Differences in Contraceptive Method Discontinuation and Contraceptive Method Preferences by Disability Status

We used data from the National Survey of Family Growth (NSFG), a nationally representative, in-person, household-based survey of non-institutionalized U.S. adults aged 15–49 years old ( National Center for Health Statistics, 2022 ). 1 The NSFG asks about a range of sexual and reproductive health topics. The NSFG oversamples Black, Hispanic, and adolescent respondents. Interviews were conducted in respondents’ homes by highly trained interviewers. The survey was administered by the National Center for Health Statistics (NCHS) continuously from 2011 to 2019, with data releases of 2-year survey waves every 2 years ( National Center for Health Statistics, 2020 ). We limited our sample to respondents who received the female respondent questionnaire. We pooled data from 2011 to 2019 to create a cross-sectional sample with an adequate number of disabled respondents for analysis. This resulted in a final unweighted analytic sample of 22,995 respondents. 2 All measures used in these analyses were self-reported and collected during the face-to-face portion of the NSFG interview ( Appendix A ). The primary exposure of interest was disability status. The NSFG implemented a standardized set of disability items endorsed by the U.S. Department of Health and Human Services ( CDC, 2019 ). Disability status was measured via the following six questions, with yes/no responses: “(1) Do you have serious difficulty hearing? (2) Do you have serious difficulty seeing, even when wearing glasses or contact lenses? (3) Because of a physical, mental, or emotional condition, do you have serious difficulty concentrating, remembering or making decisions? (4) Do you have serious difficulty walking or climbing stairs? (5) Do you have difficulty dressing or bathing? and (6) Because of a physical, mental, or emotional condition, do you have difficulty doing errands alone such as visiting a doctor’s office or shopping? ” We created a binary measure of any disability and a categorical measure (0, 1, or 2+ disabilities) and examined the responses to each of the six disability questions separately. We examined two outcomes related to satisfaction with contraceptive methods. The NSFG measures method discontinuation due to dissatisfaction with a single question asking whether the respondent has ever stopped a contraceptive method because they were dissatisfied with the method: “ Some people try a method and then don’t use it again, or stop using it, because they are not satisfied with the method. Did you ever stop using a method because you were not satisfied with it in some way?” A second measure of contraceptive method satisfaction was an unfulfilled contraceptive preference due to cost; current contraceptive users were asked, “ If you did not have to worry about cost and could use any type of contraceptive method available, would you want to use a different method? ” and nonusers were asked, “ If you did not have to worry about cost and could use any type of contraceptive method available, would you want to use a method? ” We combined these responses into a single dichotomous measure; respondents were not asked further about their preferred method or if factors other than cost were a barrier to using it. First, we examined differences in the distribution of disability for key sociodemographic, sexual activity, and contraceptive use variables. These variables were age (15–19, 20–24, 25–29, 30–34, 35–39, 40+ years), race and ethnicity (Hispanic, Non-Hispanic white, Non-Hispanic Black, Non-Hispanic other or multiple races), current health insurance status (Private; Medicaid or other state-sponsored plan; Medicare, military, or some other federal sponsored plan; No health insurance), sexual activity (oral, anal, or penile-vaginal sex) with an opposite sex partner in the past 12 months, sexual activity with a same sex partner in the past 12 months, desire for a baby/child sometime in the future, and the respondent’s fecundity status (the ability of the respondent to achieve pregnancy and carry a pregnancy to term) at the time of interview (Fecund; Surgically sterile, contraceptive; Surgically sterile, non-contraceptive; Sterile, non-surgical; Subfecund; Long interval (infertile 36+ months)). The NSFG creates the fecundity status variable from a series of questions about the respondent’s current pregnancy status, whether they or their partner received a sterilization procedure, difficulty with conceiving or delivering a baby, contraception use, and sexual activity. For the self-reported race and ethnicity variable, we used the recoded variable created by the NSFG to match the 1997 Office of Management and Budget guidance about reporting of racial and ethnic categories ( National Center for Health Statistics, n.d. ). We also examined the distribution and descriptive statistics for contraceptive use in the past 12 months, grouped by effectiveness (Highly effective, Moderately effective, Less effective, and No method). Second, to create the adjusted models, we selected individual-level covariates that prior research suggests are associated with contraceptive use and experiences of contraceptive satisfaction. These included age, race and ethnicity, current health insurance status, sexual activity with opposite sex partner, and fecundity. Additionally, we adjusted the models for survey wave (2011–2013, 2013–2015, 2015–2017, and 2017–2019) to account for potential changes over time. All data management, cleaning, and analysis were performed in Stata version 18.0 ( StataCorp, 2023 ). First, we calculated the prevalence of disability in our sample. Second, we calculated descriptive statistics for the covariates of interest by disability status. Third, we estimated unadjusted and adjusted logistic regression models of the association between the disability measures and each outcome. For each outcome, we adjusted for survey wave, sexual activity in the past 12 months, fecundity, and demographics hypothesized to affect contraceptive method use (age, race/ethnicity, and insurance status). All model results are presented with odds ratios, 95% confidence intervals, and predicted probabilities. We assessed statistical significance at an alpha level of 0.05. We accounted for the complex survey design of the NSFG using the svy command in Stata. We employed appropriate 4-year and 8-year sampling weights, depending on the years in which the outcome of interest was assessed. The method dissatisfaction question was asked of all respondents in the survey waves from 2011 to 2019; evaluating cost as a barrier to preferred contraceptive method use was included in the survey rounds from 2015 to 2019. Institutional Review Board (IRB) review was not required for this project because the NSFG data are publicly available on the NCHS website, and the NCHS’s IRB has approved the dissemination of the public-use dataset.

