Safety beliefs, [mis]information sources & contraceptive behavior among young people who can get pregnant in the U.S.: A theory-driven, mixed methods approach Restricted; Files Only

Riley, Halley (Fall 2021)

Permanent URL:


Approximately 45% of pregnancies in the U.S. are unintended, meaning that are unwanted or mistimed. Among 15-19-year-olds, 75% are unintended.1-4 Young people, people of color, those with low socioeconomic status, and people in certain regions, such as the Southeast (SE), are at heightened risk of unintended pregnancy (UIP).2, 4-9 People experiencing UIP and subsequent births are at risk of adverse health and socioeconomic outcomes, especially adolescents.10-12 Without contraceptive use, as many as 85% of people who have penile-vaginal sex will become pregnant in one year. With effective contraception, this is as low as 0.05%.13 Methods vary in effectiveness, side effects, and ease of use.14, 15 They are often characterized by method effectiveness tiers by groups like the U.S. Centers for Disease Control and Prevention.16-18 Yet, other factors, including safety, are important to users.19, 20 Clinical research and pharmacovigilance suggest that contraceptive-related serious adverse events (SAEs) are extremely rare,17 but research also shows consistent concerns about safety.17, 21-42 Gaps in the literature surround how beliefs develop and translate to behavior, and data from the SE is lacking. This three-aim mixed methods dissertation sought to address these gaps.


First, I systematically reviewed published literature on contraceptive safety beliefs to characterize this literature and determine which beliefs were prevalent. Across 48 studies, concerns about infertility, adverse pregnancy and fetal outcomes, problems with long-acting reversible contraceptive (LARC) devices, menstrual changes, pain, and weight changes were common. Misconceptions about SAEs appear throughout the literature, which consists largely of low-quality, cross-sectional studies.


In the second study, I used multinomial logistic regression to assess relationships safety and side effect concerns and method use among young people who can get pregnant in Atlanta, GA (n=148). In multivariable models using baseline data, concerns about side effects, pain, bleeding, SAEs and any AE were associated with increased odds of less effective or no method use compared with LARC use. Magnitude and statistical significance of relationships differed by concerns and method use operationalization (ever vs. current use). In longitudinal models, all concerns except bleeding were associated with less effective method use.


Finally, I conducted in-depth-interviews with 29 people who could get pregnant aged 15-24 in GA. Interviews included questions about information sources, interpersonal and social influences, and contraceptive decision-making. We used a grounded theory approach to analyze data. We found three distinct pathways. One group was exposed to negative information about safety and avoided prescription methods. Another enjoyed non-contraceptive and pregnancy preventive benefits of oral contraceptive pills and avoided other methods. A third cycled through methods, using information from peers, providers, and the Internet to make decisions.


Through this research, I characterized common safety concerns, determined how concerns impact behavior among a cohort of young people, and dove deeper into drivers of behavior. Findings highlight the need for better education and counseling to correct misconceptions and empower people to make evidence-based contraceptive decisions.

Table of Contents

Table of Contents

Chapter 1: Introductory Literature Review. 1

Introduction. 1

Unintended pregnancy in the U.S. 1

UIP among adolescents and young people who can get pregnant in the U.S.: 2

Contraceptive use as primary prevention strategy for adolescent UIP in the U.S. 4

Contraceptive use in the U.S. 5

Safety of Contraceptive Methods available in the U.S. 6

Concerns about side effects and AEs as a barrier to contraceptive use. 14

Misinformation about contraceptive safety. 16

The nocebo effect and contraceptive use. 18

Sources of [mis]information about contraceptive safety. 19

Reproductive injustice in the U.S. 24

Gaps in the Research addressed by this dissertation research. 26

Overemphasis on LARC methods in the current literature. 28

Lack of clear theoretical underpinnings of prior research. 28

A dearth of research on young people who can get pregnant in the Southeastern U.S. 29

Theoretical Framework. 29

Health Belief Model 30

The. 30

Social Ecological Model 35

Table 1.1 SAEs and Side Effects Listed in Package Labels of FDA-Approved Contraceptive Methods. 38

Table 1.2. Operationalization of HBM Constructs in Contraceptive Use Literature. 39

Table 1.3. Multi-level influences on contraceptive use. 40

Figure 1.1 Conceptual Model. 43

Chapter 2: Contraceptive Safety Beliefs among People who can get Pregnant in the U.S.: A Systematic Review. 44

