• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Abortion stigma in the United States: Quantitative and qualitative perspectives from women seeking an abortion

Dissertation
Author: Kristen M. Shellenberg
Abstract:
Objectives . Abortion stigma has been shown to be predictive of negative emotional reactions after an abortion. Limited research has been conducted on abortion stigma in the United States (US). The objectives of this dissertation were to use quantitative and qualitative methodologies to learn about stigma among abortion patients in the US. Methods . The quantitative data came from the Guttmacher Institute's 2008 Abortion Patient Survey (APS). The APS is a clinic-based survey that collects data on abortion patient characteristics. The sample consisted of 4,613 women. The proportion of abortion patients who perceive and/or internalize stigma was estimated. Stratified multivariate logistic regression was to estimate the association between women's characteristics and perceived and internalized stigma by race and ethnicity. For the qualitative component, 49 in-depth interviews were conducted with women at abortion clinics. Women were recruited from three different regions of the US. Analyses of the transcripts were conducted to assign meaning to themes and concepts, and to explore patterns of similarities and differences across interviews. Results . Two-thirds of abortion patients perceived stigma from other people. Less than half (40%) perceived stigma from friends and family, and 12% from their healthcare provider. Over half needed to keep the abortion a secret, and one-third cared about other people's opinions about abortion. The proportion of women reporting perceived and internalized stigma varied significantly by race/ethnicity. Region of residence, number of previous abortions, not having one's mind made up about the abortion at the time of making the appointment, and not having the man involved know about the abortion were found to be associated with stigma across the race/ethnicity groups. Qualitatively, many abortion patients perceived high levels of abortion stigma. Women perceived stigma from friends and family, people in their community and general society. Although most patients did not experience overt stigmatization, perceived stigma was sufficient to create negative feelings of self and need for secrecy about the abortion. Conclusion . Study results provide evidence of perceived and internalized stigma among abortion patients, as well as personal characteristics strongly associated with stigma. Future research should aim to understand the impact of abortion stigma on women's health and well-being.

TABLE OF CONTENTS DISSERTATION ABSTRACT ii COMMITTEE OF FINAL THESIS READERS iv ACKNOWLEDGEMENTS v TABLE OF CONTENTS viii LIST OF TABLES xi LIST OF FIGURES xiii DISSERTATION OVERVIEW xiv CHAPTER ONE 1 OVERVIEW OF ABORTION IN THE UNITED STATES 2 UNDERSTANDING SOCIAL STIGMA 11 ABORTION STIGMA 16 SIGNIFICANCE OF RESEARCH 22 REFERENCES 26 CHAPTER TWO 33 CONCEPTUAL FRAMEWORK 34 RESEARCH AIMS AND HYPOTHESES 36 REFERENCES 41 CHAPTER THREE 42 QUANTITATIVE COMPONENT 43 Study Design and Data Collection 43 Sample Selection 44 Analytic Sample 47 Variables 48 Weighting 50 Statistical Analyses 51 QUALITATIVE COMPONENT 56 Study Design and Sample 56 Instrument 58 Data Management and Analysis 58 ETHICAL CONSIDERATIONS 59 REFERENCES 71 Vlll

CHAPTER FOUR ,72 INTRODUCTION 73 METHODS 74 Data 74 Measures 76 Statistical Analyses 77 RESULTS 78 Survey non-response 78 Stigma item non-response 79 DISCUSSION AND IMPLICATIONS 84 REFERENCES 96 CHAPTER FIVE 97 INTRODUCTION 98 METHODS 101 Data 101 Outcome Measures 105 Independent Variables 106 Statistical Analyses 107 RESULTS 108 Sample Characteristics 108 Bivariate Analyses I l l Multivariate Analyses 112 DISCUSSION 122 Limitations 132 Implications for Future Research 134 Implications for Abortion-related care 136 REFERENCES 158 CHAPTER SIX 160 INTRODUCTION 161 METHODS 163 Instrument 164 Data Management and Analysis 165 RESULTS 166 Sample Characteristics 166 Perceived Stigma 166 Experienced Stigma 174 Impact of Stigma 177 DISCUSSION AND IMPLICATIONS 183 REFERENCES 189 I X

