Patient choice: comparing criteria for selecting an obstetrician-gynecologist based on image, gender, and professional attributes

Meeting Papers

www. AJOG.org

Patient choice: comparing criteria for selecting an obstetrician-gynecologist based on image, gender, and professional attributes Peter F. Schnatz, DO; Jessica L. Murphy, MS; David M. O’Sullivan, PhD; Joel I. Sorosky, MD OBJECTIVE: The purpose of this study was to compare gender, humanistic qualities or technical competence (HQTC), and age when an obstetrician-gynecologist is selected. STUDY DESIGN: Participants saw photographs (2 women, 2 men)

without descriptors then the same photographs with descriptors. To test whether HQTC were more important than gender, the men in the photographs were given HQTC descriptors. Female patients, visitors, and staff at Hartford Hospital along with community and outpatient sites were recruited. Demographic information was collected. RESULTS: From 901 participants, 83% chose a woman, 59% of

whom selected gender or age as the reason. Single and younger

patients were more likely to choose female and younger providers, respectively. With descriptors, 62% of the women chose a male provider. A significant number chose a different gender provider (P ⬍ .001) and made their selection for a different reason (P ⬍ .001). CONCLUSION: More women chose a female provider when no additional information is known. A significant number changed their selection when male providers were described with HQTC.

Key words: compassion, competence, education, gender, humanistic qualities, obstetrician-gynecologist, preference, provider

Cite this article as: Schnatz PF, Murphy JL, O’Sullivan DM, Sorosky JI. Patient choice: comparing criteria for selecting an obstetrician-gynecologist based on image, gender, and professional attributes. Am J Obstet Gynecol 2007;197:548.e1-548.e7.

S

electing an obstetrics and gynecology and women’s health care provider is an important decision for women. Although many studies have looked at the factors that go into this de-

From the Departments of Obstetrics & Gynecology (Drs Schnatz and Sorosky and Ms Murphy), Internal Medicine (Dr Schnatz), and Research Administration (Dr O’Sullivan), Hartford Hospital, Hartford; the Departments of Obstetrics & Gynecology (Drs Schnatz and Sorosky) and Internal Medicine (Dr Schnatz), The University of Connecticut School of Medicine, Farmington; and New Britain General Hospital, New Britain (Dr Sorosky), CT. Presented at the annual meeting of the Council on Resident Education in Obstetrics and Gynecology and the Association of Professors of Gynecology and Obstetrics, Salt Lake City, UT, March 7-10, 2007. Received Mar. 15, 2007; revised May 30, 2007; accepted July 23, 2007. Reprints: Peter F. Schnatz, DO, FACOG, Hartford Hospital; Conklin Building 203B, 80 Seymour St, Hartford, Connecticut 06102; [email protected]. 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2007.07.025

548.e1

cision,1-16 there is debate about the role that gender plays in the process of selection. Whether women prefer female providers is of more than academic interest, because it appears to be affecting decisions of current medical students. The number of men who apply to obstetrics and gynecology residency programs has been decreasing steadily.17 In 2005, only 24% of obstetrics and gynecology residents were men,18 in contrast to 53.5% in 1990 and 80.6% in 1978.3,19 Some studies have suggested that women prefer a female obstetrician-gynecologist provider.1,2 Alternatively, other studies have shown that gender is of minor importance in the selection of an obstetrician-gynecologist provider.3,9,11,15 Asking a woman if she prefers a male or female provider, for instance, is an oversimplification of a complex decision. We previously have demonstrated that many women who state that they prefer a woman provider actually will select a male when choosing a provider. This suggests that other variables are of equal or greater importance.3 Additionally, patients may assume that male and female providers

American Journal of Obstetrics & Gynecology NOVEMBER 2007

each possess certain identifiable characteristics or stereotypes.20 Other studies have identified gender bias or stereotypes as the culprit for maintaining a leadership gap between the sexes. One such study that was based on a corporate model found that women are often characterized as being more sympathetic and emotional and that men are characterized as being more achievement-oriented and competitive.21 In medical practice, the characteristics that patients seek in physicians often vary depending on the physician’s specialty. This implies that stereotypes play a role in patient choice.22 Wolosin and Gesell23 explored the idea that an increased number of women entering the field of obstetrics and gynecology would add feminine traits that are more humanistic in quality and therefore would increase patient satisfaction. The study failed to prove this hypothesis and only cemented the idea that physician gender had no impact on patient satisfaction.2,3,11,23 Another study that explored gender stereotypes of physicians found that women considered both sexes to be equal in comprehensiveness, but that female physicians were more humane than

