Abstract
Background
Medication abortion has the potential to increase abortion availability, primarily through new provider networks; however, without a better understanding of how and why women make decisions regarding both their abortion method and their provider, expansion efforts may be misguided and valuable resources may be wasted.
Study Design
We undertook an exploratory study to investigate method and provider preferences. Semistructured one-on-one interviews were conducted with 205 abortion clients at three family planning clinics.
Results
Study participants greatly preferred the clinic setting for their abortion; the majority of women in the study would not have gone to their regular physician if they had been given the option. In addition, method choice trumps provider choice for the majority of women who would have preferred their regular provider. Participants who chose the aspiration procedure were more likely to have previous knowledge about the medication method. Travel time was not a predictor of preferring one’s regular physician over the clinic.
Conclusions
Expanding provider networks via the private sector is unlikely to be a panacea. In addition to these efforts, more attention may need to be paid to addressing logistic barriers to access. Physicians offering abortion services need to let their patients know they offer such services prior to their patients’ need for them. Questions remain regarding the information being circulated about medication abortion.
Keywords: Abortion, Medication abortion, Provider preference
1. Introduction
Approximately half of all pregnancies in the United States are unintended, and >40% of these pregnancies end in termination [1]. While the abortion rate has declined slightly in recent years, this service is still sought annually by approximately 2% of women of reproductive age [2]. Although the demand remains high, there are growing concerns about the supply of providers and access to services. For example, only 13% of US counties have an abortion provider; consequently, 34% of women in the United States (49% in the Midwest and 45% in the South) live in counties with no access to abortion, creating a substantial burden for women who must travel a great distance to reach a provider [2]. This burden increases for women living in the 24 states with enforceable mandatory delay laws, many of which require multiple visits [3].
A related concern is the decline, over several decades, in the number of abortion provision sites, as well as in the number of individual clinicians performing the service [2,4]. In addition, since most abortions are performed at abortion clinics [2], women with established health care providers often must seek out a new provider at a new location in order to obtain abortion services. While these various barriers to access may cause financial, logistic and/or emotional hardships, they also contribute to delays in the abortion procedure itself, for which the associated risks have been shown to increase with gestation [5].
Medication abortion has the potential to greatly improve the availability of abortion services by expanding provider networks to include physicians who do not offer aspiration procedures. Diversifying the types of providers might be especially useful in rural areas where family physicians provide the majority of health care [6]. In the handful of states without physician-only abortion laws [3], nurse practitioners and physician assistants could also provide this service. A related benefit to expanding provider networks could be a reduction in the harassment of abortion patients, which is still a fairly common occurrence [7].
Unfortunately, such an expansion has yet to occur, as medication abortion is infrequently available from any provider not already offering aspiration abortion [8,9]. Both logistic and legislative issues quite likely play a large role in this lack of uptake [9–11]. A third possible reason, briefly addressed in prior publications [8] and requiring further exploration, is a perceived lack of demand for medication abortion from women utilizing private health care. If the latter is indeed a barrier to adding medication abortion services, it should be validated and addressed. However, little is known about women’s interest in medication abortion, let alone their willingness to forgo a choice of surgical or medical method. While a few studies of American women have explored reasons for choosing medication abortion (the majority of which were conducted prior to the approval of mifepristone in this country or were performed among women using methotrexate regimens) [12–15], very little is known about why some American women select the aspiration option when choosing to have an abortion [16]. Moreover, to our knowledge, there exist no published data exploring clients’ desires regarding the type of abortion provider (specifically the interest in one’s private or “regular” physician), nor how method and provider choices might intersect. Until there is a better understanding of the clinical decision making of abortion clients, the promise of the use of medication abortion to improve access to services through expanded provider networks is likely to remain unrealized.
2. Materials and methods
An exploratory study was undertaken to investigate how abortion clients select their method of abortion, how they select their abortion provider, and how these two choices interact. The study took place from May through July 2006 at three Planned Parenthood of Greater Iowa (PPGI) clinics that offered both aspiration and medication abortion services. The study was approved by the University of Michigan Health Sciences Institutional Review Board and by the Planned Parenthood Federation of America, Inc. All participants gave written informed consent.
Data were collected through semistructured one-on-one interviews. Interview questions addressed where and when the clients learned about medication abortion, what factors and who may have influenced the choices of method and provider, and whether clients would have preferred to obtain the abortion with their regular physicians.1 A short series of sociodemographic questions was also asked. The interviews were recorded and transcribed. Each transcription was checked in full against the corresponding digital recording to ensure accuracy.
