HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bender, D. J. Search for Related Content PubMed PubMed Citation Articles by Bender, D. J.

Patient Preference for a Racially or Gender-Concordant Student Dentist

Key words: disparities, dentistry, gender, gender concordance, race, racial concordance, patient preference, patient choice

Submitted for publication 12/18/06; accepted 03/04/07

	Abstract

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 
Disparities in health and health care due to race and ethnicity are a national problem. One commonly proposed method to address disparities is to increase the number of underrepresented minorities in the health professions to serve the needs of growing minority populations. This position is based in part on the racial concordance hypothesis, an untested assumption that minority patients prefer to be treated by providers of the same race as themselves. The purpose of this study was to test the racial concordance hypothesis. Gender preference of dental patients was also investigated. One hundred twenty male and female Caucasian, African American, Hispanic, and Asian dental patients over eighteen were recruited from the reception area of a private dental school clinic in Northern California. Participants were randomly assigned to a treatment condition that gave them a choice between two equally qualified fictitious dental students that varied on race and gender. Participants were initially blinded to the study’s purpose to preserve the authenticity of the choice decision. Results showed that 58 percent of the participants had no preference for the race or gender of their student dentist, but that some black and Hispanic patients preferred a racially concordant student dentist and some female patients preferred a gender-concordant student dentist. Hispanic females were especially likely to prefer racial and gender concordance. The findings suggest that the racial concordance hypothesis may not apply to choosing a dentist and thus may not be as strong an argument to justify efforts to increase diversity in dental schools as previously thought.

Research has shown that ethnic and racial minority patients, especially African Americans, Hispanics, and American Indians/Alaska Natives, have higher mortality rates and lower life expectancy and suffer from more chronic and disabling diseases than whites.^4,7,10 Minority patients are less likely to have health insurance¹¹ and a regular primary source of care¹² and are more likely to be treated in emergency rooms or in community or outpatient clinics.^2,13 Researchers have uncovered evidence to suggest that minority patients receive lower quality care than Caucasians¹⁴ and that they are less likely than white patients to report being treated with respect.^15,16 Even when important covariables such as health insurance, income, and access to care are controlled, health disparities persist.^2,12 Some evidence suggests they are increasing.^12,17,18

Disparities in oral health and health care between minority and nonminority patients have also been identified. The U.S. surgeon general’s report on oral health concluded that "Non-Hispanic blacks, Hispanics, and American Indians and Alaska Natives generally have the poorest oral health of any of the racial and ethnic groups in the U.S. population" (p. 74).¹⁹ The report shows, in comparison to white Americans, that Hispanics, African Americans, and American Indians have higher rates of tooth decay; African Americans and American Indians have higher rates of gum disease; and African American males have the highest incidence of oral and pharyngeal cancers but are less likely to be diagnosed and have lower survival rates than other racial groups. In addition, American Indians, African Americans, and Hispanics are more likely to rate their oral health as fair or poor, and despite having higher levels of unmet dental needs, African American and Hispanic patients are more likely than whites to report never having seen a dentist or not having seen a dentist within the last twelve months.

Formicola et al.²⁰ concluded that a major challenge facing medicine and dentistry is how to address racial and ethnic disparities in health care. Two commonly proposed methods to address the disparities problem are to a) increase the cultural competency of all health care providers so they are more able to treat patients from a variety of racial, ethnic, and cultural backgrounds, and b) increase the number of underrepresented minority (URM) students (African American, Hispanic/Latino, and Native American/Alaska Native) in health professions schools to serve the health care needs of growing minority populations. As the gatekeepers and educators of the nation’s health care workforce, health professions schools are heavily involved in both efforts.^8,21,22 With one of the least diverse practitioner populations of all health sciences,¹⁸ dentistry is arguably in particular need of a deeper understanding of the role of race in the patient-provider relationship.

The notion of increasing the number of URM students in health professions schools in order to train providers to treat growing minority populations is widely supported as one method of addressing disparities.^{2,8,9,23–27} This position appears to be based in part on the racial concordance hypothesis. This hypothesis suggests that minority patients are more comfortable with members of their own race and are therefore more likely to choose a racially concordant provider when given the choice.²⁸ This assumption persists despite the lack of substantive and direct empirical evidence to support it.^25,29 Conclusions about patient-provider racial concordance have been drawn indirectly from research showing a) minority providers in a variety of health care fields treat minority patients in greater numbers than do their Caucasian counterparts,^21,25,30 even when the percentage of African Americans and Hispanics is statistically controlled,³¹ and b) patient self-report of having a choice of providers is associated with greater patient-provider racial concordance.^28,32,33 However, the literature review for my investigation shows that the preference of adult patients for racially concordant providers has not been tested in a clinical setting in medicine, and not at all in dentistry. The primary purpose of this study was to test the racial concordance hypothesis.

Research has shown that patient-provider racial concordance exists to varying degrees in some health care fields, including medicine and dentistry.^32,34,35 However, the factors contributing to this racial matching dynamic are not clearly understood.³⁶ Geographical accessibility, increased patient satisfaction, gender, socioeconomic status, and primary language have been suggested as possible explanations as to why URM patients may seek treatment from a URM provider.^{32,33,37–39} As Saha et al.³³ and Gray and Stoddard²⁸ have pointed out, however, it is not clear what role, if any, patient choice plays in determining patient-provider racial concordance.

