Patient preferences for physician characteristics in university-based primary care clinics

259Ethnicity & Disease, Volume 13, Spring 2003

PATIENT PREFERENCES FOR PHYSICIAN CHARACTERISTICS

IN UNIVERSITY-BASED PRIMARY CARE CLINICS

Objective: To examine patient preferences forage-, gender-, and racial/ethnic-concordantprimary care physicians.

Design: Focus group interviews.

Patients and Setting: Forty-nine adults (Afri-can Americans, Caucasians, and Latinos) re-ceiving their outpatient medical care in uni-versity-based primary care clinics in NorthernCalifornia.

Main Measures and Results: Guiding ques-tions were designed to elicit information aboutwhat patients look for in establishing andmaintaining a therapeutic relationship with aprimary care physician. Patients were prompt-ed to provide examples and to discuss demo-graphic-concordance factors. Many partici-pants felt that their continuity of care was poorand that they could not choose their own pri-mary care physicians in the academic system.Most reported tolerating these inconveniencesfor what they perceived to be a higher qualityof care linked to medical innovations at aca-demic medical centers. Patients’ views regard-ing age concordance were varied and unrelat-ed to gender or racial/ethnic group. Women inall English-proficient groups described genderconcordance as important to their relation-ships with primary care physicians. Spanish-speaking participants uniformly preferredSpanish-speaking providers. African-Americanparticipants and Spanish-speaking Latino menfelt that race/ethnic concordance contributedto a practitioner’s empathy, and some wereconcerned by the lack of race/ethnic- or lan-guage-concordant doctors.

Conclusions: Examining patients’ preferencesfor patient-physician demographic concor-dance provides insight into the patient-physi-cian relationship. Addressing areas where thesepreferences affect communication patternsand perceptions of quality of care may lead tooverall improvements in patient-physician re-lationships and health outcomes. (Ethn Dis.2003;13:259–267)

Key Words: Ethnic Groups/Psychology, Pa-tients/Psychology, Physician-Patient Relations

From the Department of Internal Med-icine/Division of General Medicine and theCenter for Health Services Research in Pri-mary Care, University of California, DavisSchool of Medicine, Sacramento, California.

Jorge A. Garcıa, MD, MS; Debora A. Paterniti, PhD;Patrick S. Romano, MD, MPH; Richard L. Kravitz, MD, MSPH

INTRODUCTION

A number of studies have examinedthe subject of gender or racial/ethnicconcordance between patients and theirphysicians. These studies have linkedpatient-provider demographic concor-dance to communication and decision-making styles, patient satisfaction andquality of care perceptions, and health-care processes. With respect to gender,women express a preference for gender-concordant physicians more often thando men.1–3 In a 1993 survey of Dutchhouseholds, Kerssens et al4 found thatmost women and men expressed nogender preference for their internists orgeneral practitioners; however, whenpreferences were expressed, over twice asmany women and approximately 9times as many men preferred gender-concordant physicians. Most respon-dents with gender preferences indicatedthat it was easier to talk to a gender-concordant provider, and that they feltmore at ease during intimate physicalexaminations by such providers.

The literature also provides evidenceof patient preferences for race/ethnic-concordant providers. In a study byCooper-Patrick et al,5 patients with ra-cial/ethnic-concordant providers ratedtheir physicians’ decision-making stylesas more participatory than did patientswith race-discordant providers. Similar-ly, Saha et al6 demonstrated that Afri-

Address correspondence and reprint re-quests to Jorge Garcıa, MD, MS; Division ofGeneral Medicine; University of CaliforniaDavis Medical Center; Patient Support Ser-vices Building; 4150 V Street, Suite 2400;Sacramento, CA 95817-1498; 916-734-7004; 916-734-2732 (fax); [email protected]

can-American patients with race-con-cordant physicians were more likely torate their physicians as excellent, and toreport receiving preventive and neededmedical care during the previous yearthan those with race-discordant provid-ers. Moreover, Hispanic patients withethnic-concordant providers were morelikely to be very satisfied with their over-all health care than those with ethnic-discordant physicians. In a later study ofdata from the Commonwealth Fund1994 National Comparative Survey ofMinority Health Care,7 Black and His-panic Americans who sought care fromrace-concordant physicians reportedchoosing these providers to satisfy per-sonal and linguistic preferences, ratherthan for their greater geographic acces-sibility.

Among female patients, several stud-ies have linked patient-physician genderconcordance to specific processes ofcare. For example, compared to womenwith male physicians, women with fe-male physicians have been reported tohave higher rates of Pap testing,8,9 cho-lesterol screening,8 and mammography,9

or aggressive breast cancer screening (ie,mammography at ages 35–39 years).8 Inaddition, after adjusting for patients’preferences for using estrogen replace-ment therapy, Seto et al10 found that fe-male physicians were more likely thanmale physicians to prescribe hormonaltherapy for women.

