WOMEN’S HEALTH

Medical Students’ Intentions to Seek Abortion Training and to Provide Abortion Services in Future Practice Daniel T. Myran, MD,1 Caitlin L. Carew, MD,1 Jingyang Tang, BSc (Hons),1 Helena Whyte, BSc (Hons),1 William A. Fisher, PhD2,3 Schulich School of Medicine and Dentistry, Western University, London ON

1

Department of Psychology, Western University, London ON

2

Department of Obstetrics and Gynaecology, Western University, London ON

3

Abstract Objectives: Lack of providers is a barrier to accessing abortion in Canada. The factors influencing the number of abortion providers are poorly understood. In this study, we assessed the attitudes and intentions of medical students towards abortion training and provision to gain insight into the future supply of abortion providers. Methods: We surveyed first, second, and third year medical students at an Ontario university to determine their intentions to train in and provide abortion services during different stages of training and in future practice. We assessed students’ attitudes and intentions towards training in and providing abortions, their perceptions of social support, their perceived ability to receive training in and to provide abortion services, and their attitudes towards the legality of abortion. Results: Surveys were completed by 337 of 508 potential respondents (66.7%). The responses indicated that the students in the survey held relatively positive attitudes towards the legality and availability of abortion in Canada. Respondents had significantly more positive attitudes towards first trimester medical abortions (and a greater intention to provide them) than towards second trimester surgical abortions. Thirty-five percent of students planned to enter a specialty in which they could perform abortions, but fewer than 30% of these students planned to provide any type of abortion. Intentions to provide abortions were correlated with positive attitudes toward abortion in general and greater perceived social support for abortion provision. Conclusion: A small proportion of students sampled intended both to enter a specialty in which abortion would be within the scope of practice and to provide abortion services. Lack of perceived social Key Words: Women’s health, medical education, health policy, abortion, theory of reason action (TRA), theory of planned behaviour (TPB) Competing Interests: None declared. Received on June 2, 2014 Accepted on November 28, 2014

236 l MARCH JOGC MARS 2015

support for providing abortions and the perceived inability to obtain abortion training or to logistically provide abortions were identified as two potentially modifiable barriers to abortion provision. We propose increasing education on abortion provision and creating policies to promote medical abortion as a method of improving access to abortion across Canada.

Résumé Objectifs : Au Canada, l’accès à des services d’avortement est entravé par le manque de fournisseurs de soins étant en mesure d’offrir de tels services. Les facteurs qui influencent le nombre de fournisseurs de services d’avortement sont mal compris. Dans le cadre de cette étude, nous avons évalué les attitudes et les intentions des étudiants de médecine en ce qui concerne la formation en matière d’avortement et l’offre de tels services dans le cadre de leur future pratique, et ce, pour en connaître davantage au sujet de nos futurs effectifs dans ce domaine. Méthodes : Nous avons sondé les étudiants de médecine de première, de deuxième et de troisième année d’une université ontarienne, à différents stades de leur formation (et en leur demandant de remplir un questionnaire traitant de la future pratique qu’ils envisageaient), en vue de déterminer leurs intentions en ce qui concerne l’obtention d’une formation en matière d’avortement et l’offre de services d’avortement. Nous avons évalué les attitudes et les intentions de ces étudiants en ce qui concerne l’obtention d’une formation en matière d’avortement et l’offre de services d’avortement, leurs perceptions en ce qui a trait au soutien social, leur capacité subjective de recevoir une formation en matière d’avortement et d’offrir des services d’avortement, et leurs attitudes envers la légalité de l’avortement. Résultats : Trois cent trente-sept des 508 répondants potentiels (66,7 %) ont répondu aux questionnaires. Les réponses indiquent que les répondants adoptaient des attitudes relativement positives envers la légalité et la disponibilité de l’avortement au Canada. Les répondants adoptaient des attitudes considérablement plus positives envers les avortements médicaux au premier trimestre (et un plus grand nombre d’entre eux avaient l’intention d’offrir de tels services) qu’envers les avortements chirurgicaux menés au deuxième trimestre. Trente-cinq pour cent des étudiants avaient l’intention de choisir une spécialité dans le cadre de laquelle la

Medical Students’ Intentions to Seek Abortion Training and to Provide Abortion Services in Future Practice

tenue d’avortements serait possible; toutefois, moins de 30 % de ces étudiants avaient l’intention d’offrir quelque type de services d’avortement que ce soit. Les intentions d’offrir des services d’avortement étaient en corrélation avec les attitudes positives envers l’avortement en général et avec la perception d’un soutien social accru envers l’offre de services d’avortement. Conclusion : Une faible proportion de l’échantillon d’étudiants analysé avait l’intention de choisir une spécialité dont le champ d’activité englobe la tenue d’avortements et d’offrir des services d’avortement. L’absence subjective de soutien social envers l’offre de services d’avortement et l’incapacité subjective d’obtenir une formation en matière d’avortement ou d’offrir des services d’avortement au plan logistique ont été identifiés comme étant deux obstacles potentiellement modifiables pour ce qui est de l’offre de services d’avortement. Pour assurer l’amélioration de l’accès à l’avortement d’un bout à l’autre du Canada, nous proposons l’augmentation des efforts d’éducation au sujet de l’offre de services d’avortement et la création de politiques visant la promotion de l’avortement médical. J Obstet Gynaecol Can 2015;37(3):236–244