Results from research on patient preference for racially concordant providers are inconsistent. Some studies show that racial concordance is important to patients. In a secondary data analysis of the 1987 National Medical Expenditure Survey, Gray and Stoddard found that minority patients were more than twice as likely as nonminority patients to report having a minority provider.²⁸ This was particularly true for Hispanics and African Americans. Saha et al. analyzed data from the 1994 Commonwealth Fund Minority Health Survey to determine the extent to which racial concordance is a matter of patient choice or geographical accessibility.³³ The authors concluded that patient choice played a role in the selection of a physician for African American and Hispanic subjects, but that Caucasians were more likely to choose a provider based on geographical accessibility. Findings also showed that language concordance may be an important factor for Spanish-speaking Hispanic patients. Using the same 1994 national data set, LaVeist and Nuru-Jeter studied factors that contribute to patient-provider racial concordance, including patient choice.³² Results showed that, regardless of race, all respondents were significantly more likely to choose a racially concordant physician when given a choice and, furthermore, that patient report of having a choice predicted doctor-patient racial concordance.

Other studies have found that doctor-patient racial concordance was not an important consideration for patients. In an early study that acknowledged the untested assumption underlying the racial concordance hypothesis, Bertakis surveyed black and white patients selected at random in the reception area of a county clinic in the San Francisco Bay Area to determine if they were more satisfied with care provided by racially concordant doctors.⁴⁰ Ninety percent of respondents disagreed or strongly disagreed that they would prefer to see a racially concordant doctor. Furthermore, all but one respondent disagreed or strongly disagreed that provider race impacted the care they received. In a study using videotaped scenarios, Aruguete and Roberts investigated the effect of physician race and nonverbal communication style on patient evaluations of physicians.⁴¹ Contrary to their hypothesis, these researchers found that participants did not rate the racially concordant provider higher on any measure (trust, satisfaction, disclosure, recommendation, recall, and compliance). Saha et al. analyzed responses to the 2001 Commonwealth Fund’s Health Care Quality Survey to study the effect of race on patient satisfaction and use of health care services.⁴² One item on the survey asked if respondents preferred to be treated by a physician of the same race as themselves. Results revealed that patient-provider racial concordance was not related to better doctor-patient interactions or to increased use of services by minority patients. Furthermore, only 10 percent of subjects responded positively to the question of whether they preferred a racially concordant provider—African Americans were the least likely and Hispanics the most likely to report a preference for a racially concordant physician. In a study from dentistry, Hardie et al. measured satisfaction in a group of 462 patients at a dental school clinic in South Central Los Angeles.³⁸ These researchers hypothesized that satisfaction among nonemergency dental patients would be higher when treated by a dentist of the same racial background and that patients with high levels of dental anxiety would show a greater preference for a racially concordant provider. Results showed that satisfaction was unrelated to racial concordance, although participants with low levels of satisfaction had stronger race preferences than did patients with high levels of satisfaction. As predicted, higher levels of dental anxiety were associated with a stronger preference for a racially concordant dentist in this sample.

A third line of investigation on doctor-patient racial concordance concludes that, while race is important to some patients, it can be mediated by factors such as compassion, courtesy, or professional competence. In an early study from health care marketing, Crane and Lynch interviewed 100 randomly selected adults in Halifax, Nova Scotia, to determine the most salient choice factors in selecting a doctor and a dentist.⁴³ Courtesy and competency were the most frequently named criteria for both providers. Although race was not examined in this study, other demographic variables were consistently ranked important by less than 5 percent of respondents. In a study with similar methodology, Bornstein et al. asked 636 adults to rank twenty-three items in order of importance in the selection of a health care provider.⁴⁴ The most important factors were board certification, a recommendation from family or friends, the physician’s appearance, and area of specialization. The lowest ranked items were those related to the provider’s personal characteristics, such as gender, age, race, and marital status.

The most recent study on patient choice and doctor-patient racial concordance uncovered in the literature review was conducted at the University of California, San Francisco by investigators from the Schools of Dentistry and Medicine. Using videotaped actors portraying physicians of differing race and gender, Gerbert et al. tested the hypothesis that subjects would choose a male Caucasian physician when given a choice of doctors of different races and gender.⁴⁵ Over half of all subjects initially chose a Caucasian provider, but 50 percent of African Americans, 51 percent of Hispanics, and 66 percent of Caucasians initially chose a racially concordant provider. After viewing a forty-five-second videotaped health prevention message from each actor-physician, preference for the Caucasian provider decreased from 53 percent to 44 percent. Preference for a racially concordant provider increased from initial to final selection only for African Americans. These researchers concluded that the stereotype of the Caucasian physician may be waning and that race may be less important than other factors to some groups in the selection of a physician.

In addition to inconclusive results, much of the research on patient preference has methodological drawbacks common to studies of health care delivery and access to care that limit the generalizability of the findings. The Institute of Medicine noted that studies of patient preference often showed weakness in study design, methodology, or data analysis.² My literature review did not uncover a single study that used a randomized control design to assess patient preference in an authentic setting. The majority of research on patient choice and doctor-patient racial concordance has been descriptive or correlational. Surveys and secondary analyses of large national data sets appear to be common data collection and analysis techniques.^28,32,39,42 Survey measures may or may not reflect participants’ actual behavior and have been shown to be less reliable than other measures,^36,37 whereas secondary data analyses prohibit the manipulation of major study variables. The reliance on self-report or proxy measures for critical variables such as patient or provider race and patient choice illustrate measurement concerns.^28,33,35,39 Sampling can be nonrandom or not representative.