The study population was a convenience sample of women presenting at the clinics for either aspiration or medication abortion. Eligibility included age ≥18 years and fluency in English. All women were interviewed by the first author on the days of their procedures. Aspiration patients were interviewed prior to receiving any pain medication or anesthesia; medication patients were interviewed on the days they were to take mifepristone. Recruitment was conducted by the clinic staff. All eligible women were given an information sheet by either the ultrasound nurse or the laboratory technician. (During the final week of the study, only women seeking medication abortion were recruited for the study.) Patients were later asked by the staff if they were interested in being part of the study. All interested women met with the first author in a private room to learn more about the study and to ask any questions they might have; partners and accompanying adults were asked to wait in the waiting room. Those wishing to continue with the study signed informed consent and were interviewed. Following completion of the interview, each participant was given a US$20 gift certificate. The interviews were confidential, and no personal or identifying information was collected.
A database was created from the interview questions. Data were coded and entered by two separate readers; discrepancies were evaluated by both readers until consensus was reached. Descriptive statistics were produced for the entire study population and by clinic site. Differences by clinic site were assessed using Fisher’s Exact Test, and one-way analysis of variance was used for comparing means. Statistical significance for method-based and provider-based analyses was assessed using Fisher’s Exact Test. Odds ratios (ORs) were calculated with logistic regressions to evaluate predictors of choice/preference. For all logistic regressions, models were run stepwise. Given the exploratory nature of the study, multiple variable combinations were examined to see where there were important connections, starting with logic-based choices of independent variables and adding control variables one by one, as well as in groups. This approach was taken because no previous studies have examined these particular relationships, so none serves to suggest potential associations. One of the important goals of this project is to help determine areas (and variables) of interest to be explored more thoroughly in future studies. All statistical analyses were conducted using Stata software, version 9.2 (Stata Corp., College Station, TX).
3. Results
3.1. Descriptives
During the interview period, 638 patients aged ≥18 years had abortions at the participating clinics. Of these, 205 were interviewed. The response rate is therefore estimated at 32%; however, this percentage is somewhat conservative due to a small number of non-English speakers, estimated at three to four per week (personal communication with clinic managers), who were not eligible for the study. All but one woman who expressed interest in the study signed consent forms and were interviewed.
The average age of the subjects was 25 years, with a range of 18–43 years. The sample was largely White (87%) and predominantly Christian (71%). Three percent of the subjects self-identified as Hispanic or Latina. The age, race and ethnicity profile of the sample was very similar to the profile of the overall clientele seen at PPGI clinics. Two thirds (63%) of the subjects had attended at least some college. Ease of paying monthly expenses was fairly evenly distributed. The majority of subjects resided in Iowa (89%). One-way travel time to the clinic averaged 43 min, with a maximum time of 210 min (Table 1).
Table 1.
Descriptive statistics
| n (%) | Total [n (%)] | |
|---|---|---|
| Residency | 205 (100) | |
| Iowa | 183 (89) | |
| Illinois | 18 (9) | |
| Other | 4 (2) | |
| Race | 205 (100) | |
| White | 179 (87) | |
| Black | 14 (7) | |
| Multi | 8 (4) | |
| Other | 4 (2) | |
| Hispanic/Latina | 7 (3) | 205 (100) |
| Religion | 204 (100) | |
| Christian (nonspecified) | 59 (29) | |
| Christian (specified, Protestant) | 46 (23) | |
| Christian (specified, Catholic) | 39 (19) | |
| Other | 13 (6) | |
| None | 47 (23) | |
| Education | 205 (100) | |
| Less than high school | 11 (5) | |
| High school/general equivalency diploma | 53 (26) | |
| Trade school | 11 (5) | |
| Some college | 89 (43) | |
| Associate’s degree | 16 (8) | |
| College degree or more | 25 (12) | |
| Expenses | 205 (100) | |
| Difficult | 80 (39) | |
| In-between | 50 (24) | |
| Easy | 75 (37) | |
| Insurance that covers abortion | 205 (100) | |
| Yes | 49 (24) | |
| No | 126 (61) | |
| Unsure | 30 (15) | |
| Had previous abortion | 203 (100) | |
| Yes | 68 (34) | |
| No | 135 (67) | |
| Age in years [mean (range)] | 25 (18–43) | 205 |
| Travel time in minutes [mean (range)] | 43 (2–210) | 203 |
| Gestational age in weeks [mean (range)] | 7 (4–16) | 202 |
Approximately one fourth of the participants (24%) reported having insurance that covers abortion services. Of these women, 84% planned to use it on that day. Of the 30 women who were not sure if they had insurance that would cover the abortion (15% of the total sample), only a small number were going to try to use it. Finally, three women reported having insurance that covered aspiration, but not medication abortion.