Abstract. 44

Introduction. 45

Methods. 49

Results. 51

Overview of included articles: 51

Beliefs about SAEs. 52

Beliefs about side effects. 58

Beliefs about non-contraceptive benefits associated with method use. 64

Discussion. 67

Limitations and Strengths. 69

Conclusion. 70

Figure 2.1. PRISMA 2020 flow diagram for new systematic reviews. 71

Table 2.1. Outcome categorization. 73

Table 2.2. Characteristics of included studies (n= 47). 74

Table 2.3. Studies reporting SAEs (n=30). 100

Table 2.4. Studies reporting side effects (n=27). 105

Table 2.5. Studies reporting beliefs about non-contraceptive benefits (n=20). 110

Figure 2.2. Bar Chart of proportions reporting concerns about infertility. 113

Figure 2.3. Bar Chart of proportions reporting concerns about adverse pregnancy and fetal outcomes. 113

Figure 2.4. Bar Chart of proportions reporting concerns about adverse pregnancy and fetal outcomes. 114

Figure 2.5. Bar Chart of proportions reporting concerns about cancer and blood clots. 114

Figure 2.6. Bar Chart of proportions reporting beliefs about Pain. 115

Figure 2.7. Bar Chart of proportions reporting beliefs about sexual side effects. 116

Figure 2.8. Bar Chart of proportions reporting beliefs about protective effects of methods other than condoms or abstinence on STIs. 116

Figure 2.9. Bar Chart of proportions reporting beliefs about protective effects of OCPs on cancer and benign disease. 117

Figure 2.10. Forest Plot Depicting Relationship Between Race/Ethnicity and Odds of Beliefs. 118

Figure 2.11. Forest Plot Depicting Relationship Between Age and Odds of Beliefs. 119

Appendix 2.A. PRISMA 2020 Checklist. 120

Appendix 2.B. Search Terms. 124

References. 125

Chapter 3: Safety concerns and contraceptive method use among young people who can get pregnant in Atlanta, GA. 132

Abstract. 132

Introduction. 134

Methods. 137

Study Overview.. 137

Measures. 138

Analysis. 141

Results. 143

Baseline Univariate Results: 143

Baseline Bivariate Results. 144

Moderation Assessment 146

Baseline Multivariable Results. 147

Longitudinal Results of Baseline Contraceptive Concerns and Time-varying Method Use. 148

Discussion. 149

Conclusion. 153

Table 3.1. Demographics and reproductive history of YWSS participants who reported ever having sex with a male partner by baseline, n=148. 155

Table 3.2. Bivariate Relationships between Contraceptive Method Use (Dependent Variables) and independent variables, moderators, and covariates, n=148. 157

Table 3.3. Bivariate Relationships between Contraceptive Safety Concerns (Independent Variables) and dependent variables, moderators, and covariates, n=148. 160

Table 3.4 Multivariable Relationships between contraceptive safety concerns and method use, baseline cross-sectional analyses, n=148. 163

Figure 3.1. Forest Plots depicting odds ratios and confidence intervals for models using most effective method ever used as independent variable. 165

Figure 3.2. Forest Plots depicting odds ratios and confidence intervals for models using most effective method currently used as independent variable. 166

Figure 3.3. “Spaghetti” Plot depicting contraceptive use patterns over 11 months, n=78. 167

Table 3.5. Multivariable Relationships between contraceptive safety concerns and monthly method use over 12-months. 168

Appendix 3.A. Goodness of fit statistics for Baseline Multivariable Models. 169

Appendix 3.B. Moderation Assessment, Baseline cross-sectional analyses. 172

Appendix 3.C. Multivariable Relationships between contraceptive safety concerns and method use among Black-identifying participants, baseline cross-sectional analyses, n=60. 174

References. 176

Chapter 4: “I'm not necessarily sure if it's safe”: Perspectives on Contraceptive Safety among young people who can get pregnant in Georgia. 187

Abstract. 187

Introduction. 189

Methods. 192

Sampling and Recruitment 192

Data Collection. 193

Analysis. 195

Results. 196

Contraceptive Use Pathways. 196

Discussion. 208

Interpersonal Implications: 209

Social Implications: 209

Health Systems Implications: 210

Policy Implications: 211

Limitations and Strengths. 212

Conclusion. 213

Figure 4.1. Contraceptive Use Pathways. 215

Table 4.1 State Laws on Sexuality Education in the U.S. Southeast. 216

Appendix 4.A. Interview Guide. 217

References. 222

Chapter 5. Conclusion.. 225

Overview.. 225

Summary of findings. 226

Strengths. 231

Limitations. 232

Implications for research. 233

Implications for Policy and Practice. 234

Dissertation References. 237

About this Dissertation

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified Preview image embargoed

Primary PDF

Supplemental Files