Meeting Papers

www.AJOG.org male physicians. Conversely, they considered men to be more technically competent.4 Results obtained from analyses of pilot data leading up to this study demonstrated that the number of women who chose a male provider doubled when they learn that both the male and female providers had humanistic qualities, technical competence, or other positive attributes.20 That pilot, which comprised a convenience sample of 149 women, suggested that physician characteristics had a significant impact on the gender of the provider who was selected.20 We therefore hypothesized that women would be even more likely to change their choice to a man if the male physicians were described preferentially as technically competent or as having humanistic qualities. In an attempt to build on our pilot data, we analyzed a larger sample and assigned humanistic qualities or technical competence to the male providers to see whether these attributes were more important than gender. Our objective was to compare the importance of gender, humanistic qualities and/or technical competence, and age in the selection of an obstetrician-gynecologist.

M ATERIAL AND M ETHODS This prospective study was reviewed and approved by the Hartford Hospital Institutional Review Board. Participants, not including the pilot study sample, were enrolled from January 15 to July 15, 2006, inclusive. Our sample, who are subsequently referred to as participants, included female patients, visitors, and staff at Hartford Hospital and at several community and outpatient sites. The same 3 women, who were trained concurrently, conducted all interviews faceto-face. During all the days of recruitment and at all sites, all available women were offered inclusion. Verbal consent was obtained from all women who enrolled. Participants were given 3 tasks. First, they were asked to view photographs of fictitious obstetrician-gynecologist physicians, 2 men and 2 women (Figure 1). During the second task, they were asked

FIGURE 1

The 4 photographs of fictitious physicians with the associated descriptor(s)

“graduated from the Medical University of Kalamazoo” “graduada de la Universidad de Medicina de Kalamazoo”

“warm bedside manner and compassionate” “buen cuidad de cabecera y compasivo”

“pleasant personality” “personalidad agradable”

“competent and excellent surgeon” “competente y excelente cirujano”

The photographs were first shown to study subjects without the testimonial and subsequently with the descriptor(s). Schnatz. Patient preferences of their provider. AJOG 2007.

to review the same 4 photographs to which had been added a simple description of a professional attribute (Figure 1). In both instances, the participant was asked to select her preferred physician. There was no indication of the factors that we deemed to be important or the factors that we were studying. It was emphasized that participants did not need to change their selection but, in each case, were to choose the provider that they would prefer for general obstetrics and/or gynecological care. The photographs of men were associated with descriptions of humanistic qualities or technical competence. We previously identified these attributes to be of higher importance to patients when they are selecting an obstetrician-gynecologist physician.3,20 During the third task, without photographs, the study participants were asked to select from a list of 8 characteristics or qualities, the ones that they believed

were most important when selecting an obstetrician-gynecologist physician. The 4 photographs were selected to provide a balance of age and gender; there were 1 “older” man and 1 woman (each with glasses) and 1 “younger” man and 1 woman (each without glasses). The people in all 4 photographs were white. When tested among a subset of our study population, there was no consensus on whether photographs represented white Anglo-Saxon, white Hispanic, or white Spanish individuals. After being asked to identify her choice of a physician, a list of 8 predetermined items was provided for each participant to select the reason that she made that particular choice. The physician attributes and list of 8 characteristics, which had been identified from our previous research,3,20 were available in English and Spanish and included race/ethnicity, gender, age (younger), age (older), compassion (warm bedside