INTRODUCTION

A

bortion in Canada has not been subject to legal regulation since the 1988 R. v. Morgentaler Supreme Court decision.1 Canadian women’s access to abortion, however, remains variable and is frequently constrained.2–4 Although national Canadian data are not available, the number of abortion providers is thought to be in decline.2 The number of physicians providing abortion services in rural British Columbia dropped by over 60% between 1998 and 2005.5,6 The number of abortion providers in the United States is known to have declined steadily across recent decades. In 1992, there were 2380 abortion providers in the United States, a number that decreased by 24.9% to 1787 by 2005.7 In addition, a significant number of abortion providers in the United States have retired or are nearing the age of retirement.8 Studies from the United States and the United Kingdom have found that medical students and recent medical graduates appear to be less willing than more experienced physicians to provide abortion services.9,10 These trends suggest both a current and a future shortage of abortion providers, with the potential to further decrease women’s access to abortion. A 2006 survey of Canadian residency programs in obstetrics and gynaecology found that all programs provided either mandatory or elective opportunities to obtain training in abortion.11 Studies of obstetrics and gynaecology residents in Canada and the United States have found that the most significant predictors of providing abortions in practice include personal beliefs about abortion, having an intention before entering residency to provide abortions, and the number of abortions performed during residency.12,13 Medical students’ attitudes towards abortion provision

and the amount of exposure to abortion they receive during training are potentially important influences on their intention and capacity to provide abortion services.10 Attitudes towards an action, perceived social norms favouring or opposing the action, and perceived ability to carry out the action have been extensively validated as strong predictors of future behaviour, in research based upon the theory of reasoned action (TRA) and theory of planned behaviour (TPB).14–16 These theoretical models of the determinants of health behaviour may be applied to understand factors that contribute to medical students’ intentions to train in abortion practices and provide abortion services in future practice. In this study we assessed medical students’ interest in acquiring undergraduate and postgraduate training to provide medical and surgical abortion and their intentions to provide these services in future practice settings. We assessed students’ attitudes towards abortion, their perceptions of social support, and perceived behavioural control (perceptions of their ability to actually provide abortion services in the future) as correlates of their intentions to train in and to provide abortion services in future practice. Motivating factors that students identified as pivotal to their plans to provide or not provide abortion services in their future practice were also identified. METHODS

To better understand factors that may influence medical students’ intentions to provide abortions, we developed a survey instrument following the structure specified in two theoretical models for the prediction of health behaviour.14–16 The theory of reasoned action and the theory of planned behaviour assert that an individual’s intention to engage in a health behaviour is the strongest predictor of actually engaging in this behaviour subsequently.14–16 Intention, in turn, is a function of an individual’s attitude towards performing that behaviour, the individual’s perception of social support or opposition in relation to the behaviour, and the individual’s perception of his or her capability to actually carry out the behaviour. The TRA and the TPB have been well validated in multiple areas of health behaviour and related research,14–16 but the theories have not been employed in research to explore influences on intentions to provide abortion services. To identify concerns specific to medical students, a preliminary survey was conducted with 20 medical students studying at an Ontario university. The students were asked 10 open-ended questions about positive and negative aspects of abortion provision, their perception of social support or MARCH JOGC MARS 2015 l 237