Some investigators have collected data on patient preference by asking participants outright whether race is important to them when choosing a provider.^33,37,42,46 This approach may be unreliable because it could prompt socially desirable responses. Others have gathered data in nonauthentic or nonclinical settings, such as shopping malls and community gatherings.^44,45 In more controlled studies, researchers have used photographs of alleged providers, videotapes of actors portraying providers, or written vignettes describing a hypothetical clinical situation to study the independent effect of race on patient preference.^45,47,48 As noted by Aruguete and Roberts, however, conducting research of this type in an artificial setting could be problematic because patient behavior in a simulated medical visit may be different from that in an actual medical visit.⁴¹ Health care researchers have pushed for explicit examination of patient preference for providers,^33,44,49 and the Institute of Medicine^2,5 and others^29,50 have called for more rigor in this area. The investigation reported in this article attempted to address the methodological limitations of past research by using an experimental design to study adult patients in a clinical setting who believed they were selecting an oral health care provider.

To strengthen the study’s design, gender was included as a secondary independent variable. Combining the race and gender variables created three treatment conditions to which participants were randomly assigned. There has been considerable research on gender preference in health care, but like the race literature, results are inconclusive. Several studies in medicine and dentistry have reported that a majority of participants express no gender preference for providers.^38,51,52 However, Cooper-Patrick et al.⁵³ found that gender concordance was highly correlated with patient satisfaction, and Hardie et al.³⁸ concluded that dental patients with high levels of dental anxiety were more likely to choose a gender-concordant dentist. Gerbert et al.⁴⁵ reported that, regardless of subject race or gender, all subjects preferred the female physician shown on videotape, while Bare and Dundes⁴⁹ found that dental patients with high levels of dental anxiety preferred a male dentist. Although it appears that gender, like race, may be important to some patients, research has also indicated that gender may be mediated by factors such as effectiveness of treatment, compassion and empathy, or language ability.^37,45,51

The primary purpose of my study was to test the racial concordance hypothesis in a clinical setting on participants who believed they were choosing a student dentist. Specifically, the study sought to determine to what extent adult Caucasian, African American, Hispanic, and Asian/Pacific Islander dental patients had a preference for a racially concordant provider when given a choice between two equally qualified candidates. The study also explored whether adult dental patients had a preference for the gender of their provider. Results of a controlled study supporting the racial concordance hypothesis might provide evidence to suggest that increasing the number of underrepresented minority students in dental schools is a justifiable way to reduce disparities in oral health and health care due to race.

	Method

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 
This study used a randomized experimental between-subjects design to explore to what extent adult dental patients had a preference for the race and gender of an oral health care provider. Participants were randomly assigned to one of three treatment conditions and asked if they had a preference for a fictitious student dentist. At the time the preference was solicited, participants were unaware of the fictitious nature of the dental students. The three treatment conditions were a) same gender/same race student dentist and same gender/different race student dentist; b) same gender/same race student dentist and different gender/same race student dentist; or c) same gender/same race student dentist and different gender/different race student dentist. Participants could indicate a) preference for the racially and gender-concordant student dentist, b) preference for the student dentist of the alternative race/gender combination, or c) no preference. Participants who indicated a preference were asked in a debriefing at the conclusion of the study why they chose a particular student dentist. Restricted randomization was used to balance assignment to treatment conditions such that five male and five female subjects from all racial groups (African American, Hispanic, Asian/Pacific Islander, Caucasian) were assigned to each treatment condition. Once this requirement was met, no further assignments were made to that condition.

Sample
Stratified random sampling was used to recruit 120 first-time, nonemergency, English-speaking patients over eighteen years old seeking comprehensive dental care in the clinic of a medium-sized private dental school in the San Francisco Bay Area. Stratification was on race and gender such that fifteen males and fifteen females from each of the four racial groups under study were recruited. Data were collected from March to August 2005. At the start of data collection the student body at the study site was 36 percent female, 56 percent Caucasian, 33 percent Asian/Pacific Islander, 1 percent African American, 4 percent Hispanic, 2 percent multiethnic, and 4 percent unknown/other/decline to state. One student reported being Native American/Alaska Native.

Participants were recruited in the clinic reception area during morning and afternoon clinic sessions. All participants were presenting for an initial screening appointment and did not receive dental treatment on the study days. First-time patients were chosen as participants for several reasons: they do not receive dental treatment at the initial visit, thereby eliminating a factor that could compromise willingness to participate in the study; responses from first-time patients are not confounded by prior positive or negative experience with treatment at the school; and new patients’ lack of familiarity with the school and clinic protocol may have lessened the transparency of the study. American Indian/Alaska Native patients were not included because they comprised less than 1 percent of the school’s patient population, making the recruitment of a sufficient number of participants from this group difficult. Also excluded from the study were individuals who identified their background as bi- or multiracial, who did not report race or gender, or who could not competently communicate in English.

The participation rate was 86 percent. Participation was voluntary, and subjects were not compensated and did not receive free or reduced cost dental care. One hundred seventy first-time, nonemergency adult patients were approached by the researcher as potential subjects. Thirty (18 percent) did not meet the study’s inclusion criteria and were excluded from the final sample. In addition, thirty-seven Caucasians (twenty-one males, sixteen females), one Hispanic male, and three African American females were excluded because the required number of participants of their racial background and gender had been attained.