We performed the current study toinvestigate patient preferences for age-,gender-, and race/ethnicity-concordantprimary care providers. Our principalobjective was to examine whether pa-tients expressed preferences for demo-graphically concordant primary careproviders, and to explore whether, andhow, such preferences translated into

260 Ethnicity & Disease, Volume 13, Spring 2003

PATIENTS AND THEIR PHYSICIAN PREFERENCES - Garcıa et al

. . . compared to women with

male physicians, women with

female physicians have been

reported to have higher rates

of Pap testing,8,9 cholesterol

screening,8 and

mammography,9 or aggressive

breast cancer screening (ie,

mammography at ages

35–39 years).8

perceptions about quality of care. Weemployed a focus group methodology tomeet these aims because we anticipatedthat provider preferences would be het-erogeneous within gender and ethnicgroups. We felt that this approachwould allow us to utilize the dynamicsof small group interaction to exploremore thoroughly patients’ thoughts andexperiences, and to gain insights intothe heterogeneity of responses withinthese groups.11

METHODS

We conducted qualitative focusgroup interviews with ethnically diverse(African-American, Caucasian, and La-tino [English- and non-English-profi-cient]) male and female patients from 2ambulatory care clinics at the Universityof California, Davis Medical Center(UCDMC) in Sacramento, California.UCDMC is the largest hospital, and thesole academic center, in the north cen-tral part of the state, and serves a pop-ulation that reflects California’s econom-ic and ethnic diversity. The purpose ofthese interviews was to explore the per-spectives and experiences of diverse pa-tients regarding patient-physician con-cordance, and to understand the con-

texts in which patients define their pref-erences for physician characteristics, andwithin which they might link these pref-erences to the quality of care they re-ceive.

We conducted a total of 8 focusgroups between December 1998 andJuly 1999. Each group included 4 to 8patients of similar ethnicity and gender.Latinos who identified themselves aspredominantly English- or Spanish-speaking were interviewed separately,since we anticipated that acculturationdifferences between these groups wouldlead to significant differences in theirexperiences and preferences.

Study ParticipantsWe selected participants from com-

puterized lists of patients attendingUCDMC’s General Medicine clinic.These lists were generated by the hos-pital’s Information and CommunicationServices and stratified patients by gen-der, ethnicity, and language preference.We conducted a 1-in-k systematic sam-ple of 100 potential participants fromeach list, using random starts and esti-mating k as a number less than or equalto the total size of each list divided by100. All sampled patients were contact-ed and asked to participate in the study.Because the 4 Latino patient lists eachhad fewer than 100 patients, we con-ducted no systematic sampling for La-tino patients, and contacted each indi-vidual on these lists. Due to difficultiesrecruiting Spanish-speaking Latinasfrom the General Medicine clinic, weused special outreach efforts to recruitadditional patients from UCDMC’sFamily Practice clinic to ensure the par-ticipation of Spanish-speaking Latinasin our study.

On recruitment into the study, wecollected information on patients’ age,gender, self-reported ethnicity, primarylanguage (English or Spanish), and self-reported estimates of the frequency ofprimary care visits. We collected addi-tional information from ‘‘Spanish-speaking’’ Latinos about their perceived

need for a medical interpreter. All pro-spective participants were offered a $20incentive for their one-time participa-tion in a 2-hour focus group interview.Eligibility criteria for participation in-cluded: 1) being 18 years of age or old-er; 2) having been followed by a pri-mary care provider (seen on at least 2occasions in the 18 months precedingthe group interview); and 3) being will-ing and able to participate actively in agroup discussion about factors that in-fluence the patient-physician relation-ship. Eligibility was determined solelyby patient self-report (ie, we did not at-tempt to validate a prospective partici-pant’s reports of eligibility).

Data CollectionThis research was approved by the

UC Davis Human Subjects Committee.All participants gave their informedconsent at the beginning of the focusgroup meeting. Focus group interviewslasted between 90 and 120 minutes, andwere led by a facilitator matched togroup members’ gender, ethnicity, andlinguistic ability. Discussions for thepredominantly Spanish-speaking Lati-nos were conducted in Spanish. One in-vestigator (JAG) moderated both Latinomen’s groups (English-proficient andSpanish-speaking), and another (RLK)moderated the Caucasian men’s group.Other moderators included an African-American female physician, an African-American male facilitator, a Caucasian(White, non-Hispanic) female researchpsychologist, and a bilingual Latinapostgraduate researcher. All focus groupinterviews were audio taped and tran-scribed, with participants identified bynumbers to protect their confidentiality.

Gender- and race/ethnic-concordantmoderators used a written guide to aska series of questions that facilitated thegroup discussion. Guiding questionsused to maintain the focus of the groupinterview included: 1) What do youlook for in choosing a primary care phy-sician (PCP)? 2) What factors influenceyour decision to stay with a particular



Table 1. Demographic characteristics of focus group participants

Race/Ethnicity N (%) Gender Mean Age, Years (range)

African-American

Caucasian

11 (22.4)

14 (28.6)

FemaleMale

FemaleMale

56

77

65.5 (30–91)53.1 (22–79)

44.2 (20–68)47.2 (31–82)

Latino (English-proficient)

Latino (non-English-proficient)

12 (24.5)

12 (24.5)

FemaleMale

FemaleMale

57

84

54.3 (37–65)50.6 (45–56)

50.6 (19–72)50.1 (30–67)

Total: 49 (100) Female, N (%): 25 (51.0)Male, N (%): 24 (49.0)

Overall Mean Age: 52.0

Table 2. Summary of major themes by race/ethnic and gender group

Race/Ethnicity Gender

Congruence by

Age Gender Race/Ethnicity Language

African-American

Caucasian

FemaleMale

FemaleMale

u*

u

u

u

Latino (English-proficient)

Latino (non-English-proficient)

FemaleMale

FemaleMale

u

u

u

u

* u denotes factors generally felt to be of importance to a particular race/ethnic and gender group.

provider? 3) To what extent does phy-sician gender influence your decision?To what extent do you believe this fac-tor influences other (wo)men from yourethnic background? 4) To what extentdoes a physician’s ethnic background in-fluence your decision? To what extentdo you believe this factor influencesother (wo)men from your ethnic back-ground? Additional guiding questionsprompted patients for their perceptionsof and specific experiences with theirPCPs. In addition, the responses of par-ticipants in each group providedprompts for further discussion, gener-ating information about the range ofpersonal experience in each focusgroup.12

The primary goal of these questionswas to understand patients’ perspectivesof factors important in establishing ameaningful and effective therapeutic re-lationship with a PCP. Each interviewer

communicated this goal at the start ofeach group interview. All interviewerswere trained to use the set of guidingquestions to assist group participants inmaintaining this ‘‘focus’’ throughout thediscussion. Interviewers were alsotrained to prompt all participants toelaborate on their responses, and to pro-vide specific examples of their experi-ences, where appropriate.

