Objective
The objective of the study was to improve the understanding of long-acting reversible contraception (LARC) use patterns among unmarried, young adults at risk of unintended pregnancy.
Study Design
We performed a secondary data analysis of a national survey conducted by Guttmacher Institute of unmarried women and men aged 18-29 years. LARC is defined as an intrauterine device (IUD) or implant. Predictors of LARC use and IUD knowledge among those at risk for unintended pregnancy (n = 1222) were assessed using χ 2 analysis and logistic regression models.
Results
LARC use was associated with older age, high IUD knowledge, and earlier onset of sexual activity. Respondents with high IUD knowledge were 6 times more likely to be current LARC users (odds ratio [OR], 6.3; 95% confidence interval [CI], 1.4–28.8). Sociodemographic variables did not predict use. Respondents with lower education (OR, 1.76; 95% CI, 1.0–3.0), an external locus of control (OR, 1.6; 95% CI, 1.1–2.3), male sex (OR, 2.8; 95% CI, 1.9–4.1), and foreign language had less knowledge of IUD.
Conclusion
Increasing knowledge of IUD among certain groups may improve LARC use among young, unmarried adults and in turn decrease unintended pregnancy.
Reducing the proportion of unintended pregnancies, which currently account for nearly half of all pregnancies in the United States, has been prioritized as a national public health goal. Increasing the use of long-acting reversible contraception (LARC), including intrauterine devices and contraceptive implants, is a recognized strategy for reducing unintended pregnancy. Requiring no user compliance once inserted, LARC methods essentially eliminate the inconsistent and incorrect use that plagues other contraceptive methods. These methods are the most efficacious and cost-effective reversible forms of contraception, resulting in fewer pregnancies during the first year of use than any other reversible method.
Despite these positive attributes, LARC methods are utilized by relatively few women, with approximately 5% of women reporting ever-use of an intrauterine device or contraceptive implant. A barrier to improved use may be lack of knowledge of LARC methods. As many as two thirds of adolescent and young adult women report that they had never heard of these methods. Lack of provider knowledge and misinformation in the media may also negatively impact use.
Contraceptive decision making is likely influenced by a complex array of sociodemographic, behavioral, attitudinal, and knowledge factors as well as the influence of the social network, partner characteristics, and recommendations from health care providers. Partner influence and relationship dynamics are also emerging as influential factors in contraceptive decision making, although less research has addressed the characteristics of men whose partners use LARC.
Increasing the proportion of at risk women who use LARC will require better understanding of use patterns and development of targeted interventions. We used data collected through a nationally representative survey to further explore factors associated with LARC use among a population at risk for unintended pregnancy.
Materials and Methods
We performed a secondary analysis using data collected through a national survey of fertility and contraceptive knowledge commissioned by the National Campaign to Prevent Teen and Unplanned Pregnancy and conducted by the Guttmacher Institute. This survey gathered detailed information from a nationally representative probability-based sample of 1800 unmarried women and men aged 18-29 years. The methods have been previously described. Briefly, the sampling method included random digit dialing of land lines and cell phones as well as targeted sampling of land lines. African Americans and Hispanics were oversampled and constituted 21% and 22% of the final study population, respectively. The primary research team weighted the data prior to public release. Permission was obtained to use the data for our analysis. Approval from the Institutional Review Board of the Medical University of South Carolina was obtained prior to the analysis.
To obtain a sample deemed at risk of unintended pregnancy, we excluded respondents who had not had sex within the last year (22%), who reported previous sterilization (4%), and who reported that they were pregnant or trying to become pregnant (5%). Current contraceptive use was not used to define risk status.
Our primary outcome was the current use of LARC, defined as use reported within the last month. Male respondents were asked whether a partner had used an intrauterine device or implant when having sex with them during the past month. If they responded yes, they were considered current users of LARC in our analysis.
Knowledge of intrauterine contraception was treated as a secondary outcome. Six intrauterine device (IUD) knowledge questions with true/false answers were asked of both male and female respondents. These questions assessed IUD knowledge in the following areas: legality of IUD use in the United States, safety of concomitant use of IUDs and tampons, whether IUD insertion requires a surgical operation, safety of IUD use in nulliparous women, the ability of an IUD to move around in a woman’s body, and the ability of a male partner to feel the IUD during intercourse.
The number of correct answers to these questions was summed to create a knowledge score ranging 0–7. Using the mean number of correct answers as a cut point, a dichotomous variable was created with low IUD knowledge defined as 2 or fewer correct answers and high IUD knowledge defined as 3 or more correct answers. IUD knowledge is analyzed as both an ordinal and dichotomous variable.
The following variables were considered as potential predictors of current LARC use: sex, age, race/ethnicity, insurance status, foreign-born status, primary language, religious affiliation, education level, prior sexual education, age at first intercourse, locus of control, gravidity, parity, experience of prior unplanned pregnancy, self-rated importance of avoiding pregnancy, prior use of any birth control, and knowledge of intrauterine contraception.
Locus of control is a general construct referring to the extent to which individuals believe that they can control events that happen to them. For our analysis, we defined the external locus of control as a response of “somewhat or strongly agree” to the following statement: “It does not matter whether you use birth control or not; when it is your time to get pregnant, it will happen.”
Descriptive analyses were conducted to describe the distribution of these variables among the population at risk for unintended pregnancy. Because the survey used weighted sampling methods, all results are reported as weighted proportions, not frequencies. Current LARC users were compared with nonusers through a bivariate analysis using a Rao-Scott χ 2 test. Significance was set at P < .05. All variables with P < .1 in bivariate analysis were considered for inclusion in a multivariable logistic regression model with current LARC use as a dichotomous outcome.
