Women's Preference for Receiving Abortion in Primary Care Settings

Abstract

Background:

Abortion is increasingly being incorporated into primary care training programs. It remains unclear if women, given the option, would accept and access these services with their primary care provider.

Methods:

An anonymous survey was administered to 299 women at two abortion clinics in New York and Chicago.

Results:

One hundred seventy-four (58.2%) respondents theoretically would choose their primary care clinic for their abortion procedure, with more Chicago participants than New York participants choosing primary care. After adjusting for other variables, only being very or somewhat comfortable with one's primary care provider regarding issues of pregnancy prevention and birth control predicted the choice of primary care for abortion services.

Conclusions:

The majority of women surveyed theoretically would choose to have a procedure at their primary care clinic. There were some regional differences between New York and Chicago. Although many variables used in this study did not explain the respondents' preferences, the primary care clinic as site for abortion care appears to be an important potential option for many women. Primary care training programs should teach full-spectrum reproductive healthcare to help meet the needs of their patients.

Introduction

Unintended pregnancy in the United States is common, contributing to approximately 1.2 million abortion procedures annually.¹ By age 45, >40% of U.S. women have experienced at least one abortion.² Despite the high frequency of abortion, 87% of U.S. counties lack even a single abortion provider, and the number of providers overall continues to decline.¹ In response to this provider shortage, abortion training is increasingly being included, or offered, in a variety of healthcare training programs, including those for obstetrics and gynecology, internal medicine, family medicine, and advance practice clinicians.^3

–10 Primary care providers (PCPs) who choose to provide early abortion procedures within the services they offer may be well positioned to help meet the core values of the discipline, including comprehensive, compassionate, and personal care, qualities that have been noted to be important to the U.S. public.¹¹

As of 2005, 69% of abortion procedures in the United States were performed in specialized abortion clinics, where at least half of patient visits are for abortion services.¹ However, a growing number of nonspecialized physician offices have introduced abortion services since 2001.¹ This trend can be attributed to the simplification of early abortion through use of the manual vacuum aspirator (MVA) and local anesthetic techniques, but also through the use of medications with mifepristone, also known as RU-486.¹

It remains unclear if women would prefer to receive such services within or separate from the setting where they receive their routine healthcare. A survey of women seeking abortion at three Iowa Planned Parenthood clinics found that only 20% of family physician patients and 36% of obstetrician/gynecologist patients would choose their regular doctor for their abortion procedure.¹² On the other hand, the majority of female patients surveyed at three family medicine clinics in New York supported the integration of early abortion services.¹³ This present study sought to (1) quantify the percentage of women seeking first-trimester abortion services from two different U.S. regional areas who would theoretically choose, if available, to have their abortion procedure at their primary care clinic, (2) determine if any of these characteristics predict clinic choice for abortion service, (3) compare the reasons behind women's responses for seeking care with their primary care doctor prior to their current abortion procedure, and (4) compare the reasons behind women's preferences for clinic for their first-trimester abortion service.

Materials and Methods

Design

This study used a mixed methods design with both qualitative and quantitative data to assess factors influencing a woman's choice of primary care as site for abortion care. We invited women seeking first-trimester abortion at an abortion clinic in either Chicago or New York City to participate in a survey study. Women aged 18–45 who stated they had a regular source of medical care, were within the first trimester of pregnancy, and spoke either English or Spanish in New York or English in Chicago were eligible to participate. Each woman was approached before her abortion procedure using a recruitment script regarding family planning and unplanned pregnancy. If eligible, participants were taken to a private room to give consent. A research assistant at each respective regional site administered an anonymous 29-question survey using both closed-ended and open-ended questions.

