Prevalence, attitudes and knowledge of misoprostol for self-induction of abortion in women presenting for abortion at Midwestern reproductive health clinics.

In the setting of increasing restrictions to legal abortion in the United States, reports have emerged of self-induced termination of pregnancies with misoprostol, obtained without a prescription or provider. This study seeks to describe the prevalence of women seeking or employing misoprostol for self-induced abortion, and how they access information. In a cross-sectional study of women immediately following their abortion at three reproductive health clinics in the Midwestern United States, an anonymous survey queried gestational age, barriers, online investigation about self-induction and opinions concerning the availability of medical abortion. From June to September 2016, 276 women participated out of 437 presenting to the clinics during the study period. One hundred and ninety-one (74.6%) women had investigated abortion options online, and of those, 58 (30.9%) investigated misoprostol online, for home use. Women who investigated online options were less likely to have had a prior abortion than those who did not investigate online (29.3% vs. 63.1%, p < .01). They were also more likely to report prior home attempts to end this pregnancy (8.6% vs. 0%, p = .05). Overall, four (1.6%) of the respondents purchased misoprostol and three (1.2%) used it. A majority of women seeking an abortion sought online information prior to their clinic appointment, and almost a third of those had investigated misoprostol for home use. Women are accessing information regarding misoprostol for self-induction of abortion on the Internet. As barriers to legal abortion increase, women may be more likely to self-induce abortion.

Introduction

Increasing numbers of laws restricting abortion access, aimed at distancing women from the right to abortion, have been passed in the United States, including 334 restrictions from 2010 to 2016.[1] Currently, nearly half of all states are hostile to abortion rights (with greater than four major abortion restrictions in place) and only 20% of women live in a state supportive of abortion rights.[1] These abortion restrictions have driven some women to engage in unsafe, self-induced abortions similar to what occurred prior to legalisation of abortion in the United States in 1973.[2,3]

Misoprostol has been studied as a single agent for a medical abortion but is less efficacious than the combination of misoprostol and mifepristone when compared for use for medical abortion prior to 63 days gestation.[4,5] The current standard, and the Food and Drug Administration approved regimen for medical abortion, is 200 mg mifepristone followed in 24–48 hours by 800 mcg misoprostol.[6] Even prior to the availability of medical abortion in the US, an urban study demonstrated that 37% of women were aware of misoprostol and its use as an abortifacient.[7] Due to its efficacy and over-the-counter availability in many countries, misoprostol used alone has become a common agent for self-induction worldwide.[8-10]

In a 2009 US national survey of women at abortion clinics, 1.7% of women reported attempting self-induction by ingesting herbs or medications, and another 1.2% of women used misoprostol to self-induce abortion.[11] A study in Texas noted that 1–4% of women attempted self-induction.[8] Rates reach 7.4% in countries where abortion is legally restricted, such as in Brazil, where abortion is restricted to cases of rape or threat to maternal life, and in Saudi Arabia, where parental or spousal consent is required.[9,12] The data also suggest that immigrants from countries with heavily restricted abortion access know more about misoprostol and its use as an abortifacient.[13]

There is good evidence on the safety of using misoprostol and mifepristone for medical abortion with established telemedicine models, including video conferencing, demonstrated by the success of Planned Parenthood in Iowa, as well as online telemedicine use by Women on Web in Ireland. Both demonstrate similar rates in complications of medical abortion in a clinic setting.[14,15] Additionally, in Australia, where mifepristone is now available by prescription in pharmacies, but where medical abortion has been uncommon, a foundation provides the medications to patients by mail after verifying the pregnancy and performing counselling over the phone.[16] However, safety is not ensured when the source of the medication and instructions provided are questionable.

Given the volume of anti-abortion legislation that has been successfully enacted in the US within recent years, including waiting periods, gestational age limits, building requirements that current clinics cannot support and physician admitting privileges, all of which have caused numerous clinics to close, we suspect that women are left in a position where abortion with a physician is becoming inaccessible. However, there is no specific data in the Midwest US on this phenomenon despite it being an area with increasing restrictions, but with fewer immigrants and less access to medications from other countries, such as were present in the studies above. Overall, self-sourced abortion is an area with many research gaps, including regarding women’s experiences and desires surrounding this practice.[10] Our objective was to identify if there is a trend in increasing women’s investigation of, or actual, self-termination; and if so, how women obtain this information. In addition, we investigated how women would react to the potential availability of medical abortion without a provider, especially with misoprostol since it is inexpensive and has been found to be readily available globally. We particularly focused on the availability of information online, since the dissemination of information via the Internet is common and relieves women of many of the barriers they face accessing typical care. Awareness of a practice taking place outside the traditional realm, with associated potential medical implications, is necessary for women’s healthcare providers and for policymakers.