WOMEN’S HEALTH

rejection of abortion provision, and their perceived ability to train in and provide abortions. From these responses, a 74-item survey was developed in consultation with an expert in the field of reproductive health research and survey design. The resulting survey was piloted a second time with 10 medical students to ensure that the survey assessed the constructs of interest and was revised accordingly. This approach to survey design has been used in previously validated TRA and TPB research.14–17 Attitudes towards abortion training and provision were assessed using statements that included “For me, observing or assisting in abortion provision during clerkship would be . . .” followed by a series of Likert scales on which participants indicated their agreement that it would be “good or bad,” “wrong or right,” “necessary or unnecessary,” to personally carry out the abortion-related activity in question. Perceptions of social support were assessed using a seven-point Likert scale that stated “Most people who are important to me think that I (1 = definitely should not, 7 = definitely should) assist in abortion provision during [various stages of medical training].” Additionally, participants rated their agreement (on a seven-point Likert scale) with statements such as “If I provide abortions . . . I will face social stigma”; “. . . I will be subjected to threats of violence and harassment in the community”; “. . . my family would approve/disapprove.” Perceived capability to train in and provide abortions (perceived behavioural control) was assessed by the responses to statements that included “For me, observing or assisting in abortion provision during clerkship would be (1 = difficult, 7 = easy),” with responses rated on a seven-point Likert scale. Additionally, participants rated their agreement on a binary (yes or no) scale with statements such as “I will not have access to the correct facilities or equipment” and “I will not be able to find support staff to help in my practice.” Finally, intentions to train in abortion provision and to provide abortion services were assessed by the responses to statements that included “I intend to observe or assist in abortion provision during my clinical clerkship” (1 = strongly disagree, 7 = strongly agree). This set of statements (attitudes, perceptions of social support, perceived behavioural control, and intentions) was repeated with respect to abortion training during clerkship, residency, and projected future practice, with variations for the topics of medical abortion, first trimester surgical abortion, and second trimester surgical abortion. All first, second, and third year medical students enrolled at Western University in Ontario were invited to participate in 238 l MARCH JOGC MARS 2015

this survey study during April and May 2013. Students were recruited by email notification and class announcements. Data collection took place via completion of an online or paper copy of the 74-item survey. Consent was obtained from all participants. Responses were anonymous, with consent forms being separate from survey responses. Respondents received a token gift for participating. All surveys that were returned in paper format were included in the analysis. Inclusion of electronic surveys was determined a priori by time limit. Electronic surveys that were completed in less than five minutes or were less than 50% complete were removed from all data analysis. Participants were instructed to respond to all statements with the assumption that they would be entering residency programs and specialties in which abortion provision was within the scope of practice. Therefore, a student intending to enter orthopedic surgery could still respond that they strongly intended to provide medical abortions. Subsequently, medical students’ career plans were assessed by respondents’ estimates (between 0% and 100%) indicating their likelihood of entering specialties in which abortion is or is not within the scope of practice. Medical students’ intentions to provide abortions in their future practice were used to categorize them for analysis as “intended providers” or “intended non-providers.” We decided a priori that only medical students who indicated a very strong intention (those who responded “7” on a 7-point Likert scale with 1 = strongly disagree, 7 = strongly agree) to provide abortion services would be labelled as “intended providers.” “Intended providers” and “non-providers” were than subdivided on the basis of future plans to enter a residency in which abortion is within or outside the scope of practice. We determined abortion as being within the scope of practice for residencies in family medicine, family medicine with an extra year of training in obstetrics, and obstetrics and gynaecology. We decided a priori that students who estimated a 60% or higher likelihood of entering one of these specialties would be categorized in the analysis as intending to enter a residency in which abortion provision was within the scope of practice. Together we used intention and future residency plan to create an estimate of the number of medical students likely to become abortion providers in future practice. Descriptive statistics were used to examine student characteristics and preferences. Pearson’s chi-square test and Fisher exact test were used to compare medical students’ endorsement of the legality of abortion as a function of their intention to provide or not provide abortion services in future practice. Linear regression was employed to predict medical students’ intentions to

36 (18.1)

31 (29.2) 16 (66.7) 7 (30.4) 18 (21.2)

Intended residency program Programs in which abortion provision is NOT within the scope of practice, n = 199 (64.6%)

Programs in which abortion provision is within the scope of practice, n = 106 (34.4%)

Obstetrics and gynaecology, n = 24 (7.8%)

Family medicine plus one year specialization in obstetrics, n = 23 (7.5%)

Family medicine, n = 85 (27.6%)

13 (15.3)

5 (21.7)

11 (45.8)

22 (20.8)

30 (15.1)

52 (16.8)

Early surgical abortion

6 (7.1)

1 (4.3)

8 (33.3)

14 (13.2)

19 (9.5)

33 (10.7)

Late surgical abortion

Intended providers, n (%)†

11 (12.9)

3 (13.0)

1 (4.2)

11 (10.4)

19 (9.5)

59 (19.2)

17 (20)

4 (17.4)

2 (8.3)

17 (16)

22 (11.1)

78 (25.3)

Early surgical abortion

26 (30.6)

5 (21.7)

2 (8.3)

26 (24.5)

29 (14.6)

95 (30.9)

Late surgical abortion

No/very low intention to provide, n (%)‡ Medical abortion

4.61 (1.42) 3.95 (1.99) 5.11 (1.92)

Perceived social support for abortion provision

Perceived control over access to abortion training

Intention to train in/provide abortion provision

4.92 (1.80)

4.44 (1.84)

4.66 (1.49)

5.91 (1.58)

Residency*

4.68 (1.98)

4.40 (2.05)

4.47 (1.58)

5.69 (1.71)

MA

4.27 (1.98)

4.10 (2.05)

4.22 (1.55)

5.55 (1.81)

ES

Future practice

*The means for attitudes, perceived social norms, and intentions for clerkship and residency are expressed as an average for all types of abortion.