NOVEMBER 2007 American Journal of Obstetrics & Gynecology

548.e2

Meeting Papers manner and communicates well), quality of care (experienced, competent, able, and good surgical skills), knowledge, and efficiency (does all that is needed without wasting time). Translation was available for these and any other language that was requested. In addition to these questions, demographic information was recorded. Patient ages were calculated as survey date minus patient date of birth. For the purpose of statistical analysis, nonsingle was defined as anyone who was married, divorced, widowed, or separated at the time of the survey. Each participant’s choice of a provider was not to be related to her current condition, situation, or reason for a visit; rather, it was supposed to reflect her choice of a new obstetrician-gynecologist for general care. However, to control for the potential confounding variable of stress, we evaluated participant stress level, based on their self-reported level of stress related to any medical conditions they had at the time of the survey. The stress level was reported with an ordinal scale: low, moderate, or high. While the pilot study of 149 women showed significant differences in the numbers who changed their choice from one gender to the other, the objective for collecting additional data was to acquire as large of a convenience sample as possible to provide sufficient power to detect differences in choice of a male or female provider between subgroups (eg, age and marital status). A power analysis demonstrated that 80% power to detect a conservative difference in proportions of 0.10 between subgroups would have been afforded with a sample size of 816. All analyses that resulted in a P value of ⬍.05 were considered statistically significant. Descriptive statistics for continuous variables (eg, age, educational background) included median, mean, standard deviation (SD), and 95% confidence interval (CI). Categoric variables were described with frequencies and expressed as a percentage. Inferential statistics were performed with both univariate and multivariate comparisons. The ␹2 statistic was used to evaluate differences in distribution between groups, and McNemar’s test was used to evaluate before548.e3

www.AJOG.org and-after differences by participant. Logistic regression was used for forward conditional (inclusion of variables at P ⬍.05, exclusion at P ⬍.10) models to generate odds ratios (ORs) and 95% CIs. SPSS software (v 14.0; SPSS, Inc, Chicago, IL) was used for all analyses.

R ESULTS A total of 997 women were invited to participate in this study. From these, 901 women (90.4%) agreed to participate. The mean age of participants (n ⫽ 893) was 39.3 ⫾ 16.3 years, with a range of 14.3-99.6 years. The median age was 34.6 years, with an interquartile range of 26.250.8 years. Minors were included because there was no risk to participate, and they are representative of some of the patients who were seen at Hartford Hospital and the community sites where the surveys were administered. To make the findings of the study more generalizable, subjects were categorized as inpatients, outpatients, visitors/staff members, or “mixed” (a combination of patients, visitors, and staff members), based on the location at which the survey was administered. A broad educational background was represented, with a range of no formal education to 24 years of educational training. The median educational background (n ⫽ 894 participants) was 14.0 years, with a mean of 14.2 ⫾ 3.3 years. The remaining demographics can be seen in Table 1. A total of 854 women completed the first task of selecting a provider from the 4 photographs. Of these, 83% (n ⫽ 706 women) chose a female provider and 17% (n ⫽ 148 women) chose a male provider (Figure 2). Specifically, 55% (n ⫽ 472 women) chose the younger woman; 28% (n ⫽ 234 women) chose the older woman; 5% (n ⫽ 47 women) chose the younger man; and 12% (n ⫽ 101 women) chose the older man. Single women were 35% more likely than nonsingle women (OR, 1.35; 95% CI, 1.131.62) to choose a female provider. Among those who gave a single response to the question about the reason that they made their initial choice, “compassion” was the reason most often given

American Journal of Obstetrics & Gynecology NOVEMBER 2007

(25%; n ⫽ 161 women), followed by “gender” (21%; n ⫽ 132 women) and “quality care” (20%; n ⫽ 131 women). The remaining responses (34%; n ⫽ 216 women) included age, knowledge, efficiency, and race/ethnicity. Although the stress questionnaire was not validated previously, there was no significant difference in the selection of a male vs a female provider based on participant reported medical stress level. When the photographs were reevaluated with descriptive attributes, data were available for 876 women (some women refused to do the first task). With descriptive information, 38% (n ⫽ 331 women) chose a female provider and 62% (n ⫽ 545 women) chose a male provider (Figure 2). More specifically, 23% (n ⫽ 198 women) chose the younger woman; 15% (n ⫽ 133 women) chose the older woman; 21% (n ⫽ 187 women) chose the younger man, and 41% (n ⫽ 358 women) chose the older man. The number who chose a women was significantly less than the number who chose a man after the descriptions (P ⬍ .001). Furthermore, the number of patients who chose a man significantly increased with the descriptors (P ⬍ .001), although the number who chose a woman significantly decreased (P ⬍ .001). Among those who gave a single response to the reason that they made their choice with descriptions, “quality care” was given most often as the reason (45%; n ⫽ 345 women), followed by “compassionate” (35%; n ⫽ 275 women) and “knowledge” (7%; n ⫽ 53 women). The remaining responses (13%; n ⫽ 99 women) included race/ethnicity, gender, age, and efficiency. Although there were statistically significant differences between the mean ages of “mixed” subjects (41.9 ⫾ 15.4 years) vs inpatients (37.5 ⫾ 17.0 years; P ⫽ .003) and visitors/staff members (33.9 ⫾ 11.5 years; P ⫽ .032), there were no significant differences among any of the 4 participant categories in the proportion of subjects who first chose a woman and then (after descriptions) chose a man. Of the 706 women who initially chose a women, 53% (n ⫽ 376 women) selected a man after receiving additional information. Among those who gave a

