MCHB/EPI Miami Conference — December 7 - 9, 2005

Unplanned Parenthood: Why I Choose Not to Use — Transcript

CORRINE WILLIAMS: Thank you very much. My talk is Contraceptive Use and Intimate Partner Violence. As was just stated contraceptive use is highly prevalent in the United States with 62 percent of women having used contraception in the last month, prior to their interview. This is data from the National Survey of Family Growth. But as was stated, there is still a significant portion at risk for an unwanted pregnancy. So my area of interest is in violence and so here's some background on intimate partner violence in the United States .

According to a survey that was done in 1994, approximately 1.5 million females are raped and/or physically assaulted by an intimate partner each year and approximately 25 percent of women in the United States report a lifetime history of partner abuse. One study found that women with an unwanted pregnancy were 4 times more likely to have experienced physical violence by a partner and domestic violence has be hypothesized as a factor that may be associated with contraceptive, noncompliance in contraceptive use.

So the research questions for this paper are: Do abused women utilize different forms of birth control then non-abused women, and I'm particularly interested in those reporting no use of contraception. And also do abused women report more discrepancy between their preferred method of contraceptive and what they're actually using, which again may be related to compliance and continuation of method.

This was sort of a two-tiered study. We administered written surveys to approximately 2,500 women in 8 hospitals around the Boston area in multiple different types, so pediatrics, OB/GYN, primary care and then from the results of this screening survey we selected women to participate in an in person interview. The screening survey had 10 items on violence, 6 that asked about violence in the last year and 4 that asked about lifetime history of violence. To be eligible for the follow-up study they had to have had a male intimate partner within the last 12 months and for what we call our index group, those who were abused, they had to have had intimate partner violence in the past year. And we said that either they had to have said yes to one of the six items in the last 12 months or with the lifetime items we asked for a date of the most recent event and if that event happened in the last year they were also counted as abused. The control women had to have said no to all 10 items. So we wanted them to have had no abuse in the last year or have a lifetime history.

Because of the low number of women that were able to get through the survey we recruited an additional 73 women through posters that were placed around the community, in the hospitals that we screened at, and in other locations to supplement that group. So our final sample was 303 women, 162 of whom were abused. The interviews took place in our office and lasted about two hours on average. We provided transportation, compensation for their time and a snack or lunch while they were in our office. And the interview asked several questions about several topics: demographics, childhood information, experiences with violence, current health status, social support, et cetera.

In terms of the measurement of contraception, women were provided with a list of several methods and asked to check the ones that they had used in the past 12 months. They were asked separately about tubal ligation and hysterectomy. And then there was one item that asked about their preferred method of birth control. And because of the way the question was asked they were only allowed to pick their one favorite, and so contraceptive discrepancy is defined as a woman not using the method that she preferred.

In terms of our sample, what the analysis done here, looking actually at 233 women who were less than 50 years old and had not had a tubal ligation or hysterectomy, the mean age, all of these variables are different between the index and control groups, those women who were abused are slightly older, about three years older. There were a few of them who are not U.S. born. We also see differences in race and education. So that women who were abused are the number on the right, so 35 percent of them were white, 49 percent were black and 15.8 percent were other.

And it's the same pattern for the education. Women who were abused, 24 percent of them had less than a high school education, 46 had greater at a high school diploma and 29.2 had graduated from college or had gone to school beyond that.

These are the 3 most common methods used in the sample. And we can see a difference so that women who were abused are much less likely to be using oral contraceptives. They were more likely however to be using condoms and there were no differences in their use of Depo Provera. These 3 methods accounted for 90 percent of the methods used, 90 percent of the women used one of these methods. And in looking at those who reported that they had not used any form of birth control in the last 12 months we see that among the control approximately 12 percent of our sample was not using any form of birth control, whereas approximately a quarter of the sample of those who were abused were not using any form of birth control.

And then when you look at contraceptive discrepancy, about a quarter of the sample overall were not using the method of contraception they preferred. And it's approximately 17 percent among those women experiencing no violence and almost 30 percent among those who are abused. And this has a p-value of 0.08.

When I ran adjusted logistic regression models looking at these two variables, there were no differences by violence status when I controlled for age, education, and race, I put these up here though because it's a small sample size there's some issues with power, but we can see that in both cases women who are experiencing violence the odds ratio is slightly elevated. So they are more likely though not statistically significant.

And the only variable that did come out significant was age, with older women being less likely to be using any contraception.

One of the strengths of the study, it's a new research area. When I was conducting my literature review I couldn't find any real quantitative data that looked at the methods of birth control used by abuse status and thinking about how this might be related to abuse status.

There are limitations with this study. This study was not designed to look at this question, since it was a small size and there may be, it may be a biased sample.

So in conclusion, these results do appear to show a trend in the relationship between contraceptive use and violence. And I think there may be other ways of looking at contraception in relation to violence. I took this just as a very, sort of preliminary look. But I think looking at, you know, methods of birth control, that there might be potential for partner interference with, and also looking at violence status, one of the analyses that I ran, but didn't present here was looking at how those women in the abuse category, it might break down by those experiencing only physical violence and those who also experience emotional violence. And those who experience emotional violence and thinking about that as a marker for more controlling behaviors in the relationship were more likely to report not using any contraception and also to report a higher contraceptive discrepancy. So I think this relationship should be explored in other data sets to get a handle on what's going on here.

I just wanted to acknowledge my dissertation chair Laura McKlesky who is currently at the University of Pennsylvania and the study was funded by the Agency for Healthcare Research and Quality. Thank you.