Feelings about abortion at time of care: Findings from an Ohio abortion facility

Abstract

Introduction:

Prevailing abortion stigma may contribute to how people feel prior to receiving an abortion, and these feelings may influence healthcare decision making. We analyzed data from a patient intake questionnaire regarding feelings at the time of first abortion appointment. We described responses, co-occurrence of sentiments, and associations between responses and abortion characteristics.

Methods:

We abstracted data from a random 20% sample of medical charts at an abortion facility in Ohio for patients who sought abortions from 2014–2018 (N=762). We analyzed data from an intake questionnaire used to assess patients’ sentiments prior to their abortion. The questionnaire had 10 intake items to which patients could respond “yes,” “maybe,” or “no.” The questionnaire also asked, “How are you feeling today?” and listed several emotions for patients to select.

Results:

About 37% of patients responded with exclusively positive emotions, 27% with mixed emotions, and 27% with exclusively negative emotions. Reporting mixed (adjusted odds ratio [95%CI]: 0.33[0.16–0.71]) and negative only (0.38[0.18–0.32]) sentiments was associated with a decreased odds of receiving an abortion. Responding “yes” (vs. “no”) to being afraid an abortion will hurt was associated with a decreased odds of having a procedural abortion vs. a medication abortion (0.62[0.40–0.96]). Responding “yes” to “I am not sure if I am making the right decision” was associated with a longer time between consultation and the abortion (adjusted incident rate ratio [95%CI]: 2.16[1.48–3.16]).

Conclusions:

Sentiment toward abortion is complicated and deserves nuanced attention, rather than being grouped into a strictly positive or negative experience.

Policy Implications:

Assessing patient sentiment prior to an abortion procedure may be valuable for providing patient-centered abortion care.

Introduction

Abortion is one of the most heavily regulated medical practices in the world (Ranji et al., 2022), and remains particularly heavily regulated in the United States (US) following the Dobbs v. Jackson Women’s Health 2022 Supreme Court decision (Guttmacher Institute, 2023a). Examples of state laws that regulate the abortion decision-making process include mandatory waiting periods, forced ultrasound viewing, and counselling sessions with unnecessary or misleading information (Guttmacher Institute, 2022, 2023b). A prevailing narrative exists in anti-abortion rhetoric that abortion regret is common; however, this narrative is incorrect, and fewer than 1% of abortion patients regret their decision to obtain an abortion (Doan et al., 2018; Rocca et al., 2015, 2020).

These medically unnecessary, restrictive laws that aim to regulate the abortion decision-making process create additional barriers for abortion seekers, which may impact patients’ thoughts and feelings prior to an abortion in a negative way. For example, mandatory waiting periods create logistical barriers, such as having to arrange transportation, arrange childcare, or take extra time off of work, which may increase stress for abortion seekers (Chakraborty et al., 2022; Smith et al., 2022). Forced ultrasound viewing and mandated counselling sessions with misleading information can lead to stress, doubt, and heightened internalized stigma about obtaining an abortion (Weitz & Kimport, 2015). These unnecessary abortion regulations create unique stressors around seeking an abortion, compared to seeking other forms of medical care that are not heavily regulated (Fox & Cole, 2021; C. Joffe & Schroeder, 2021; Serpico, 2021). Because of the adverse effects of unnecessary regulations on abortion, abortion seekers may have unique counselling needs at the time of abortion care seeking.

Societal abortion stigma—which contributes to restrictive abortion policies and is reinforced by such policies (A. Norris et al., 2011)—can also impact people’s thoughts and feelings when seeking abortion care. Research shows that abortion stigma is most often directed towards abortion seekers and providers (Jozkowski et al., 2023). Having to overcome barriers to abortion care, such as driving long distances to obtain care, is associated with higher levels of individual-level stigma among abortion seekers (Dahl et al., 2023). Furthermore, experiencing abortion-related stigma is linked to psychological distress among abortion seekers (Biggs et al., 2020).

A core component of any form of medical care is the provision of patient-centered care. Patient-centered care acknowledges the biopsychosocial model of health care and includes emotional support in health care encounters. Provision of patient-centered care is associated with numerous positive outcomes among patients, such as improved health outcomes, increased ability to cope with psychological distress, and reduced anxiety (Epstein & Street, 2011; Kuipers et al., 2019). Given that abortion seekers and providers are uniquely targeted by abortion-related stigma and restrictions, there are unique stressors around abortion care seeking and provision compared to other forms of medical care, and thus a unique role of patient-centered care in alleviating some of these stressors. In order to tailor interventions and create clinical support tools for providing patient-centered care for abortion seekers, more research is needed on the abortion counselling needs of abortion seekers, including on the emotional state of abortion seekers prior to obtaining an abortion.