At the time we designed this study, no prior studies had assessed patient preferences for first-trimester clinic for abortion service; therefore, we based our sample size on feasibility. We chose Chicago and New York as study sites because they were similar as metropolitan cities but different in terms of using states' public funds to pay for abortion. New York State Medicaid covers all abortion care for eligible women; Illinois uses state funds to pay for abortion only in cases of rape or incest or to save the life of the pregnant woman.¹⁴

Survey instrument

The survey instrument was previously described in a study by Rubin et al.¹³ and was adapted for women seeking an abortion at an abortion clinic. We determined source of care by asking: What kind of doctor or doctors do you go to? If the patient named more than one doctor, for the rest of the survey, we asked her to consider the doctor who performs her Pap test as her PCP, and the clinic where the PCP worked was considered the primary care clinic (PCC). Continuity of care was assessed using length of time and consistency with the PCP or PCC, shown to be important measures of continuity.^15,16 We determined a woman's comfort in talking to her physician about pregnancy prevention-related issues by asking: A lot of women have questions about preventing pregnancy and contraception. How comfortable are you talking with your doctor about these types of questions? Responses were recorded on a 4-point Likert scale ranging from very comfortable to very uncomfortable.¹⁷ Additionally, we asked if the respondent had spoken to her regular doctor or another provider at the PCC prior to this current abortion procedure to determine the role of the PCP when a woman is faced with an unplanned or unwanted pregnancy. Health and demographic information was assessed using patients' self-reported number of previous abortions, age, type of insurance, and birthplace. Education attainment assessment was adapted from the 2000 U.S. Census.¹⁸

The survey was translated into Spanish by a woman of Puerto Rican heritage with prior translation experience and was subsequently reviewed by a team with reproductive health experience. The team approach has been shown to be an acceptable method of translation as compared to back-translation techniques.¹⁹ We piloted the survey in New York and Chicago at specialized abortion clinics with similar patient populations as the clinics used for this present study. The survey was revised accordingly prior to full-scale implementation.

Data analysis

Responses were entered into a password-protected database managed by each respective study site (University of Illinois at Chicago and Montefiore Medical Center in New York). Site-specific assistants cleaned the data entries. Discrepancies were resolved by checking responses against the written questionnaires. The principal investigator conducted a random check of 20% of the data entries against the completed questionnaires to assure the accuracy and completeness of the data.

SAS statistical software version 9.1 (SAS Institute Inc., Cary, NC) was used to calculate frequency and summary statistics. We examined the following factors: age, educational status, health insurance type, provider type (the provider who performs the respondents' Pap smears), length of time respondent has been attending current PCC, reported comfort talking with PCP about birth control, contact with PCP (or other provider in PCC) prior to their abortion procedure, previous abortions, immigrant status, and regional site (Chicago or New York). We performed bivariate analyses using t tests for continuous variables and Fisher's exact test for categorical variables to determine statistical differences between groups. A two-tailed p value of <0.05 was considered significant. Because of the exploratory nature of this study, we used a conceptual framework of variables shown to be associated with risk of unintended pregnancy (and, thus, subsequent risk of abortion) and any variable with a p < 0.20 in the bivariate analysis to predict the odds of choosing PCC in the multivariate logistic regression. We recategorized some variables before analysis because of small sample size; specifically, we grouped health insurance with either public or private insurance and combined respondents reporting “grade school” with “less than a high school diploma,” as well as those reporting comfort with their PCP as “very uncomfortable” with “somewhat uncomfortable.” The dependent variables in the analysis included age, educational status, insurance status, prior abortion, time with PCC, comfort with PCP, whether the respondent spoke to PCP (or other provider in PCC) about the current pregnancy, and regional site. A stepwise backwards logistic regression was run after checking for interactions and correlations of all potential covariates, including regional site.

The open-ended responses regarding clinic choice for abortion service and reasons for seeking care (or not) with one's PCP (or other provider in PCC) before their abortion procedure were analyzed and coded into themes by the research team in Chicago (E.M.G. and E.J.S.). The majority of responses included only one reason for women's choices; in cases in which women had multiple responses for their choice, all responses were coded. No a priori codes were used. The institutional review boards of the University of Illinois at Chicago and Montefiore Medical Center approved the study for each respective research site.