About one in five (19.5%) reproductive-age women reported a disability in the NSFG from 2011–2019. 12.8% reported one disability, and 6.7% reported multiple disabilities; serious difficulty concentrating, remembering, or making decisions was the most common (13.4%) ( Figure 1 ). We identified demographic differences between disabled and non-disabled individuals in the sample ( Table 1 ). There were significant differences by age but not race/ethnicity (age: p<0.001, race/ethnicity: p= 0.96). Disabled individuals were more likely to have Medicaid or other state-sponsored health plans as insurance (31.4%) and less likely to have private insurance (43.1%) compared to non-disabled individuals (16.7% Medicaid, 64.4% private insurance, p=0.01). Over a quarter of the disabled sample (27.1%) were in the 2017–2019 survey round, compared to 22.2% in 2011–2013 (p <.001). Recent sexual activity with an opposite-sex partner was less common among disabled than non-disabled individuals (87.8% vs. 92.4%, p<0.001), but recent sexual activity with a same-sex partner was equal between the two groups (31.9%, p=1.00). The two groups of respondents had different patterns of contraceptive method mix, with disabled respondents having greater use of highly effective methods (including sterilization, IUDs, and implants) (35.5% vs. 32.4%, respectively) and greater prevalence of no method use (21.7% vs. 20.0%, respectively, p=0.003). Disabled individuals had substantially lower overall reports of fecundity than non-disabled people (71.1% vs. 58.8%, p <.001), with higher reports of surgical sterilization for contraceptive reasons (19.7%) and higher reports of subfecunity (14.6%), as defined by the NSFG, than non-disabled individuals (17.6% and 8.3%, respectively). The desire for a baby or child in the future was not significantly different by disability status (p=0.67). In unadjusted models, disabled respondents had 1.3 times the odds of having stopped a method due to dissatisfaction compared to respondents who were not disabled (OR=1.3, 95% CI: 1.2–1.5) ( Figure 2 ), with 52.3% (95% CI: 49.9–54.8) of disabled respondents reporting that they had been dissatisfied with a method compared to 45.8% (95% CI: 44.3–47.2) of non-disabled respondents ( Appendix B ). The relationship between disability status and method dissatisfaction was not attenuated in the adjusted models, which adjusted for survey wave, sexual activity in the past 12 months, demographic factors, and fecundity. Individuals with a disability had 1.4 times the odds of being dissatisfied with their contraceptive method than those without disabilities in the final adjusted model (aOR = 1.4, 95% CI: 1.2–1.6). In the adjusted model, the predicted probability of having stopped a method due to dissatisfaction with it was 54.4% (95% CI: 51.9–56.9) for disabled respondents and 46.5% (95% CI: 45.1–47.9) for non-disabled respondents ( Appendix B ). In unadjusted models, respondents with serious difficulty hearing (OR=1.4, 95% CI: 1.1–1.8), serious difficulty seeing (OR=1.2, 95% CI: 1.0–1.5), or serious difficulty concentrating, remembering, or making decisions (OR=1.3, 95% CI: 1.1–1.5) all had greater odds of having stopped a method due to dissatisfaction with the method compared to respondents without these disabilities ( Figure 3 , Appendix C ). After adjusting for the survey wave, sexual activity in the past 12 months, demographic factors, and