Research surrounding the emotional state of abortion seekers primarily investigates emotions and mental health after receiving an abortion (Bradshaw & Slade, 2003). These studies show that an overwhelming majority of abortion seekers feel relieved after their abortion and believe that obtaining an abortion was the right decision for them (Rocca et al., 2015, 2020). However, limited research has been conducted on the emotions and sentiments individuals feel before their abortion (Allanson & Astbury, 1995; Bradshaw & Slade, 2003; Rocca et al., 2015; Steinberg & Russo, 2008). The few studies that have examined emotions prior to an abortion have found that sentiments vary and can be contradictory. Some may feel completely positive, or completely negative, but also both positive and negative, ambivalent, or neutral (Allanson, 2007; Allanson & Astbury, 1995; Husfeldt et al., 1995; Kero & Lalos, 2000; Steinberg et al., 2016). Differences in emotional states between individuals may be influenced by having previously had an abortion, as those who have had an abortion report lower levels of stress and doubt surrounding their abortion decision (Husfeldt et al., 1995; Steinberg & Russo, 2008), though these individuals on average also report higher levels of stress in general, which may be related to pre-existing mental health issues and violence (Steinberg & Russo, 2008).

Feeling negative emotions before making a medical decision does not imply that it is the incorrect decision for that individual. Importantly, weighing one’s options is a critical component of the natural decision-making process, even if those emotions are negative (Kero & Lalos, 2000; Purcell, 2015; Rocca et al., 2015, 2020). Both pregnancy and medical procedures that may be physically painful can be stressful experiences in and of themselves (Bradshaw & Slade, 2003; Steinberg & Russo, 2008). Furthermore, the stigma surrounding abortion can amplify preexisting negative emotions. Those who have higher levels of perceived abortion stigma and lifetime violence report higher negative mental health symptoms before their abortion (Rocca et al., 2015, 2020; Steinberg et al., 2016; Steinberg & Russo, 2008).

Overall, restrictive abortion policies and abortion-related stigma can lead to increased negative feelings and emotions during abortion care seeking, which can impact patients’ abortion care decisions, preferences, and delays (Supplemental Figure 1). To address the different counseling needs a patient has before an abortion, some abortion facilities incorporate a questionnaire targeting the patient’s thoughts and feelings into their intake paperwork. Using responses to one such form from an abortion facility in Ohio, we analyzed the prevalence and patterns around different emotions and sentiments among abortion patients who sought an abortion from 2014 to 2018. Ohio is a state in the Midwestern region of the US that passed numerous abortion restrictions during this time, including a ban on abortion after 20 weeks gestation, restrictions on use of public funds for abortion, judicial bypass requirements, bans on performing nontherapeutic abortions in public facilities, requirements that abortion facilities have transfer agreements with nonpublic hospitals within 30 miles of the abortion facilities, and an in-person 24-hour waiting period between consultation and abortion that necessitates two separate visits (A. H. Norris et al., 2020). Ohio also has a 6-week ban that is currently blocked (ACLU of Ohio, 2022). While Ohio passed a constitutional amendment protecting the right to abortion in November 2023, several restrictions are still in place. Thus, in this context of an abortion restrictive state, we described the complex emotional states individuals feel while making their abortion decision. We also determined if these emotional states led to delays in obtaining care or impacted the outcome of this heavily regulated decision-making process, and if these outcomes differ among those who had prior abortion compared to those having their first abortion. We hypothesized that (1) abortion seekers would report a range of complex feelings and emotions at the time of care seeking, (2) these feelings and emotions would be associated with receiving an abortion, type of abortion, and timing of receiving abortion care, and (3) those who had a prior abortion would differ in these outcomes due to increased familiarity with abortion care.

Materials and Methods

Study design and population

We sampled medical charts using a 20% annual random sample of charts from consultation visits from 2014 to 2018 at an independent abortion facility located in a metropolitan area in Ohio (N=762). During the study period, Ohio had a mandatory 24-hour waiting period—that is, patients were required to have an in-person consultation visit and then had to wait a minimum of 24 hours before they could receive an abortion, necessitating two visits to an abortion facility to obtain an abortion (Guttmacher Institute, 2023b). The sample is representative of individuals intending to have an abortion at this facility from 2014 to 2018. Charts were dual-extracted by two data extractors from a pool of five extractors, and then reconciled to account for any errors that took place during extraction. Data extractors were trained prior to data collection to ensure consistent and proper use of the extraction tool. Data were collected via REDCap (Research Electronic Data Capture) (Harris et al., 2009). The clinic received $1500 for participation in the study. Data collection took place in 2020–2021.