CHAPTER SEVEN 190 FINDINGS AND IMPLICATIONS 192 STRENGTHS AND LIMITATIONS 201 RECOMMENDATIONS 203 REFERENCES 209 APPENDICES APPENDIX A. 2008 ABORTION PATIENT SURVEY - MODULE B 210 APPENDIX B. NON-STRATIFIED QUANTITATIVE ANALYSES 218 APPENDIX C. IDI PARTICIPANT DEMOGRAPHIC SHEET 233 APPENDIX D. 2008 APS IDI GUIDE 235 APPENDIX E. CODING STRUCTURE FOR ANALYSIS OF IDIS 242 APPENDIX F. IN-DEPTH INTERVIEW INFORMED CONSENT FORM 247 APPENDIX G. ITEMS FOR A SCALE MEASURING ABORTION STIGMA 248 BIBLIOGRAPHY 250 CURRICULUM VITAE 258 x

LIST OF TABLES Table 1.1 Abortion laws and regulations in the United States 24 Table 3.1. Abortion stigma stems for 2008 APS 65 Table 3.2. Independent variables for quantitative analysis 66 Table 3.3. Imputed independent variables and percent missing data 68 Table 3.4. Percentage of missing data and "not applicable" for stigma items 69 Table 3.5. Coding of outcome variables 70 Table 4.1. Stigma items from 2008 APS 88 Table 4.2- Percent distribution of survey responders and non-responders on age, race/ethnicity, and health insurance status measures 88 Table 4.3. Percentage of missing data and "not applicable" for stigma items 89 Table 4.4. Pearson's chi-square for missing vs. not applicable responses on stigma items 90 Table 4.5. Pearson's chi-square for not answering vs. answering stigma items 91 Table 4.6. Adjusted odds ratios for factors associated with not answering stigma items.92 Table 4.7. Frequency and percentage of items unanswered by survey respondents 94 Table 4.8. Adjusted odds ratios for not answering 4 or more stigma items 95 Table 5.1. Coding of Outcome Variables 138 Table 5.2. Characteristics of study population by race and Hispanic ethnicity 139 Table 5.3. Percent of abortion patients reporting perceived and internalized stigma by race and Hispanic ethnicity 141 Table 5.4. Adjusted odds ratios for perceived stigma measures by race and Hispanic ethnicity 142 Table 5.5. Adjusted odds ratios for perceived stigma (general measure) by race and Hispanic ethnicity 145 xi

Table 5.6. Adjusted odds ratios for perceived stigma (from friends and family) by race and Hispanic ethnicity 147 Table 5.7. Adjusted odds ratios for perceived stigma (from health care providers) by race and Hispanic ethnicity 149 Table 5.8. Adjusted odds ratios for internalized stigma measures by race and Hispanic ethnicity 151 Table 5.9. Adjusted odds ratios for keeping abortion a secret by race and Hispanic ethnicity 154 Table 5.10. Adjusted odds ratios for caring about other people's opinions about abortion by race and Hispanic ethnicity 156 Table 6.1. Qualitative sample characteristics by site location 188 xn

LIST OF FIGURES Figure 1.1. The social process of abortion stigma 25 Figure 2.1. Conceptual framework for understanding abortion stigma 39 Figure 2.2. Analytical framework 40 Figure 3.1. 2008 APS Sampling steps 64 xin

DISSERTATION OVERVIEW Chapter 1 provides an overview of published literature relevant to the issue of abortion stigma in the United States (US). First, it provides an overview of abortion in the United States, including socio-demographics, short- and long-term safety, access issues, and laws regulating abortion. Second, it defines and describes the concept and potential negative impact of social stigma. Third, it describes and reviews research on abortion stigma in the US. And fourth, it discusses the significance of this research. Chapter 2 presents the conceptual and analytical frameworks that guided the research and the specific quantitative and qualitative aims and hypotheses tested. Chapter 3 describes the data sources and the methods used to address the quantitative and qualitative study aims. Chapter 4 reports the results of an empirical analysis of survey and item non-response among women recruited to participate in Guttmacher's 2008 Abortion Patient Survey (APS). Chapter 5 reports the results on the prevalence and correlates of perceived and internalized stigma among abortion patients in the US by race and Hispanic ethnicity. Chapter 6 presents the results of the qualitative component of the study exploring the perceptions, experiences and impact of stigma among abortion patients in the US. Chapter 7 draws conclusions about the quantitative and qualitative components included in this dissertation, review the strengths and limitations of the research, and discuss recommendations for future research and implications for abortion-related care. xiv