The average self-reported gestation duration was 7.4 weeks, with a range of 3.5–16 weeks. Three quarters of the women (74%) had a gestational length of ≤8 weeks, and 95% had a gestational length of ≤12 weeks. Approximately one third (34%) of the participants had had a prior induced abortion, and 8% of the participants had had a prior medication abortion. In addition, 45% of the women had heard about medication abortion prior to the current pregnancy, and 25% had known someone who had used it.
The participation response rate did not vary significantly by clinic: 27% of abortion patients at the Quad Cities clinic were interviewed, 36% in Iowa City, and 32% in Des Moines (p=1.0). Descriptive statistics were also compared by site. Overall, the differences between study populations by clinic were minimal. The only variable with a strong statistical difference was state of residence; Des Moines and Iowa City participants live almost entirely in Iowa (98% and 90%, respectively), while over a third of Quad Cities participants (38%) live in Illinois (p<.001).
3.2. Selection of abortion method
The study population was divided almost evenly between the two method options; 104 women had an aspiration abortion, and 101 women had a medication abortion. However, over one third (39%) of the aspiration subjects came to the clinic too late to choose medication abortion (>8 weeks’ gestation, as determined by ultrasound), and 5% of the medication subjects were too early to choose surgery (sac was too small for aspiration, as also determined by ultrasound). Women who came to the clinic too late to choose medication abortion were less likely to have had a previous abortion (and none had had a medication abortion), less likely to report having insurance that covered abortion services and more likely to report that covering monthly expenses was difficult. However, these women were just as likely to have previously heard about medication abortion as women who had their choice of methods.
Among the women who had their choice of methods and selected aspiration, almost three quarters (71%) reported doing so because the method is faster. Other common reasons included believing aspiration to be a less painful method (35%) and not wanting to be at home for the abortion (27%) (Table 2). Patients who had their choice and selected medication abortion commonly did so to be home for the abortion (34%), to avoid the aspiration procedure (20%) and/or because they believed that the method was less invasive (20%), less painful (20%), less frightening (19%) or more natural (19%) (Table 2).
Table 2.
Reasons for method selection a
| Within category [n (%)] b | Total [n (%)] | |
|---|---|---|
| Choice of methods: aspiration | 63 (31) | |
| Faster | 45 (71) | |
| Less painful | 22 (35) | |
| Not at home | 17 (27) | |
| Physician present | 11 (17) | |
| Safer | 10 (16) | |
| No follow-up visit | 7 (11) | |
| Choice of methods: medication | 96 (47) | |
| At home | 33 (34) | |
| Emotional reasons | 20 (21) | |
| To avoid surgery | 19 (20) | |
| Less invasive | 19 (20) | |
| Less painful | 19 (20) | |
| Less scary | 18 (19) | |
| More natural | 18 (19) | |
| Safer | 16 (17) | |
| Cheaper | 15 (16) | |
| Easier | 15 (16) | |
| Quicker at clinic | 12 (13) | |
| Preferred method: aspiration c | 77 (38) | |
| Faster | 50 (65) | |
| Less painful | 24 (31) | |
| Not at home | 19 (25) | |
| Physician present | 14 (18) | |
| Safer | 11 (14) | |
| No follow-up visit | 8 (10) | |
| Preferred method: medication c | 122 (60) | |
| At home | 38 (31) | |
| Less invasive | 27 (22) | |
| Less painful | 26 (21) | |
| To avoid surgery | 25 (20) | |
| Easier | 25 (20) | |
| Emotional reasons | 24 (20) | |
| Less scary | 20 (16) | |
| Safer | 20 (16) | |
| More natural | 19 (16) | |
| Cheaper | 17 (14) | |
| Quicker at clinic | 15 (12) | |
| Preferred method unknown | 6 (3) |
Only reasons given by ≥10% of each subgroup are included.
Percentages add up to >100 because participants gave multiple responses.
Includes women who preferred this method, regardless of what they actually used.
Women who did not have a choice of methods (based on their ultrasound results) were asked which method they would have preferred and why, had they been given a choice. Of the 41 aspiration clients who came in too late to have a choice of methods, over half (59%) would have preferred the medication option, about one quarter (27%) would have chosen aspiration and the rest were unsure. Of the five medication clients who came in too early to have the aspiration option, three would have preferred aspiration, and two would have stayed with medication. The reasons for method preference given by the “no-choice” women were combined with the reasons given by the “choice-of-methods” women to create an overall “preferred-method” category. The breakdown of “preferred-method” reasons is similar to that of the “choice-of-methods” group (Table 2).