Of the 140 eligible participants, twenty did not grant consent (14 percent). Seven African Americans (three males, four females), five Caucasians (two males, three females), six Asian/Pacific Islanders (two males, four females), and two Hispanic females did not agree to participate in the study. Among the eighteen participants who reported a reason for not granting consent, lack of time was the most common (n=7; 39 percent).

Instruments
Two researcher-designed instruments were used in the study: a dental student profile and a new patient survey.

A dental student profile was created for eight fictitious dental students, one for each gender within the four racial groups included in the study (Appendix A). Fictitious dental student information was used to allow the researcher to manipulate the gender and race of the students presented to participants. Each profile contained a centered line at the top of the page where the name of the fictitious student provider was neatly handwritten in block letters. Below the line were checklists for race, gender, and three distracter variables (year in program, rank in dental class, and undergraduate major). Participants were not told initially of the fictitious nature of the profiled students to preserve the authenticity of their choice decision.

View larger version (17K): [in this window] [in a new window]

At the start of the recruitment process, the researcher told each potential participant he was surveying new patients to the school. For this purpose, a one-page, ten-item survey was developed on possible experiences of new patients to the school (Appendix B). The survey was reviewed for face validity and edited by the school’s associate dean for clinical services. Reliability and validity testing of the instrument was unnecessary, because survey responses were not used in the study. The sole purpose of the survey was for participants to identify their race and gender. To mask this intent, participants were also asked their age range (18–28, 29–39, 40–50, 51–60, 61–70, over 70) and their county of residence. All demographic information was solicited in a "Tell Us About You" section at the conclusion of the instrument.

View larger version (52K): [in this window] [in a new window]

The study protocol was implemented by the researcher and did not disrupt established clinic procedures for first-time, nonemergency, comprehensive care patients. The protocol was approved by the chair of the Department of Diagnostic and Emergency Services (the department responsible for new patient intake) and the associate dean for clinical services. Total time required of each participant was five to ten minutes. The study procedures and instruments were evaluated in a two-day pilot test conducted at the study site. One minor adjustment was made to the protocol as a result of the pilot test.

All new clinic patients complete intake forms and are entered by the intake receptionist into the clinic database. After the patient returned the paperwork to the receptionist, the researcher approached the patient, introduced himself, and explained that he was conducting a survey of new patients. The researcher was dressed professionally with a school identification badge clearly visible. All potential subjects were told they were under no obligation to fill out the survey and that their responses were anonymous, would not be stored in their patient chart, and did not affect their ability to receive quality treatment at the school. The impact of the race and gender of the investigator, a Caucasian male, on participant responses is unknown, but Lopez et al.⁴⁶ have reported that researcher race is unrelated to subject response.

Willing patients were handed the ten-item survey. They were asked to fill it out in their seat and to turn in the completed survey to the researcher. Upon receipt of the completed survey, the researcher thanked participants and directed them to return to their seat. The researcher then recorded the participant’s gender and race as reported in the "Tell Us About You" portion of the survey.

Participants were then randomly assigned to a treatment condition. The three treatment conditions were a) same gender/same race student dentist and same gender/different race student dentist; b) same gender/same race student dentist and different gender/same race student dentist; or c) same gender/same race student dentist and different gender/different race student dentist. Out of view of the participant, the researcher pulled a small slip of paper from a container with an equal number of identical slips individually labeled "1," "2," or "3." The number on the slip of paper corresponded to the treatment condition to which the participant was assigned. Random assignment continued until there was only one treatment condition to which participants of a given racial and gender background could be assigned. For participants assigned to one of the two different race treatment conditions, the different race component was determined by following a sequential scheme, skipping the race of the participant: a) African American, b) Hispanic, c) Caucasian, and (d) Asian/Pacific Islander. For example, an African American subject assigned to a different race condition received a profile of a fictitious Hispanic student. The next participant assigned to a different race condition received a profile of a fictitious Caucasian student (provided the subject was not Caucasian, in which case the subject received an Asian/Pacific Islander student profile).

The researcher collated the two profiles corresponding to the participant’s assigned treatment condition and approached participants to tell them they had a choice between two student dentists. The researcher handed participants the two fictitious student provider profiles and asked if they had a preference for one of the students. New patients may have been unaware this was not standard protocol, thereby diminishing the transparency of the study. Participants could indicate a) preference for the racially and gender-concordant student dentist, b) preference for the student dentist of the alternative race/gender combination, or c) no preference. At this stage participants were unaware that the two students profiled were fictitious. After the participant responded, the researcher collected the profiles and reminded the participant that due to fluctuations in patient case load no guarantees could be made that he or she would be seen by the student chosen. The researcher returned to his desk and recorded the subject’s preference.

Each participant was debriefed within five minutes of responding to the preference question. The researcher approached participants individually and spoke to them privately. The true purpose of the study was disclosed at that time. Participants were told that the study’s intent was not revealed earlier to preserve the authenticity of their choice, a critical element of the research. Three subjects (3 percent) acknowledged they were surprised to be given a choice of student providers, but reported this did not impact their selection. The investigator answered any questions the participant had and solicited the participant’s consent. If a participant denied consent, the researcher offered to return the survey responses and record of provider choice to the participant. Those who granted consent and who indicated a preference were asked why they chose a particular student dentist.