CHAPTER ONE BACKGROUND AND SIGNIFICANCE

OVERVIEW OF ABORTION IN THE UNITED STATES Abortion statistics Each year, more than six million women in the United States - one in every 10 women of reproductive age (15-44) - become pregnant, and almost half of those pregnancies are unintentional (Finer & Henshaw, 2006). Approximately 48% of unintended pregnancies end in abortion and as a result, two out of every 100 women aged 15-44 have an abortion every year (Finer & Henshaw, 2005). In 2005, 1.21 million pregnancies were terminated by abortion, making abortion one of the most common surgical procedures in the US (Jones et al, 2008a). The abortion rate, defined as the number of abortions per 1,000 women of reproductive age in a given year, was 19.4 in 2005 which represents a 9% decline over five years and was the lowest rate since 1974. The abortion ratio, defined as the number of abortions per 1000 live births indicates that 22% of pregnancies (excluding those ending in miscarriages) ended in abortion in 2005 (Jones et al., 2008a). Almost 90% of abortions are performed in the first trimester of pregnancy and six in ten abortions are performed within eight weeks gestation (Strauss et al, 2007).* The proportion of abortions performed very early in pregnancy (at six weeks or before) increased from 14% in 1992 to 28% in 2004, most likely due to increased access and use of medication abortion. In 2005, early medication abortion accounted for 13% of all abortions, an increase from 6% in 2001 (Jones et al, 2008a). The distribution of abortion across the US population is not equal. The largest group having abortions are non-Hispanic White women (36%), and Black and Hispanic women together make up more than half (55%) of women having abortions (Jones, 2010). This * All of the Strauss, Gamble et al (2007) data has been adjusted by Stanley K. Henshaw using data from the Guttmacher Institute. 2

proportion is greater than their proportion in the population because they have a higher rate of unintended pregnancy, and, among Black women, because they are more likely to resolve an unintended pregnancy through abortion (Guttmacher Institute, 2009). Non-White women have much greater rates of abortion than White women, with Black women having two times the national average and three times the rate of White women; Hispanic women have two times the rate of White women (Strauss et al, 2007). Poverty plays a large role in who has an unintended pregnancy, with unintended pregnancies and abortions becoming increasingly concentrated among poor women. Between 1996 and 2000, while abortion rates for all other groups fell, abortion rates among poor and low-income women increased. Women below the federal poverty level have abortion rates almost four times those of higher-income women (Jones et al, 2002). Many people think of unintended pregnancy and abortion as issues associated with young women or teenagers, but the majority of women (56%) having abortions are in their 20's; only 16% of women are age 15-19 and 26% are between the ages of 30 and 44 (Strauss et al, 2007). Women between the ages of 20-29 do have the highest abortion rates, followed by 15-19 year olds and then 30-34 year olds (Strauss et al, 2007). Abortion numbers and rates decline with age because fecundity declines, use of contraceptive sterilization increases and more women are married (which makes it easier to use contraceptives effectively and to continue an unintended pregnancy if it occurs) (Guttmacher Institute, 2009). Approximately 86% of abortions occur among unmarried women (this includes never married and formerly married) (Strauss et al, 2007). The proportion of unintended pregnancies terminated by abortion ranges from 67% among formerly married women and 57% among never-married women to 27% among currently married women (Finer & Henshaw, 2006). A commonly 3