Results

Demographics and response rates

Between March and August 2006 in New York and July and November 2006 in Chicago, 299 women completed the survey. Table 1 shows the participants' demographic and PCP characteristics. Table 2 shows the frequencies of preferences for clinic for first-trimester abortion service. Overall, 174 (58.2%) of the respondents stated they would theoretically choose their PCC as clinic for their abortion service. Chicago respondents were significantly more likely to choose the PCC than the New York respondents (p < 0.001).

Table 1.

Demographic and Primary Care Provider Relationship Characteristics of Respondents, Stratified by Regional Site

Variable	Total (n = 299)	Chicago (n = 149)	New York (n = 150)
Age,^** years
Mean (SD)	27.0 (5.9)	28.0 (5.8)	26.0 (6.0)
Birthplace,^* n (%)
United States	240 (80.3)	140 (93.9)	100 (66.7)
Outside United States	54 (18.1)	8 (5.4)	46 (30.7)
Missing	5 (1.7)	1 (0.7)	4 (2.7)
Education,^* n (%)
Completed grade school or less	2 (0.6)	0 (0.0)	2 (1.3)
Some high school	42 (14.1)	6 (4.0)	36 (24.0)
Completed high school	70 (23.4)	27 (18.1)	43 (28.7)
Some college	104 (34.8)	61 (40.9)	43 (28.7)
College graduate or higher	79 (26.4)	55 (36.9)	24 (16.0)
Missing	2 (0.7)	0 (0.0)	2 (1.3)
Source of health insurance,^* n (%)
Job	106 (35.5)	89 (59.7)	17 (12.2)
Parent's job	18 (6.0)	12 (8.1)	6 (4.3)
Partner's job	15 (5.0)	12 (8.1)	3 (2.2)
Public aid	122 (40.8)	22 (14.8)	100 (71.9)
No insurance	25 (8.4)	12 (8.1)	13 (9.4)
Other	2 (0.7)	2 (1.3)	0 (0.0)
Missing	11 (3.7)	0 (0.0)	11 (7.3)
Insurance used for procedure,^* n (%)
Yes	206 (68.9)	84 (56.4)	122 (81.3)
No	67 (22.4)	55 (36.9)	12 (8.0)
Missing	26 (8.7)	10 (6.7)	11 (7.3)
Previous abortion
Yes	156 (52.2)	63 (42.2)	93 (62.0)
No	140 (46.8)	85 (57.1)	55 (36.7)
Missing	3 (1.0)	1 (0.7)	2 (1.3)
Type of PCP, n (%)
Family medicine	142 (47.5)	89 (59.7)	53 (35.3)
Internal medicine	88 (29.4)	40 (26.9)	48 (32.0)
Obstetrician/gynecologist	35 (11.7)	18 (12.1)	17 (11.3)
Nurse practitioner	3 (1.0)	2 (1.3)	1 (0.7)
Other	1 (0.3)	0 (0.0)	1 (0.7)
Missing	30 (10.0)	0 (0.0)	30 (20.0)
Years with PCP or PCC
Mean (SD)	6.0 (5.5)	5.88 (5.0)	6.14 (6.0)
Comfort with PCP,^* n (%)
Very comfortable	196 (65.6)	110 (73.8)	86 (57.3)
Somewhat comfortable	69 (23.1)	27 (18.1)	42 (28.0)
Very/somewhat uncomfortable	32 (10.7)	12 (8.1)	20 (13.3)
Missing	2 (0.7)	0 (0.0)	2 (1.3)

p < 0.05, using chi-square test.

p < 0.05 using Student's t test.

SD, standard deviation; PCP, primary care provider; PCC, primary care clinic.

Table 2.

Preference of Clinic for First-Trimester Abortion, Stratified by Regional Site

	Total n = 299 (%)	Chicago n = 149 (%)	New York n = 150 (%)
Preferred clinical site for abortion care:
Primary care clinic	174 (58.2)	103 (69.1)^*	71 (47.3)
Abortion clinic	118 (39.5)	43 (28.9)	75 (50.0)
Don't know	7 (2.3)	3 (2.0)	4 (2.7)

p < 0.001 using two-sided chi-square test comparing the preferred clinic type by regional site.