Measures

We analyzed data from a questionnaire given to patients during intake at the ultrasound and consultation appointment; patients filled out the form prior to meeting with the medical staff. Facility staff developed the form for the purposes of guiding abortion counseling and had been using it prior to our data collection process. The questionnaire has 10 intake questions to which patients could respond with “Yes,” “Maybe” or “No”: (1) I would like to learn more about how abortion is done, (2) I’m scared that having an abortion is dangerous, (3) I’m afraid it will hurt, (4) I’m wondering what the pregnancy looks like—how developed it is, (5) I’m worried that I won’t be able to have children later on if I wanted to, (6) I am not sure if I am making the right decision, (7) I know I will regret having an abortion, (8) I would like to know more about your adoption services, (9) Someone is forcing me or pushing me to have an abortion, and (10) I am pregnant as a result of sexual assault. The questionnaire also asks, “How are you feeling today?” and lists several emotions for patients to mark all that apply: confident, angry, happy, trapped, relieved, afraid, mean, strong, worried, curious, relaxed, sad, guilty, peaceful, ashamed, numb, and resolved. Of note, these questionnaires were filled out by patients prior to and separate from other support and counseling they would have received during their abortion. The responses represent the feelings and thoughts patients had at intake of their first appointment.

We also abstracted several patient and abortion characteristics from the medical charts, including whether the patient got the abortion, type of abortion, and dates and purposes of each appointment; patients’ race, age, education, and number of living children; and whether the patient had a prior abortion. We calculated the time in days between the first appointment and the abortion appointment using the date of the first consultation appointment and date of the abortion. If one of the dates was missing, the time variable was recorded as missing. If the patient did not receive an abortion, we recorded the time between first and abortion appointment as missing.

Analytic strategy

To describe the range of responses from the sentiment questionnaire, we reported the percentage of patients who selected each answer choice for each intake question. We reported the percentage of patients who selected each feeling word, and then created categories to report the percentage of patients who selected strictly positive words (confident, happy, relieved, strong, curious, relaxed, peaceful, or resolved), strictly negative words (angry, trapped, afraid, mean, worried, sad, guilty, ashamed, and numb), or some combination of positive and negative words. We categorized “curious” as positive because curiosity is often defined as a “positive emotional-motivational system” associated with novel experiences, and measures of curiosity have been shown to be correlated with positive affect (Wagstaff et al., 2021). We reported overall percentages as well as percentages stratified by whether the patient had an abortion in the past.

To examine co-occurrence of responses, we calculated polychoric correlation coefficients, which is a measure of association for ordinal variables, between each of the intake questions and feeling words. We plotted the coefficients in a cluster graph, where more strongly correlated items are grouped together.

Additionally, we examined unadjusted and adjusted associations between the responses in the intake questionnaire and abortion characteristics. We had eight exposures of interest using responses from the intake questionnaire: one variable representing the categories we created using the feeling word selections (positive, negative, and ambivalent) and seven variables from the intake questions, specifically questions 1–7 (questions 8–10 were excluded from this analysis after we determined there was not enough variation in the responses for statistical analysis). We had three outcomes of interest: (1) whether the patient received the abortion (yes/no), (2) type of abortion (medication/procedural), and (3) time between the first appointment and abortion appointment (days). We used logistic regression to examine associations between each of the eight exposures and whether the patient had the abortion. Among those who had the abortion, we used logistic regression to examine associations between the intake questionnaire responses and type of abortion, and we used zero-truncated negative binomial regression to examine associations between intake questionnaire responses and time between the first appointment and the procedural appointment. Our adjustment set included race, age, education, number of living children, and whether the patient previously had an abortion. Additionally, we stratified adjusted and unadjusted estimates by whether the patient previously had an abortion because having a previous abortion may be an effect modifier.

We cleaned and analyzed data using Stata 16 (StataCorp, College Station, Texas, USA). We ran the correlation analysis and created all figures in R (R Foundation, Vienna, Austria).

Results

Of the 770 charts extracted, we excluded charts with missing intake questionnaires, leading to an analytic sample of 762.