Data AnalysisTwo investigators (JAG, DAP) in-

dependently reviewed focused grouptranscripts for the natural languagestatements (ie, words and phrases in theparticipants’ own language) that de-scribed patient preferences for physiciancharacteristics and aspects of care. Theinvestigators agreed on the natural lan-guage statements to be classified underphysician characteristics (eg, age, gen-der, ethnicity) and aspects of care (eg,

empathy, communication, continuity,accessibility, quality). Through an iter-ative process, the investigators induc-tively grouped similar statements intothe same conceptual category. Whereparticipants expressed differing opin-ions, investigators noted the range ofperspectives, including positive and neg-ative statements expressed about thesame phenomenon. Disagreementamong the investigators regarding thecategorical grouping of natural languagestatements was resolved by splittingstatements into different categories (orby creating sub-categories). In the sub-sequent section, we report the results ofpatients’ discussions about their experi-ences with PCPs, and their preferencesfor demographic concordance.

RESULTS

A total of 49 individuals participatedin the 8 focus groups. Table 1 providesthe demographic characteristics of thissample, and Table 2 summarizes themajor themes discussed in each group.

Choosing a Primary CareProvider (PCP)

When asked what they look for inchoosing a PCP, most focus group par-ticipants believed they ‘‘didn’t have achoice’’ of personal physician under theacademic system of care. A participantin the English-proficient Latino men’sfocus group stated,

‘‘I don’t know about the rest of the fel-lows here, but I don’t really think thatyou have a choice of picking your doc-tor. [Several people comment, ‘Right.’]I’ve been coming [here to the medicalcenter] probably . . . about thirty years. . . and during that time, I’ve had doz-ens of primary care doctors.’’

The inability to choose a PCP didnot always translate into dissatisfactionwith one’s provider, however. An En-glish-proficient Latina commented onher experience with her PCP: ‘‘I did notchoose [my PCP]. I had another doctorand he left, and then I was given her.



But I was glad because I really like her.I really like her.’’

In spite of some satisfaction, Afri-can-American patients not only felt aninability to choose their PCP, but alsoperceived that they had a limited rangeof options. One African-American manremarked, ‘‘. . . it would be nice if I hadthe luxury where I could say, yeah, giveme the brother [the African-Americanphysician], you know, but the brothersain’t showin’ up in the medical schools. . . . I don’t see ’em. So we don’t havethat luxury, you know, of a choice.’’ AnAfrican-American woman expressed thesame sentiment by saying, ‘‘If I had topick, I would prefer to have an African-American [doctor], but . . . you can’t re-ally decide who you can have. You knowthey choose everything for us . . . .’’ An-other woman immediately reacted tothis statement saying, ‘‘I haven’t evenseen any [African-American doctors]over here [at the medical center].’’

Spanish-speaking Latinos uniformlypreferred Spanish-speaking providers,reportedly because of communicationissues or concerns. One man recountedhow he specifically requested a Spanish-speaking doctor at his first clinic visit:‘‘. . . I told them that I spoke and un-derstood English but that it was occa-sionally difficult—particularly in medi-cine and in the courts, English is verydifficult—so I prefer a doctor thatspeaks Spanish, male or female . . .’’ An-other man described a different experi-ence in the clinic: ‘‘. . . when I camehere, no one asked me if I wanted aSpanish-speaking provider. They justtold me, ‘You have an appointment withDr. X.’ I don’t even remember his name. . . . I would have preferred someonewho spoke Spanish.’’ Another man em-phasized his preference for a Spanish-speaking provider by expressing his frus-tration with medical interpreters whohad failed to convey all of his concerns:‘‘. . . I understand many things that Isay to [the interpreter] and he does notsay what I want him to tell [the doctor]. . . the interpreter needs to know every-

thing that the patient says in order totell it to the doctor. . . . in fact, wherewill we find all the necessary Spanish-speaking doctors, right?’’ Spanish-speak-ing women expressed a similar prefer-ence for Spanish-speaking providers. Indescribing her doctor, one Latina com-mented: ‘‘. . . I have a doctor who’s verygood. She’s very concerned about myhealth and is always attentive, and I cancall her for any matter.’’ This womanadded that she could contact her PCP(at her home or in the clinic) whenevershe felt the need to speak to her. Sheexplained, ‘‘I’m very satisfied with mydoctor—primarily because she speaksSpanish.’’ Another woman illustratedher preference for a Spanish-speakingdoctor by describing the awkwardness ofhaving an interpreter present during abreast or vaginal exam, particularlywhen the interpreter was a man. Shereasoned that having a Spanish-speakingprovider would mitigate her discomfortby reducing the number of people pre-sent during these sensitive examinations.