Odds of theoretically choosing primary care as clinic for abortion service

We performed a bivariate analysis to determine associations between demographics and other characteristics and preference of clinical site for abortion service. Being very comfortable discussing pregnancy prevention with PCP (odds ratio [OR] 5.62, 95% CI 2.45-12.88) and being a respondent in Chicago (OR 2.53, 95% CI 1.56-4.10) were significantly associated with theoretically choosing the PCC as site for abortion. In the multivariate model, when adjusted for regional site, age, education, insurance status, previous abortions, time with current PCC, and having spoken with PCP (or other provider at PCC) before their abortion procedure, only being very comfortable (OR 5.91, 95% CI 2.29-15.27) or somewhat comfortable (OR 2.87, 95% CI 1.03-7.89) with one's regular provider predicted the theoretical choice of PCC as clinic site for abortion (Table 3).

Table 3.

Odds of Theoretically Choosing Primary Care Clinic for Abortion Service

Variable	Women theoretically choosing PCC for abortion n = 174 n (%)	Crude odds ratio OR (95% CI)	Adjusted odds ratio^ OR (95% CI)*
Regional site
Chicago	103 (59.1)	2.53 (1.56-4.10)	1.37 (0.87-3.22)
New York	71 (40.8)	Referent	Referent
Mean age, years (SD)^a	27.5 (±6.3)	1.14 (1.00-1.08)	1.07 (0.93-1.38)
Education
Less than high school diploma	20 (11.5)	0.50 (0.23-1.08)	0.85 (0.32-2.24)
Completed high school	38 (21.8)	0.68 (0.35-1.32)	0.9 (0.44-2.15)
Some college	67 (38.5)	1.03 (0.55-1.91)	1.28 (0.62-2.63)
College graduate or higher	49 (28.2)	Referent	Referent
Source of health insurance
Private	95 (54.6)	Referent	Referent
Public/no insurance	74 (42.6)	0.49 (0.30-0.80)	0.88 (0.44-1.76)
Missing	5 (2.9)
Previous abortion	82 (47.1)	0.59 (0.37-0.95)	0.73 (0.42-1.28)
Mean years with PCP or PCC (SD)^b	5.6 (±5.3)	1.01 (0.97-1.05)	1.08 (0.85-1.37)
Talked to PCP or other provider in PCC prior to abortion procedure	60 (34.7)	2.45 (1.39-4.32)	1.57 (0.84-2.95)
comfort with PCP
Very comfortable	132 (72.9)	5.62 (2.45, 12.88)	5.91 (2.29, 15.27)
Somewhat comfortable	33 (19.0)	2.41 (0.97, 5.96)	2.87 (1.03, 7.89)
Very/somewhat uncomfortable	9 (8.1)	Referent	Referent

Multivariate logistic model adjusts for all covariates in table: F(12,234), p ≤ 0.001, r² = 0.142.

Age and years in primary care clinic are reported as mean (SD) and modeled in 5-year increments in crude and adjusted analysis.

OR, odds ratio; CI, confidence intervals; SD, standard deviation; PCP, primary care provider; PCC, primary care clinic.

Reasons for seeking care with PCP before abortion procedure

We reviewed reasons why women may have sought care with their PCP before their abortion procedure. In Chicago, 37.6% of women spoke to their PCP (or other provider at PCC), whereas in New York, only 17.3% did so; the most common reasons reported were (1) needing advice or options counseling (46%) and (2) confirming their positive pregnancy test or seeking a referral (42%).

For women who did not speak to the PCP (or other provider at PCC) before their current abortion procedure, common themes were (1) fear that their decision would not be supported (23%), (2) concerns of feeling embarrassed or judged by the PCP (22%), and (3) being sure of her decision and not feeling the need to discuss it further with the PCP (17%). The Chicago and New York respondents differed in their reasons for not seeking care with the PCP before their current abortion procedure. Chicago participants reported actively avoiding the PCP because of concerns of not getting support for their choice. Many New York participants, however, did not consider talking with the PCP as being a part of care because their decision about having an abortion was already made.