held myth is that women have abortions before they begin childbearing but actually 6 in 10 women seeking an abortion are already mothers; approximately half of women seeking an abortion have already had at least one prior abortion (Jones et al., 2002). Public opinion polls show that many Americans are tolerant of one abortion, but do not approve of women having more than one abortion or "using it as birth control." When U.S. women having abortions are asked their religious affiliation, 37% say they are Protestant. This is a lower proportion than the approximately 50% of women 18-44 in the U.S. population who identify themselves as Protestant, which means that their abortion rate is lower than that of all women; the proportion of abortion patients who are Catholic (27%) is slightly higher than the Catholic proportion of the population (Guttmacher Institute, 2010). Thus, the abortion rate of Catholics (abortions per 1,000 female Catholics aged 18-44 in the population) is slightly higher than that of all women. Approximately 13% of abortion patients identify as "born again or evangelical Christians," which confirms that people's personal or religious beliefs about abortion do not necessarily dictate their own behavior. Women's reasons for terminating a pregnancy vary although over 70% of women report terminating a pregnancy out of concern for or responsibility to other people (e.g. they already have children to take care of) or because they cannot afford a baby at that time (Jones et al., 2008b); other common reasons are related to an interruption of school or career, not wanting to be a single mother, or having already completed childbearing (Finer et al, 2005). Safety of abortion Short-term safety Abortion is one of the safest surgical procedures for women when performed by a trained medical professional and under sanitary conditions. The risk of death associated with 4

safe abortion is low—approximately 0.6 deaths per 100,000 abortions—and the risk of major complications is less than 1% (Grimes, 2006). An abortion is safer the earlier in pregnancy it is performed; when an abortion is performed under 12 weeks gestation, the risk of death is 0.4 per 100,000 abortions but that number dramatically increases to 8.9 deaths per 100,000 abortions when an abortion is performed at 21 weeks gestation or beyond (Bartlett et ah, 2004). Fortunately, fewer than 2% of abortions are performed after 20 weeks. The most common causes of abortion-related death are infection and hemorrhage, accounting for over 50% of deaths (Bartlett et ah, 2004). On average, eight women each year die in the US from induced abortion, compared with about 280 who die from pregnancy and childbirth, excluding abortion and ectopic pregnancy. The risk of death when a pregnancy is continued to birth is about 12 times as great as the risk of death from induced abortion; however, this calculation of mortality from childbirth omits deaths from miscarriage and ectopic pregnancy (Grimes, 2006). Long- term safety For 35 years, researchers have investigated the long-term effects of voluntarily terminating a pregnancy, and the preponderance of evidence indicates that abortion is safe over the long term and carries little or no risk of fertility-related problems, cancer or psychological illnesses (Boonstra et al., 2006). Studies indicate that vacuum aspiration - the method most commonly used during first-trimester abortions - poses virtually no long-term risks of future fertility-related problems, such as infertility, ectopic pregnancy, spontaneous abortion or congenital malformation (Hogue et al, 1982; Atrash & Hogue, 1990; Hogue et al, 1999). Several well-designed and methodologically sound studies have found no association between women's abortion histories and future pregnancy outcomes (Frank et al, 5

1991; Chen et al., 2004; Yimin et al., 2004). In 1996, a meta-analysis of several studies suggested that there was a significant positive association between abortion and breast cancer, but since then several reviews by experts have concluded that there is in fact no association between the two (Beral et al, 2004; Boonstra et al, 2006). In 2003, the National Institute's of Health declared with certainty that the evidence shows that "induced abortion is not associated with an increase in breast cancer risk." One of the most contested and controversial abortion-related issues is the association with women's mental health. Opponents of abortion have claimed that abortion is bad for women's mental health and leads to negative psychological outcomes. In 1987, at the request of President Ronald Reagan, Surgeon General Everett C. Koop reviewed the evidence linking abortion to negative mental health outcomes and stated that he could not come to a conclusion because of the serious methodological flaws of the studies, but that he perceived psychological problems related to abortion to be "miniscule from a public health perspective" (Koop, 1989). Since that time numerous studies have attempted to link abortion to a range of conditions, including psychiatric treatment, depression, anxiety, substance abuse and death (Coleman et al, 2002; Reardon & Cougle, 2002; Reardon et al, 2002; Cougle et al, 2003; Reardon et al, 2003; Cougle et al, 2005); many of these studies, however, have serious methodological flaws (e.g. failure to control for important confounders such as preexisting psychological and contextual factors) that make it impossible to infer any type of causal relationship (Boonstra et al, 2006). Several research studies and reviews have found no significant association between abortion and women's long term mental health (Rogers et al, 1989; Adler et al, 1990; Bradshaw & Slade, 2003; Thorp et al, 2005; Charles et al, 2008) but the most conclusive evidence comes from a large scale (13,000 women), prospective 6