Materials and methods

We developed a survey to investigate age, race, marital status, educational status, income, gravidity, parity, maternal estimation of gestational age and actual gestational age measured by ultrasound, money spent to obtain an abortion excluding medical fees, and distance travelled. The survey asked questions on women’s position on abortion rights, familiarity with misoprostol as an abortifacient, how information about abortion care and misoprostol was obtained, and barriers to care. The first 16 questions were about demographics and pregnancy history, followed by seven questions about the investigation of misoprostol online and use of misoprostol in past and present pregnancies, and six questions about barriers to care and whether a hypothetical alternative of over the counter misoprostol would be acceptable. The survey was comprised of multiple choice questions, some of which allowed multiple responses. There were also a few options to write in responses if the multiple choices did not include their response. Questions were developed by the study authors as no validated study about self-induced abortion was available. The survey was available in English and validated for readability and easy language level by a third party in the College of Public Health with expertise in survey design.

Following the University of Iowa Institutional Review Board approval, survey administration commenced at three non-profit independent reproductive health clinics, identifying all women presenting for medical or surgical abortion procedures. The maximum gestational age for procedures was 21 weeks, confirmed by ultrasound. Women presenting for induced abortions were included. Women under the age of 18 years, currently incarcerated, or not proficient in English, were excluded. A letter containing elements of consent was provided to women by a research assistant prior to or following their abortion procedure, while in the recovery room. Subjects completed the survey either on paper or computer tablet. Surveys were returned to a secure box with the research assistant whether completed or not, in order to maintain confidentiality. No compensation was provided.

Study data were collected and managed using REDCap (Research Electronic Data Capture) secure, web-based, electronic data capture tools hosted by the University of Iowa Institute of Clinical and Translational Science,[17] supported by the National Institutes of Health (NIH) Clinical and Translational Science Award (CTSA) programme, grant U54TR001356. Statistical analysis was done using SAS version 9.3 (SAS Institute Inc. Base SAS®, Cary, NC, USA). Descriptive statistics included means and percentages. Test of symmetry was used to assess change for the variables with ≥3 nominal response categories. A statistically significant result was defined as p-value <.05. Comparisons between subject groups utilized Pearson-Chi square tests for categorical variables and Wilcoxon-Rank Sum for ordinal data.

Results

From June through September 2016, 437 women obtained abortions and 276 (63.2%) completed the questionnaire. Demographic data is included in Table 1, with 228 (90.1%) at less than 14 weeks gestation. Response rates to individual questions varied and denominators are therefore noted. Gestational age averaged 8.4 weeks by patient estimate and 8.7 weeks by ultrasound dating (p = .12) with a median of 8 weeks gestation for both groups. However, the mean difference in a single participant’s estimate was 0.91 weeks. Most respondents (191 of 256 [74.6%]) investigated misoprostol and other abortion options online. Of the 188 women who specified how they investigated online, 58 (30.9%) reported investigating misoprostol self-sourced online for home use. Women who investigated misoprostol online were less likely to have had a prior abortion (29.3% vs. 63.1%, p < .01) and more likely to report prior home attempts to end this pregnancy (8.6% vs. 0%, p = .049) (Table 2). Over one-third of women (97 [36.2%]) had had a prior abortion. Overall, four (1.6%) respondents purchased misoprostol, one from a store, one from a clinic, and two who had purchased it from outside the United States.

Table 1.

Demographic data of women presenting to abortion clinics in the Midwest participating in a survey on self-sourcing abortion (N = 276)