Sample characteristics

About 61% (n=465) of patients were White and 24% (n=185) were Black (Table 1). About 81% of the sample reported being not married and 14% reported being married. Most patients (91%) received an abortion for the pregnancy for which they were sampled. Of those who had an abortion, 45% (n=312) of the abortions were procedural and 55% (n=379) were medication. The average patient age was 28 years (SD=6.7, range=15–47 years), and the average time between first appointment and abortion was 5 days (SD=5.3, range=1–79 days).

Table 1.

Sample characteristics of patients seeking an abortion from an Ohio abortion facility from 2014–2018, n=762.

Variable	N	%
Race
White	465	61%
Black	185	24%
Another	21	3%
Missing	91	12%
Marital Status
Not married	617	81%
Married	109	14%
Missing	36	5%
Education
Less than high school	42	6%
High school or equivalent	257	34%
Greater than high school	414	54%
Missing	49	6%
Type of abortion
No abortion	71	9%
Medication abortion	379	50%
Procedural abortion	312	41%
Missing	0	0%
Has living children
No	297	39%
Yes	460	60%
Missing	5	1%
Had an abortion in the past
No	474	62%
Yes	283	37%
Missing	5	1%
Variable	Mean	SD
Age (years)*	28	6.7
Time between first and abortion appointment (days)†	5	5.3

N=number; %=percent, SD=standard deviation

^*The variable, age, had 761 non-missing values, with a minimum value of 15 and a maximum value of 47 years.

^†The variable, time between first and abortion appointment, had 689 non-missing values with a minimum value of 1 and a maximum value of 79 days. Those who did not get an abortion have a missing value for this variable.

Distribution of sentiment question responses

Most patients responded “no” to most of the intake questions (Figure 1), especially often responding “no” to questions asking whether they would like to know more about adoption services, whether someone was forcing or pushing them to have the abortion, and whether the pregnancy was the result of rape. Of the questions with more mixed responses, 33% said they would like to learn more about how abortion is done, 32% were afraid the procedure will hurt, and 19% said they were afraid the procedure is dangerous. Among those who previously had an abortion, responses were less mixed: only 20% were afraid the procedure will hurt, 17% said they would like to learn more about how abortion is done, and 10% said they are afraid the procedure is dangerous, compared to 38%, 43%, and 23% respectively, among those who never had an abortion.

Figure 1.

Distribution of responses for the intake questions, n=762.

Of the feelings offered in response to the question “How are you feeling today?” the three most endorsed were: worried (31%), relieved (27%), and confident (22%) (Figure 2). Fewer patients felt angry, happy, trapped, mean, or peaceful. Once we grouped each patient’s responses into positive, negative, mixed, or no response (Figure 3), we saw that the largest group of patients had only positive sentiments (37%), followed by mixed (27%) and only negative (27%). Eight percent of patients did not circle any feeling questions. Sixty-five percent of patients expressed at least one positive feeling, and 54% of patients expressed at least one negative feeling.

Figure 2.

Distribution of responses for the feeling words, n=762.

Figure 3.

Distribution of responses of feeling words grouped into three categories: negative only, positive only, and mixed (Panel A, n=762) and stratified by whether the patient ever had an abortion in the past* (Panel B, n=757)

*Five patients had missing information on if they previously had an abortion

Correlation analysis

We next examined the polychoric correlation coefficients in a cluster graph (Figure 4; the values of these coefficients are shown in Supplemental Table 1). In Figure 4, questions with answers that are more correlated to each other are situated more closely together, and questions with answers that are less correlated to each other are further apart. Thicker lines represent stronger correlations, blue lines represent positive correlations, and red lines represent negative correlations. While very few patients (n=5) responded yes to intake question 9 (“someone is forcing or pushing me to have an abortion”) it clusters at the center of several feeling words. It is negatively correlated with reporting feeling confident, relieved, resolved, strong, peaceful, and happy. It is positively correlated with reporting feeling angry, ashamed, and sad. It is also positively correlated with intake questions 6 (“I am not sure if I am making the right decision”) (n=37) and 7 (“I know I will regret having an abortion.”) (n=26). Taken together, these questions illuminate a small cluster of patients who would have needed more counseling given their feelings at intake, and several of them did not have an abortion.

Figure 4.

Cluster graph of intake questions and feeling words using polychoric correlations, n=762*

*Red lines represent negative correlations and blue lines represent positive correlations. Closer distances between nodes and thicker lines represent stronger correlations.