Meeting Papers

www.AJOG.org single response about the reason they made the new choice of a man, “quality care” was given most often as the reason (62%; n ⫽ 219 women), followed by “compassionate” (37%; n ⫽ 132 women) and “younger” (1%; n ⫽ 3 women). Of the 148 women who initially chose a man, 3% (n ⫽ 5 women) changed to a woman after receiving additional information. When we analyzed the change in reason for physician selection based on testimonial, there were 817 responses. Based only on the photographic image, 82% of women (n ⫽ 673) chose a female provider. After seeing the written testimonial, 41% (n ⫽ 275 of the 673) indicated the same characteristic; 59% (n ⫽ 398) picked a different reason for their provider selection. Based only on the photographic image, 17.6% of women (n ⫽ 144) chose a male provider. After seeing the written testimonial, 50% (n ⫽ 72 of the 144) indicated the same characteristic, and 50% (n ⫽ 72) picked a different reason for their provider selection. Therefore, more than one-half of women who initially chose a female provider based on photograph alone chose a man after the testimonials were introduced. At the same time, fewer than 5% of the women who initially chose a male provider chose a female provider after reading the testimonials (P ⬍ .001). Similarly, nearly 60% of the women who initially chose a female provider made their posttestimonial choice for a different reason. Among those women who initially chose a male provider, one-half of the women picked the same reason after the testimonial was introduced (P ⬍ .001). When we looked at those women who changed their provider gender preference after seeing the testimonials, few demographic variables had significant effects. For these analyses, demographic variables were dichotomized as either yes/no (white, Hispanic, single, children) or stratified (age ⱖ40 or ⬍40 years, education ⱕ12th or ⬎12th grade, and income ⱕ$40,000 vs ⬎$40,000 annually). The women who changed their preference to a male provider differed in their age, marital status, and stress level. Older patients were more likely to select

TABLE

Demographic data for the study participants Variable

N*

%

Ethnicity

892

99.0

Hispanic

257

28.8

Non-Hispanic

635

71.2

..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................

Race

880

97.7

White

504

57.3

African American

145

16.5

18

2.0

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

Asian/Pacific Islander

.....................................................................................................................................................................................................................................

Native American

2

0.2

Mixed

8

0.9

Other

203

23.1

899

99.8

English

870

96.8

Spanish

28

3.1

Chinese

1

0.1

896

99.4

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..............................................................................................................................................................................................................................................

Primary language spoken

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................

Marital status

.....................................................................................................................................................................................................................................

Married

453

50.6

Single

322

35.9

Widowed, divorced, or separated

121

13.5

870

96.6

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..............................................................................................................................................................................................................................................

Children (n)

.....................................................................................................................................................................................................................................

None

141

16.2

ⱖ1

729

83.8

894

99.2

326

36.5

..................................................................................................................................................................................................................................... † .............................................................................................................................................................................................................................................. ‡

Highest level of education (y)

.....................................................................................................................................................................................................................................

High school or lower (ⱕ12)

.....................................................................................................................................................................................................................................

College (13-16)

421

47.1

Graduate/professional (⬎16)

147

16.4

Household income

809

89.8

⬍$15,000

153

18.9

$15,000-$40,000

225

27.8

$40,001-$75,000

247

30.5

⬎$75,000

184

22.7

884

98.1

Protestant

269

30.4

Roman Catholic

365

41.3

Jewish

30

3.4

Muslim

10

1.1

91

10.3

119

13.5

..................................................................................................................................................................................................................................... .............................................................................................................................................................................................................................................. ..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..............................................................................................................................................................................................................................................