Two potential confounding factors, researcher and response bias, were identified and addressed. To control for researcher bias, scripts were used to standardize presentations to participants. The scripts were reviewed and approved by the chair of the school’s Department of Diagnostic and Emergency Services, the intake manager, and the associate dean for clinical services. Minor changes were made to the scripts to improve clarity as a result of the pilot test. Response bias was investigated by computing separate chi-squares on the race and gender of non-consenting participants to determine if participation was equal across groups. Seven African Americans (three males, four females), five Caucasians (two males, three females), six Asian/Pacific Islanders (two males, four females), and two Hispanic females declined to participate in this study. Results suggested equal participation of all racial groups and both genders in the study, race ²=2.80 (3, N=20), p=.42; gender ²=1.80 (1, N=20), p=.18.

Analysis
The study examined relationships among two categorical variables (race, four levels, and gender) and three dichotomous outcome variables: overall preference (yes, no), racial concordance (same, different), and gender concordance (same, different). Chi-square was used to assess relationships among variables in 2x2 and larger contingency tables and to analyze variables with more than two categories. Two-tailed binomial tests were used to analyze variables with two categories. Binomial tests were chosen for two reasons. First, in samples larger than twenty-six, the statistical software used to analyze the data applies a contingency correction to binomial tests but not to chi-square, and second, in samples less than twenty-five, the p value of a binomial test is more accurate because it is based on a binomial distribution.⁵⁴

Logistic regression was used to investigate whether race and gender predicted the selection of a racially or gender concordant student dentist. Three dichotomous dummy variables were created to represent race in the regression analysis: BLACK (0=black, 1=non-black), HISPANIC (0=Hispanic, 1=non-Hispanic), and ASIAN (0=Asian, 1=non-Asian).

All analyses were done with SPSS Version 11.5. Alpha was set at .05. Whenever multiple chi-squares were computed, the Bonferroni technique was used to control for Type I error.

	Results

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 
The basis for all analyses was frequency of participant response on the three main outcome variables: overall preference (yes, no), racial concordance (same, different), and gender concordance (same, different). Frequency of participant responses is shown in Table 1.

Although the majority of participants in this study had no overall preference for a provider, 43 percent of participants indicated they preferred a particular dental student. To investigate the preferences of this subgroup, chi-square was used to analyze the frequency of participant response among the four choice options (same gender/same race, same gender/different race, different gender/same race, different gender/different race).

Table 1 shows that, of the fifty-one participants with a preference, twenty-nine chose a student dentist of the same gender and same race as themselves (57 percent), nine chose a student dentist of the same gender but a different race (18 percent), ten chose a student dentist of the opposite gender but same race (20 percent), and three chose a student dentist of a different gender and race (6 percent). A chi-square test indicated that subjects with a preference were significantly more likely to chose the racially and gender-concordant student dentist when given a choice: ²=29.86 (3, n=51), p<.001. Among racial groups, only Hispanics were significantly more likely to choose the race- and gender-concordant student dentist: ²=14.80 (2, n=15), p=.001. Hispanic females were significantly more likely than Hispanic males to choose a same gender/same race student dentist (p=.002).

The analysis of the individual race and gender preference variables of participants with a preference confirms these findings. Thirty-eight participants from this subgroup chose a gender-concordant student dentist (75 percent), and thirty-nine chose a racially concordant student dentist (76 percent). Results of two-tailed binomial tests showed that the observed proportions of .75 for gender and .76 for race were significantly different from .50, a value indicating the absence of a preference (p=.001 and p<.001, respectively), suggesting that subjects with a preference were significantly more likely to choose a gender-concordant or a racially concordant student dentist. Analysis by participant gender found that, among those with a preference, females were significantly more likely than males to choose a gender-concordant and a racially concordant student dentist. Eighty-five percent of females preferred a gender-concordant dentist compared to 67 percent of males (p=.006), and 78 percent of females preferred a racially concordant dentist compared to 71 percent of males (p<.001).

Race and Preference
Data on the overall preference of participants by race are presented in Table 2. A chi-square was computed and showed that participant race was unrelated to overall preference in this sample: ²=2.83 (3, N=120), p=.42. However, binomial tests comparing the proportion of responses within each racial group to .50 supports the earlier finding that African Americans were significantly more likely to indicate no preference when given a choice between two equally qualified candidates (p=.04).

	Discussion

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 
The primary purpose of this study was to test the racial concordance hypothesis that suggests patients—in particular, racial minority patients—prefer to be treated by an oral health care provider of the same race as themselves. The study also investigated whether dental patients had a preference for the gender of their provider. The racial concordance hypothesis has been supported indirectly by research showing that underrepresented minority providers treat larger numbers of underrepresented minority patients than do Caucasian providers,⁵⁵ that under-represented minority patients are generally more satisfied when treated by a URM provider,³⁹ and that URM patients participate more actively in the clinical encounter when paired with a racially concordant doctor.¹⁷ However, the racial concordance hypothesis had never before been empirically tested in a clinical setting on subjects who believed they were choosing a provider. Findings from a controlled study in support of the racial concordance hypothesis could provide more convincing evidence that increasing the number of underrepresented minority students in dental schools is a justifiable way to meet the oral health care needs of minority populations and reduce the health and health care disparities due to race.