Themes explaining preference of first-trimester clinic for abortion service

Table 4 shows the frequency of reasons behind women's preferences for site for abortion care. Women theoretically opting for the PCC (n = 174) stated “comfort” as the most common reason, followed by the doctor “knowing their [medical] history.” For women theoretically choosing the abortion clinic (n = 118), the most common themes mentioned were “specialization” and the “privacy and anonymity” when the procedure is “separate” from one's regular source of care.

Table 4.

Themes Explaining Preference of Clinic for First-trimester Abortion, Stratified by Regional Site ^a

Theme	Total n (%)	Chicago n (%)	New York n (%)
Would theoretically choose abortion at PCC	174 (58.2)	103 (69.1)^b	71 (47.3)
Reasons
More comfortable at PCC	89 (51.1)	49 (47.6)	40 (56.3)
Medical clinic knows medical history	51 (29.3)	31 (30.1)	20 (28.1)
More convenient	22 (12.6)	16 (15.5)	6 (8.5)
Trust the physician more/feeling of safety	18 (10.3)	12 (11.7)	6 (8.5)
Other (clinic takes insurance, woman provider, parents not aware)	10 (5.7)	6 (5.8)	4 (5.6)
Would theoretically choose abortion at abortion clinic	118 (39.5)	43 (28.9)	75 (50.0)
Reasons
The clinic is specialized in the procedure	37 (31.4)	12 (27.9)	25 (33.3)
The clinic offers more privacy and anonymity	33 (28.0)	17 (39.5)	16 (21.3)
More comfortable, lack of judgment	26 (22.0)	7 (16.3)	19 (25.3)
Would like to keep this procedure separate	15 (12.7)	10 (23.3)	5 (6.7)
More convenient	8 (6.8)	0 (0.0)	8 (9.3)
Other (No memory of event, clinic reputation, recommended by friend, quality service)	6 (5.1)	2 (4.6)	4 (5.3)

Column totals may be >100%, reflecting multiple responses given by some respondents. Percentages calculated by number of respondents, not number of responses.

Although statistically more women at the Chicago site would theoretically choose the PCC as site for abortion care, there were no significant differences among the themes listed in this table using chi-square test (p > 0.05).

Discussion

First-trimester abortion in the United States is common. Such technologies as the introduction of medication abortion with mifepristone are making early abortion more accessible in a primary care setting. Yet in an era of patient-centered care where recent evidence suggests that patient preferences matter,²⁰ preferences of women who seek first-trimester abortion regarding where and why they would prefer to receive such services are largely unknown. Understanding how preferences may differ regionally is also unknown.

Our results indicate that the majority of women would theoretically choose their PCC as a site for abortion. These preferences appeared to differ regionally, with more Chicago participants than New York participants favoring the PCC option for site for abortion. Our study is unique because we used a similar survey instrument to that used by Rubin et al.¹³ but among a different population: participants with an identified PCP who were seeking abortion services at present. We showed that New York participants in this present study viewed the PCC as site for abortion less favorably than New York participants seeking primary care services within several family medicine clinics (47% vs. 67%). This may imply the inherent biases of the participant pool; women at an abortion clinic choose that option because it already is their preferred site for care. What this study adds, however, is an exploration of participants' interactions with their PCP (or other provider at the PCC) regarding their unplanned pregnancy. Interestingly, more than twice the number of women in Chicago reported seeking care with their PCP (or other provider at PCC) before their current abortion procedure compared with women in New York (37% vs. 17%). Many of these women indicated a need for advice about what to do or where to obtain an abortion, suggesting that some women rely on their PCP for thoughts about unplanned pregnancy and where to seek further care.