cohort study in the UK which compared women who terminated a pregnancy to women who carried the pregnancy to term. The findings indicate no difference in the psychiatric outcomes between the two groups of women; in fact, women who carried the pregnancy to term had slightly higher odds of a psychiatric episode (Gilchrist et al, 1995). In 2008, the American Psychological Association's (APA) Task Force on Mental Health and Abortion released a report reviewing published literature on abortion and mental health in the US. The APA concluded that women who have an abortion are at no greater risk of mental health problems than women who deliver, and that mental health prior to an abortion is most likely the strongest predictor of post-abortion mental health. Additionally, the task force identified several factors that may be predictive of psychological responses after an abortion - perceived stigma and need for secrecy about the abortion were two of the factors (Major et al, 2009). Adjustment after abortion The decision to have an abortion is often a difficult and complex one, and one that occurs in the context of experiencing an unintended pregnancy, which is often considered to be a stressful life event. While women may experience distress (anxiety and situational depression) prior to having an abortion (Henshaw et al, 1994; Slade et al., 1998), many women report feelings of relief or happiness after the abortion (Major et al, 1990; Adler et al, 1990; Miller et al, 1998; Major et al, 2000). Some women do report experiencing feelings of anxiety, guilt, shame or sadness associated with the abortion (Major et al, 1990; Adler et al, 1990; Major et al, 2000); however, feelings of anxiety and depression tend to decrease dramatically in the first few weeks after the procedure (Urquhart & Templeton, 1991; Zolese et al, 1991; Henshaw et al, 1994; Slade et al, 1998; Lauzon et al, 2000). 7

What is not well explained in the literature is whether lingering feelings of guilt and shame are caused by a woman's personal conflict about the abortion or by her concerns about how others view her decision (i.e. feeling stigmatized by the abortion). In a study of women's emotional health 1-month post-abortion, results indicated that the well-being of women was "relatively high" (Major et al, 1997). Sociological research indicates that perceived social support is directly related to feelings of well-being and may reduce or buffer adverse consequences of a stressful life event (Major et al, 1990). Perceived social support can be defined as information or actions (real or potential) leading individuals to believe they are cared for, valued, or in a position to receive help when they need it. Several studies have found a positive relationship between perceived social support (from partner, family and friends) and post-abortion well-being; women who perceive their family and friends to be supportive of their decision consistently rate higher on measures of well- being than women who perceive their friends and family to be less than supportive (Major et al, 1990; Major et a/.,"1997; Cozzarelli et al, 1998). Perceived social support may be particularly important for successful adjustment to abortion because of the strong moral sanctions against abortion in this culture. Family and friends attitudes about abortion (as well as those of the general public) have the potential to influence a woman's perceptions of stigmatization, which may influence disclosure about abortion and access to social support, which is known to be protective against psychological distress after an abortion. Access to abortion services In 2005, there were an estimated 1,787 abortion providers in the United States. Abortion clinics - defined as facilities where half or more of patient visits are there for abortion services - provide 71% of all abortions. Other clinics, including some group 8

practices and clinics that provide an array of reproductive health services, provide most of the rest (Jones et ah, 2008a). Only about 5% of abortions are performed in hospitals and about 2% are performed by physicians who provide fewer than 400 abortions per year (Guttmacher Institute, 2009). Although the number of abortion providers in the US is steadily decreasing, with the provision of medication abortion by non-specialized clinics and physicians' offices, the decline between 2000 and 2005 was only 2% instead of 8% (Jones et ah, 2008a). The decline in providers is often attributed to fear (or experiences) of harassment; non-hospital providers report experiencing harassment in the form of picketing (80%), picketing with physical contact with patients (28%), vandalism of the clinic (18%), picketing of providers homes (14%) and bomb threats (15%) (Henshaw & Finer, 2003). This type of harassment, aggressive anti-abortion picketing at clinics in particular, has been found to emotionally upset many abortion patients and be associated with postabortion distress (Cozzarellie/aZ.,2000). Despite the raw numbers of independent abortion providers, 87% of U.S. counties had no abortion provider in 2005; in non-metropolitan areas, 97% of counties had no provider (Jones et al., 2008a). As a result, many women must travel substantial distances to access abortion services. About one in four women who have an abortion travel 50 miles or more for the procedure, a significant distance and a documented barrier to timely care (Henshaw & Finer, 2003). The proportion of women living in counties without a facility that provides even one abortion a year has steadily increased since 1978 and reached 35% in 2005 (Guttmacher Institute, 2009). The proportion of women in "unserved" or underserved counties would be much higher if not for the efforts of nonprofit organizations to establish and maintain clinics in areas without other providers. 9