Religion

..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... ..................................................................................................................................................................................................................................... .....................................................................................................................................................................................................................................

Other

.....................................................................................................................................................................................................................................

No religion

..............................................................................................................................................................................................................................................

* Total participants, 901. †

Range, 1-9 children; mean, 2.3 children.

‡

Range, 0-40 years; mean ⫾ SD, 14.2 ⫾ 3.3 years; median, 14 years.

NOVEMBER 2007 American Journal of Obstetrics & Gynecology

548.e4

Meeting Papers

Although many studies have suggested that female patients prefer a female medical provider,1,2 other data have suggested that patients change their selection when they choose who they will see3 or when they are given additional information besides gender.20 Some women may prefer a certain gender when considering sensitive obstetrics and/or gynecological care. These patients are unlikely to change their selection to the opposite gender when given new information. For other women, certain characteristics or qualities may carry high degrees of importance during the selection process. These patients are more likely to change their provider when given new information, regardless of gender, because gender is of less importance. Our pilot data revealed that, after the fictitious providers were all given positive qualities, the choice of a male provider went from 20% to 40%.20 Because all providers were given potentially desirable qualities, it is hard to tell what proportion of those women who kept their selection of a female provider was due to gender vs the quality description. For this reason, the current study gave the female providers qualities that were 548.e5

Patient’s choice: before and after a written descriptor 100

90

female

male

female

male

O/F

O/M

80

70

O/F 60

50

40

30

20

Y/F

Y/F O/M

10

Y/M Y/M 0 picture only (n=854)

picture and info (n=876)

Y/F – younger female Y/M – younger male O/F – older female O/M – older male

Graphs show the distribution of the providers who were selected by the subjects based on photograph alone (left side) and based on photograph plus description (right side). Schnatz. Patient preferences of their provider. AJOG 2007.

less desirable. It should be pointed out, however, that, in the current study, the quality of having attended a nonspecific school (of less importance from our previous data3,20) was viewed by many women as a desirable attribute. Participants concluded that, because this person had “graduated from medical school” (Figure 1; physician in the upper left), they must be intelligent and knowledgeable, which is one of the highly desirable qualities. Previous data have shown that most patients who have a gender preference prefer a female provider.3 Most patients appear not to have gender preferences and desire providers who are knowledgeable, experienced, and capable.3 Patients may learn of these qualities by provider’s reputation or word of mouth. Without further information, however, patients may be left to make assumptions that are based on stereotypes. Specifically, patients tend to assume that female providers are compassionate and exhibit

American Journal of Obstetrics & Gynecology NOVEMBER 2007

more humanistic qualities,4,16,23 although they assume male providers are more technically competent and domineering.4,10 There are also data to suggest measurable differences in the way male FIGURE 3

Frequencies of characteristics and attributes that subjects chose as the 3 most important in their choice of an obstetrician-gynecologist provider 100 90 80 70 60 % in Top 3

C OMMENT

FIGURE 2

percentage of selections

an older physician (OR, 2.05; 95% CI, 1.18-3.57; P ⫽ .011). Single patients were less likely to pick an older physician (OR, 0.48; 95% CI, 0.25-0.92; P ⫽ .028). The patients with lower stress levels also were less likely to pick an older physician (OR, 0.54; 95% CI, 0.32-0.93; P ⫽ .027). Of those women who initially chose a younger physician without glasses or an older physician with glasses and then changed after seeing the testimonials, none of the demographic variables played significant roles. There were not enough women who changed their preference from a male to a female provider to analyze this variable. During the third task, when we asked the women to rank the 3 most important factors for selecting an obstetrician-gynecologist provider, “quality care,” “compassion,” and “knowledge” all were ranked in the top 3 ⬎50% of the time; “gender” was chosen in the top 3 ⬍20% of the time (Figure 3).

www.AJOG.org

50 40 30 20 10 0 quality of care

compassionate

knowledge

efficiency

gender

older

younger

characteristic / attribute

Schnatz. Patient preferences of their provider. AJOG 2007.