Of those women who did not seek care with their PCP before their abortion procedure, women in Chicago mostly expressed concern that their decision would be viewed unfavorably by their PCP, whereas women in New York mostly expressed no role for their PCP when their decision about abortion was already made. Such findings suggest that there may be regional differences about abortion stigma and that separate abortion clinics where women can have anonymity and avoid potential judgment from their PCP are important. These findings also imply that expansion of abortion services in the PCC in New York may be viewed more favorably but that considerable patient education about these services existing within this setting would be critical.

Shochet and Trussell¹² reported a lower proportion of women at several Iowa Planned Parenthood clinics theoretically choosing their PCC for abortion care (20% of participants of family physicians and 36% of participants of obstetricians/gynecologists vs. 58% of all participants in this present study). This Iowa-based study included women who did not necessarily identify a regular source of medical care and used only demographic characteristics in the logistic model to predict choice of PCC for abortion care, so potential variables may have been limited. Our study's strength is the greater overall sociodemographic variability of our populations in New York and Chicago and our inclusion of several PCP relationship characteristics as variables to predict preference for PCC as site for abortion care. Although many of our PCP characteristics did not show an association with the choice of PCC for abortion care, being somewhat or very comfortable with one's regular provider regarding issues of pregnancy prevention and birth control did. Being comfortable, although not well defined in this study, suggests that factors relating to the patient-doctor relationship may play an important role in preferences for PCC for abortion care.

Our results suggesting that many women would prefer to receive abortion services at their PCC may reassure physicians and clinic managers who are contemplating the addition of these services within primary care. Plans to integrate early abortion care into a PCC must address barriers regarding billing and reimbursement, malpractice coverage, ordering mifepristone, and possible additional funds needed to initiate services and training.^4,21
–23

This study has several limitations. Because of its preliminary nature, we did not use a validated questionnaire. The results show that variables included in the survey instrument and subsequent analysis did not adequately explain the differences in choice of site for abortion services. Chicago as a regional site showed an association with theoretical choice of PCC in the bivariate analysis, and participants at this site appeared to be from a higher socioeconomic group than women from New York, yet the demographic factors used in this survey study did not show an association with the theoretical choice of PCC as site for abortion care. Therefore, it would be important to include more demographic factors in future study, such as the woman's economic status, occupation, religion, or marital status. In this study, we intentionally opted not to ask women about race and ethnicity because of the inherent complexity within this category and the lack of clarity of how such information relates to patient preferences for abortion care.²⁴ For a future study, however, race/ethnicity may be important. Other factors to include might be those elicited from our open-ended responses, such as confidence about PCP skill level, PCC appointment availability, and more questions about respondents' relationship with their PCPs.

Although our open-ended questions may have assisted us with question design for a future study, they have limitations. For example, had the respondents had several options to choose from for possible reasons for their theoretical choice about preferred site for abortion care, the themes in Table 4 might have been listed differently. This study included the use of open-ended questions because of its exploratory nature and the paucity of information that exists regarding patient preferences. Home ZIP code would also have helped to understand whether women resided in these urban areas or traveled from outlying communities and how travel may have influenced their preferences for clinic for their first-trimester abortion service. Because of the limited sample size in this study, the findings may be applicable to women in these two clinics and other similar urban communities. They may not be generalizable to the population at large.

Conclusions

The majority of women surveyed would theoretically choose to have a procedure at their PCC. There were some regional differences between New York and Chicago. Although many of the variables used in this study did not explain the respondents' preferences, the PCC as site for abortion care appears to be an important potential option for many women. Primary care training programs should teach full-spectrum reproductive healthcare to help meet the needs of their patients.

Footnotes

Acknowledgments

We acknowledge the survey study staff at Montefiore Medical Center and the University of Illinois for facilitating this research. Special thanks to the women who participated in this study and to Family Planning Associates Medical Group, Limited, Chicago, and to the family planning clinic in New York, who allowed us to conduct this research. Thanks to Dr. Louis Keith for his critical feedback on the manuscript. We thank Kevin Grandfield for editorial assistance.

The study was supported by an anonymous foundation. The findings were presented as an Abstract at the North American Primary Care Research Group (NAPCRG) Annual Meeting, Vancouver, British Columbia, Canada, October 2007.