The final model was constructed using stepwise subtraction of variables with retention of variables that remained significant with P < .05. A separate logistic regression model was constructed in the same fashion using our dichotomous IUD knowledge variable as the outcome. Results are reported as odds ratios (ORs) with 95% confidence intervals (CIs). All statistical analyses were conducted using SAS 9.2 (SAS Institute, Cary, NC), taking into account the weighted nature of the data.
Results
Among those at risk for unintended pregnancy (n = 1222), only 4% of participants were currently using a LARC device (IUD 52; implant 4). Baseline characteristics of respondents in our sample are outlined in Table 1 . Males represented 54% of the total respondents in our sample and 41% of those who reported use of LARC within the last month. The population was predominantly white (59%), English speaking (84%), and had insurance (77%). The majority of the respondents had at least a high school education (83%). Most reported that avoiding pregnancy was important (very important, 77%; somewhat important, 11%). Approximately half experienced sexual debut at age 16 years or younger. Approximately one-quarter of patients had given birth or fathered a child and a quarter had experienced an unintended pregnancy. More than half of the respondents were classified as having an external locus of control.
| Characteristic | Total population | Current LARC users | Current LARC nonusers | P value b |
|---|---|---|---|---|
| (n = 1222) | (n = 56) | (n = 1156) | ||
| Sex, % | .14 | |||
| Male | 54 | 41 | 54 | |
| Female | 46 | 59 | 46 | |
| Age, % | .003 | |||
| 18-19 y | 26 | 3 | 27 | |
| 20-24 y | 40 | 52 | 39 | |
| 25-29 y | 35 | 45 | 34 | |
| Race, % | .3 | |||
| Non-Hispanic white | 59 | 64 | 59 | |
| Non-Hispanic black | 16 | 13 | 16 | |
| Hispanic | 18 | 22 | 18 | |
| Asian | 6 | 1 | 6 | |
| Insurance, % | .47 | |||
| Medicaid only | 15 | 25 | 14 | |
| Medicaid and private | 9 | 9 | 9 | |
| Private only | 51 | 49 | 51 | |
| Other Insurance | 2 | 2 | 2 | |
| Uninsured | 23 | 16 | 24 | |
| Education, % | .2 | |||
| Less than high school | 17 | 9 | 17 | |
| High school graduate/GED | 30 | 41 | 30 | |
| Greater than high school | 53 | 50 | 53 | |
| English as primary language, % | 84 | 91 | 84 | .25 |
| Have children, % | 25 | 46 | 24 | .03 |
| Prior unplanned pregnancy, % | 27 | 49 | 26 | .002 |
| Age at first intercourse younger than 17 y, % | 54 | 73 | 53 | .04 |
| External locus of control, % | 60 | 65 | 35 | .55 |
| Stated very/somewhat important to avoid pregnancy, % | 88 | 94 | 88 | .2 |
| High IUD knowledge, % | 64 | 91 | 63 | .005 |
a All data are presented as column percent. No frequencies are reported given the weighted nature of the data;
b P values are based on χ 2 test for comparison of categorical variables between current users and nonusers of LARC.
In the univariate analysis ( Table 2 ), current LARC use was less likely among 18-19 year olds when compared with 25-29 year olds (OR, 0.1; 95% CI, 0.02–0.4) and was less likely among those who classified their race as Asian or other (OR, 0.1; 95% CI, 0.03–0.5). For each year older a respondent was at the time of sexual debut, they were 10% less likely to be a current LARC user (OR, 0.9; 95% CI, 0.8–0.98). Those who had given birth or fathered a child were 2.7 times more likely to report current LARC use (95% CI, 1.3–5.6).
| Variable | Unadjusted OR (95% CI) | P value | Adjusted OR (95% CI) a | P value b |
|---|---|---|---|---|
| Race/ethnicity | ||||
| White non-Hispanic | Referent | Referent | ||
| Black non-Hispanic | 0.7 (0.3–1.8) | .5 | 0.6 (0.2–1.5) | .3 |
| Hispanic | 1.1 (0.4–2.6) | .8 | 1.4 (0.5–3.7) | .5 |
| Asian/other | 0.1 (0.03–0.5) | .003 | 0.1 (0.03–0.4) | .002 |
| Age | ||||
| 18–19 | 0.1 (0.02–0.4) | .0009 | 0.1 (0.01–0.2) | .0001 |
| 20–24 | 1.002 (0.5–2.1) | 0.8 (0.4–1.8) | ||
| 25–29 | Referent | .99 | Referent | .6 |
| Age at sexual debut c | 0.9 (0.8–0.98) | .02 | 0.8 (0.7–0.9) | .0007 |
| Have children d | 2.7 (1.3–5.6) | .01 | — | — |
| Prior unplanned pregnancy d | 2.8 (1.3–5.8) | .01 | — | — |
| Total IUD knowledge e | ||||
| Low | Referent | Referent | ||
| High | 5.8 (1.5–23.0) | .01 | 6.2 (1.4–28.5) | .02 |
a n = 1203 for the adjusted model, which is smaller than the entire sample size, n = 1222, because of missing responses for age at first intercourse;
b OR, 95% CIs, and P values were generated through univariate and multivariate logistic regression modeling;
c In the model, age at sexual debut was treated as a continuous variable, with the OR representing the probability of current LARC use for each additional year older at the time of sexual debut;
d These variables were no longer significant in the adjusted analysis;
e IUD knowledge was defined as low if 2 or fewer of 6 questions were answered correctly and high if 3 or more of 6 were answered correctly.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
