In this study, which involved the smallest geographic unit available through the US census, assumed no barriers to interstate travel, and examined a variety of potentially acceptable drive times, we found that changes to TMAB programming and policy could expand abortion accessibility in the United States. Removing all state TMAB bans and expanding TMAB services to all Planned Parenthood health centers that did not offer abortion care in 2018 would result in more than 3.5 million additional US women 15 to 44 years old living within 30 minutes of an abortion-providing facility. Our findings are consistent with results demonstrating increased access in one clinic network after the implementation of a TMAB program and further illustrate how TMAB could affect accessibility among all reproductive-aged women in the United States.
Although the policy and programming changes we considered would universally increase abortion accessibility, there was variation in the magnitude of the increase, with some states seeing little effect. This variation appeared to be a result of a combination of factors including current accessibility, whether the state or neighboring states have a TMAB ban, the number of health centers offering care, state size, and rural versus urban census block classification.
Our findings suggest that TMAB programming or policy changes could have larger benefits for rural communities. These results are in line with a large body of research indicating that rural US residents face many health disparities relative to urban US residents, in part as a result of poor access to health care. Furthermore, because changes to provisions in one state can affect accessibility in another and TMAB bans are concentrated in the Southeast and middle areas of the United States, future research should estimate changes in abortion accessibility assuming policy or programming changes in only certain states or regions. Such studies would help to determine where changes could have the most dramatic impact on accessibility.
The TMAB programming expansion we considered could help ensure that a range of highly acceptable abortion provision options are accessible to abortion seekers, including those who wish to visit a clinic in person. However, even in a scenario in which all Planned Parenthood health centers offered abortion care, in some states less than half of the female population 15 to 44 years of age lived within 90 minutes of an abortion-providing facility. In other words, these changes alone do not ensure accessibility for all US women.
There is growing use of direct-to-client telemedicine abortion services in the United States. In December 2021, the Food and Drug Administration made policy changes that permanently allow for remote provision of the medication abortion drug mifepristone. As a result, direct-to-client telemedicine abortion services that arose during the COVID-19 pandemic are likely to remain available to abortion seekers in some settings. Although these services are an important addition to the abortion care provision landscape, it is vital that in-clinic options remain available for interested clients. Our study thus focused on in-clinic abortion care provision. Because TMAB bans also apply to direct-to-client telemedicine models, the policy changes we considered are relevant not only for TMAB expansion but for the expansion of other telemedicine in medication abortion provision models, including those that do not require clients to visit a facility. As a result, ban removal could result in greater increases in accessibility than those reported here.
A major limitation of our study is possible misclassification of abortion accessibility; the study is susceptible to the ecological fallacy, as a population-level measure of accessibility stands in for an individual-level measure. Our analyses assumed that women had access to a vehicle at their home location and traveled to abortion care from that location. In addition, we used 3 dichotomous drive times (30 minutes or less, 60 minutes or less, and 90 minutes or less) to represent a range of reasonable distances.
Misclassification of accessibility may also arise because we calculated accessibility using data from the ANSIRH facility database. Although the combination of the ANSIRH facility database and Planned Parenthood data represents an attempted census of abortion providers in the United States, the facility database may not be a complete census of abortion-providing facilities and could result in an undercount of facilities and, consequently, accessibility. However, because the database was constructed through Internet search terms mimicking those of people seeking services, we believe that our definition of accessibility closely represents the lived experience of US women of reproductive age.
Our analyses also considered only a single element of access, accessibility operationalized as driving time, which alone cannot ensure access to abortion care. Other factors that influence abortion access include hours of operation, out-of-pocket cost of care, and the cultural competency of staff and providers. Accessibility alone likely misclassifies abortion access; however, the population did not change across the 3 scenarios, and thus the estimates of percentage point changes in population access should have been unbiased.
Furthermore, in the United States, medication abortion is approved for use up to 10 weeks' gestation, although some providers offer care at later gestational ages through off-label use of the medications. Although nearly 80% of abortion care is provided at or before 9 weeks' gestation, our measure of accessibility was misclassified for pregnancies beyond 10 weeks' gestation. Similarly, some abortion seekers, even if within the gestational age limit for medication abortion, have contraindications for medication abortion or prefer procedural abortion and cannot or would not use medication abortion. Our accessibility measure was also misclassified for these individuals.
Finally, we considered only 2 changes in abortion provision scenarios, programming and policy changes that resulted in care expansions to additional Planned Parenthood health centers. Given resource constraints and laws that target abortion provision (i.e., targeted regulation of abortion provider laws), it is unlikely that one provider network would expand services so dramatically, even in response to policy changes. Furthermore, other abortion providers, including independent providers, might choose to establish additional clinics should these changes occur. Also, given the politicization of abortion in the United States, it is unlikely that all states with TMAB bans would remove these bans simultaneously.
Our analyses estimated the upper limit of expansion if only Planned Parenthood made service changes; however, other program and policy changes (e.g., clinic openings and closures) over time within and outside the Planned Parenthood system are likely and would affect the exact proportion of the population with access to abortion care. Our estimates serve as an example of how policy and program shifts could affect accessibility.
Public Health Implications
Our findings point to areas where increased abortion provision in the form of TMAB would have the greatest impact on one domain of abortion access, accessibility, as defined by the number of reproductive-aged women within a given drive time of an abortion-providing facility. These increases in accessibility could have meaningful public health effects given that obtaining wanted abortion care is a determinant of health and well-being.[1–9] For abortion seekers in states with mandatory preabortion counseling or waiting periods between counseling and the abortion visit, TMAB expansion that included TMAB options for counseling (either direct to client or site to site) could further reduce barriers to care. However, even with these changes, large numbers of women would remain without adequate abortion accessibility according to our measures. Our data can be used by health care advocates and funders as they consider where to invest policy- and program-specific resources to improve abortion accessibility in the United States.
We used a broadly applicable framework to measure abortion accessibility. Our study design can easily be adapted to assess the effects of different abortion service expansions or restrictions on the same measure of accessibility. In addition, rather than the road networks used for our drive time analysis, future studies could employ public transportation networks to assess accessibility among populations without access to a car. In the future, both projections that are reactive to policy changes and those that proactively assess the effects of policy changes may be particularly useful to help determine resource and funding allocation as the abortion provision and policy landscape shifts. Furthermore, such measures of accessibility must be incorporated into future studies that consider multiple domains of access simultaneously. Doing so will help ensure that access is accurately measured and that results inform a multifaceted response to improve abortion access in the United States.
A version of this study was a chapter of J. W. Seymour's doctoral dissertation, which was published through the Boston University Library. Findings from this work were also presented at 2021 National Abortion Federation annual meeting in May 2021.
We thank Ushma Upadhyay for generously sharing the Advancing New Standards in Reproductive Health facility database used in this study. Also, we are grateful for Kenneth Rothman's support of this work and feedback on article drafts.
Human Participant Protection
In consultation with the Boston University Medical Campus and Boston Medical Center institutional review board staff, it was determined that this study did not involve human participants and therefore did not require institutional review board oversight.
Am J Public Health. 2022;112(8):1202-1211. © 2022 American Public Health Association