Study participants were also asked whether anyone other than themselves was involved in selecting the method. The vast majority (70%) reported that they made the method decision on their own. Nineteen percent said that their husband or partner was involved, and 16% reported the involvement of a friend or family member, almost all of whom were mothers or sisters. Only 2% reported that clinic staff, including physicians, were involved in the method decision.
3.3. Predictors of method selection
Sociodemographic and behavioral factors related to method selection were evaluated using both the “choice-of-methods” and the “preferred-method” definitions of selection. Only two variables were statistically significant. Women who chose/preferred aspiration abortion were more likely to use their medical insurance (if they reported having it) to cover the procedure than were those who chose/preferred medication abortion (p=.02 and p=.03, respectively). Aspiration patients reported significantly greater gestations (p<.001), even when the comparison was confined to those having terminations at ≤8 weeks (p=.01).
Predictors of method choice were then assessed using logistic regression after correlations of all potential covariates had been run to check for collinearity. In the first model, the odds of preferring aspiration abortion over medication abortion were twice as high for women who knew about medication abortion before their pregnancy than for those who did not [OR=2.02; 95% confidence interval (95% CI)= 1.06–3.85] (Table 3). In a second model, for each week’s increase in gestational length, the odds of women choosing aspiration abortion over medication increased by nearly threefold (OR=2.69; 95% CI=1.80–4.03) (Table 3). Restricting the method choice model to include only women at ≤8 weeks’ gestation does not substantially alter these results.2
Table 3.
Method selection logistic regressions
| Model | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) |
|---|---|---|---|---|
| Method preference: aspiration | ||||
| Knew about medication abortion | 1.64 (0.92–2.91) | 2.02 (1.06–3.85) | 2.08 (1.09–3.99) | 2.21 (1.11–4.37) |
| Did not know about medication abortion (reference) | 1.00 | 1.00 | 1.00 | 1.00 |
| Previous medication abortion | 0.53 (0.17–1.72) | 0.54 (0.16–1.77) | 0.44 (0.10–1.87) | |
| No previous medication abortion (reference) | 1.00 | 1.00 | 1.00 | |
| Previous aspiration abortion | 0.90 (0.46–1.76) | 0.84 (0.42–1.68) | 0.87 (0.42–1.79) | |
| No previous aspiration abortion (reference) | 1.00 | 1.00 | 1.00 | |
| Age (continuous) | 1.02 (0.97–1.08) | 1.01 (0.95–1.07) | ||
| White | 0.80 (0.32–1.99) | 1.03 (0.39–2.74) | ||
| Non-White (reference) | 1.00 | 1.00 | ||
| Some college or more | 0.70 (0.37–1.32) | 0.66 (0.34–1.27) | ||
| High school diploma/general equivalency diploma or less (reference) | 1.00 | 1.00 | ||
| Expenses difficult | 0.81 (0.44–1.49) | 0.78 (0.42–1.37) | ||
| Expenses easy or in-between (reference) | 1.00 | 1.00 | ||
| Knows someone who had a medication abortion | 0.75 (0.41–1.37) | |||
| Does not know anyone who had a medication abortion (reference) | 1.00 | |||
| Method choice: aspiration | (≤8 weeks’ gestation) | |||
| Gestational age in weeks | 2.69 (1.80–4.03) | 2.93 (1.89–4.56) | 3.03 (1.94–4.74) | 2.76 (1.69–4.51) |
| Previous abortion | 1.60 (0.70–3.68) | 1.29 (0.53–3.17) | 1.30 (0.53–3.15) | |
| No previous abortion (reference) | 1.00 | 1.00 | 1.00 | |
| Age (continuous) | 1.00 (0.93–1.09) | 1.01 (0.93–1.09) | 1.001 (0.93–1.09) | |
| White | 1.32 (0.40–4.38) | 1.38 (0.42–4.58) | 1.44 (0.42–4.91) | |
| Non-White (reference) | 1.00 | 1.00 | 1.00 | |
| Some college or more | 1.25 (0.55–2.82) | 1.21 (0.53–2.73) | 1.24 (0.54–2.81) | |
| High school diploma/general equivalency diploma or less (reference) | 1.00 | 1.00 | 1.0 | |
| Expenses difficult | 0.53 (0.24–1.17) | 0.52 (0.23–1.15) | 0.53 (0.24–1.19) | |
| Expenses easy or in-between (reference) | 1.00 | 1.00 | 1.00 | |
| Knew about medication abortion | 1.70 (0.73–3.95) | 1.69 (0.73–3.92) | ||
| Did not know about medication abortion (reference) | 1.00 | 1.00 | ||
3.4. Accessing “regular” physicians for abortion services
Almost two thirds (62%) of the participants had a PC/FP, and approximately one fourth (27%) had an obstetrician gynecologist (OB/GYN) (Table 4). Seventy-two percent had one or both of these types of physicians, and these women were more likely to be older and to have a health insurance that covers abortion than women with neither of these regular physicians. None of the women with a PC/FP and only one woman with an OB/GYN reported that this regular physician provides abortion. Of the women with a PC/FP, 69% said that their physician did not provide abortion, and the rest (31%) said they did not know. Of the women with an OB/GYN, 57% said that their physicians did not offer abortions, and 41% did not know.
