Determinants of trends in condom use in the United States, 1988-1995.

CONTEXT: Although overall condom use has increased substantially over the past decade, information is needed on whether dual method use has also become more common. In addition, there is little information on which characteristics of women influence condom use and dual method use, and on whether these characteristics have changed over time.
METHODS: Data from the 1988 and 1995 National Surveys of Family Growth are examined to evaluate trends in condom use--either use alone or use with another highly effective method (dual method use). Logistic and multinomial regression analyses are presented to analyze the influence of women's characteristics on condom use.
RESULTS: Current condom use rose significantly between 1988 and 1995, from 13% to 19% of all women who had had sex in the past three months. Dual method use increased from 1% in 1988 to 3% in 1995, still a very low level. In both years, current condom use was higher among women younger than 20 (32-34% in 1995) than among those aged 30 or older (less than 20% in 1995). Likewise, current condom use was most common among never-married women who were not cohabiting in both 1988 (20%) and in 1995 (34%). Multivariate analyses showed that women in the early stage of a relationship (six months or less in duration) were much more likely than those in a long-standing relationship (five years or more in length) to use the condom (odds ratio, 1.5). In both 1988 and 1995, younger women and better educated women were more likely to be currently using the condom than were older or less-educated women. For example, in 1995, women younger than 18 were 1.8 times as likely as 40-44-year-olds to be using condoms, and college graduates were 1.5 times as likely as high school graduates to do so. Further, women who were not in a union and either had never been married or were formerly married were more likely to be current condom users in 1995 than were married women (odds ratios, 1.5-1.9). Poor women were less likely than higher income women to be condom users in 1995 (odds ratios, 0.7-0.8), but poverty had made little difference in 1988. Groups likely to be dual method users were those also likely to be at greater risk of sexually transmitted disease: women in a union of less than six months duration (2.8), women younger than 20 (4.6-6.8), unmarried women (2.8-7.5) and women with two or more partners in the past three months (1.7).
CONCLUSIONS: While the increase in condom use, especially among unmarried and adolescent women, is encouraging, condom use overall is substantially less than that needed to protect women and men against sexually transmitted diseases (including HIV). Moreover, steps need to be taken to understand why levels of dual method use are low and how they may be increased.