Intake questions 1–5, which are about the abortion (e.g., afraid it will hurt, worry about future fertility) had more “yes” responses and had stronger correlations to feeling curious, worried, and afraid, compared to the rest of the positive and negative feeling words.

Associations between abortion characteristics and abortion sentiment, overall

Outcome: Receiving an abortion

We next assessed associations between sentiments and having, or not having, an abortion (71 patients did not have an abortion) using logistic models. Those who reported mixed or negative feelings had lower odds of receiving an abortion relative to those who reported positive feelings (Table 2). The only intake questions significantly associated with whether the patient received an abortion in both the unadjusted and adjusted analyses were “I know I will regret having an abortion” and “I am not sure if I am making the right decision”—those who responded “yes” to the former question had a lower odds of receiving an abortion (adjusted odds ratio [aOR], 95% confidence interval [CI]: 0.25, 0.08–0.77]) and those who responded “maybe” (0.32, 0.15–0.69) or “yes” (0.25, 0.10–0.66) to the latter question also had a lower odds of receiving an abortion. Responding “yes” to wanting to know more about how abortion is done was also associated with a lower odds of receiving an abortion in the unadjusted model, but was not significant after adjustment.

Table 2.

Associations between sentiment questions and procedural characteristics

	Received an abortion (yes vs. no)*		Type of abortion (medication vs. procedural)*		Time between first appointment and procedure†
	OR (95% CI)	aOR‡ (95% CI)	OR (95% CI)	aOR‡ (95% CI)	IRR (95% CI)	aIRR‡ (95% CI)
Feeling words category (ref: Positive)
Mixed	0.46 (0.24, 0.89)	0.33 (0.16, 0.71)	1.06 (0.73, 1.54)	0.90 (0.58, 1.39)	1.11 (0.92, 1.34)	1.06 (0.86, 1.30)
Negative	0.43 (0.22, 0.82)	0.38 (0.18, 0.82)	1.19 (0.81, 1.74)	1.06 (0.68, 1.66)	1.15 (0.95, 1.39)	1.04 (0.85, 1.28)
I am afraid it will hurt (ref: No)
Maybe	0.67 (0.36, 1.25)	0.61 (0.30, 1.28)	0.76 (0.52, 1.13)	0.65 (0.40. 1.04)	0.97 (0.80, 1.17)	0.95 (0.77, 1.17)
Yes	0.71 (0.40, 1.26)	0.60 (0.31, 1.17)	0.63 (0.45, 0.90)	0.62 (0.40, 0.96)	1.05 (0.89, 1.24)	0.92 (0.76, 1.12)
I would like to know more about how abortion is done (ref: No)
Maybe	0.52 (0.26, 1.03)	0.54 (0.25, 1.18)	1.13 (0.73, 1.76)	1.16 (0.67, 1.99)	1.04 (0.84, 1.29)	0.92 (0.72, 1.17)
Yes	0.53 (0.30, 0.91)	0.65 (0.34, 1.25)	0.83 (0.59, 1.16)	0.65 (0.42, 1.00)	1.13 (0.96, 1.33)	1.07 (0.89, 1.30)
I’m scared that having an abortion is dangerous (ref: No)
Maybe	0.55 (0.31, 1.00)	0.55 (0.28, 1.07)	0.95 (0.65, 1.39)	0.80 (0.51, 1.27)	0.84 (0.70, 1.01)	0.81 (0.66, 1.00)
Yes	0.56 (0.30, 1.04)	0.52 (0.25, 1.06)	0.87 (0.58, 1.31)	0.87 (0.53, 1.42)	1.17 (0.97, 1.42)	1.04 (0.83, 1.29)
I’m worried that I won’t be able to have children later on if I wanted to (ref: No)
Maybe	0.60 (0.31, 1.17)	0.58 (0.27, 1.24)	0.67 (0.42, 1.05)	0.78 (0.46, 1.35)	1.01 (0.82, 1.25)	1.08 (0.85, 1.36)
Yes	0.55 (0.29, 1.03)	0.55 (0.27, 1.13)	1.02 (0.66, 1.58)	1.32 (0.78, 2.26)	1.10 (0.90, 1.35)	1.06 (0.84, 1.33)
I’m wondering what the pregnancy looks like—how developed it is (ref: No)
Maybe	1.12 (0.46, 2.70)	1.45 (0.50, 4.25)	1.05 (0.63, 1.76)	1.24 (0.67, 2.29)	0.86 (0.67, 1.11)	0.85 (0.64, 1.13)
Yes	0.91 (0.46, 1.81)	1.19 (0.53, 2.69)	0.68 (0.44, 1.05)	0.60 (0.36, 1.00)	1.24 (1.02, 1.52)	1.18 (0.94, 1.48)
I know I will regret having an abortion (ref: No)
Maybe	0.63 (0.33, 1.20)	0.66 (0.32, 1.35)	1.36 (0.88, 2.10)	1.04 (0.63, 1.74)	1.41 (1.15, 1.72)	1.38 (1.10, 1.72)
Yes	0.24 (0.09, 0.59)	0.25 (0.08, 0.77)	0.58 (0.22, 1.53)	0.42 (0.11, 1.63)	0.90 (0.58, 1.41)	0.85 (0.51, 1.42)
I am not sure if I am making the right decision (ref: No)
Maybe	0.38 (0.20, 0.75)	0.32 (0.15, 0.69)	1.60 (0.94, 2.71)	1.50 (0.79, 2.86)	0.90 (0.70, 1.16)	1.00 (0.75, 1.32)
Yes	0.29 (0.13, 0.67)	0.25 (0.10, 0.66)	0.55 (0.25, 1.24)	0.47 (0.16, 1.40)	2.19 (1.59, 3.02)	2.16 (1.48, 3.16)

