Trends and determinants of IUD use in the USA, 2002-2012

What factors and subgroups have propelled the recent increase in intrauterine device (IUD) use in the USA?

The increase in IUD use, from 1.8 to 9.5% in the USA between 2002 and 2012, was driven primarily by a marked uptake among parous women who intended to have more children.

Recent data suggest an unprecedented increase in IUD use among women in the USA, yet less is known about how this increase has affected the overall proportion of women, at risk of unintended pregnancy, who are using contraception and which social and economic groups are involved.

Data are drawn from the 2002 and 2011-2013 National Surveys of Family Growth. The surveys were based on cross-sectional, national samples of women of 15-44 years of age in the USA. Women responded to in-person interviews, which lasted an average of 80 min. The response rate was 80% in 2002 and 73% in 2011-2013. The sample included 7643 completed interviews in 2002 and 5601 interviews in 2011-2013.

This study was limited to women at risk of unintended pregnancy, i.e. women who were sexually active in the previous 3 months (using contraception or not); it excludes women who were sterile, currently pregnant or trying to conceive. Altogether, 5181 women were at risk in the 2002 sample and 3681 were at risk in the 2012 sample. We used descriptive statistics to investigate trends in contraceptive use patterns by women's sociodemographic characteristics between 2002 and 2012 and used logistic regression to identify current predictors of IUD use in 2012.

IUD use increased from 1.8% in 2002 to 9.5% in 2012 (P < 0.001). The surge was especially marked among parous women who intended to have more children (4.2% in 2002 to 19.3% in 2012; P < 0.001); it occurred to a lesser extent among parous women who did not intend to have more children (2.0-9.7% P < 0.001), suggesting that IUDs are more often used for spacing than for ending childbearing in the USA. The most important predictors of IUD use in 2012 were age, parity and intent to have children. Dissatisfaction with a previous method was also associated with IUD use (adjusted odds ratio = 1.89, P < 0.001).

As with all cross-sectional studies, causal inference is limited. Data are self-reported, but the survey had a high response rate and rigorous quality controls.

This study shows promising trends in the use of highly effective contraceptive methods in the USA, which may help to explain recently reported declines in unintended pregnancy in the USA.

Caroline Moreau was supported by the William Robertson endowment funds. The work of Hannah Lantos and William Mosher on this analysis was supported by the Department of Population Family and Reproductive Health, The Johns Hopkins Bloomberg School of Public Health. The authors declare that no conflict of interest exists.