アンケート調査実施群と非実施群の肺がん検診受診率の推計値の検討

本研究は，がん検診が始まる前に行われたアンケート調査が受診率に与える影響を検討することと，はがき調査回答者の受診状況のデータを基に，母集団の受診率を推計するときの問題点を検討することを目的とした．われわれが行った先行研究の，A市の肺がん検診対象者（地域検診と任意型・職域検診）の４０歳から５９歳までの１９００６人を対象とし２０００人を無作為抽出（アンケート群）し，アンケート調査およびはがき調査を行った結果を解析した．アンケート群と非抽出群における実際の肺がん地域検診受診率の差を検討した．さらに，受診後に行われたはがき調査から任意型・職域検診の受診率の推定を行った．地域検診受診率は，アンケート群では１４．５％で非抽出群の６．７％と比較して有意に高かった（p＜０．０１）．任意型・職域検診受診率は，はがき調査から推定でき，アンケート群で９．０％となり，非抽出群の４．２％と比較して有意に高かった（p＜０．０１）．はがき調査による地域検診受診者数は６０人で，A市による調査実数６１人とほぼ一致した．その時の受診率は５４．１％で，実際の受診率と比較して約４倍となった．アンケート群の受診率が非抽出群の受診率よりも有意に高くなったことは，アンケートを行ったこと自体が受診率に影響をもたらしたと考えられた．またアンケートに回答する人は，受診者に多くみられたことから，アンケートによる受診率の推計は回収率に影響された．アンケート結果の解析方法によっては，任意型・職域検診受診率を推計できた．

The purposes of this study are to evaluate the effect on the screening rate by questionnaire survey conducted before the cancer screening begins, and to clarify the problem when estimate the screening rate of the population on the data of the examination situation of postcard survey respondents. The data of questionnaire and postcard survey obtained from our previous research where men and women aged ４０ to ５９ years old for lung cancer screening （population screening by municipality “A” city, and occupation-related and opportunistic screening）of１９，００６ were targeted and ２０００ （questionnaire group） were selected by random sampling method, was newly analyzed again. The difference of population screening rate between a questionnaire group and a non-selected group was compared. And occupation-related and opportunistic screening rate was estimated by postcard survey after screening. The population screening rate of １４．５％ for questionnaire group was significantly higher than that for non-selected group of ６．７％ （p＜０．０１）. The occupation-related and opportunistic screening rate was estimated by the postcard reply, and the rate was９．０％ for questionnaire group. The rate was significantly higher than ４．２％ for non-selected group （p＜０．０１）. The consultation number of population screening by postcard reply （６０ persons） was almost corresponding to the actual number by municipality survey （６１persons）. Then the rate by postcard reply was ５４．１％ and was about ４ times compared with the actual screening rate. It might be indicated that only a questionnaire survey affected screening rate because the screening rate for questionnaire group was significantly higher than that for non-selected group. And because almost every consulted people replied to the postcard, the estimated screening rate by postcard survey was affected by collection rate. Occupation-related and opportunistic screening rate could be estimated depending on the analysis of questionnaire result.