Data are presented as means of ratings on a 7-point Likert scale, where 7 indicates the most positive/supportive and 1 indicates the least positive/supportive.

MA: medical abortion; ES: early surgical abortion; LS: late surgical abortion; SD: standard deviation

5.63 (1.70)

Attitudes re training and provision

Clerkship*

Anticipated stage in career, mean (SD)

3.83 (1.93)

3.52 (2.00)

4.22 (1.55)

5.20 (1.90)

LS

Table 2. Medical students’ attitudes, perceptions of social support, and intentions to train in and to provide abortions at current and anticipated future stages of training and practice

‡“No/very low intention” was defined as students endorsing a 1/7 on the following scale: “I intend to acquire medical abortion provision training during residency”; 1 = strongly disagree – 7 = strongly agree.

†“Intended providers” was defined as students endorsing a 7/7 on the following scale: “I intend to acquire medical abortion provision training during residency”; 1 = strongly disagree – 7 = strongly agree.

*When answering questions regarding intention to provide abortion during practice medical students were asked to assume that they would be in a specialty in which abortion would be within scope of practice. We defined intention to specialize as an estimate of a 60% or higher likelihood of entering a specialty.

67 (21.7)

All medical students, N = 308 (100%)

Medical abortion

Table 1. Medical students’ intention to provide abortion services in their future practice, stratified by planned future specialty*

Medical Students’ Intentions to Seek Abortion Training and to Provide Abortion Services in Future Practice

MARCH JOGC MARS 2015 l 239

WOMEN’S HEALTH

acquire training in and to provide abortions as a function of attitudes, social norms, and perceived behavioural control during different stages of medical training and for different types of abortions. All analyses were conducted using SPSS Statistics 19 (IBM Corp., Armonk NY). These procedures were approved by the University of Western Ontario Research Ethics Board. RESULTS

A total of 337 surveys were returned. Twenty-nine of the surveys were excluded from all analyses for being less than 50% complete or having been completed in less than five minutes. The final analysis included 308 of 508 (61%) potential respondents (152 males and 154 females; 2 students did not designate their gender), with a mean age of 23.8 years (SD 0.85). Eighty-two students (27%) were in their first year of study, 133 (43%) were in their second year, 85 (27.5%) were in their third year, five (2.6%) were completing a combined MD/PhD degree, and three (1%) declined to answer. A total of 69.4% of respondents had completed their reproductive medicine pre-clerkship education (including a one hour long lecture on abortion provision) and 66 of the 85 third-year students (21.5% of total respondents) had completed their obstetrics and gynaecology clerkship rotation, in which exposure to abortion provision was offered but not mandated. Forty-eight of 308 respondents (15.6%) had observed or participated in an abortion. The intention of medical students to provide abortions in future practice according to their plans for residency are shown in Table 1. Although we had determined a priori to define “intended providers” as students who responded “7” (strongly agree) on a 7-point Likert scale when asked if they intended to provide abortion services, reanalysis to include respondents who responded “6” or “7” on the Likert scale did not change observed trends. Ten percent of medical students reported both a greater than 60% probability of training in a specialty that included abortion in its scope of practice and together with very strong intentions to provide medical abortions; 7% would provide surgical first trimester abortions, and 4.5% would provide surgical second trimester abortions in their future practices. Medical students’ attitudes, perceptions of social support, perceived behavioural control, and intentions to train in and to provide medical abortions, first trimester surgical abortions, and second trimester surgical abortions during medical school, future residency, and future practice were assessed (Table 2). This analysis included all 308 complete 240 l MARCH JOGC MARS 2015

responses and was independent of anticipated field of practice. Students were instructed to respond to these questions with the assumption that abortion provision would be within their scope of practice. Attitudes, perceived social norms, and perceived ability to provide abortion services were similar and relatively positive for anticipated clerkship and residency, but became less positive when provision of abortions in future practice was envisioned. In addition, a uniform decline in positive attitudes, perceived social support, and perceived ease of providing abortions was observed when moving from medical abortions to first trimester surgical abortions and to second trimester surgical abortions. We explored medical students’ attitudes towards the legality of abortion, and their intentions concerning the provision of abortion, under specific circumstances (Table 3). Data from the 308 survey respondents were included in the analysis, regardless of future career plans. Medical students’ most positive attitudes towards abortions and strongest intentions to provide abortions were for pregnancies of less than eight weeks’ gestation. Very few students endorsed refusing abortion under all circumstances. Uniformly, medical students expressed significantly greater support for the legality of abortion than for willingness to provide the service personally under the same circumstances. We conducted a linear regression analysis to determine the independent influence of medical students’ attitudes, perceived social norms, and perceived behavioural control on their intentions to provide abortions. All 308 respondents were included in the analysis, which indicated that attitudes, social norms, and perceived control each contributed significantly to the prediction of intentions to provide abortion services (Table 4). This form of analysis does not take into account the overlap in influence that the three independent factors (attitudes, perceived social norms, and perceived behavioural control) have on the dependent variable (intentions). Attitudes were the primary determinant of intentions; these were followed by social norms, and subsequently by perceived behavioural control. These factors accounted for a substantial proportion of the variance (ranging from 41% to 72%) in intentions to train in and provide abortions. For male medical students, the strength of attitudes as a predictor of intentions diminished and the strength of perceived social norms as a predictor of intentions increased when students envisioned progressing through training into future practice. This pattern was not observed for female medical students. Perceived behavioural control had a larger influence during training years than during practice uniformly.