Disclosure Statement

No competing financial interests exist.

References

Jones

, Zolna

, Henshaw

, Finer

. Abortion in the United States: Incidence and access to services, 2005. Perspect Sex Reprod Health, 2008; 40:6–16.

Guttmacher Institute. Facts on induced abortion in the United States. New York: Guttmacher Institute, 2008.

Paul

, Nobel

, Goodman

, Lossy

, Moschella

, Hammer

. Abortion training in three family medicine programs: Resident and patient outcomes. Fam Med, 2007; 39:184–189.

Dehlendorf

, Brahmi

, Engel

, Grumbach

, Joffe

, Gold

. Integrating abortion training into family medicine residency programs. Fam Med, 2007; 39:337–342.

Bennett

, Johnson

, Wu

et al. A family medicine training collaborative in early abortion. Fam Med, 2007; 39:164–166.

Eastwood

, Kacmar

, Steinauer

, Weitzen

, Boardman

. Abortion training in United States obstetrics and gynecology residency programs. Obstet Gynecol, 2006; 108:303–308.

Schwarz

, Luetkemeyer

, Foster

, Weitz

, Lindes

, Stewart

. Willing and able? Provision of medication for abortion by future internists. Womens Health Issues, 2005; 15:39–44.

Joffe

, Yanow

. Advanced practice clinicians as abortion providers: Current developments in the United States. Reprod Health Matters, 2004; 12:198–206.

Steinauer

, Landy

, Jackson

, Darney

. The effect of training on the provision of elective abortion: A survey of five residency programs. Am J Obstet Gynecol, 2003; 188:1161–1163.

10.

Samora

, Leslie

. The role of advanced practice clinicians in the availability of abortion services in the United States. J Obstet Gynecol Neonat Nurs, 2007; 36:471–476.

11.

Martin

, Avant

, Bowman

et al. The future of family medicine: A collaborative project of the family medicine community. Ann Fam Med, 2004; 2:S3–S32.

12.

Shochet

, Trussell

. Determinants of demand: Method selection and provider preference among U.S. women seeking abortion services. Contraception, 2008; 77:397–404.

13.

Rubin

, Godfrey

, Shapiro

, Gold

. Urban female patients' perceptions of the family medicine clinic as site for abortion care. Contraception, 2009; 80:174–179.

14.

Guttmacher Institute. State funding of abortion under Medicaid. State policies in brief. New York: Guttmacher Institute, 2007.

15.

Fortney

, Sullivan

, Williams

et al. Measuring continuity of care for clients of public mental health systems. Health Serv Res, 2003; 38:1157–1175.

16.

Fan

, Burman

, McDonell

et al. Continuity of care and other determinants of patient satisfaction with primary care. J Gen Intern Med, 2005; 20:226–233.

17.

Nutting

, Goodwin

, Flocke

, Zyzanski

, Stange

. Continuity of primary care: To whom does it matter and when? Ann Fam Med, 2003; 1:149–155.

18.

Census 2000 Questionnaire Questions. www.census.gov/dmd/www/content.htm. 2008 February 6.

19.

Rose

. Guidelines implementation: Translating a new Canadian/U.S. health survey into Spanish. Arlington, VA: Federal Committee on Statistical Methodology (FCSM), 2003.

20.

Preference Collaborative Review Group. Patients' preferences within randomised trials: Systematic review and patient level meta-analysis. BMJ, 2008; 337:1–8.

21.

Bennett

, Aguirre

, Burg

et al. Initiating abortion training in residency programs: Issues and obstacles. Fam Med, 2006; 38:330–335.

22.

Leeman

, Espey

. “You can't do that ‘round here”: A case study of the introduction of medical abortion care at a University Medical Center. Contraception, 2005; 71:84–88.

23.

Prine

, Lesnewski

, Bregman

. Integrating medical abortion into a residency practice. Fam Med, 2003; 35:469–471.

24.

Kaplan

, Bennett

. Use of race and ethnicity in biomedical publication. JAMA, 2009; 289:2709–2716.