能降低子宮頸癌死亡率，提高篩檢涵蓋率是國家癌症防治計畫重要政策，因此本研究探討：(一)人口學變項、鄉鎮內醫療資源及子宮頸抹片篩檢史與四種不同介入方式的關係；(二)人口學變項、鄉鎮內醫療資源及子宮頸抹片篩檢史與完成邀約與否的關係；(三)四種不同介入方式在邀約與到檢的關係；(四)完成邀約者在四種不同介入方式對到檢之影響。
方法：本研究為隨機分派研究（randomized clinical trial, RCT），邀約名單是依據行政院衛生署國民健康局「子宮頸抹片追蹤管理子系統」，2012年7月至9月提供之30-69歲3年以上未檢資料18,651，與彰化縣戶籍人口資料檔串聯，扣除戶籍不在彰化的483人後有18,168人，經隨機分派為四組進行電話邀約介入，最後有十二鄉鎮9,797人進入研究分析。四種介入方式分別是護理組(二次都由護理人員撥打)、語音組(二次皆由自動語音辨識撥打)、護理語音組(第一次護理人員撥打與第二次自動語音辨識撥打)及語音護理組(第一次自動語音辨識撥打與第二次護理人員撥打)。本研究採用Microsoft Excel 及SAS 9.3版進行資料處理與描述性統計、雙變項分析及羅吉斯迴歸分析。
結果：研究樣本經四種方式介入後，完成邀約之到檢率分別為，護理組14.6%、語音組9.8%、護理語音組13.3%及語音護理組12.4%，相較未完成邀約之平均到檢率0.4%，可見介入之成效。在2次皆由自動語音辨識撥打的成效(9.8%)雖然不及2次皆由護理人員撥打(14.6%)，但如果兩者互相搭配，其撥打成效就與2次皆由護理人員撥打接近(13.3%及12.4%)。
另外在邀約成本上，護理組的到檢成本是每案89元、護理語音組75元，語音護理組53元，語音組成本是24.5元最低。而在介入方式、人口學變項、子宮頸抹片篩檢史及鄉鎮內醫療資源與子宮頸抹片到檢分析結果顯示，年齡在30-39歲、婚姻狀況為已婚、教育程度為高中及以上、篩檢行為為應篩檢年滿9年以上，篩檢次數≧3次，且為3年未檢及鄉鎮內有醫療資源到檢情形較佳，p<0.05皆呈現統計相關。
結論：雖然語音組邀約到檢成效不及護理組，但如果當成輔助工具與護理人員搭配邀約其成效就與單純護理人員撥打接近。在從邀約成本來看，任何1種的語音介入其邀約到檢成本皆明顯低於單純護理人員撥打，如果再加上語音撥打時間不設限、可重複多次撥打、撥打功效大等因素，將來自動語音辨識電話會是很好的篩檢邀約輔助工具。
Background:
Cervical cancer is a major public health concern in Taiwan. From the Cancer Registry Report in 2009, the age-standardized incidence of cervical cancer is 9.9 per 100,000 ,it’s the second leading cause of women’s related-cancers. Its age-standardized mortality is 4.2 per 100,000, also as the seventh place within women’s ten leading death causes of cancers.
No matter the incidence and mortality of cervical cancer in Taiwan are lower than that of Southeast Asia, Africa, and Central and South America and other developing or underdeveloping countries. The statistics reveals that they are still higher than Europe and the United States and other developed countries.
In Taiwan, due to the expand period for cervical cancer screening began on the year of 1995, the attendance rate was far behind the rates of Europe and the United States. How to extensively use automated telephone with speech recognition as one useful cancer screening tactics in health plane is the study inquiry.
Objective:
High screening rate can reduce the high mortality of cervical cancer. How to improve the screening coverage rate is an important policy of the National Cancer Control Program in Taiwan. The study objectives are: (a)to study the associations between sociodemographic characteristics and the accessibility of medical resources for four interventions of telephone and automatic speech recognitions .(b)to investigate the associations of Pap Smear’s attendance rate between women’s demographic characteristics and the accessibility of medical resources for screening. (c)To compare the impact of four different telephone and automatic speech recognition on cervical cancer screening attendance. (d)To understand the factors associated with participant and non-participate to the screening program.
Methods:
This is a randomized clinical trial (RCT) by using the Changhua County databank of “Management System of Cervical Screening (MSCS)” from the Bureau of Health. The MSCS registers all cervical cancer. This study will seized women of 30-69 years old who have not registered smear for the past three or/more years from July to September 2012, there were 18,651 women in the initial listing. After excluding the 483 who were not census registers in our study county, 18,168 individuals were random assigned to four different intervention programs. Another excluding factors such as not have a telephone, error number, no answered call for three times, a total of 9,797 women from twelve townships completed baseline telephone reminding invitation. Intervention-Women were assigned randomly to one of four reminder groups:(1)two calls by nurse(Nurse); (2)two calls by automatic speech recognition(ASR); (3) first call by nurse and by ASR later(Nur- ASR);(4)first call by ASR and by Nurse later(ASR-Nur).Descriptive statistical analysis and chisquare test will be proceed before multivariate logistic regression in this study.
Results:
Each intervention got adherence proportion was 14.6% in NURSE, 9.8% in ASR, 13.3% in NUR-ASR, and 12.4% in ASR-NUR separately. The complete reminded attendances were significantly higher than not-complete attendance 0.4%. About reminded screening cost, it was 89 NT dollars in NURSE group, and 75 dollars in NUR-ASR, 53 dollars in ASR-NUR, and 24.5 dollars in ASR group. Attendance was positively associated with age 30-39, being married, high-school education, screening interval greater than nine years, screened more than three times, and women without screening history within three years but lived at the townships of high accessibility of medical resources.
Conclusions:
Although attendance was lowest in the group of ASR in comparing with the other three interventions, but less cost spent in reminded invitation. In two time’s reminded calls, intervention one time by ASR can be used as an auxiliary tool followed by nurse’s phone call to increase smear attendance proportion in the situation of less cost. Moreover, phone call by automatic speech recognition system can be executed at anytime and repeatable under the conditions of default. |