The majority of participants in this study expressed no preference for a student dentist, including 50 percent or more of participants in each racial group. This suggests that race was not an important consideration in the selection of a provider for most participants, a finding that contradicts the racial concordance hypothesis. Results from previous research showing African American and Hispanic patients significantly more likely to have a racially concordant physician,²⁸ members of all major racial groups more likely to choose a racially concordant doctor if given the chance,³² and having a choice of providers predicted patient-provider racial concordance⁵⁶ were not corroborated in this study. On the contrary, most participants did not choose a racially concordant student dentist when given the choice, and black or Hispanic race did not individually predict the selection of a racially concordant student dentist. In fact, African American males and females, Hispanic males, and Asian females were significantly more likely to express no preference than to select one of the available options. Having a choice of providers was unrelated to choosing a student dentist of the same racial background for all groups except a minority of Hispanic and African American patients.

However, race preferences were identified in some groups. Participants with a preference were significantly more likely to choose a racially and gender-concordant student dentist. Results also show that Hispanics and African Americans with a preference, in particular Hispanic females, were significantly more likely to choose a racially concordant provider, despite the fact that race was not specifically mentioned by any study participant as a reason for making his or her choice. These findings offer some support for the racial concordance hypothesis and suggest that race may be important to some black and Hispanic patients when choosing a dentist.

Results of the gender analysis revealed that the majority of males and females in this study had no preference for their student dentist. The absence of a gender preference among most participants is consistent with previous gender research in medicine and dentistry,^{38,51,57–59} but contradicts studies that found most subjects preferred a health care provider of the same gender.^37,47,48,60 The increased internal and external validity of the present investigation compared to previous research may support the conclusion that gender is not an important consideration for most dental patients when choosing a provider.

Findings from this study also show that participants with a preference were significantly more likely to choose a gender-concordant student dentist. Females, in particular Hispanic females, had stronger gender preferences than males and were significantly more likely to choose a female student dentist when given the choice. These findings support research in medicine that has repeatedly shown women are more likely than men to have a preference and that they are more likely to prefer a gender-concordant health care provider^{37,47,51,57,58,60} and suggest that gender may be important to some women when selecting a dentist. A finding from dentistry that showed males had stronger gender preferences than females³⁸ was not supported.

Some gender researchers have reported that patients prefer a provider of the opposite gender^38,45,49,57 and that gender preference may be explained by the type of problem for which treatment is sought.^51,57 The results of my investigation do not indicate an opposite gender preference. On the contrary, the majority of patients had no gender preference, and most subjects with a preference selected a gender-concordant student dentist. Type of problem was not considered in this study.

Noteworthy among the findings are the preferences of Hispanic females. Results clearly show that Hispanic women were more likely to choose a gender- and racially concordant student dentist when given a choice and that doctor-patient language concordance was important to some members of this group. This pattern is generally consistent with previous research. Garcia et al.³⁷ reported that language concordance was more important to non-English-speaking Hispanic females than gender and racial concordance, but Lopez et al.⁴⁶ showed that female Mexican American college students consistently preferred a racially concordant mental health counselor. The results of this study appear to support the conclusion that, in dentistry as in other health care fields, Hispanic females have a preference for the race, gender, and language ability of the provider. This may suggest the importance to some groups of cultural similarity between patients and providers and could illustrate how cultural norms and expectations play into the selection of a health care provider. Culture was not considered in the present investigation, but in light of the emphasis on racial and ethnic diversity in dental schools and in the profession, it is a concept worthy of further discussion and research. The findings reported here may have been different if the study had been conducted in a less racially diverse or less culturally tolerant location.

Previous research on patient preference has led some to conclude that personal characteristics of health care providers may be less important to patients than professional qualifications^43,61,62 and that empathy and good communication skills are more important to patients than demographics.^{38,41,42,44,45,51,63} It is unclear to what extent the pattern of preferences uncovered here support these conclusions. Professional qualifications were not considered—the lone qualification indicator on the student profile, class rank, was held constant to avoid a confounding effect of ability—and the study design did not allow patients to assess empathy or other human characteristics of the providers. Still, nine participants with a preference reported selecting a female dental student because they perceived women to be more empathetic, kinder, and to have a greater attention to detail, providing tentative support for the results of previous research in this area.

Two conclusions may be drawn from the results presented here. First, the evidence may suggest that the preference of minority patients for racially concordant oral health care providers is not as widespread as previously reported or that the race and gender preferences observed in medicine and other fields^46,47 do not apply to dentistry or do not apply to the same degree. Patients may perceive dentistry differently than they do medicine or other health professions where distinct patient preferences have been observed. Fennema et al.⁵¹ reported that patients have stronger preferences when the problem for which treatment is sought is of an intimate nature, suggesting that if dental patients perceive oral health care to be less personal than medical or other care, they may demonstrate less pronounced preferences. If true, dental patients may want a different relationship with the dentist than with other providers. Although the results presented here support the continued recruitment of underrepresented minority students and women into dentistry to accommodate the preferences of some black, Hispanic, and female patients, additional research on patient perception of dentistry and its relationship to patient preference may be warranted.

The second conclusion suggested by the results of this study is that the racial concordance hypothesis may not be as strong an argument for increasing the enrollment of historically underrepresented minority students in U.S. dental schools as previously thought. If demographic characteristics of dentists are not important to most patients or are less important than other attributes, even among those patients with pre-existing preferences, increasing the number of URMs in dental schools may not be an effective way to meet the needs of underserved populations or to reduce disparities. Recent studies on doctor-patient racial concordance have suggested racial congruity has limited impact on reducing disparities in medical care for children and adults.^29,42 Future research should examine the preferences of dental patients for provider characteristics beyond the demographic and investigate to what extent dentist-patient racial concordance impacts oral health.