race/eth

Meeting Papers

www.AJOG.org and female providers interact with patients.5-7 Interestingly, many women do not believe that there is a difference in the caring qualities of male vs female physicians,10 and patient satisfaction consistently has been equal, regardless of provider gender.2,3,11,23 A recent study showed that, without extra information, competency is much harder to assume than compassion; therefore, patients, unless explicitly told of competency, will choose physicians who they assume have humanistic qualities.8 Because patients stereotype women as having greater humanistic qualities and traits, it is not surprising that 83% of the patients in our study, given only photographs, chose a female provider. These patients also stated that this initial selection was based on gender or age 59% of the time, which are factors that could be the basis of a stereotype. Because most data suggest that gender is of minor importance when one must select a provider, it is not surprising that a significant number of patients (53%) changed their selection from a female to a male provide when given additional information. Furthermore, only 17% of the participants said that they made their new selection on the basis of gender or age; 83% made their selection on the basis of the stated qualities of compassion, knowledge, or efficiency. Limitations of this study include the multiple variables and factors that are involved in choosing a physician, which makes it difficult to study this topic. Showing photographs causes a limitation in that there are many variables and thought processes for which we cannot control. Interestingly, however, this is much closer to a real life situation, where there are multiple variables, compared with simply asking a woman whether she prefers to see a male or a female physician. One way we attempted to control for the many variables and to create objectivity was to compare participant choices on the basis of knowing the provider’s gender and perceived qualities vs knowing the provider’s gender along with provided qualities. Additional limitations include the appearance of the fictitious providers. Although all 4 individuals in the photographs were white, 3

of them had off-white skin tone. Among a subset of our survey population, there was no consensus on which photograph represented a white Anglo-Saxon, white Hispanic, or white Spanish individual. Furthermore, the introduction of age among the providers adds another variable to be factored into comparisons. Although this survey was administered only by female study personnel, the initial results (phase I) were almost identical to those that were obtained from our pilot data, which were collected by a man.20 This is consistent with another study that showed no difference in selection based on the gender of the interviewer.2 In addition, recent similar methods has been reported.24 The potential study population was recruited from greater Hartford, CT, which is a metropolitan area of 1.2 million people. In addition, the diverse recruitment of patients, the broad demographic backgrounds (many nonwhite individuals and a broad representation of religions), and a diverse socioeconomic representation suggests that these results may be generalizable to other populations. The following are the clinical implications of this study: First, male medical students who are interested in obstetrics and/or gynecological should be encouraged to understand that, for most women, gender is of minor importance. Although many female patients may say they prefer a woman provider, this is often based on stereotypes. When a patient realizes or believes that a male provider is technically competent or possesses humanistic qualities, she is unlikely to see gender biases. Second, male practicing physicians should also be encouraged and may choose to use this information to help advertise. Whether by word of mouth, yellow page, newspaper advertisements, or other marketing strategies, male obstetrician-gynecologist providers may benefit by letting patients know that they are technically competent or that they possess humanistic qualities. Knowledge about an obstetrician-gynecologist provider appears to influence significantly a patient’s selection of her provider. Gender appears to significantly influence choice when considered alone. Younger patients have a greater ten-

dency than do older patients to prefer younger female providers. Nonsingle patients are more likely than single patients to prefer male providers. The humanistic qualities and technical competence of a provider, for many women, appears to be more important than gender. Most of the women in this study believe that quality of care, compassion, and knowledge are the most important factors when selecting an obstetrician-gynecologist provider. Concordantly, when male providers were described preferentially as being technically competent or as having humanistic qualities, most participants chose a man. These findings suggest that women tend to project certain qualities or technical skills, based on gender stereotypes, when no additional information is given. These data also confirm what many physicians have hoped, that women primarily want a physician who cares about their well-being and has the training and experience to make a meaningful difference in their f life. ACKNOWLEDGMENTS The authors thank Laura Kubica and Verónica Y. Schmidt Terón for assistance with data collection and J. David Schnatz, MD, and William Metheny, MD, for reviewing the manuscript.