Characteristic	All survey participants(n = 276)	Investigated DIY misoprostol online(n = 58)	Online but not DIY misoprostol(n = 130)	Did not investigate online(n = 65)	p-value
Age-(years), mean [SD]	26.7 [6.2]	26.8 [6.2]	26.9 [6.4]	26.5 [6.0]	.93
Race/ethnicity, count (%)	(n = 262)
White	191 (72.9)	44 (75.9)	96 (73.1)	40 (65.6)	.510
Black	31 (11.8)	5 (8.6)	12 (9.6)	12 (19.7)
Hispanic	16 (6.1)	3 (5.2)	7 (5.6)	3 (4.9)
Other	24 (9.2)	6 (10.3)	10 (8.0)	6 (9.8)
Level of education	(n = 269)
≤High school	89 (33.1)	17 (29.3)	42 (32.6)	22 (33.8)	.163
Some college	110 (40.9)	30 (51.7)	44 (34.1)	33 (50.8)
College diploma	55 (20.5)	10 (17.2)	31 (24.0)	9 (13.8)
Post college	15 (5.6)	1 (1.7)	12 (9.3)	1 (1.5)
Household income	(n = 252)
<$10 K	72 (28.6)	20 (36.4)	25 (20.2)	19 (33.9)	.017
$10 K–<$50 K	138 (54.7)	26 (47.3)	77 (62.1)	29 (51.8)
≥$50 K	42 (16.7)	9 (16.4)	22 (17.7)	8 (14.3)
Travelled >50 miles	(n = 273)116 (42.5)	24 (41.4)	56 (43.1)	36 (56.2)	.963
Marital status	(n = 270)
Single	126 (46.7)	22 (37.9)	57 (44.9)	37 (56.9)	.313
Divorced/widowed	14 (5.2)	3 (5.2)	7 (5.5)	3 (4.6)
Married/with partner	130 (48.2)	33 (56.9)	63 (49.6)	25 (38.5)
Gravidity, median {IQR}*	(n = 265)2 {0–3}	2 {0–4}	1 {0–3}	3 {1–5}	.0004
Gestational age (weeks)	(n = 262)
Estimate	8.0 {6.0–10}	7.5 {6.0–10.5}	7.5 {6.0–10.0}	8.0 {6.0–10.0}	.870
By ultrasound	7.0 {6.0–10}	8.0 {6.0–10.5}	7.0 {6.0–10.0}	8.0 {6.0–10.0}	.671
Type of abortion	(n = 264)
Surgical	137 (51.9)	33 (56.9)	63 (50.0)	34 (54.0%)	.664
Medical	127 (48.1)	25 (43.1)	63 (50.0)	29 (46.0)
History prior abortion	(n = 268)
	97 (36.2)	17 (29.3)	33 (25.8)	41 (63.1)	<.0001

*Interquartile range in brackets {}. DIY = do it yourself.

Table 2.

Comparison of those that did not investigate online prior to appointment versus each of the two groups that investigated online in terms of abortion experiences and preferences for an over the counter option for abortion (OTC = over the counter, DIY = do it yourself)

	Did not investigate online prior to appt(n = 65)	Investigated DIY misoprostol online(n = 58)	Online but not DIY misoprostol(n = 130)	No online prior to appointment vs.
				DIY misoprostol online (p-value)	No DIY misoprostol online (p-value)
Had abortion later in gestation than wanted	(n = 57)45.6% (26)	(n = 56)46.4% (26)	(n = 124)43.6% (54)	.99	.95
Previous abortion	(n = 65)63.1% (41)	(n = 58)29.3% (17)	(n = 128)25.8% (33)	.0005	<.0001
Encountered barriers to care	(n = 62)64.5% (40)	(n = 56)64.3% (36)	(n = 123)63.4% (78)	.99	.98
In the past ended pregnancy on your own	(n = 62)14.5% (9)	(n = 56)16.1% (9)	(n = 126)19.8% (25)	.96	.56
Prior to today tried to end pregnancy on your own	(n = 61)0.0% (0)	(n = 58)8.6% (5)	(n = 128)4.7% (6)	.05	.36
In a past or this pregnancy tried to end pregnancy on your own	(n = 60)15.0% (9)	(n = 56)17.9% (10)	(n = 126)23.0% (29)	.87	.33
If there was a safe pill, it should be available OTC or online	(n = 57)50.9% (29)	(n = 55)65.4% (36)	(n = 119)58.8% (70)	.20	.50
Comfortable using OTC abortion option without physician	(n = 57)28.1% (16)	(n = 54)42.6% (23)	(n = 122)32.0% (39)	.19	.81
Likely to use OTC abortion option if needed abortion in future	(n = 58)27.6% (16)	(n = 55)40.0% (22)	(n = 121)35.5% (43)	.27	.45
OTC option would improve abortion access	(n = 59)	(n = 54)	(n = 119)	.38	.007
Yes	37.3% (22)	53.7% (29)	59.7% (71)
Maybe	38.0% (23)	25.9% (14)	31.9% (38)
No	23.7% (14)	20.4% (11)	8.4% (10)

Prior to coming to the clinic, 11 of 252 women (4.4%) tried to end their pregnancy, 3 (1.2%) with a medication. There were 17 instances reported of trying to end a pregnancy in the current pregnancy, or in a prior pregnancy with a variety of means, including Vitamin C, black cohosh (an herbal remedy), dong quai (a medicinal herb), and other herbal remedies, although these may include overlap of a single woman trying multiple methods. Nine of 258 women (3.5%) reported that they had a friend who used misoprostol for abortion at home.