Although the findings presented here suggest that preference for doctor-patient racial concordance may not be as pronounced in dentistry as in other fields, there are several compelling reasons to continue the emphasis on increasing the racial diversity in U.S. dental schools. Some believe that expanding the number of URM students in the health professions is a moral issue—one of fairness and social equity that affirms the commitment of health care providers to the notion of a good society.^21,33,64 Others believe that increasing diversity in health care is necessary to correct past injustices and inequalities and to provide a health care environment where all patients feel supported and welcome, a distinctly different experience for some URM groups.^2,65 Recent court decisions have determined there is a clear and compelling government interest in ensuring racial diversity in higher education. The educational advantages of a diverse student population in U.S. colleges and universities have been documented^66,67 and their relevance to professional health care education advanced.⁶⁸

A stronger argument in support of increasing the underrepresented minority student population in dental schools may be the preferences of URM practitioners. Research has shown that URM physicians and dentists are predisposed to treating members of their own racial group or other underserved populations.^30,35,69 In an early study in dentistry, Montoya et al. showed that most URM dental students returned to serve their communities upon graduation.⁷⁰ The most recent survey of dental student seniors shows this pattern largely persists.⁷¹ Responses revealed that 34 percent of Native American, 38 percent of Asian, 54 percent of Hispanic, and 69 percent of black dental students in the class of 2005 reported one of the main reasons for pursuing a career in dentistry was to serve their own racial group. The comparable statistic for Caucasians was 16 percent. Furthermore, 6 percent of Native American, 20 percent of Hispanic, and 28 percent of African American members of the class of 2005 expected their patient population to consist of at least 50 percent underrepresented minority or other underserved groups, compared to 3 percent of Caucasians and 9 percent of Asians. If dental practitioners from underrepresented minority groups prefer to practice in communities where they can provide care to members of their own race or to other underserved populations, dental school admissions committees are well served to continue to actively and aggressively recruit them into dental education programs.

The finding that doctor-patient racial concordance was not important to most dental patients in this sample seems to caution against the "simplistic assumption" (p. 91) that racial congruity alone is sufficient to bridge the gap in health and health care in the United States.⁶¹ Jackson suspected that some African American patients want African American providers, some want Caucasian providers, but that the majority want competent providers.⁷² The Institute of Medicine has reminded educators and others that not every URM provider works well with every URM patient and has warned against assuming that nonminority providers cannot adequately and competently treat minority patients.⁵ Aruguete and Roberts wrote that the declining enrollments of URM students in some health care professions make the assumption underlying the racial concordance hypothesis implausible.⁴¹ The results of this study appear to support these positions.

The findings may also indicate that the continued emphasis on cultural competency in health care education is necessary. Research has shown that compassion, empathy, and communication skills are more important to some patients than race or gender when choosing a provider.^41,42,45 Kleinman et al.⁷³ were the first to suggest that illness is a cultural construct that varies across groups and impacts the doctor-patient relationship, and subsequent studies have shown that cultural sensitivity plays a role in the selection of a health care professional for some non-white patients.^39,47 This may account for the preferences of Hispanic females observed in this study. The need for culturally sensitive providers has gained acceptance among health care researchers and educators, and many in dental education have drawn attention to the need for culturally competent oral health care providers.^8,24,74,75 Culture was not considered in the present investigation, but it is conceivable that results may have been different if the study had been conducted in a less diverse and multiculturally aware location.

Unfortunately, research in cultural competency remains a relatively young area, and cultural competency curricula in health professions schools need to be carefully designed and evaluated.^20,76 Practitioners, researchers, educators, and public health officials must identify valid outcome measures and rigorously assess the effectiveness of cultural competency instruction on those outcomes.^9,76–78 No single effort is likely to reduce or eliminate disparities in health and health care between whites and non-whites in the United States. At present, the best strategy to address disparities in oral health care may be to educate culturally competent dental students in a racially diverse learning environment, combined with ongoing rigorous research on the provider preferences of underserved minority groups.

This study has several important limitations. Participant preference was limited to the selection of a student dentist being trained in general dentistry. It is unknown whether the results are applicable to the selection of a practicing general dentist or to other oral health care professionals, such as hygienists or specialists. Furthermore, the data were collected in a single dental school clinic in a politically liberal, multiculturally sensitive, and racially diverse city. It is unclear whether the results generalize to patients at other school clinics, to patients outside of a dental school setting, or to other geographical locations. Additionally, because the study focused on race and gender, other participant characteristics that may influence or explain preference, such as insurance status or access to care, were not explored. Also, some participants may have surmised the purpose of the study and not responded truthfully. First-time patients were intentionally selected as study participants to reduce transparency, and less than 3 percent of participants acknowledged being suspicious when given the choice option.

Important population subgroups—in particular, American Indian/Alaska Natives, non- and limited-English speakers, and subjects who reported their race to be bi- or multiracial—were not included in the study. The preferences of these groups, some of which are especially impacted by health care disparities, were not examined. Vietnamese names were used to represent Asian/Pacific Islanders on the fictitious dental student profiles given to each subject because Vietnamese students comprise the largest subgroup of Asians at the school. The use of Vietnamese names may have influenced the choice of some Asian subjects. Lastly, given the small sample size within racial groups and the categorical nature of the variables, non-parametric tests were used in most data analyses, limiting the statistical power of some analyses. However, the study’s experimental design, the clinical setting in which it was conducted, and the authenticity of the subjects’ choice decision strengthened the study’s internal and external validity.