Table 4.
Reasons for provider preference a
| n (%) b | Total [n (%)] | |
|---|---|---|
| Participant with a PC/FP or an OB/GYN | 147 (72) | |
| PC/FP | 128 (62) | |
| Would want abortion services | 26 (20) | |
| Would not want abortion services | 83 (65) | |
| Unsure about wanting abortion services | 19 (15) | |
| Would be comfortable talking about abortion | 56 (47) | |
| Would not be comfortable talking about abortion | 54 (45) | |
| Unsure about feeling comfortable | 9 (8) | |
| PC/FP: would want services | 26 (20) | |
| Knows physician | 17 (65) | |
| Comfortable with physician | 13 (50) | |
| Physician has records/knows medical history | 4 (15) | |
| Trusts physician | 3 (12) | |
| Physician is closer to home | 3 (12) | |
| PC/FP: would not want services | 83 (65) | |
| Privacy/confidentiality | 25 (30) | |
| Comfort level | 25 (30) | |
| Physician knows family | 19 (23) | |
| Planned Parenthood’s reputation/expertise | 10 (12) | |
| Fear of being judged | 10 (12) | |
| Lives in a small town | 10 (12) | |
| OB/GYN | 56 (27) | |
| Would want abortion services | 20 (36) | |
| Would not want abortion services | 23 (42) | |
| Unsure | 12 (22) | |
| Would be comfortable talking about abortion | 31 (58) | |
| Would not be comfortable talking about abortion | 19 (36) | |
| Unsure | 3 (6) | |
| OB/GYN: would want services | 20 (36) | |
| Comfortable with physician | 10 (50) | |
| Knows physician | 9 (45) | |
| Physician has records/knows medical history | 4 (20) | |
| Physician is closer to home | 3 (15) | |
| OB/GYN: would not want services | 23 (42) | |
| Comfort level | 9 (39) | |
| Privacy/confidentiality | 7 (30) | |
| Fear of being judged | 5 (22) | |
| Physician is pro-life | 3 (13) |
Only reasons given by ≥10% of each subgroup are included.
Percentage within the category; percentages add up to >100 because of multiple responses.
When asked if they would have preferred to go to their PC/FP (assuming that the PC/FP in fact provided the service), 20% of the women said yes, 65% said no and 15% were unsure. Moreover, less than half of these women (47%) said they would have felt comfortable talking about abortion and abortion services with this physician (Table 4). Of those who would have gone, the most common reasons stated were that they knew their physician (65%) and that they were comfortable with their regular physician (50%). The most common reasons for not wanting to go to one’s PC/FP were privacy and/or confidentiality (30%), not feeling comfortable (30%) and that the physician knew the woman’s family (23%) (Table 4).
Thirty-six percent of women with an OB/GYN would have preferred going to that physician, 42% would not have wanted to go and 22% were unsure. In addition, only slightly more than half (58%) felt comfortable enough to talk about abortion with this physician. Of those who would have gone, the main reasons included feeling comfortable with this regular physician (50%) and knowing this physician (45%). The reasons given for not wanting to seek abortion from one’s OB/GYN included not feeling comfortable (39%), issues of privacy and/or confidentiality (30%) and not wanting to be judged (22%) (Table 4).
3.5. Predictors of regular physician preference
Sociodemographic and behavioral factors related to whether women would have wanted to access abortion through their regular physician were evaluated separately for PC/FPs and for OB/GYNs. Women who would have preferred going to their PC/FP were less likely to report having a health insurance that covers abortion (p=.01) and more likely to have looked into other options for where to have the abortion (p=.05). Women who would have preferred their OB/GYN were more likely to be older (p=.01) and to have spent a longer amount of time traveling to the clinic (p=.05). They were also less likely to have chosen Planned Parenthood clinics because they had been there before (p=.02).