Quality of CareMost focus group participants ex-

perienced high turnover rates with theirPCPs, primarily related to the physi-cians’ completion of residency training.In addition, some patients perceived alack of continuity in the teaching clinic,as noted by the concern of this African-American woman: ‘‘I don’t like to be inthe assembly line, if you know what Imean. I like to have my own doctor. . . . I do not like any doctor coming into see me. I like my own doctor.’’ Otherpatients reported that they didn’t have a‘‘basic doctor’’ to help them, and Afri-can-American women noted that theydidn’t know ‘‘who my primary care doc-tor is’’ more frequently than did indi-viduals in other groups.

Focus group participants reported awillingness to sacrifice continuity of carefor what they perceived as a higher qual-ity of clinical care linked to medical in-novations at academic medical centers.One participant from the Caucasian

men’s focus group stated, ‘‘. . . my in-terest in [this medical center] is throughwatching Pulse [a local television pro-gram about medical center physiciansand their research], which I think is awonderful program and really sells thisplace.’’ An African-American manclaimed,

‘‘. . . what I like the most is that theyounger students are learning right now,so they know more, then you’ve gotthese experienced doctors who’ve beendoin’ it [caring for patients] for years,and if they need new treatments, theylearn it right then and there ‘cause it’s aschool . . .’’

By contrast, medical treatment byphysicians-in-training led some Spanish-speaking Latinas to believe they were re-ceiving substandard, or ‘‘experimental,’’treatment. One of these women stated,

‘‘. . . first the student has to put his handin to check us and depending on whatthe student sees, the doctor arrives andalso puts his hand in, and then they sendyou home in worse shape than before.It’s as if they grab you like a rabbit toexperiment on you. First the studentsand then, if they see that it’s necessary,they call the doctor. And, it should bethat if someone comes in with a severepain, the doctor should be first.’’

However, a Latina in the English-proficient group enthusiastically com-mented about her experiences with phy-sicians at the medical center: ‘‘No, they[patients outside of the medical center]said that the doctors here experiment onyou. You’re like a guinea pig. But I don’tfeel like that. Because if I was a guineapig, at least I like the [way the medicalcenter doctors] reconstructed my faceand my eye and everything else.’’

Age Concordance and PerceivedQuality

While some focus group participantsbelieved that younger physicians atmedical centers learn ‘‘state of the art’’medicine, patients’ general discussionsregarding preference for age concor-dance with their PCPs were varied, anddid not relate to gender or racial/ethnicgroup composition. One African-Amer-



ican man claimed, ‘‘age . . . wouldn’tmatter because no matter what, by thisbeing a school, everybody’s gettin’ all thesame knowledge. Just the older doctors. . . are more experienced . . . .’’ An En-glish-proficient Latino stated, ‘‘. . . theyounger doctor, he’s not gonna make adecision himself. He’s gonna have one[doctor] that’s been there ten, fifteenyears down the road that he’s gonna askhim some questions . . . [get] his ad-vice.’’ Most focus group participantsequated age with experience and/or ed-ucation. Many were certain that youn-ger physicians would enthusiastically ab-sorb any and all new clinical informa-tion and, at the same time, would havea structured system of supervision pro-vided by older, more experienced med-ical center physicians. Nonetheless,most participants did not perceive astrong link between the overall qualityof care that they received and the age oftheir PCPs. As one Caucasian womanremarked, ‘‘. . . age doesn’t matter. It’s[the physicians’] experience, their char-acteristics, and their compassion [thatmatters]. Period.’’

Gender Concordance andPerceived Quality

Women in the African-American,Caucasian, and English-proficient Lati-na groups all described gender concor-dance as important to their relationshipswith PCPs. Describing her method forselecting a PCP, a Caucasian woman ex-plained, ‘‘. . . first of all, I look for awoman doctor, if possible, because I feelthat a woman doctor has more under-standing, and she listens better . . . theydo have understandings of things we aregoing through with our special sets ofproblems, and the way that our bodiesand minds and souls and everything aremade up . . . .’’ Based on her previousexperiences with medical care, an En-glish-proficient Latina said, ‘‘. . . some-times I feel more comfortable to talkabout certain things, like being a bat-tered housewife . . . because she [awoman physician] . . . knew how I re-

ally felt . . . and made me feel a littlemore comfortable.’’ Most women feltthat gender-concordant practitionerscould understand important life experi-ences that they considered ‘‘part of be-ing a woman.’’ Further, many female fo-cus group participants noted that a gen-der-concordant practitioner would ac-commodate the needs and demands ofa patient’s family members and otherdependents. A woman in the African-American focus group stated, ‘‘. . . oneof the things that I would really like tohave in my primary care physician is afemale, and what I’ve discovered recent-ly is if I schedule my appointments withmy mom, you know, it [coordinating ajoint family visit] kind of takes cama-raderie . . . it’s a nice way to spend timetogether.’’ Later in the discussion, thisparticipant proclaimed her unequivocalpreference for female physicians:‘‘Hands down female . . . . I mean a fe-male doctor completely . . . . I mean forexaminations, for everything . . . . . . .there is an understanding you don’t haveto worry about. Theirs is a kind of com-fort, a kind of nurturing that goes be-tween you, you know.’’ In contrast,Spanish-speaking Latinas stated thatthey had no preference regarding theirprovider’s gender. In fact, one womanargued that in her experience with a fe-male provider, the provider was ‘‘roughand discourteous.’’