OR=odds ratio, IRR=incidence rate ratio, aOR=adjusted odds ratio, aIRR=adjusted incidence rate ratio, CI=confidence interval, ref=reference

^*ORs obtained from logistic regression models.

^†IRRs obtained from zero-truncated negative binomial regression models

^‡adjusted for race, age, education, marital status, whether the patient has living children, and whether the patient ever had an abortion in the past

Outcome: Abortion method

Among those who had an abortion, those who responded that they were afraid an abortion will hurt were more likely to have a medication abortion, with decreased odds of having a procedural abortion compared to a medication abortion (0.62, 0.40–0.96) (Table 2). The feeling word categories and the other intake questions were not statistically significantly associated with abortion method.

Outcome: Time to abortion

Among those who had an abortion, the feeling word categories were not significantly associated with time between the first appointment and the procedure. Responding “yes” to wanting to know more about what the pregnancy looks like was associated with longer time between intake and abortion in the unadjusted, but not the adjusted model. Responding “yes” to “I am not sure if I am making the right decision” was statistically significantly associated with longer time between intake and abortion (adjusted incidence rate ratio [aIRR], 95% CI: 2.16, 1.48–3.16).

Associations between abortion characteristics and abortion sentiment, stratified results

In the analysis stratified by whether patients had an abortion in the past, the results varied compared to the unstratified results.

Abortions among those with prior abortions

Among respondents who had an abortion in the past, those who reported mixed and exclusively negative feelings and responded “yes” compared to “no” to all except two of the intake questions (wanting to know more about how abortion is done, and wondering what the pregnancy looks like) were significantly less likely to receive an abortion (Figure 5 and Supplemental Table 2). Those who responded “maybe” compared to “no” for “I would like to know more about how abortion is done,” “I’m scared that having an abortion is dangerous,” and “I am not sure if I am making the right decision” also had decreased odds of receiving an abortion. Among those who had an abortion for their current pregnancy, responding “yes” to “I am afraid it will hurt” and to “I would like to know more about how abortion is done” was statistically significantly associated with decreased odds of receiving medication abortion (versus procedural) in the unadjusted, but not adjusted, model. Responding “yes” to “I’m worried that I won’t be able to have children later on if I wanted to” was statistically significantly associated with longer time between intake and abortion appointment in the unadjusted, but not adjusted, model. Responding “yes” to “I am afraid it will hurt” was associated with less time between the first appointment and the abortion only in the adjusted analysis (aIRR, 95% CI: 0.62, 0.43–0.90).

Figure 5.

Associations between sentiment questions and procedural characteristics stratified by whether the patient ever had an abortion in the past

*Odds ratios obtained from logistic regression models. Incident rate ratios obtained from zero-truncated negative binomial regression models. Models are adjusted for race, age, education, marital status, and whether the patient has living children. Values of estimates and confidence intervals that are plotted in this figure are shown in Supplemental Table 2. Results from the question, “I’m wondering what the pregnancy looks like—how developed it is” are omitted from this figure due to small sample sizes but are shown in Supplemental Table 2.

