Contraceptive Choice: How Do Oral Contraceptive Users Differ from Condom Users and Women Using No Contraception?

Abstract

Objective

To determine whether contraceptive choice is influenced by social and reproductive characteristics in a cohort of high-risk women.

Study Design

This is a cross-sectional analysis of baseline date from a randomized clinical trial. We evaluated characteristics associated with oral contraceptive use, male condom use, or use of no contraceptive method.

Results

Women using OCs were less likely to have less than a high school education, to be African American or Hispanic, and to pay out of pocket for medical services compared to women not using any form of contraception. Women using OCs differed compared to women using condoms in that they were less likely to pay out of pocket for medical services. Finally, while number of sexual partners was associated with contraceptive choice, other reproductive characteristics, were not.

Conclusion

Among this cohort of women at high-risk for STDs and unintended pregnancy, sociodemographic characteristics influenced contraceptive choice.

Introduction

Unintended pregnancy is a significant public health problem in the United States. Nearly half of all pregnancies are unplanned and over half of these pregnancies end in abortion.¹ Unintended pregnancy is a frequent occurrence despite the availability of numerous methods of effective contraception. Contraceptive decision making is influenced by a number of factors including a woman’s personal health behaviors, perceived risks and benefits, convenience, socioeconomic factors, and a provider’s willingness to prescribe a method.

This study sought to determine whether the choice of contraception correlates with various sociodemographic and reproductive characteristics. We hypothesized that characteristics of oral contraceptive (OC) users would differ from women using no contraception and women reporting male condom use.

Materials and Methods

We analyzed baseline data from a randomized controlled trial, Project PROTECT, an NIH-funded clinical trial to promote dual contraceptive method use to prevent unplanned pregnancy and sexually transmitted diseases (STDs). A complete description of the methods of Project PROTECT have been published.² Briefly, women age 13 – 35 who were not interested in conceiving in the next twelve months were randomly assigned to two computer-based interventions to promote dual methods. This analysis is based on the baseline characteristics of participants and their reported contraceptive method prior to the intervention.

Women who reported no contraceptive use or single method use (OCs or male condoms) were included in this analysis (n=422). Because contraceptive use was common in the study population, we used multivariable robust Poisson regression to obtain relative risk estimates and 95% confidence intervals while controlling for sociodemographic characteristics and reproductive and sexual histories. All analyses were conducted using SAS.

Results

Of all patients in this analysis, 26% of women reported OC use, 34% reported condom use, and 41% reported no use of contraception at baseline.

After adjusting for potential confounders, women using OCs were less likely to have less than a high school education (RR=0.48, 95% CI: 0.30–0.75), to be African American (RR=0.60, 95% CI: 0.39–0.92) or Hispanic (RR=0.66, 95% CI: 0.44–0.44), and pay out of pocket for medical services (RR=0.65, 95% CI: 0.48–0.88) compared to women not using any form of contraception. Women using OCs were 52% less likely to have multiple partners during the previous month (RR=0.48, 95% CI: 0.24–0.96). Women using OCs differed compared to women using condoms in that they were less likely to pay out of pocket for medical services (RR=0.77, 95% CI: 0.67–0.90) and have multiple sexual partners during the previous month (RR=0.73, 95% CI: 0.60–0.89).

Having a new main sex partner was also associated with women’s contraceptive choices. Women with a new sex partner in the last six months were less likely to use OCs than condoms (RR=0.83, 95% CI: 0.72–0.95) but more likely to use condoms than no contraception (RR=1.26, 95% CI: 0.99–1.61). Other reproductive characteristics (parity, history of STD or unplanned pregnancy) were not associated with contraceptive choice.

Comment

A recent study by Santelli attributed the declining US adolescent pregnancy rates to improved contraceptive use, emphasizing the need to understand factors influencing the decision to use contraception.³ Our study sought to provide this understanding. We found a significant correlation between race and educational level among OC users and women using no contraception. We also noted a significant correlation between insurance coverage and reported contraceptive use. These finding were consistent with previous studies which have shown African American and Hispanic women are less likely to use oral contraceptive pills.^4,5

Our results suggest persistent barriers, such as lack of access, lack of knowledge about contraception, logistical or transportation issues, etc., may influence contraceptive choice for African American and Hispanic women and women without insurance. Despite the presence of organizations that provide discount contraception, there is still a significant population of women who are not benefiting from OCs, one of the more effective means of preventing unwanted pregnancy. Alternative strategies may be necessary in order to provide adequate access for all women and to eliminate the disparity for those without insurance.

The second important finding from this study pertains to sexual history demographics. Women with a new primary partner in the prior six months were more likely to report male condom use than no contraception; they were less likely to report OC use than either of these. Previous studies have demonstrated a similar correlation regarding sexual history and the use of barrier methods versus hormonal methods.⁶ Women with a new partner are not only at an increased risk for STDs, but also are at an increased risk for unwanted pregnancy yet these women were less likely to use OCs versus no contraception. One explanation is that women with a new sexual partner may be less likely to be in a long-term relationship and therefore do not perceive a need for daily contraception. Indeed, sixteen percent of all women having abortions in 2000–2001 became pregnant because they were not expecting to have sexual intercourse.⁷

These findings, along with those of prior studies, stress the importance of introducing dual contraception during counseling. Women engaging in high-risk behavior may not understand the elevated risk of unwanted pregnancy conferred by less effective methods of contraception such as male condoms. This study stresses the importance of obtaining a thorough sexual history in order to appropriately counsel women who may perceive their personal risk of unintended pregnancy to be low.

One of the strengths of this study is the ethnic and socioeconomic diversity of the study participants. Thus, our observations may reflect typical urban patient populations. Although the findings from this study reflect similar findings from previous studies, many of these studies are from data more than a decade ago. Given these findings, that ability to pay, race/ethnicity and level of education, impact contraceptive choice, studies should be completed to further disentangle whether institutional barriers, such as cost, logistical challenges and knowledge, can improve rates of effective contraception among women at risk for unintended pregnancy.