In general, women’s preferences forfemale providers were linked to greaterexpectations for the kinds of care theyhoped to receive from their PCPs.While men in all groups denied PCPgender preferences, they maintained aninterest in having a doctor who would‘‘listen to me and try to relate to whatI’m explaining,’’ and who would ‘‘showsome interest in me and my health.’’Spanish-speaking Latinos stated thatthey preferred a relationship with theirPCP that was like a friendship. OneSpanish-speaking man summarized thisby stating, ‘‘the key issue is the human-ness of the person [the doctor].’’ In gen-eral, men, unlike women, did not linkthese characteristics to gender.

Ethnic Concordance andPerceived Quality

African-American women and menaffirmed that race/ethnic concordanceenhanced patient-provider communica-tion by contributing to a practitioner’sability to be more empathetic. Echoingthe sentiments of other women in theAfrican-American focus group, a partic-ipant explained, ‘‘. . . an African-Amer-ican [PCP] would kind of be a little bitmore attentive or understanding aboutwhat our needs really are. Not just whatthey think we should have.’’ African-American men agreed. One man de-scribed his previous experiences with Af-rican-American physicians:

‘‘. . . it makes it nicer, ’cause like whenyou go in there . . . you talk to ’em likeon a different level. Just like if I was Jew-ish and I went in and seen a Jewish doc-tor; a Jewish doctor’s gonna talk aboutJewish stuff. So, you feel more comfort-able . . . when you got it the same race.It is a difference. It’s a big difference.’’

Another man stated, ‘‘. . . if we hadmore Black doctors when we go in, wecan sit down and say ‘Hey brother!’ Youknow, ‘What’s happening?’ . . . you canreally be comfortable . . . maybe someof us [African-American patients] wouldbe a little bit more comfortable aboutopening up to a Black doctor than wewould to a White doctor.’’

Because of this, some African-Amer-ican men and women said they wereconcerned, since they saw few African-American doctors at the medical center:‘‘. . . I ain’t gonna see no . . . Black in-terns over at the primary care. . . . Theonly thing I see is White and Chineseor Japanese. I don’t even see one . . .even one [African American] dressedlike a doctor. I wonder where they’reat.’’

Spanish-speaking Latino men per-ceived that a provider who shared theirethnic and cultural background wouldbe better able to care for them, as re-vealed in this exchange between three ofthe group’s participants:

[Participant 1]: ‘‘There is a preference.With the [non-Latino, Spanish-speaking



doctor] there are no similar customs. Iprefer the Mexican [doctor]. There is aperson-to-person union.’’

[Participant 2]: ‘‘Perhaps, as he said, be-cause of the customs. For example, if Icome because something has harmedme, the doctor would already know. Idon’t know, perhaps it’s the same roots.’’

[Participant 3]: ‘‘Yes, one’s roots are mostimportant . . . a Latino [physician] al-ready knows the illnesses that Latinossuffer with . . . he knows what we eat,how we live, et cetera.’’

Spanish-speaking Latinas did notshare this belief; however, in dismissingthe relevance of ethnic concordance,they emphasized their strong preferencefor Spanish-speaking providers. Onewoman summarized the group’s senti-ments: ‘‘I also don’t have a preferencefor [a doctor] of any given race . . . butthe only thing . . . I’d like [my doctor]to speak my language so that I couldunderstand him better.’’ Caucasians andEnglish-proficient Latinos did not be-lieve that race/ethnic concordance wasessential to good care. A Latino fromthe English-proficient focus group said,‘‘I don’t have a problem with any raceor anything. I mean, all I’m lookin’ foris a good doctor . . . period. You know?I mean, he could be purple with orangespots on him . . . as long as he gave megood medical care I could care less.’’ ACaucasian woman remarked, ‘‘The bestdoctor I had here was a gay, Black doc-tor . . . . I almost followed him to hisnew place of practice because, and I toldhim, ‘You’re the best doctor I’ve had.’’

DISCUSSION

This study was conceived and con-ducted to inform our understanding ofpatient preferences for demographicallyconcordant PCPs, including how thesepreferences might translate to patients’quality of care perceptions. Anticipatingthat provider preferences might vary asmuch within as between gender andethnic groups, we utilized a focus groupmethodology. Our results demonstrate

that participants indeed held a widerange of opinions in this regard. Whenasked about the factors that influencetheir decision to establish a relationshipwith, or remain with, a particular PCP,many focus group participants reportedthat they could not choose a personalphysician in our academic system. Mostidentified housestaff graduation as a keyreason for their lack of choice, and somedescribed their continuity of care aspoor. These results are not surprising,given the constant turnover of residentphysicians in the continuity clinic, theconstraints of their limited outpatientschedule, and the absence of explicitmechanisms by which patients can se-lect their own PCPs. The findings mir-ror resident physicians’ perceptions ofthe lack of continuity in ambulatoryteaching clinics,13 as well as the resultsof other studies documenting the lim-ited continuity of care in academic out-patient clinics.14–17 Lichstein18 com-mented on the significance of ‘‘the an-nual problem’’ of resident graduation,suggesting that the separation can evokegrief responses in both patients and res-idents that can lead to maladaptive be-haviors if overlooked.

Participants explained that they tol-erated the drawbacks of the academicpractice setting in order to take advan-tage of the perceived higher quality ofcare at the medical center, which thesepatients linked to the center’s diagnosticand therapeutic innovations. In survey-ing the parents of children whose resi-dent pediatricians were graduating andleaving their hospital-based continuityclinic, Serwint et al19 found that mostparents were willing to make a similartrade-off. In particular, 96% of respon-dents stated that they would continueto have their children receive medicalcare at the current site, because theyknew that the children would receive‘‘good care’’ in this system.