Abortions among those with no prior abortion

Among respondents with no prior abortion, none of the sentiment questions were significantly associated with receiving an abortion. Among those who had the abortion for their present pregnancy, responding “yes” to wondering what the pregnancy looks like was significantly less likely to have a medication abortion (versus procedural) (aOR, 95% CI: 0.50, 0.28–0.92). Furthermore, responding “yes” to this question, as well as to “I am not sure if I am making the right decision” were both significantly associated with a longer time between intake and abortion (aIRR, 95% CI: 1.40, 1.09–1.80; and aIRR, 95% CI: 2.61, 1.78–3.83, respectively.)

Discussion

In this study, we found that the greatest proportion (37%) of patient sentiment was exclusively positive, while just over a quarter of patients responded with conflicting emotions – e.g., feeling worried and relieved – and the same proportion reported exclusively negative sentiments. These findings suggest that sentiment toward abortion is complicated and deserves nuanced attention, rather than being grouped into a strictly positive or negative experience. This finding is in line with previous scholarship that reports emotional response to abortion being separate from certainty about an abortion: patients may have conflicted or negative feelings about the abortion itself, and at the same time be very certain about their choice to get an abortion, and not experience regret about the decision later (Rocca et al., 2015, 2020). Another study that examined people who considered, but did not ultimately have an abortion, found that respondents were very similar in demographics and reasons for considering abortion as those who ultimately chose an abortion, further highlighting the complexity within which people seeking abortion care experience their decision-making (Kimport, 2021). Furthermore, this study provides insight into how patients’ views or concerns about their abortions correlate with emotional states, which is informative for pre-abortion counseling.

Importantly, we also found that patient sentiments before the abortion may have influenced care decisions surrounding their abortions. For example, patients who reported being unsure about obtaining an abortion, or those who felt that they will regret getting an abortion, were less likely to get an abortion. Among patients who received abortions, those who reported being afraid an abortion will hurt were more likely to receive medication abortions. Additionally, patients’ sentiments around abortion were not associated with time between intake and abortion, except if patients reported being unsure about their decision, suggesting that these patients may have taken more time to weigh their pregnancy options before deciding to obtain an abortion.

The natural decision-making process involves the weighing of options, which can bring about negative, neutral, and positive emotions. These emotions are dynamic and may change throughout the decision-making process. Many factors contribute to the emotions and feelings people experience before their abortions. First, people may feel anxious or stressed because abortion, like other medical procedures, carries a risk of physical pain and feeling not in control of one’s body. Second, other underlying stressors (e.g., financial insecurity, intimate partner violence, etc.), may also heighten stress around abortion decision-making. Experience of violence, specifically, is associated with increased risk of unintended pregnancy and abortion, as well as adverse mental health outcomes (Cahill & Doyle, 2021; Grace et al., 2022; Miller et al., 2014).

Policy implications

In the specific case of abortion, anti-abortion advocates and legislators create a false narrative that patients will experience negative mental health outcomes following their abortion or that they will later regret their abortion (Doan et al., 2018; Rowlands, 2011). These narratives of regret and adverse physical and mental health risks created by restrictive policies and misinformation campaigns can increase abortion seekers’ feelings of uncertainty and fear; in our cluster analysis, we found that reporting uncertainty about abortion was correlated to reporting feeling curious, worried, and afraid. Little research exists on the impact of abortion restrictions and misinformation campaigns related to abortion on psychological distress among abortion seekers. One recent study showed that a greater number of state-level abortion restrictions were correlated with higher suicide rates among reproductive-aged women (Zandberg et al., 2023).

The relationships between pregnancy decisions, individual-level life circumstances in the context of pregnancies, mental health and emotional well-being, feelings about abortion, and the narratives and stressors that result from abortion restrictions are complex. Furthermore, structural, cultural, and social constraints impact abortion decision-making (Kimport, 2021). Currently, there is no formal scale for measuring how patients feel about their abortion prior to receiving it. However, asking patients these types of questions during intake may help guide clinicians in offering more patient-centered care. This is especially important now that several states have either banned or severely restricted abortion care after the June 2022 Dobbs vs. Jackson Women’s Health Organization ruling by the U.S. Supreme Court, and abortion seekers face additional barriers to abortion access.