Medical Students’ Intentions to Seek Abortion Training and to Provide Abortion Services in Future Practice

Table 3. Medical students’ attitudes and intentions concerning abortion: proportions endorsing statements about legality of abortion and willingness to provide abortion* under specified circumstances I believe abortion should be legal in Canada

I would perform an abortion

P

Under any circumstance until 8 weeks’ gestation using medication only

0.86

0.80

< 0.001

Under any circumstance until 13 weeks’ gestation using manual vacuum aspiration technique

0.84

0.71

< 0.001

Under any circumstance until 23½ weeks’ gestation using dilatation and evacuation

0.65

0.45

< 0.001

Under emergency circumstances in which the pregnancy endangers the life of the woman

0.97

0.93

< 0.001

When the pregnancy is the result of rape

0.94

0.83

< 0.001

When the pregnancy is the result of incest

0.90

0.78

< 0.001

When the fetus has non-lethal congenital abnormalities (e.g., Down syndrome)

0.76

0.65

< 0.001

When the fetus has lethal fetal abnormalities (e.g., anencephaly)

0.96

0.86

< 0.001

Under no circumstances

0.007

0.055

0.1

Students indicated their endorsement on a dichotomous scale of 0 or 1, with 0 representing disagreement with the statement, and 1 representing agreement. *For willingness to provide, respondents were told to assume that they had the training and competence to do so.

Table 4. Linear regression on the variance of intentions to provide abortions, as predicted by attitudes, social norms, and perceived behavioural control, divided by gender and anticipated stage of training or practice Intentions Male (n = 151)

Female (n = 153)

R2

P

R2

P

Attitudes

0.712

< 0.001

0.707

< 0.001

Social norms

0.252

< 0.001

0.370

< 0.001

Perceived behavioural control

0.261

< 0.001

0.453

< 0.001

Attitudes

0.664

< 0.001

0.581

< 0.001

Social norms

0.445

< 0.001

0.314

< 0.001

Perceived behavioural control

0.433

< 0.001

0.338

< 0.001

Attitudes

0.520

< 0.001

0.587

< 0.001

Social norms

0.457

< 0.001

0.398

< 0.001

Perceived behavioural control

0.127

< 0.001

0.144

< 0.001

Attitudes

0.430

< 0.001

0.576

< 0.001

Social norms

0.383

< 0.001

0.373

< 0.001

Perceived behavioural control

0.100

< 0.001

0.166

< 0.001

Attitudes

0.407

< 0.001

0.576

< 0.001

Social norms

0.349

< 0.001

0.356

< 0.001

Perceived behavioural control

0.081

< 0.001

0.150

< 0.001

During clerkship

During residency

Providing medical abortion in practice

Providing early surgical abortion in practice

Providing late surgical abortion in practice

A significant P value is obtained when an independent variable (attitudes, social norms, or perceived behavioural control), contribute significantly to a change in intentions. R2 indicates the degree of variance in the intentions (dependent factor) that is accounted for by attitudes, social norms, and perceived behavioural control (independent factors).