Patients’ discussions regarding ageconcordance were varied, and did notrelate to gender or ethnic group com-position. Patients most frequently asso-

ciated a physician’s age with perceptionsof her/his overall competence, includingthe ability to stay on top of medical in-formation and to ask questions, or toadmit not knowing ‘‘the answer’’ with-out the fear of being perceived nega-tively by patients. Medical sociologistshave described uncertainty in medicineto be based on 3 factors: the limits ofmedical science, the limits of a physi-cian’s knowledge-base, and the ability totell the difference between the limits ofmedical science and personal knowl-edge.20,21 Although most of the patientsin our focus group interviews did notexpress a preference for PCP age-con-cordance, many expressed a desire fortheir PCP to recognize areas of uncer-tainty in medical knowledge, and in thePCP’s own stock of knowledge, and tohave the ability to distinguish the un-certainty of medicine from personal un-certainty. In these instances, patients feltthat the quality of care they received atthe medical center was enhanced bytheir ability to discuss uncertainties withphysicians of varying ages, levels of en-thusiasm, and experience.

The male participants in our studyexpressed no specific preference for gen-der-concordant physicians. Nonetheless,all groups of English-proficient womendescribed patient-PCP gender concor-dance as desirable because they felt thatfemale physicians would have more em-pathy for patients. In general, these par-ticipants perceived that female physi-cians, being women themselves, couldbetter ‘‘relate’’ to female patients’ life-styles and experiences in the femalebody. In addition, these patients feltthat, relative to male physicians, femalephysicians had a broader understandingof relevant social issues, such as childrearing, caregiving, and battering. In ex-pressing these preferences, female pa-tients shared the notion that an en-hanced patient-provider relationshipwould directly contribute to improve-ments in the overall quality of care thatthey received. These results are consis-tent with those in previous studies,



. . . those who did express a

preference for an ethnic-

concordant provider clearly

linked it to the expectation

or perception that they would

receive higher quality medical

care by enhancing their

communication, and possibly

their relationships, with such

providers.

which have shown that female and malePCPs have distinct practice and com-munication styles,9,22–27 leading some toposit that patients who prefer femalephysicians may be seeking a more sen-sitive or empathetic style of medicalcare.27 Surprisingly, Spanish-speakingLatinas did not share these sentiments.To our knowledge, their lack of PCP-gender preference has not been previ-ously described, and may reflect thesewomen’s prior experiences with provid-ers, or their perception that linguisticbarriers are more important.

African-American women and men,and Spanish-speaking Latino men, re-ported a preference for ethnic-concor-dant PCPs, perceiving that such con-cordance would lead to more empathet-ic and/or effective therapeutic relation-ships. Of note, not every participantexpressed such a preference, which wewould expect from previous studies, anda cursory consideration of human diver-sity. Nonetheless, those who did expressa preference for an ethnic-concordantprovider clearly linked it to the expec-tation or perception that they would re-ceive higher quality medical care by en-hancing their communication, and pos-sibly their relationships, with such pro-viders. As a dimension of care,patient-physician communication hasbeen widely recognized and utilized asan important process measure of health-care quality in a number of stud-ies,6,7,23,25,33 including the benchmarkMedical Outcomes Study,35 and theCommonwealth Fund’s study on qualityof care for minority patients.32 Osten-sibly, patients felt that these more robustrelationships would improve the qualityof care they received.

Further, some participants expressedconcern at the lack of race/ethnic- orlanguage-concordant doctors at themedical center, while others resignedlydiscussed this as an unpleasant ‘‘reality’’that they had come to accept. African-American participants more generallyfelt that African-American physicianswould be less prescriptive in their pro-

vision of care, and more able to elicitgenuine communication from their pa-tients.

These findings reflect those of priorstudies from the behavioral sciences andmedical literature. Research in the fieldof counseling psychology has providedsome evidence to support the existenceof client preferences, albeit not univer-sal, for ethnically concordant counsel-ors,28–30 and one study even linked cli-ent-therapist ethnic and linguistic con-cordance to the length or outcome ofmental health treatment for ethnic mi-nority patients.31 Recent survey findingsindicate that, compared to Caucasians,Latino, Asian, and African Americansare more likely to experience problemscommunicating with their physicians.32

Studies of patient-physician communi-cation demonstrate that physicians havepoorer interpersonal skills33 and are lesslikely to use a participatory decision-making style with non-Caucasian pa-tients, compared to Caucasians.34–36

However, all of these studies involvedpredominantly White physicians. Onlyone previous study compared physician-patient communication in racially/eth-nically concordant and discordant dy-ads5; concordance was associated with aclear perception that physicians em-

ployed a more participatory style of de-cision-making.