fecundity, the association with previous method discontinuation due to dissatisfaction was attenuated for people with serious difficulty seeing (aOR = 1.2, 95% CI: 1.0–1.5). The association remained significant and only slightly attenuated for individuals with serious difficulty hearing (aOR = 1.3, 95% CI: 1.0–1.7) and those with serious difficulty concentrating, remembering, or making decisions (aOR = 1.4, 95% CI: 1.2–1.6). After adjusting, individuals who reported having serious difficulty performing errands alone due to physical or mental limitations had 1.3 times the odds of having discontinued a method due to dissatisfaction (aOR = 1.3, 95% CI: 1.0–1.6). Respondents with one disability and two or more disabilities both had greater odds of having discontinued a method due to dissatisfaction compared to those with no disability in unadjusted and adjusted models (one disability: OR=1.3, 95% CI: 1.1–1.5, aOR = 1.4, 95% CI: 1.2–1.6; two or more disabilities: OR=1.3, 95% CI: 1.1–1.6, aOR = 1.5, 95% CI: 1.2–1.8). In unadjusted models, disabled respondents had 1.6 times the odds of wanting to choose a different method or any method if cost were not a barrier compared to non-disabled respondents (OR=1.6, 95% CI: 1.3–1.9) ( Figure 2 ), with a predicted probability of 30.1% (95% CI: 26.4–33.8) of disabled individuals identifying that they would choose a different method if the cost were not a barrier compared to 21.1% (95% CI: 19.3–23.0) of non-disabled individuals ( Appendix D ). The association was slightly attenuated in magnitude and statistical significance after adjusting for survey wave, sexual activity in the past 12 months, demographic factors, and fecundity. After adjustments, disabled respondents had 1.5 times the odds of wanting to start a new method or switch methods if cost were not a barrier compared to non-disabled respondents (95% CI: 1.2–1.8). The predicted probability of wanting to start a new method or switch methods if cost were not a barrier was 28.8% (95% CI: 25.2–32.3) for disabled respondents and 21.5% (95% CI: 19.8–23.2) for non-disabled respondents. Respondents with serious difficulty seeing (OR=1.8, 95% CI: 1.3–2.4), serious difficulty concentrating, remembering, or making decisions (OR=1.6, 95% CI: 1.3–2.0), or serious difficulty doing errands alone due to physical or mental limitations (OR=1.7, 95% CI: 1.2–2.2) all had greater odds of wishing to switch methods or try a method if cost were not a barrier compared to respondents without each of these functional limitations ( Figure 3 , Appendix E ). In adjusted models, these associations hold and are only slightly attenuated (serious difficulty seeing: aOR=1.5, 95% CI: 1.1–2.1; serious difficulty concentrating, remembering, or making decisions: aOR=1.5, 95% CI: 1.2–1.8; serious difficulty doing errands alone due to physical or mental limitations: aOR=1.7, 95% CI: 1.3–2.2). When compared to respondents with no disability, respondents with one disability and two or more disabilities had 1.6 times the odds of wishing to switch methods or start a new method if cost were not a barrier in unadjusted models (OR=1.6, 95% CI: 1.3–2.0; OR=1.6, 95% CI: 1.3–2.1, respectively). The associations are similar in the models adjusted for survey wave, sexual activity in the past 12 months, demographic factors, and fecundity (one disability: aOR=1.5, 95% CI: 1.2–1.8; two or more disabilities: aOR=1.6, 95% CI: 1.2–2.1).