MARCH JOGC MARS 2015 l 241

WOMEN’S HEALTH

DISCUSSION

Prior studies have found medical students having relatively positive attitudes towards abortion. Medical students from the United States, the United Kingdom, India, and South Africa showed 72.1%, 62%, 85%, and 30% support, respectively, for the legality of abortion.9,17–19 A study of Canadian medical students in British Columbia found that 38% of medical students would be willing personally to provide an abortion to a patient requesting one, and a further 34% would be willing to refer the patient to another physician.20 As in prior studies, medical students in our study held mostly positive attitudes towards the availability and legality of abortion, with 86% of respondents supporting first trimester abortions and 65% supporting second trimester abortions. However, medical students’ intentions to provide abortion services themselves were considerably weaker than their attitudes supporting the availability of abortion. Obstetrician-gynaecologists perform a significant proportion of abortions in Canada, particularly second trimester abortions. In British Columbia, 50% of all abortions are performed by obstetrician-gynaecologists.7 Medical students interested in obstetrics and gynaecology (6.5% of our respondents) had the strongest intentions (65%) to provide medical abortions, with reduced intention to provide first trimester and second trimester surgical abortions (45% and 40% respectively). These results indicate a greater willingness to provide abortion services than has previously been reported for Canadian residents in obstetrics and gynaecology; Roy et al. reported that 38% of Canadian residents planned to provide medical abortions and 21% planned to provide second trimester surgical abortions.11 Family physicians are frequently the first-line abortion providers in Canada, particularly in rural settings. In the past three years, approximately 40% to 50% of graduating Canadian medical students entered a family medicine residency.21 We found that 20% of medical students planning to specialize in family medicine intended to provide any type of abortion service. Most students felt abortion provision would be outside their future scope of practice. Respondents’ perceptions of social support or opposition regarding abortion training and provision were variable, but remained on average neutral as they progressed through envisioned training to envisioned practice. Respondents who perceived social opposition to abortion had significantly weaker intentions to pursue training in abortion provision and weaker intentions to provide abortion services in future practice compared with respondents who did not perceive social opposition. Forty-five percent of medical students stated that anticipated harassment towards themselves or 242 l MARCH JOGC MARS 2015

their family had negatively influenced their desire to provide abortions, regardless of whether or not they intended to provide abortions in future practice (data not shown). Respondents’ perceived some difficulty in obtaining training in abortion provision during medical school and residency. Respondents who perceived that obtaining abortion training would be difficult had significantly weaker intentions to pursue training in and to provide abortion services in future practice. Sixty-five percent of medical students felt that they would not have access to the necessary equipment and facilities to provide abortions, and 48% felt that they would not be able to find support staff willing to assist in abortion provision (data not shown). These findings suggest that medical students perceive a high degree of logistical difficulty in setting up a practice that includes abortion provision. Our findings may overestimate the probable number of future abortion service providers, because medical students who intend to provide abortion services may face the currently perceived barriers as well as other unanticipated barriers. Previous studies have identified local hospital policy, a lack of operating room time, a lack of supportive colleagues and administrative staff, or high levels of social stigma as having a negative effect on provision of abortion services.2,4,7,8 In addition, physicians who plan to provide only medical abortions may be limited by a lack of regional physicians who are able to perform surgical abortion if a medical abortion fails. A logistical regression analysis (data not shown) demonstrated that medical students’ attitudes towards personal involvement in abortion training and provision were the strongest predictors of their intention to train in and provide abortion services, followed by perceived social norms and perceived behavioural control. All medical students were significantly influenced by perceived social norms, and for male medical students the influence of social norms grew as they envisioned progressing through training to practice. Medical students also anticipated high rates of harassment or threats towards themselves or their family if they provide abortions. Together, these results suggest medical students perceive a high degree of professional and societal disapproval for the provision of abortions and that male medical students are more likely to be deterred from abortion provision by these perceptions. Further, the existence of these perceptions, whether based on true societal and professional norms or not, indicates an ongoing need to de-stigmatize abortion and abortion provision. Including abortion services as an expected part of medical education, with an option for conscientious objectors

Medical Students’ Intentions to Seek Abortion Training and to Provide Abortion Services in Future Practice

to decline participation, could de-stigmatize training. Similarly, creating collective agreements between members of family medicine practices or obstetrics and gynaecology departments so that more or all members participate in abortion provision would decrease the stigma of provision and the visibility of providers. In Canada, physicians working in collective clinics report feeling well supported by their hospitals and colleagues.22 Similarly, many report personal and professional isolation to be challenges in providing abortion services, and that workshops to discuss experiences and offer support can help clinicians deal with stigma.22,23 Medical students in our study were more supportive of medical abortions than of first trimester or second trimester surgical abortions. However, the majority of abortions performed in Canada are surgical. In 2010 only 3.4% of abortions performed in Canadian were medical abortions, while 60% of abortions in comparable European countries and 20% of abortions in the United States were medical.24–26 This substantial difference may be partially explained by the fact that mifepristone, a drug used in the World Health Organization’s recommended protocol for medical abortions, is approved for use in Europe and the United States but not in Canada.27,28 Instead, Canadian physicians use a combination of methotrexate and misoprostol, which is less effective than mifepristone.29 A submission has been made to Health Canada to approve use of mifepristone for medical abortion in Canada.28 In light of our finding of substantial support for providing medical abortion in the future practices of medical students, approval of this submission could increase access to abortion in Canada, particularly in rural settings. Providing medical abortion requires less specialized training and is within the scope of practice of essentially all family physicians and obstetrician-gynaecologists. Furthermore, the use of medical abortion, by reducing the visibility of abortion for both care providers and women seeking abortions, would reduce the social stigma associated with abortion in general. A limitation of the current study is that it involved students at a single medical school; thus, it may have failed to capture regional or institution-specific differences. The response rate was robust, but non-responders (33%) may nonetheless have held significantly different views from those who responded. Fourth year medical students were unavailable to complete the survey because of licensing examination schedules. Their views about abortion practice may differ as a result of their greater clinical experience. Additionally, the medical students surveyed were not given the opportunity to characterize their level of exposure to