Logistic regressions were used to determine the predictors of preferring a regular physician after collinearity had been assessed by correlating all potential covariates. In the PC/FP model, having a health insurance that covers abortion decreased the odds of wanting abortion services from a regular physician by over fivefold (OR=0.19; 95% CI=0.05–0.76). In addition, being Catholic increased the odds by over threefold (OR=3.68; 95% CI=1.00–13.49) (Table 5). In the OB/GYN model, age was a significant predictor of regular physician preference; for each year increase, the odds of women wanting to access abortion services through their regular physician increased by 18% (OR=1.18; 95% CI=1.01–1.36) (Table 5).
Table 5.
Provider preference logistic regressions
| Model | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) |
|---|---|---|---|---|
| Preference for PC/FP | ||||
| Has insurance that covers abortion | 0.19 (0.05–0.71) | 0.19 (0.05–0.76) | 0.19 (0.05–0.76) | 0.20 (0.05–0.86) |
| Does not have insurance that covers abortion (reference) | 1.00 | 1.00 | 1.00 | 1.00 |
| Catholic | 3.24 (1.01–10.40) | 3.68 (1.00–13.49) | 4.78 (1.19–19.19) | |
| Any religion other than Catholic or nonreligious (reference) | 1.00 | 1.00 | 1.00 | |
| Age (continuous) | 0.93 (0.83–1.04) | 0.90 (0.79–1.01) | ||
| White | 2.04 (0.38–11.07) | 2.72 (0.45–16.37) | ||
| Non-White (reference) | 1.00 | 1.00 | ||
| Some college or more | 1.64 (0.54–5.03) | 1.35 (0.37–4.94) | ||
| High school diploma/general equivalency diploma or less (reference) | 1.00 | 1.00 | ||
| Expenses difficult | 1.40 (0.47–4.15) | 1.59 (0.48–5.27) | ||
| Expenses easy or in-between (reference) | 1.00 | 1.00 | ||
| Previous abortion | 2.47 (0.69–8.90) | |||
| No previous abortion (reference) | 1.00 | |||
| Method preference: aspiration | 0.45 (0.13–1.53) | |||
| Method preference: medication (reference) | 1.00 | |||
| Preference for OB/GYN | ||||
| Age (continuous) | 1.18 (1.04–1.34) | 1.21 (1.05–1.39) | 1.18 (1.01–1.36) | 1.20 (1.02–1.41) |
| White | 6.36 (0.90–44.76) | 5.31 (0.71–39.66) | 4.91 (0.65–37.08) | |
| Non-White (reference) | 1.00 | 1.00 | 1.00 | |
| Catholic | 2.05 (0.29–14.65) | 1.93 (0.23–16.30) | ||
| Any religion other than Catholic or nonreligious (reference) | 1.00 | 1.00 | ||
| Some college or more | 1.92 (0.39–9.38) | 2.19 (0.39–12.47) | ||
| High school diploma/general equivalency diploma or less (reference) | 1.00 | 1.00 | ||
| Expenses difficult | 0.60 (0.13–2.83) | 0.76 (0.14–4.15) | ||
| Expenses easy or in-between (reference) | 1.00 | 1.00 | ||
| Previous abortion | 0.79 (0.13–4.63) | |||
| No previous abortion (reference) | 1.00 | |||
| Method preference: aspiration | 0.21 (0.04–1.24) | |||
| Method preference: medication (reference) | 1.00 | |||
3.6. Intersection of method selection and provider preference
After separately exploring method selection and provider preference, we examined the intersection of the two. Differences in wanting to obtain an abortion from one’s regular physician were evaluated by method choice and method preference groups. A similar comparison regarding comfort level in discussing abortion with one’s regular physician was conducted. Within this study sample, neither interest in utilizing one’s regular physician for abortion nor comfort in talking about abortion with this provider differed significantly according to which abortion method was chosen or which method was preferred.
We also explored the possibility of one selection issue (method or provider) “overcoming” the other. That is, we tested the possibility that the desire to use a particular method might outweigh the interest in potentially going to one’s regular physician, or that the desire to utilize one’s regular physician might outweigh the preference for one particular method. Approximately two thirds of participants who would have opted to utilize their regular physician (67% of participants with an OB/GYN and 61% of participants with a PC/FP) said that specific method availability from the regular physician would affect their decision to go to that physician. In other words, method preference would trump provider choice. Cross-tabs were run to see whether this hierarchy was more likely to be established by women preferring a particular abortion method; it was not.