Clinical implications

Given the individual-level experiences and varying state-level abortion policy landscapes in the US, abortion providers face unique challenges when counseling their patients, compared to other health care encounters. Pre-abortion counseling itself is heavily influenced by legal restrictions placed on clinics (C. Joffe, 2018). Clinic staff in many states are required to read patients negative, medically inaccurate scripts during their consultation visit, or to force patients to view their ultrasounds and listen to descriptions of fetal development (Guttmacher Institute, 2022). In contrast, pre-abortion counseling may include discussions around patients’ lived experiences that influence their abortion decisions. We also found that the intake questions reflecting concerns about the abortion had stronger correlations to feeling curious, worried, and afraid. These concerns may stem from lack of knowledge about abortion and/or misinformation narratives from anti-choice advocates; thus, it possible that selecting “yes” on these questions may correlate with feeling curious and wanting to know more about abortion as well as feeling worried or afraid. Thus, measuring such concerns in patient intake forms and addressing them during abortion counseling may assist with reducing patients’ anxiety about abortion care. Furthermore, those who have previously had an abortion may require different counseling because of their prior experience in navigating abortion stigma or experiencing additional stigma for multiple abortions. For example, among those who had a prior abortion, we found that negative feelings played more of a role in whether they received an abortion, whereas among those with no prior abortion, the feeling words did not appear to play a role in receiving an abortion.

To provide the best counseling before an abortion, the patient-provider conversation must be tailored to the individual. This patient-centered approach moves away from “either/or” thinking that the abortion decision is either completely easy or completely challenging, and instead to “and/both” where some aspects may be easy to decide, and others are more difficult (Allanson, 2007). Counselors must be prepared to use trauma-informed conversation techniques as the context of the abortion decision is situated in a patient’s lifetime of experiences (Cahill & Doyle, 2021). Pre-existing stressful life circumstances such as poverty, mental health issues, and violence can greatly influence the mental well-being of an individual before their abortion (Husfeldt et al., 1995; Purcell, 2015; Steinberg & Russo, 2008). Small interventions, like eliciting information about a patient’s emotional state prior to meeting with them, can go a long way in facilitating conversations that meet the diverse needs of patients seeking abortion care. Thus, more research is needed to develop evidence-based validated tools to support abortion providers in assessing patients’ emotional states prior to their abortion.

Strengths and limitations

One limitation of this study is that we could only examine relationships between the questionnaire items that were reported on this intake form. Thus, we were not able to examine other aspects of abortion sentiment that may be relevant to abortion decision-making or how this questionnaire facilitated conversations during patient encounters. We were also not able to assess what patients were thinking about as they were answering the questions; for example, if participants selecting “confident” were feeling confident about the abortion or as people. Another limitation is the extent to which these findings can be generalized to a larger population, given the unique characteristics of Ohio as an abortion-restrictive state. Based on the U.S. Centers for Disease Control and Prevention abortion reports, the study sample has a higher distribution of White patients compared to the those receiving abortion in Ohio as a whole from 2014 to 2018 (Jatlaoui, 2018a, 2018b, 2019; Kortsmit, 2020). Most abortion patients in Ohio were also 20–29 years old; the mean age in the study sample was 28 years old (Jatlaoui, 2018a, 2018b, 2019; Kortsmit, 2020). Nevertheless, our sampling approach means our findings are representative of abortion patients at this facility. Our study does not have the statistical power needed to investigate how abortion sentiment, and its relationship with various abortion characteristics, differs by patient characteristics such as age, race, and geography, or financial situations of patients. Future research should focus on how those experiencing the negative effects of poverty and structural racism, or those having to navigate parental consent laws, may feel differently about having an abortion. Many patients across the country have abortions in states with restrictions like Ohio, even more so following the Dobbs decision, so our findings are relevant for hundreds of thousands of abortion seekers in the US.

Conclusions

People making abortion decisions are situated within their life experiences, in addition to perceived abortion stigma stemming from national- and state-level narratives around abortion (Kimport, 2021). Thus, abortion care providers face unique challenges in caring for their patients, including in pre-abortion counseling (C. E. Joffe, 1996). We found that while 37% of patients reported exclusively positive feelings prior to their abortion, about 25% of patients reported mixed feelings. Feeling worse during abortion care decisions (e.g., due to pregnancy-related circumstances, or stigma rooted in narratives used by state policies) does not necessarily change patients’ abortion decisions, and thus there is unique importance of counseling in abortion care to support a patients’ health and wellbeing throughout the abortion process. Currently, there is no validated scale for helping providers assess their patients’ feelings about their abortions prior to the procedures. Our findings from a metropolitan abortion facility, located in a state with severe abortion restrictions, show preliminary evidence of the value of such an approach. Future research is needed to create a validated tool to assist providers, especially those in abortion-hostile states, in helping their patients navigate their abortion decisions and care.

Data availability:

The data are not available due to their sensitive nature.