Spanish-speaking participants uni-formly preferred Spanish-speaking pro-viders, consistent with their perceptionof the central role of communication inthe patient-physician relationship. Span-ish-speaking Latinas reported that hav-ing a physician who spoke their lan-guage meant better care, irrespective ofphysician race/ethnicity. Spanish-speak-ing Latinos, however, believed that bothethnic and linguistic concordance con-tributed to good quality, empatheticcare. Perceptions of illness and responsesto it vary by race/ethnicity37–39 and cul-ture,40 and possibly ethnically concor-dant physicians seem more empatheticdue to their greater sensitivity to thesevariations. Unfortunately, we were un-able to convene additional focus groupsto confirm or refute this unexpected dif-ference between Spanish-speaking menand women. Our findings suggest thatacculturation has a significant effect onthe views of Latino patients, as English-proficient Latinos were more likely thanAfrican Americans or Spanish-speakingpatients to express the same sentimentsas Caucasians on the relevance of ethnicconcordance and the risks of experimen-tation at the medical center. This find-ing is consistent with previous surveyresearch demonstrating significant dif-ferences between the healthcare experi-ences of English-speaking and Spanish-speaking Latinos, as well as between En-glish-speaking and non-English-speak-ing Asian Americans.41

Our qualitative study has 2 impor-tant limitations. First, the data representa survey of attitudes and beliefs thatmay or may not correlate with the focusgroup participants’ actual actions andbehaviors. Therefore, in establishing ormaintaining patient-PCP relationships,participants may actually behave in waysthat would not be predicted by theirstatements regarding certain physiciancharacteristics. Second, the participantsclearly are not representative of all pa-tients, nor of patients seen at the aca-



demic medical center where this studywas conducted. As such, our results maynot generalize to other groups of pa-tients or to other clinical settings. Theselimitations highlight the need for addi-tional studies, using both qualitativeand quantitative methods, to validateand extend our findings.

Nonetheless, this examination of pa-tients’ preferences for demographicallyconcordant PCPs provides new insightsinto the complexity of the patient-PCPdynamic by demonstrating how thesepreferences can influence patients’ qual-ity of care perceptions. Attention to pa-tient preferences for demographic con-cordance with their PCPs may improvepatient-physician communication, andhave the potential to strengthen thetherapeutic relationship. This, in turn,may lead to improvements in healthoutcomes, particularly for women, Af-rican Americans, and Spanish-speakingLatino patients.

Although our study does not directlylink patient preferences for demograph-ically concordant PCPs to outcomes orprocesses of care, our findings do bringus a step closer to understanding howdiverse communities of patients perceivetheir relationships with their PCPs andthe impact these relationships have onthe quality of care these patients receive.In particular, we found that some pa-tients expressed clear preferences forgender- and/or ethnically concordantPCPs, whom they felt could offer betterhealthcare quality, relative to demo-graphically discordant providers. If con-firmed, these findings have potentialpolicy implications. For example,healthcare systems may respond to con-sumer preferences by hiring providerswho match the demographic composi-tion of the communities they serve. Inturn, medical school and residencytraining programs might face increasingpressures to train a physician workforcethat more closely matches the genderand ethnic distribution of the US pop-ulation.

ACKNOWLEDGMENTSWe are indebted to the patients who gener-ously shared their time and personal expe-riences with us. We also acknowledge thetremendous contributions of Carol E. Franz,PhD, Steve Harrell, Jann L. Murray-Garcıa,MD, MPH, and Marisa I. Ramos, whoserved as focus group moderators. This workwas supported by a grant from the HibbardE. Williams, MD, Research Fund Program,University of California, Davis School ofMedicine. Presented in part at the 17th an-nual meeting, Association for Health Servic-es Research, Los Angeles, California, June25–27, 2000.

REFERENCES1. McPherson A, Small J. Women general prac-

titioners in Oxfordshire. J R Coll Gen Pract.1980;30:108–111.

2. Weyrauch KF, Boiko PE, Alvin B. Patient sexrole and preference for a male or female phy-sician. J Fam Pract. 1990;30:559–562.

3. Fennema K, Meyer DL, Owen N. Sex of phy-sician: patients’ preferences and stereotypes. JFam Pract. 1990;30:441–446.

4. Kerssens JJ, Bensing JM, Andela MG. Patientpreference for genders of health professionals.Soc Sci Med. 1997;44:1531–1540.

5. Cooper-Patrick L, Gallo JJ, Gonzales JJ, et al.Race, gender, and partnership in the patient-physician relationship. JAMA. 1999;282:583–589.

6. Saha S, Komaromy M, Koepsell TD, Bind-man AB. Patient-physician racial concordanceand the perceived quality and use of healthcare. Arch Intern Med. 1999;159:997–1004.

7. Saha S, Taggart SH, Komaromy M, BindmanAB. Do patients choose physicians of theirown race? Health Aff (Millwood). 2000;19:76–83.

8. Kreuter MW, Strecher VJ, Harris R, KobrinSC, Skinner CS. Are patients of women phy-sicians screened more aggressively? A prospec-tive study of physician gender and screening.J Gen Intern Med. 1995;10:119–125.

9. Lurie N, Margolis KL, McGovern PG, MinkPJ, Slater JS. Why do patients of female phy-sicians have higher rates of breast and cervicalcancer screening? J Gen Intern Med. 1997;12:34–43.

10. Seto TB, Taira DA, Davis RB, Safran C, Phil-lips RS. Effect of physician gender on the pre-scription of estrogen replacement therapy. JGen Intern Med. 1996;11:197–203.

11. Kitzinger J. Qualitative research. Introducingfocus groups. BMJ. 1995;311:299–302.

12. Lofland JF, Lofland LH. Analyzing Social Set-tings: A Guide To Qualitative Observation andAnalysis. 3rd ed. Belmont, Calif: Wadsworth;1995.

13. Steiner E, Stoken JM. Overcoming barriers togeneralism in medicine: the residents’ per-spective. Acad Med. 1995;70:S89–S94.

14. Rulin MC, Hayashi TT, Badway DM. Con-tinuity of ambulatory care in an obstetricsand gynecology residency program. ObstetGynecol. 1988;71:787–790.

15. Irby DM. Teaching and learning in ambula-tory care settings: a thematic review of theliterature. Acad Med. 1995;70:898–931.