Finally, it is worth noting that 14% of all study participants included method availability in their list of reasons for choosing the Planned Parenthood clinic. All but one of these women (96%) both chose and preferred the medication method.
4. Discussion
This research study is exploratory and, therefore, primarily descriptive. Given the paucity of studies that have investigated abortion method preferences and the absence of published work on abortion provider preferences, these data may offer some important insights into these arenas. Additionally, while the study was conducted in only one state, there are implications for many parts of the country.
Expanding provider networks into the private sector (primarily via provision of medication abortion) is a commonly discussed solution for improving access to abortion services. However, this study demonstrates that recruiting providers not associated with abortion clinics is unlikely to solve the problem; many women in this sample reported not wanting to access abortion services through their regular family physician or gynecologist. Therefore, more attention may need to be paid to addressing logistic barriers to service at family planning clinics; increasing days and hours of service, providing transportation to and from the clinics, helping to defray the expenses and even setting up additional branches might be more effective ways to help women access services more easily.
In addition, physicians who want to offer this service will need to let their patients know that they are offering such service. Given the lack of comfort in discussing abortion with one’s physician, many women are unlikely to ask about its availability. As was evident with emergency contraception, the “don’t-ask, don’t-tell” policy of medical services does not work, particularly around stigmatized products [17]. While physicians may fear (with good reason) being labeled an abortion doctor, being trained and ready to provide medication (or aspiration) abortion may be a waste of time and money if their patients do not know about its availability prior to their need for it.
Method choice trumps provider choice for the majority of women who would have preferred their regular provider. This effect is equally robust for users of medication and aspiration abortions, suggesting that method availability is more important than provider preference. Clinicians who currently offer only one method may want to consider partnering with someone who offers the other, in order to best serve their clientele.
Participants who chose the aspiration procedure were more likely to have previous knowledge of the medication method than those who chose medication abortion. This counterintuitive finding suggests that the information that women receive about medication abortion may be primarily negative or disconcerting in some way. Given the high levels of satisfaction reported by medication abortion users in past studies [12,18,19], this result raises questions about the information being provided in medical facilities, on health Web sites and in the media. It would be useful to further investigate the roots of this adverse reaction to gain a better understanding of how their knowledge of medication abortion led these women to prefer aspiration.
Surprisingly, travel time was not a predictor of preferring one’s regular physician for the abortion services. This finding is particularly interesting given that studies discussing abortion access have often focused on the distance that women must travel in order to obtain the services and on the hardship that this might entail. Moreover, travel time was not a factor in method preference, even though women using the medication method are required to return for a follow-up visit. These results do not necessarily minimize the inconvenience of long travel times; rather, they suggest that other issues may be of greater importance to many women seeking an abortion.
There are limitations to this study. First, the participants were not selected randomly and, thus, the findings might not be generalizable to the clinics’ abortion population as a whole, to women attending other clinics and/or to women outside the clinic setting, outside this region of the Midwest or in more urban locales. The omission of minors is meaningful, as teens under the age of 18 years may face greater barriers than older women in accessing abortion services for reasons such as cost, confidentiality and travel time. Similarly, non-English speakers may have a harder time navigating access to health care, including abortion services, and, thus, their opinions and attitudes might differ significantly from English speakers. The power of the study is limited. Due to a relatively small sample size, small differences in common outcomes were hard to detect. However, the results do provide initial insights into how women choose their abortion method and provider, and serve as a good starting point for future broader empirical work.
Acknowledgments
This study was funded by Danco Laboratories, LLC. Thanks are expressed to Rachel Snow for comments on an earlier draft of this study.
Footnotes
Participants who reported having any “regular physicians” (allowing each woman her own interpretation of “regular”) were asked detailed questions about any primary care/family physicians (PC/FPs) and/or gynecologists whom they might have. Other clinician types were not discussed, as Iowa and Illinois have physician-only laws for abortion provision.
Only women who reported not having a choice of methods were coded as such. It is possible that aspiration clients at >8 weeks’ gestation who indicated that they had a choice (n=13) made their method decision prior to the medication deadline.