Our study identified differences in contraceptive satisfaction and preferred method choice among US women with and without disabilities aged 15–49 years in nationally representative data from 2011–2019. These issues are directly relevant to the nearly one in five reproductive-age women who were identified as disabled in the NSFG sample. Most disabled NSFG respondents were sexually active, and almost two-thirds wanted a child in the future. Several of our findings highlight unmet needs for high-quality contraceptive care for people with disabilities. First, in line with findings of earlier studies ( Haynes et al., 2018 ; J. Wu et al., 2018 ; J. P. Wu et al., 2017 ), we found that disabled individuals are more likely to use highly effective methods than those without disabilities. Higher female sterilization rates among disabled people drive much of this difference. Research suggests that the use of sterilization is influenced by provider behavior ( J. Wu et al., 2019 ). Social norms related to pregnancy and reproduction may also constrain the choices of disabled individuals when thinking about pregnancy and contraception ( Hassan et al., 2023 ; LaPierre et al., 2017 ; Tarasoff, 2017 ; J. Wu et al., 2019 ). We found that disabled respondents were less satisfied with a contraceptive method they previously used than their non-disabled peers. We estimated greater odds of stopping a method due to dissatisfaction and desire to use a different contraceptive method if cost were not a barrier among disabled individuals compared to non-disabled individuals; these results held in adjusted models. These findings may indicate that disabled individuals have more difficulty receiving contraceptive methods and care that align with their desires than non-disabled individuals. However, the NSFG did not ask further questions of participants about their preferred contraceptive methods or if other barriers influenced their contraceptive choices. Provider bias, ableism in medical settings, and lack of accessible care may contribute to these differences in method dissatisfaction for disabled individuals, but these factors cannot be directly investigated using NSFG data. Deficiencies in medical training in the U.S. about providing reproductive healthcare for disabled individuals (e.g., IUD placement for people with mobility issues) may contribute to this population not receiving contraceptive care that aligns with their wants and needs ( Becker et al., 1997 ; Hassan et al., 2023 ; Horner-Johnson et al., 2021 ). Additionally, future research should consider if disabled individuals face higher contraceptive costs due to care occurring in more costly settings (i.e., hospital versus family planning clinics) and how this may constrain their method choices. The broader context in which disabled people are navigating reproductive health care is informed by the history of stratified reproduction in the U.S., where the fertility of non-disabled populations is privileged over the fertility of disabled people and others with marginalized identities (e.g., people of color and people with lower incomes). This context may inform the contraceptive method mix observed in this study and differences in method dissatisfaction and preferences ( Harris & Wolfe, 2014 ; J. Wu et al., 2019 ). Additionally, current patterns of sexual and reproductive health care for disabled people occur in the context of the U.S.’s history of eugenics and contraceptive experimentation on marginalized populations and the ongoing devaluing of their reproduction and reproductive autonomy ( National Partnership for Women & Families, 2021 ; Greenwood et al., 2014 ; Hassan et al., 2023 ; Lee, n.d. ; A. Stern, 2020 ; A. M. Stern, 2005 ). These experiences in the healthcare system also likely lead to delaying needed reproductive healthcare and constraining the reproductive choices of disabled individuals. These delays and constraints likely affect decision-making related to pregnancy and contraception and lead to decisions that are not aligned with the individuals’ desires. These difficulties with accessing wanted and needed reproductive health care are even more heightened for disabled people after the Dobbs decision led to curtailed abortion access for disabled people living in restrictive abortion states ( Hassan et al., 2022 ; Powell, 2023 ). Our study is limited by several factors related to the data source. The NSFG likely underestimated the prevalence of disabilities in the general population because the survey does not include individuals who are institutionalized ( National Center for Health Statistics, 2022 ). Additionally, the design and implementation of the NSFG interviews required the ability to hear, speak, and type on a laptop; no specific accommodations are available. Future designs of the NSFG should consider incorporating community engagement to gather insights on adaptability and potential measures. The disability items used in the 2011–2019 NSFG were developed more than a decade ago and should be vetted again in the current environment with disabled individuals ( Livermore et al., 2011 ). The standardized approach used by the NSFG to measure disability status employs a functional perspective, encompassing limitations in hearing, vision, cognition, mobility, self-care, and independent living. Respondents with these functional limitations may not self-identify as disabled. Additionally, not all individuals who identify as having a disability will be included in these functional measures ( Hall et al., 2022 ). For example, individuals with chronic health conditions that may influence contraceptive decisions and use patterns (e.g., conditions with estrogen contraindications or that pose elevated risks for adverse pregnancy and infant outcomes) may not have functional limitations across the six measured domains, effectively leading to underreporting of disability. Furthermore, disability can change over time, and the current measure does not ask about temporal changes in functional limitations (e.g., cyclical symptoms of endometriosis). Another limitation is that cost is the only barrier to preferred method use mentioned in the questions, yet emerging research indicates other