16. Dumont-Driscoll MC, Barbian LT, PollockBH. Pediatric residents’ continuity clinics:how are we really doing? Pediatrics. 1995;96:616–621.

17. Garfunkel LC, Byrd RS, McConnochie KM,Auinger P. Resident and family continuity inpediatric continuity clinic: nine years of ob-servation. Pediatrics. 1998;101:37–42.

18. Lichstein PR. The resident leaves the patient:another look at the doctor-patient relation-ship. Ann Intern Med. 1982;96:762–765.

19. Serwint JR, Wasserman BJ, Chernoff RG.The impact on families of pediatric residentdeparture from a continuity clinic practice.Arch Pediatr Adolesc Med. 1997;151:679–683.

20. Atkinson P. Training for certainty. Soc SciMed. 1984;19:949–956.

21. Fox RC. Essays In Medical Sociology: JourneysInto The Field. 2nd ed. New Brunswick, NJ:Transaction Books; 1988.

22. Franks P, Clancy CM. Physician gender biasin clinical decisionmaking: screening for can-cer in primary care. Med Care. 1993;31:213–218.

23. Roter D, Lipkin M Jr, Korsgaard A. Sex dif-ferences in patients’ and physicians’ commu-nication during primary care medical visits.Med Care. 1991;29:1083–1093.

24. Bertakis KD, Helms LJ, Callahan EJ, AzariR, Robbins JA. The influence of gender onphysician practice style. Med Care. 1995;33:407–416.

25. Hall JA, Roter DL. Medical communicationand gender: a summary of research. J GenderSpecif Med. 1998;1:39–42.

26. Lurie N, Slater J, McGovern P, Ekstrum J,Quam L, Margolis K. Preventive care forwomen. Does the sex of the physician matter?N Engl J Med. 1993;329:478–482.

27. Roter DL, Hall JA. Why physician gendermatters in shaping the physician-patient re-lationship. J Womens Health. 1998;7:1093–1097.

28. Atkinson DR. Ethnic similarity in counselingpsychology: a review of research. CounsPsychol. 1983;11:79–92.

29. Ponce FQ, Atkinson DR. Mexican-Americanacculturation, counselor ethnicity, counselingstyle, and perceived counselor credibility. JCouns Psychol. 1989;36:203–208.

30. Lopez SR, Lopez AA, Fong KT. MexicanAmericans’ initial preferences for counselors:the role of ethnic factors. J Couns Psychol.1991;38:487–496.

31. Sue S, Fujino DC, Hu L, Takeuchi DT.Community mental health services for ethnicminority groups: a test of the cultural respon-



siveness hypothesis. J Consult Clin Psychol.1991;59:533–540.

32. Collins K, Hughes D, Doty M, Ives B, Ed-wards J, Tenney K. Diverse Communities,Common Concerns: Assessing Healthcare Qual-ity for Minority Americans. Findings from theCommonwealth Fund 2001 Healthcare QualitySurvey. New York, NY: The CommonwealthFund; 2002. Available at: http://www.cmwf.org. Accessed March 10, 2002.

33. Bartlett EE, Grayson M, Barker R, LevineDM, Golden A, Libber S. The effects of phy-sician communications skills on patient sat-isfaction, recall, and adherence. J Chronic Dis.1984;37:755–764.

34. Blendon RJ, Aiken LH, Freeman HE, CoreyCR. Access to medical care for Black andWhite Americans. A matter of continuingconcern. JAMA. 1989;261:278–281.

35. Kaplan SH, Gandek B, Greenfield S, RogersW, Ware JE. Patient and visit characteristicsrelated to physicians’ participatory decision-making style. Results from the Medical Out-comes Study. Med Care. 1995;33:1176–1187.

36. Kaplan SH, Greenfield S, Gandek B, RogersWH, Ware JE Jr. Characteristics of physicianswith participatory decision-making styles.Ann Intern Med. 1996;124:497–504.

37. Oddone EZ, Horner RD, Diers T, et al. Un-derstanding racial variation in the use of ca-rotid endarterectomy: the role of aversion tosurgery. J Natl Med Assoc. 1998;90:25–33.

38. Shepardson LB, Gordon HS, Ibrahim SA,Harper DL, Rosenthal GE. Racial variationin the use of do-not-resuscitate orders. J GenIntern Med. 1999;14:15–20.

39. Diala CC, Muntaner C, Walrath C, Nicker-son K, LaVeist T, Leaf P. Racial/ethnic differ-ences in attitudes toward seeking professionalmental health services. Am J Public Health.2001;91:805–807.

40. Harwood A. Ethnicity and Medical Care.Cambridge, Mass: Harvard University Press;1981.

41. Weech-Maldonado R, Morales LS, Spritzer K,Elliott M, Hays RD. Racial and ethnic dif-ferences in parents’ assessments of pediatriccare in Medicaid managed care. Health ServRes. 2001;36:575–594.

AUTHOR CONTRIBUTIONSDesign and concept of study: Garcıa, Paterniti,

RomanoAcquisition of data: Garcıa, KravitzData analysis and interpretation: Garcıa, Pa-

terniti, KravitzManuscript draft: Garcıa, Paterniti, RomanoAcquisition of funding: Garcıa, Romano,

KravitzAdministrative, technical, or material assis-

tance: Garcıa, Paterniti, KravitzSupervision: Garcıa, Paterniti

Patient preferences for physician characteristics in university-based primary care clinics

Documents

Transcript of Patient preferences for physician characteristics in university-based primary care clinics