or training in abortion services, and they indicated only generally if they had observed or participated in abortion provision (which was not found to have a significant impact on attitudes or intentions). A further limitation of our study is the uncertainty medical students may have towards future residency and career plans. Twenty-eight percent of students in our study estimated their likelihood of entering family practice as 60% or higher, and 6.5% estimated their likelihood of entering obstetrics and gynaecology as 60% or higher. This represents a substantial deviation from final residency plans, as the proportion of medical students matching to a family medicine residency in the 2014 Canadian residency match was 43.9%, and to an obstetrics and gynaecology residency it was 3.4%.30 Consequently, our analysis of future providers will likely have underestimated the number of family physician abortion providers and overestimated the number of obstetrician-gynaecologist abortion providers. In addition, our study did not examine actual abortion provision but rather medical students’ intentions to train in and provide abortions. However, there is substantial evidence for the strength of intentions in predicting future behaviour.14–16,31,32 The results from our study suggest a higher degree of intention to provide abortions in future practice among current medical students than among current residents and practising physicians. Further studies assessing the attitudes and intentions of current family medicine and obstetrics and gynaecology residents towards abortion provision to better understand this discrepancy are indicated. Further studies also should be conducted to understand how and under what specific conditions the intentions of medical students to provide abortions translate into future actions; they should also assess the effectiveness of our proposed policy changes in reducing barriers to abortion training and provision. CONCLUSION

Most of the medical students in this study supported the legality and availability of abortion. Comparatively fewer expressed an intention to provide abortions personally in future practice, with even fewer students intending both to train in a specialty in which abortion is within the scope of practice and to provide abortions. We do not know whether sufficient numbers of current medical students will go on to provide abortions in practice to meet current and future demands. However, some trends such as perceptions of negative social stigma regarding abortion provision are reducing medical students’ intentions to provide abortions. These issues should be addressed using interventions aimed at MARCH JOGC MARS 2015 l 243

WOMEN’S HEALTH

decreasing the visibility of individual abortion providers and the social stigma encountered by both trainees and providers. Finally, our findings suggest that the approval of mifepristone combined with interventions aimed at promoting the use of medical abortion could increase the number of providers and the overall availability of abortion in Canada. ACKNOWLEDGEMENTS

This research was supported by the University of Western Ontario University Student Council Grant, and the Hippocratic Council Innovator Grant. REFERENCES 1. R. v. Morgentaler(1988), 37 C.C.C. (3rd) 449 (S.C.C.). 2. Sabourin JN, Burnett M. A review of therapeutic abortions and related areas of concern in Canada. J Obstet Gynaecol Can 2012;34(6):532–42. 3. Sethna C, Doull M. Far from home? A pilot study tracking women’s journeys to a Canadian abortion clinic. J Obstet Gynaecol Can 2007;29:640–7. 4. Canadian Abortion Rights Action League (CARAL). Protecting abortion rights in Canada: special report to celebrate the 15th anniversary of the decriminalization of abortion. Ottawa: CARAL; 2003. 5. Norman WV. Abortion in British Columbia: trends over 10 years compared to Canada. Contraception 2011;84(3):316. 6. Norman WV, Soon JA, Maughn N, Dressler J. Barriers to rural induced abortion services in Canada: findings of the British Columbia Abortion Providers Survey (BCAPS). PLoS One [Internet] 2013 Jun 28 [cited 2013 Dec 12]; 8(6): e67023. doi:10.1371/journal. pone.0067023. 7. Jones RK, Zolna MR, Henshaw SK, Finer LB. Abortion in the United States: incidence and access to services, 2005. Perspect Sex Reprod Health 2008;40:6–16. 8. Grimes DA. Clinicians who provide abortions: the thinning ranks. Obstet Gynecol 1992;80:719–23. 9. Eastwood KL, Kacmar JE, Steinauer J, Weitzen S, Boardman LA. Abortion training in United States obstetrics and gynecology residency programs. Obstet Gynecol 2006; 108(2):303. 10. Gleeson R, Forde E, Bates E, Powell S, Eadon-Jones E, Draper H. Medical students’ attitudes towards abortion: a UK study. J Med Ethics 2008;34(11):783–7. 11. Roy G, Parvataneni R, Friedman B, Eastwood K, Darney PD, Steinauer J. Abortion training in Canadian obstetrics and gynecology residency programs. Obstet Gynecol 2006;108:309–14. 12. Steinauer J, Landy U, Filippone H, Laube D, Darney PD, Jackson RA. Predictors of abortion provision among practicing obstetriciangynecologists: a national survey. Am J Obstet Gynecol 2008;198(1):39.1–6. 13. Allen RH, Raker C, Steinauer J, Eastwood KL, Kacmar JE, Boardman LA. Future abortion provision among US graduating obstetrics and gynecology residents, 2004. Contraception 2010;81:531–6. 14. Fishbein M, Ajzen I. Belief, attitude, intention, and behavior: an introduction to theory and research. Reading: Addison-Wesley; 1975. 15. Ajzen I, Madden TJ. Prediction of goal-directed behavior: attitudes, intentions, and perceived behavioral control. J Exp Soc Psychol 1986;22:453–74.

