A modified rational decision model incorporating salient events and social influences (particularly from sexual partners) is used to analyze adolescent women's consistent use of oral contraceptives (OCs) over a six-month period. Data are taken from a panel study of 308 clients of an inner-city family planning clinic. Expected OC use was computed for each subject on the basis of subjective expected utility (SEU) theory, and is found in multivariate analyses to be a significant predictor of actual OC use. In addition, variables representing baseline and follow-up partner influences, the salience of pregnancy for the subject, and positive side effects of OCs during the first months of use are found to predict OC use. Partner's support of OC use during follow-up and positive side effects of OCs are found to predict OC use among subjects for whom OC use was not the expected decision according to baseline SEU. Implications of the findings for models of adolescents' contraceptive behavior and for clinicians are discussed.
PIP: Adolescent women's contraception decision making is studied from the perspective of modified decision making theory, specifically subjective expected utility (SEU) theory. It is assumed that individuals choose contraceptive use on the basis of perceived costs and benefits, susceptibility (risk), and severity of outcomes. A comparison is made the SEUs of using versus not using contraception where the decision may be a rational one. There is a discussion of how this study is different from other SEU studies. The study population included 430 OC users, who were 18 years and not pregnant, at a Planned Parenthood Clinic in Baltimore City in 1988 and interviewed at 3 and t months after their OC visit. 89% of the baseline population completed both followup interviews. Attrition was due to inability to locate respondents. The analysis pertained to 308 respondents who had at least 1 act of coitus in the followup period. Respondents were primarily inner city and sexually active with a high risk of teen pregnancy. Consistency of OC use is the dependent variable, and expected OC use is a key independent variable if the subject is a rational decision maker. Other independent variables were the selection of who were influential members of her household, the number of stability of sexual relationships, having an older sex partner, and having a sex partner supportive of her consistent OC use. Other than partner changes, the frequency of intercourse and side effects of OC use, and the age, race, and mother's education were included. The multiple repression analysis revealed that for those subjects with the expected choice in SEU terms of OC use were 1.59 times more likely to use OCs consistently than those for whom OC use was not expected. SEU concepts were helpful in predicting consistency of OC use in a 6 month period, and intervening events and social influences affect OC use in predictable ways. Inconsistent contraceptors and partners who were older, nonsupportive of birth control, or unstable or nonexclusive. Partners may have more influence than conventionally accepted. Breakups increased the odds of consistent use, but not frequency of intercourse. Having had an abortion is a significant predictor of OC use. The implications are that adolescents are capable of rational decision making and could be advised of the risk factors for consistent use. The limitations are that clinic clients may be self selected on the level of partner support, may not be representative of all seeking contraceptive services, and did not receive the current AIDs prevention and condom recommendations. Future research might use the SEU model with a non-OC method, other outcomes such as STD or abortion prevention, further refinement of the partner's SEU, and the framework in which OC use is promoted over other methods as the clinic effect