	Conclusion

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 
This study found that race was not a major consideration for most patients when choosing a student dentist, a finding that contradicts the racial concordance hypothesis. Gender was also not an important factor when selecting a student dentist for the majority of participants in this sample. However, some African American and Hispanic patients prefer to be treated by a racially concordant provider, and some female patients prefer a gender-concordant provider. U.S. dental schools are encouraged to continue to recruit URM and female students into dental education programs to meet the preferences of these groups, but the results presented here may suggest the racial concordance hypothesis does not apply to the choice of an oral health care provider and thus is not the strongest argument to justify targeted recruitment efforts.

	APPENDIX A

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 

	APPENDIX B

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 

	Acknowledgments

 
I wish to thank the members of my dissertation committee in the School of Education at the University of San Francisco—Dr. Robert Burns, Dr. Susan Evans, and Sr. Mary Peter OP—under whose supervision and guidance this study was conducted. I also wish to acknowledge Drs. Arthur Dugoni, David Chambers, Alan Budenz, and Richard Fredekind, as well as the students, staff, faculty, and patients at the Arthur A. Dugoni School of Dentistry, who helped me complete this research and who encouraged me throughout.

	Footnotes

 
Dr. Bender is Director of Academic Affairs, University of the Pacific, Arthur A. Dugoni School of Dentistry. Direct correspondence and requests for reprints to him at the Office of Academic Affairs, University of the Pacific, Arthur A. Dugoni School of Dentistry, 2155 Webster Street, San Francisco, CA 94115; 415-929-6420 phone; 415-929-6654 fax; dbender{at}pacific.edu.

	REFERENCES

Top Abstract Method Results Discussion Conclusion Appendix a Appendix b References

 

1. Bollinger LD. The need for diversity in higher education. Acad Med 2003; 78(5):431–6.[Medline]
2. Institute of Medicine. Unequal treatment: confronting racial and ethnic disparities in health care. Smedley BD, Stith AY, Nelson AR, eds. Washington, DC: National Academies Press, 2003.
3. Lillie-Blanton M, Brodie M, Rowland D, Altman D, McIntosh M. Race, ethnicity, and the health care system: public perceptions and experiences. Med Care Res Rev 2000; 57(Suppl 1):218–35.[Abstract/Free Full Text]
4. National Institutes of Health. Strategic research plan and budget to reduce and ultimately eliminate health disparities. Vol. I. Fiscal years 2002–2006. Washington, DC: National Institutes of Health, 2002.
5. Institute of Medicine. The right thing to do, the smart thing to do: enhancing diversity in the health professions. Smedley BD, Stith AY, Colburn L, Evans CH, eds. Washington, DC: National Academies Press, 2001.
6. Institute of Medicine. In the nation’s compelling interest: ensuring diversity in the health care workforce. Smedley BD, Butler AS, Bristow LR, eds. Washington, DC: National Academies Press, 2003.
7. U.S. Department of Health and Human Services. Healthy people 2010: understanding and improving health. 2^nd ed. Washington, DC: U.S. Government Printing Office, 2000.
8. American Dental Education Association 2003–04 annual proceedings. J Dent Educ 2004;68(7).
9. American College of Physicians. Racial and ethnic disparities in health care. Position paper of the American College of Physicians. Ann Intern Med 2004; 141(3):226–32.[Abstract/Free Full Text]
10. The Sullivan Commission on Diversity in the Healthcare Workforce. Missing persons: minorities in the health professions. At: www.amsa.org/advocacy/Sullivan_Commission.pdf. Accessed: March 8, 2007.
11. Monheit AC, Vistnes JP. Race/ethnicity and health insurance status: 1987 and 1996. Med Care Res Rev 2000; 57(Suppl 1):11–35.[Abstract/Free Full Text]
12. Weinick RM, Zuvekas SH, Cohen JW. Racial and ethnic differences in access to and use of health care services, 1977 to 1996. Med Care Res Rev 2000; 57(Suppl 1):36–54.[Abstract/Free Full Text]
13. Cornelius LJ. Ethnic minorities and access to medical care: where do they stand? J Assoc Acad Minor Phys 1993; 4(1):16–25.[Medline]
14. Hall JA, Roter DL, Katz NR. Meta-analysis of correlates of provider behavior in medical encounters. Med Care 1998; 26(7):657–75.
15. Blanchard J, Lurie N. R-E-S-P-E-C-T: patient reports of disrespect in the health care setting and its impact on care. J Fam Med 2004; 53(9):721–30.
16. LaVeist TA, Nickerson JJ, Bowie JV. Attitudes about racism, medical mistrust, and satisfaction with care among African American and white cardiac patients. Med Care Res Rev 2000; 57(Suppl 1):146–61.[Abstract/Free Full Text]
17. Cooper LA, Roter DL, Johnson RL, Ford DE, Steinwachs DM, Powe NR. Patient-centered communications, ratings of care, and concordance of patient and physician race. Ann Intern Med 2003; 139(11):907–16.[Abstract/Free Full Text]
18. Mertz E, O’Neil E. The growing challenge of providing oral health care services to all Americans. Health Aff 2002; 21(5):65–77.[Abstract/Free Full Text]
19. U.S. Department of Health and Human Services. Oral health in America: a report of the surgeon general. Rockville, MD: U.S. Department of Health and Human Services, National Institute of Dental and Craniofacial Research, National Institutes of Health, 2000.
20. Formicola AJ, Stavisky J, Lewy R. Cultural comp