上海市婚检青年性行为调查资料一致性的影响因素分析

Analysis on influential factors of consistency of data on premarital couples' sexual behavior in Shanghai

目的:根据两次调查中对象对'您与现在的男友/女友在婚检之前是否发生过性行为?'问题的回答,比较不同特征的男女对象回答的一致性并分析其可能的影响因素.方法:在上海市6所妇幼保健院(所)招收接受婚检的男女青年718对,随机分为自填问卷与面对面调查两种方式开展调查,4～6周后随访了其中的市区对象399对,调查方式与初访相同.结果:对象本人前后回答的一致率为90.11%,Kappa值为0.762;初访时男女友之间的回答一致率为88.75%,Kappa值为0.733;对象本人前后的回答及男女友之间回答的不一致率及Kappa值因对象特征的不同而有所不同;多元Logistic回归分析发现,接受调查时有他人在场的女性对象,倾向于做出与其'爱人'不一致的回答;自填问卷、职业为技术管理人员、未同住者则有更大的可能做出前后不一致的回答.结论:对象对性相关问题的回答受到其本人特征和调查方法的影响.在敏感问题的调查中,应充分估计各种因素对资料可靠性的影响,以对结果做出合理的解释,才能得出可靠、正确的结论.

Objective: To evaluate the influential factors of consistency of data on sexual behavior based on the consistence of answers in relation to sex-related information. Methods:718 pairs of premarital couples received antemarital examination in six MCH centers were randomly divided two groups: self-completed and face-to-face questionnaires. 399 pairs in urban areas were followed-up for 4~6 weeks later. The same questionnaire in the initial survey was applied in the follow-up survey. Results:Consistent rate and Kappa between two visits for the question “Have you ever had sex intercourse with you girl/boy friend before marriage” were 90.11% and 0.762, respectively, whereas they were 88.75% and 0.733 correspondingly for the same question between the “couples”. Consistent rate and Kappa between two visits or between the couples were different between the subjects with different characteristics. Multivariable logistic regression analysis revealed that the female who were interviewed in more open environment inclined to respond inconsistently with her “husband”, while those who completed the self-administered questionnaire or those who were technical or manageable staffs or those who did not cohabit were prone to respond incongruously with themselves.Conclusion:Consistency of sex-related data was influenced by the content and the sensitivity of questions. Therefore, for studies concerning sensitive topics, it is necessary to take into account the potential influences on data reliability in order to explain the results reasonably and draw a reliable and correct conclusion.