individual, interpersonal, and structural influences on preferred method use ( Burke & Lindberg, 2024 ; Gomez et al., 2024 ). An investigation into the new disability measures used in the 2022–2023 NSFG survey round is needed. Asking more questions about method preferences and barriers in future NSFG waves would allow for a more nuanced and complete investigation of the challenges of receiving contraceptive methods and care for disabled individuals. Our results suggest that being disabled was associated with having discontinued a contraceptive method due to dissatisfaction and wanting to have made a different method choice if cost were not a barrier. These findings highlight that disabled individuals are likely experiencing barriers to getting wanted or needed reproductive healthcare, in addition to not receiving adequate information about their contraceptive methods when counseled. More research is needed to understand how clinical care can be improved and provided to this population, with a focus on their reproductive desires and needs. Provider training and resources about specific counseling considerations for disabled individuals, as well as improvements in accessibility at reproductive health clinics, would improve counseling and the dissemination of contraceptive information to this population. The American College of Obstetricians and Gynecologists’ recent guidelines on access to services for people with disabilities can be a central resource for providers and clinics wishing to improve care delivery to this population ( American College of Obstetricians and Gynecologists, 2025 ). Future research examining contraceptive care outcomes and patient-reported experiences of contraceptive use may benefit from focusing on a smaller sample of disabled individuals, where more in-depth and retrospective or prospective information can be collected from the population. These study designs would be advantageous for exploring questions of contraceptive care quality and person-centeredness from a causal framework. Researchers in the field of reproductive healthcare delivery to people with disabilities and chronic conditions have encouraged the use of a justice-based and intersectional approach ( J. Wu et al., 2019 ). The next step for research using the NSFG would be to incorporate intersectionality analytical methods to understand how people with multiple marginalized identities, including disability, may have unique contraceptive care and use patterns. Continued surveillance using nationally representative surveys that accurately measure disability is vital for understanding ongoing patterns in contraceptive access inequities and quality issues for the disabled population ( Landes et al., 2023 ; Swenor, 2021 ). Disabled individuals will continue to experience greater dissatisfaction with their contraceptive methods and barriers to accessing preferred methods without intentional interventions to improve the delivery of these services.

Disability status and experiences of chronic health conditions may influence whether individuals receive wanted and needed reproductive healthcare, including contraceptive care. National studies estimate that nearly 20% of reproductive-age individuals live with a disability involving difficulty with hearing, vision, cognition, mobility, self-care, and independent living ( Horner-Johnson et al., 2020 , 2022 ). Experiences of disability may constrain reproductive health care access and reproductive health choices ( Alhusen et al., 2021 ; Biggs et al., 2023 ), which occur in light of the historical context of forced sterilization and eugenics. In addition, the ongoing impact of structural ableism in medicine and public health often devalues the experiences and needs of disabled people ( Lundberg & Chen, 2024 ). To support reproductive autonomy, disabled people and people with chronic health conditions should have access to the full spectrum of contraceptive methods and comprehensive sexual and reproductive health care ( Hassan et al., 2023 ; Powell, 2023 ). Research using nationally representative and state-level data provides evidence that contraceptive use and contraceptive service utilization differ by disability status. Women with disabilities are more likely to be sterilized and less likely to use moderately effective methods or long-acting reversible contraception (e.g., intrauterine devices (IUD) and implants) compared to non-disabled women ( Haynes et al., 2018 ; J. Wu et al., 2018 ; J. P. Wu et al., 2017 ), as well as less likely to receive family planning services ( Mosher et al., 2017 ). However, this narrow focus on contraceptive mix does not provide information about contraceptive users’ satisfaction with these methods or the use of non-preferred methods. Despite a growing body of research examining unmet contraceptive preferences ( Burke et al., 2020 ; Burke & Potter, 2023 ; Gomez et al., 2024 ; He et al., 2017 ; Kavanaugh et al., 2022 ), to our knowledge, this issue has not been examined by disability status. Further research is needed to identify current patterns in method satisfaction and motivations for switching or discontinuing methods among disabled individuals. Recent changes in contraceptive service guidelines and an increased emphasis on person-centered and justice-informed sexual and reproductive healthcare may have helped reduce the disparities by disability status in contraceptive care and behaviors observed in earlier studies ( American College of Obstetricians and Gynecologists, n.d. ; Dehlendorf et al., 2021 ). In this study, we used a nationally representative household survey to assess potential differences in contraceptive experiences for disabled and non-disabled individuals. We examined differences by disability status in two key outcomes related to contraceptive satisfaction: report of stopping a contraceptive method because of dissatisfaction with the method, and report of a desire to switch to another contraceptive method or try a new method if cost were not a barrier to use. Understanding these experiences can help identify needs and provide the research and clinical care communities with potential mechanisms for interventions to improve contraceptive care for the disabled population.