Legislation and regulation of abortion Abortion was legal at the time the US was formed, but by the early 1900's it was illegal in all states. In 1973, the Supreme Court cases of Roe v. Wade and Doe v Bolton declared abortion to be a "fundamental right" that is rooted in a woman's constitutional right to privacy. The Supreme Court also recognized the State's interest in potential life, and in the cases of Webster v. Reproductive Health Services (1989) and Planned Parenthood v. Casey (1992) the court modified the original trimester framework established by Roe and gave states the right to regulate abortion as long as it did not put an "undue burden" on the woman. Since that time, states have constructed a lattice work of abortion law, codifying, regulating and severely limiting whether, when and under what circumstances a woman may obtain an abortion (Guttmacher Institute, 2008). In 2000, the Supreme Court struck down a Nebraska ban on so-called "partial-birth abortions" which made 2nd trimester abortions illegal without providing exceptions to protect the life of the mother. Three years later, the US Congress passed the "Partial Birth Abortion Ban" which lead to the case of Gonzales v. Carhart (2007) in which the Supreme Court upheld the Partial Birth Abortion Ban of 2003 - effectively making it illegal to perform an "intact dilatation and extraction" even though medical professionals maintain that it is one of the safest methods for performing a late term abortion. Table 1.1 presents the types of laws regulating and limited access to abortion in the US. Some of these laws have been passed and implemented based on incorrect information and research results from methodologically weak studies. Despite the legal status and high prevalence of abortion in the US, abortion is often portrayed as an abnormal and uncommon event, one that needs excessive restrictions and regulations. This creates a hostile environment that fosters judgment and criticism about 10

Full document contains 278 pages
Abstract: Objectives . Abortion stigma has been shown to be predictive of negative emotional reactions after an abortion. Limited research has been conducted on abortion stigma in the United States (US). The objectives of this dissertation were to use quantitative and qualitative methodologies to learn about stigma among abortion patients in the US. Methods . The quantitative data came from the Guttmacher Institute's 2008 Abortion Patient Survey (APS). The APS is a clinic-based survey that collects data on abortion patient characteristics. The sample consisted of 4,613 women. The proportion of abortion patients who perceive and/or internalize stigma was estimated. Stratified multivariate logistic regression was to estimate the association between women's characteristics and perceived and internalized stigma by race and ethnicity. For the qualitative component, 49 in-depth interviews were conducted with women at abortion clinics. Women were recruited from three different regions of the US. Analyses of the transcripts were conducted to assign meaning to themes and concepts, and to explore patterns of similarities and differences across interviews. Results . Two-thirds of abortion patients perceived stigma from other people. Less than half (40%) perceived stigma from friends and family, and 12% from their healthcare provider. Over half needed to keep the abortion a secret, and one-third cared about other people's opinions about abortion. The proportion of women reporting perceived and internalized stigma varied significantly by race/ethnicity. Region of residence, number of previous abortions, not having one's mind made up about the abortion at the time of making the appointment, and not having the man involved know about the abortion were found to be associated with stigma across the race/ethnicity groups. Qualitatively, many abortion patients perceived high levels of abortion stigma. Women perceived stigma from friends and family, people in their community and general society. Although most patients did not experience overt stigmatization, perceived stigma was sufficient to create negative feelings of self and need for secrecy about the abortion. Conclusion . Study results provide evidence of perceived and internalized stigma among abortion patients, as well as personal characteristics strongly associated with stigma. Future research should aim to understand the impact of abortion stigma on women's health and well-being.