Reliability of adolescents' self-reported sexual behavior: A comparison oftwo diary methodologies

Purpose: To evaluate techniques for measuring highrisk sexual behaviors by comparing the reliability and acceptability of two daily sexual behavior di ary modes: a written calendar and an automated telephone interview. Methods: This randomized controlled study included 105 sexually active fema le adolescents aged 15-19 years recruited from among teens seeking reproduc tive health care services at a family planning clinic in the San Francisco Bay Area. Participants completed a standardized sexual behavior questionnai re each day for 4 weeks. Contraceptive use by method type was recorded. Rep orting differences between the two diary modes were assessed using generali zed estimating equations, concordance of diary and retrospective interview responses was evaluated using kappa statistics, and contingency table analy sis and Poisson regression models were constructed to examine mode acceptab ility, Results: Respondents randomized to the telephone diary cohort reported less frequent use of barrier contraceptive methods, specifically less spermicid e use (odds ratio 0.27, 95% confidence interval 0.08, 0.95), and decreasing male condom use over time, whereas reports of male condom use increased fo r written diary respondents (p = .007). Participant characteristics associa ted with diary acceptability, defined as the frequency of diary completion, were assessed and teens classified as higher risk provided fewer diary rep orts (p < .01). Regardless of mode completed, 65% of respondents believed t he telephone diary would be preferable to the written diary for most teens. Conclusions: The automated telephone diary offered an acceptable, even pref erred, methodologic alternative to the written diary calendar and elicited more accurate reporting of selected contraceptive behavior. (C) Society for Adolescent Medicine, 2001.