References
- 1.Finer LB, Henshaw SK. Disparities in rates of unintended pregnancy in the United States, 1994 and 2001. Perspect Sex Reprod Health. 2006;38:90–6. doi: 10.1363/psrh.38.090.06. [DOI] [PubMed] [Google Scholar]
- 2.Finer LB, Henshaw SK. Abortion incidence and services in the United States in 2000. Perspect Sex Reprod Health. 2003;35:6–15. doi: 10.1363/3500603. [DOI] [PubMed] [Google Scholar]
- 3.NARAL Pro-Choice America Foundation, NARAL Pro-Choice America. Who decides? The status of women’s reproductive rights in the United States. 16. Washington, DC: NARAL Pro-Choice America; 2007. [Google Scholar]
- 4.Grimes DA. Clinicians who provide abortions: the thinning ranks. Obstet Gynecol. 1992;80:719–23. [PubMed] [Google Scholar]
- 5.National Abortion Federation (NAF) Clinical policy guidelines. Washington, DC: National Abortion Federation (NAF); 2005. [Google Scholar]
- 6.Prine LW, Lesnewski R. Medication abortion and family physicians’ scope of practice. J Am Board Fam Pract. 2005;18:304–6. doi: 10.3122/jabfm.18.4.304. [DOI] [PubMed] [Google Scholar]
- 7.Harper CC, Henderson JT, Darney PD. Abortion in the United States. Annu Rev Public Health. 2005;26:501–12. doi: 10.1146/annurev.publhealth.26.021304.144351. [DOI] [PubMed] [Google Scholar]
- 8.Coeytaux F, Moore K, Gelberg L. Convincing new providers to offer medication abortion: what will it take? Perspect Sex Reprod Health. 2003;35:44–7. doi: 10.1363/3504403. [DOI] [PubMed] [Google Scholar]
- 9.Joffe C, Weitz TA. Normalizing the exceptional: incorporating the “abortion pill” into mainstream medicine. Soc Sci Med. 2003;56:2353–66. doi: 10.1016/s0277-9536(02)00240-x. [DOI] [PubMed] [Google Scholar]
- 10.Jones BS, Heller S. Providing medication abortion: legal issues of relevance to providers. J Am Med Women’s Assoc. 2000;55(Suppl 3):145–50. [PubMed] [Google Scholar]
- 11.Borgmann CE, Jones BS. Legal issues in the provision of medication abortion. Am J Obstet Gynecol. 2000;183(Suppl 2):S84–S94. doi: 10.1067/mob.2000.108229. [DOI] [PubMed] [Google Scholar]
- 12.Winikoff B, Ellertson C, Elul B, Sivin I. Acceptability and feasibility of early pregnancy termination by mifepristone–misoprostol. Arch Fam Med. 1998;7:360–6. doi: 10.1001/archfami.7.4.360. [DOI] [PubMed] [Google Scholar]
- 13.Creinin MD, Park M. Acceptability of medical abortion with methotrexate and misoprostol. Contraception. 1995;52:41–4. doi: 10.1016/0010-7824(95)00122-q. [DOI] [PubMed] [Google Scholar]
- 14.Creinin MD, Burke AE. Methotrexate and misoprostol for early abortion: a multicenter trial. Acceptability Contraception. 1996;54:19–22. doi: 10.1016/0010-7824(96)00114-x. [DOI] [PubMed] [Google Scholar]
- 15.Elul B, Pearlman E, Sorhaindo A, Simonds W, Westoff C. In-depth interviews with medical abortion clients: thoughts on the method and home administration of misoprostol. J Am Med Women’s Assoc. 2000;55(Suppl 3):169–72. [PubMed] [Google Scholar]
- 16.Harvey SM, Beckman LJ, Satre SJ. Choice of and satisfaction with methods of medical and surgical abortion among US clinic patients. Fam Plann Perspect. 2001;33:212–6. [PubMed] [Google Scholar]
- 17.Trussell J, Ellertson C, Stewart F, Raymond EG, Shochet T. The role of emergency contraception. Am J Obstet Gynecol. 2004;190:S30–8. doi: 10.1016/j.ajog.2004.01.063. [DOI] [PubMed] [Google Scholar]
- 18.Creinin MD, Fox MC, Teal S, Chen A, Schaff EA, Meyn LA MOD Study Trial Group. A randomized comparison of misoprostol 6 to 8 hours versus 24 hours after mifepristone for abortion. Obstet Gynecol. 2004;103:851–9. doi: 10.1097/01.AOG.0000124271.23499.84. [DOI] [PubMed] [Google Scholar]
- 19.Shannon CS, Winikoff B, Hausknecht R, et al. Multicenter trial of a simplified mifepristone medical abortion regimen. Obstet Gynecol. 2005;105:345–51. doi: 10.1097/01.AOG.0000152003.94320.59. [DOI] [PubMed] [Google Scholar]