REFERENCES 1. Chandler PJ, Chandler C, Dabbs ML. Provider gender preference in obstetrics and gynecology: a military population. Mil Med 2000;165:938-40. 2. Zuckerman M, Navizedeh N, Feldman J, McCalla S, Minkoff H. Determinants of women’s choice of obstetrician/gynecologist. J Womens Health Gend Based Med 2002;11:175-80. 3. Johnson AM, Schnatz PF, Kelsey AM, Ohannessian CM. Do women prefer care from female or male obstetrician-gynecologists? A study of patient gender preference. J Am Osteopath Assoc 2005;105:369-79. 4. Fennema K, Meyer DL, Owen N. Sex of physician: patients’ preferences and stereotypes. J Fam Pract 1990;30:441-6. 5. Roter DL, Geller G, Bernhardt BA. Effects of obstetrician gender on communication and patient satisfaction Obstet Gynecol 1999;93: 635-41. 6. van Dulmen AM, Bensing JM. Gender differences in gynecologist communication. Women Health 2000;30:49-61. 7. Hall JA, Irish JT, Roter DL, Ehrlich CM, Miller LH. Gender in medical encounters: an analysis of physician and patient communication in a pri-

NOVEMBER 2007 American Journal of Obstetrics & Gynecology

548.e6

Meeting Papers mary care setting. Health Psychol 1994;13: 384-92. 8. Bendapudi NM, Berry LL, Frey KA, Parish JT, Rayburn WL. Patients’ perspectives on ideal physician behaviors. Mayo Clin Proc 2006;81:338-44. 9. Plunkett BA, Kohli P, Milad MP. The importance of physician gender in the selection of an obstetrician or a gynecologist. Am J Obstet Gynecol 2002;186:926-8. 10. Elstad JI. Women’s priorities regarding physician behavior and their preference for a female physician. Women Health 1994;21:1-19. 11. Howell EA, Gardiner B, Concato J. Do women prefer female obstetricians? Obstet Gynecol 2002;99:1031-5. 12. Haar E, Halitsky V, Stricker G. Factors related to the preference for a female gynecologist. Med Care 1975;13:782-90. 13. Schmittdiel J, Selby JV, Grumbach K, Quesenberry CP Jr. Women’s provider preferences for basic gynecology care in a large health maintenance organization. J Women’s Health Gend Based Med 1999;8:825-33.

548.e7

www.AJOG.org 14. Fang MC, McCarthy EP, Singer DE. Are patients more likely to see physicians of the same sex? Recent national trends in primary care medicine. Am J Med 2004;117:575-81. 15. Mavis B, Vasilenko P, Schnuth R, Marshall J, Jeffs MC. Female patients’ preferences related to interpersonal communications, clinical competence, and gender when selecting a physician. Acad Med 2005;80:1159-65. 16. Kerssens JJ, Bensing JM, Andela MG. Patient preference for genders of health professionals. Soc Sci Med 1997;44:1531-40. 17. Lyon DS. Where have all the young men gone? Keeping men in obstetrics & gynecology. Obstet Gynecol 1997;90:634-6. 18. Brotherton SE, Etzel SI. Graduate medical education, 2005-2006. JAMA 2006;296:1154-69. 19. The Resident Analysis Report, Council on Resident Education in Obstetrics and Gynecology, The American College of Obstetrics and Gynecologists, 1978 and 1990. 20. Guile M, Schnatz PF, O’Sullivan DM. Relative importance of gender in patients’ selection

American Journal of Obstetrics & Gynecology NOVEMBER 2007

of obstetrics and gynecology provider. Conn Med 2007;71:325-32. 21. Catalyst Research and Advisory Organization. Women “take care,” men “take charge:” stereotyping of U.S. business leaders exposed. Sponsored by the General Motors Corporation. Catalyst Publication Code D62; ISBN # 0-89584-252-1. Last accessed: May 30, 2007. Available at: http://www. catalystwomen.org/files/full/Women%20 Take%20Care%20Men%20Take%20Charge. pdf. 22. Harris MB, Conley-Muth MA. Sex role stereotypes and choice of physician. J Am Med Womens Assoc 1983;38:149-54. 23. Wolosin RJ, Gesell SB. Physician gender and primary care patient satisfaction: no evidence of “feminization.” Q Manage Health Care 2006;15:103. 24. Trilla A, Aymerich M, Lacy AM, Bertran MJ. Phenotypic differences between male physicians, surgeons, and film stars: comparative study. BMJ 2006;333:1291-3.