In total, 163 of 251 (64%) women reported barriers to care including those in Figure 1, as well as considerations including childcare and time off work. If these barriers were removed, 110 of 152 (72.4%) respondents reported that they could have presented to care more than one week earlier. This is further demonstrated as women who reported barriers to care presented at a mean gestational age of 8.6 weeks and those without barriers presented at a mean of 7.6 weeks (p = .03). Women who reported barriers to care had higher out of pocket expenses ($412 vs. $260, p = .01) but were similar in terms of race, income, highest level of education, parity, and history of prior abortion. There was also no difference in numbers of women who looked online or found information on misoprostol between those who did and did not report barriers to care.

Figure 1.

Barriers to accessing abortion care cited by women undergoing abortions in the Midwest between May to September 2016. 159 (64.9%) of all women reported barriers to care, denoted below. Women were able to choose all applicable categories

Respondents were of the opinion that an over-the-counter option would have improved access to care in 128/240 (53.3%) of respondents’ opinions. If a safe, over-the-counter abortion option were available, 80/242 (33.0%) reported they would be comfortable trying it and 82/240 (34.2%) would likely use it with a future unwanted pregnancy. Those who had looked online for misoprostol tended towards being more comfortable with a theoretical safe over-the-counter abortion option than those not looking online (42.6% vs. 28.1%, p = .19).

Discussion

The public health successes of abortion legalisation in the US remain tenuous in the current political and healthcare environment where self-induction is re-emerging.[18] With barriers that eliminate access to standard abortion providers and procedures, the use of misoprostol self-administration without a provider has proven to be a safe and efficacious alternative.[15,19]

Our study shows that women in a rural Midwestern state are aware of and investigate these options. The rate of 1.6% of women who bought misoprostol before the presentation is similar to rates reported previously.[8,11] However, at the time the survey was administered, state law was fairly permissive regarding abortion access. Follow-up studies are needed to determine if increasing restrictions and closure of several clinics will increase self-induction rates. The use of misoprostol is likely to be more widespread than our study and similar previous studies suggest, given that it only encompasses women accessing abortion services. Those who obtained misoprostol illicitly and succeeded in self-termination would not likely access clinic care. Given the efficacy of single-agent misoprostol regimens of 76%,[4,5] the four patients who obtained misoprostol prior to coming in may represent a much larger group who obtained misoprostol and never presented to care or had used misoprostol solely because they could not access an in-clinic abortion. The inability to capture the full number of those using misoprostol is a major limitation of our study and other studies which are limited to an abortion clinic setting, introducing participation bias. Another limitation is that we only surveyed English-speaking patients; it is estimated that less than 10% of those who did not participate were because of a language barrier, consistent with the typical patient population at the study clinics. Anecdotally, lower response rates were noted on days when study personnel were unavailable, and the survey was given by clinic staff only.

Women who sought information about misoprostol online were less likely to have had a previous abortion than those who had not looked online and were also more interested in the potential for the availability of over-the-counter options for abortion. This indicates those with a prior termination possess knowledge about the abortion clinic process and are less likely to seek out other options.

This study demonstrates that women investigate and are knowledgeable of means to end pregnancies other than physician supervised abortions. As women gain knowledge of self-induction, demand for self-sourced abortion medication as well as high-quality information, such as what is provided by the Plan C organisation (http://plancpills.org), Women on Web (https://www.womenonweb.org/), or Safe2Choose (https://safe2choose.org/), will likely increase. Given the studied safety and efficacy of medical abortion regimens, this is an opportunity to increase access to safe abortion, but it also raises questions about the safety of taking medications that may not be from regulated sources, the ability of women to estimate gestational age, and the complications that could arise without proper instructions. Studies on online instructions, label comprehension, complication rates, and actual use are needed to guide women who are unable or unwilling to seek traditional abortion services to provide the care they need in a safe environment with legitimate medications and evidence-based protocols.[20] In addition, ensuring that information necessary to find a clinic with abortion services is readily available online is necessary, since that is a primary source where women get their information. While maintaining access to family planning services is essential to ensure women’s safety, this study demonstrates that non-traditional methods of abortion cannot be ignored as part of women’s reproductive health care.