244 l MARCH JOGC MARS 2015

16. Fisher WA, Kohut T, Salisbury C, Salvadori M. understanding human papillomavirus vaccination intentions: comparative utility of the theory of reasoned action and the theory of planned behavior in vaccine target age women and men. J Sex Med 2013; 10(10);2455–64. 17. Shotorbani S, Zimmerman FJ, Bell JF, Ward D. Attitudes and intentions of future health care providers toward abortion provision. Perspect Sex Reprod Health 2004;36(2) 58–63. 18. Sjostrom S, Essen B, Syden F, Gemzell-Danielsson K, Klingberg-Allvin M. Medical students’ attitudes and perceptions on abortion: a cross-sectional survey among medical interns in Maharastra, India. Contraception 2014;90(1):42–6. 19. Wheeler SB, Zullig LL, Bryce BR, Buga GA, Morroni C. Attitudes and intentions regarding abortion provision among medical students in South Africa. Int Perspect Sex Reprod Health 2012;38(3):154–63.  20. Cessford TA, Norman W. Making a case for abortion curriculum reform: a knowledge-assessment survey of undergraduate medical students. J Obstet Gynaecol Can 2011;33(1):38–45. 21. Canadian Resident Matching Service (CaRMS). R1 match reports—2014. Table 59: Family Medicine Positions 2013–2014 Match Comparison. Ottawa: CaRMS; 2014. Available at: http://www.carms.ca/assets/upload/ Match%20reports/2014%20R-1%20match/EN/Table%2059%20-%20 Family%20Medicine Royal%20College%20Positions%202012–2013%20 Match%20Comparison_English.pdf. Accessed October 18, 2014. 22. Dressler J, Maughn N, Soon JA, Norman WV. The perspective of rural physicians providing abortion in Canada: qualitative findings of the BC Abortion Providers Survey (BCAPS). PLoS One 2013;8(6):e67070. doi:10.1371/journal.pone.0067070 23. Harris LH, Debbink M, Martin L, Hassinger J. Dynamics of stigma in abortion work: findings from a pilot study of the Providers Share workshop. Soc Sci Med 2011;73(7):1062–70. 24. Canadian Institutes for Health Information. Induced abortions performed in Canada in 2012. Table 7: number and percentage distribution of induced abortions reported by Canadian hospitals (Excluding Quebec) in 2010, by Method of Abortion. Ottawa: CIHI; 2014. Available at: http://www.cihi.ca/CIHI-ext portal/pdf/internet/TA_11_ ALLDATATABLES20140221_EN. Accessed July 24, 2014. 25. Raymond EG, Shannon C, Weaver MA, Winikoff B. First-trimester medical abortion with mifepristone 200 mg and misoprostol: a systematic review. Contraception 2013;87:26–37. 26. Jones RK, Henshaw S. Mifepristone for early medical abortion: experiences in France, Great Britain and Sweden. Perspect Sex Reprod Health 2002;34:154–61. 27. World Health Organization. Safe abortion: technical and policy guidance for health systems. 2nd ed. Geneva (Switzerland): WHO Press; 2012. Available at: http://apps.who.int/iris/bitstream/10665/70914/1/ 9789241548434_eng.pdf. Accessed July 24 2014. 28. Dunn S, Cook R. Medical abortion in Canada: behind the times. CMAJ 2014; 186(1):13–4. 29. Wiebe E, Dunn S, Guilbert E, Jacot F, Lugtig L. Comparison of abortions induced by methotrexate or mifepristone followed by misoprostol. Obstet Gynecol 2002;99:813–9. 30. Canadian Resident Matching Service (CaRMS). R1 match reports—2014. Table 27: Match results of CMGs by school of residency & discipline. Ottawa: CaRMS; 2014. Available at: https://www.carms.ca/assets/ upload/Match%20reports/2014%20R1%20match/EN/Table%2027%20 %20Match%20Results%200f%20CMGs%20by%20School%200f%20 Residency%20-%20Discipline_English.pdf. Accessed October 18, 2014. 31. Sheeran, P. Intention—behavior relations: a conceptual and empirical review. Eur Rev Soc Psychol 2001;12(1):1–36. 32. Armitage CJ, Conner M. Efficacy of the theory of planned behavior: a meta-analytic review. Br J Soc Psychol 2001;4(40):471–99.