Objectives. A growing body of evidence suggests that provider decisionmaking contributes to racial/ethnic disparities in care. We examined the factors mediating the relationship between patient race/ethnicity and provider recommendations for coronary artery bypass graft surgery.

Methods. Analyses were conducted with a data set that included medical record, angiogram, and provider survey data on postangiogram encounters with patients who were categorized as appropriate candidates for coronary artery bypass graft surgery.

Results. Race significantly influenced physician recommendations among male, but not female, patients. Physicians’ perceptions of patients’ education and physical activity preferences were significant predictors of their recommendations, independent of clinical factors, appropriateness, payer, and physician characteristics. Furthermore, these variables mediated the effects of patient race on provider recommendations.

Conclusions. Our findings point to the importance of research and intervention strategies addressing the ways in which providers’ beliefs about patients mediate disparities in treatment. In addition, they highlight the need for discourse and consensus development on the role of social factors in clinical decisionmaking.

Dozens of empirical studies have documented significant and widespread racial/ethnic disparities in health care,13 and the most heavily documented disparities involve treatment for coronary artery disease.424 While insurance status, site of care, and patient preferences account for a portion of overall racial/ethnic variations in care, significant and substantial effects persist independent of these factors.12,2528

The lack of evidence supporting other causal pathways has led to an increased focus on the role of provider behavior. As a result, there is a growing body of evidence suggesting that provider behavior and decisionmaking contribute significantly to racial/ethnic disparities in care9,25,2937 (see van Ryn38 for a review). However, little is known about why patient race or ethnicity influences the clinical decisionmaking process, and thus, there is an inadequate evidence base for guiding intervention priorities and strategies.

One approach to understanding the processes by which patient race/ethnicity influences provider decisionmaking can be guided by the extensive amount of social psychological research demonstrating how unconscious stereotypes may lead to bias even among well-intentioned, egalitarian people.2,3840 This evidence suggests that the effects of race/ethnicity on provider behavior may be mediated by unconscious stereotypes or beliefs. In other words, providers’ beliefs about patients may vary according to patients’ race/ethnicity, and these beliefs, in turn, may influence their behavior and decisionmaking.

We tested this hypothesis regarding predictors of physician recommendations using a data set that included medical record and provider survey data collected from coronary artery disease patients who had an angiogram performed at one of 8 New York hospitals. Understanding provider recommendations is essential for understanding disparities in care, given that previous published results from the present data set indicate that such recommendations are the proximal cause of almost all observed disparities in receipt of coronary artery bypass graft surgery (CABG).30

We examined the factors associated with provider recommendations for CABG among the subset of patients who were appropriate candidates for this procedure but not for other aggressive treatments. The data used in this study had unique advantages in that they included provider survey information on the postangiogram encounter at which a treatment determination was made. In their survey responses, providers reported on their recommendations and rated patients on a number of social and behavior characteristics. Previous analyses of the dataset used in the present study revealed that physicians’ perceptions of patients on these measures varied according to patient race, independent of patient age, gender, race, frailty/sickness, mental health status, mastery, and social assertiveness and physician characteristics. Specifically, physicians rated non-White patients as less likely than their White counterparts to participate in cardiac rehabilitation, to desire a physically active lifestyle, and to have a sufficient amount of social support and more likely to be non-adherent and to abuse drugs.39

In another report involving the present data, we found that patients who had already been rated either as appropriate candidates for CABG or as patients for whom CABG was necessary were more likely to undergo CABG if they had specific clinical characteristics (left main coronary artery disease, 3-vessel disease, and older age).30 This set of findings suggests the possibility that providers overapply certain clinical characteristics when making treatment recommendations. If these characteristics are also associated with patient race/ethnicity, they may mediate a portion of the observed effect of patient race/ethnicity on provider recommendations. The present data provided an opportunity for a meaningful examination, after control for physician characteristics, patient clinical characteristics, and insurance status, of the degree to which perceptions of patients mediate the observed relationship between patient race/ethnicity and provider recommendations for CABG among patients who are appropriate candidates for the procedure.

Data Collection

The methods used to collect the data for this study have been described in detail elsewhere.30,39 Briefly, a stratified random sample of 4907 coronary artery disease patients undergoing an angiogram at 1 of 8 New York State hospitals between 1996 and 1998 was selected for the medical record abstraction portion of the study. Criteria developed by the RAND Corporation4144 were applied to medical record and angiogram data to classify each patient as an appropriate, inappropriate, or uncertain candidate for CABG and, separately, percutaneous transluminal coronary angioplasty (PTCA).

Of the 1261 patients who were classified as appropriate candidates for CABG but not for PTCA, an alternate revascularization procedure, a random sample of 614 were recruited for the survey portion of the study. Those patients who also were appropriate candidates for PTCA were eliminated from the analyses to provide a reasonable and conservative test of physician recommendations for CABG. In the case of each of the 614 patients who were appropriate for CABG, the physician (or physicians) identified as involved in the treatment decision by the hospital, patient, or another physician was sent a self-administered survey focusing on the postangiogram encounter in which the treatment determination was discussed. Of the 792 encounters identified as relevant to the 614 patients, physicians returned surveys with data on 570 (72%) postangiogram encounters among 462 (75%) patients.

Physicians’ likelihood of response was unrelated to the race/ethnicity of the patient. Of the 570 encounters with patients who were appropriate candidates for aggressive treatment, 38 involved patients who were found to have had previous CABG surgery and thus were eliminated from the analyses, resulting in 532 physician reports on encounters. As a result, 532 encounters were included in the present analyses. Of these encounters, 305 involved male patients (57%) and 227 involved female patients (43%). Only encounters with White (n = 182; 34%), Black (n = 161; 30%), and Hispanic (n = 189; 36%) patients were sampled for this study.

Dependent variable.

The dependent variable was whether or not the physician reported recommending CABG. Physicians were asked “How strongly did you recommend CABG/PTCA for this patient?” Response options were exclusively as only appropriate treatment (1), best of the options (2), neutrally as one possible treatment (3), and recommended against CABG (4). A dichotomous dependent variable was created in which responses 1 and 2 were combined as “physician recommended CABG” (coded 1) and responses 3 and 4 were combined as “physician did not recommend CABG” (coded 0).

Independent variables.

Data on patient race/ethnicity and gender were obtained from medical records. We chose to use medical record data rather than patient self-reports because our hypotheses focused on the effect of providers’ perceptions regarding patients’ race/ethnicity.

We used 4 categories of potential mediators of the effects of patient race/ethnicity on physicians’ recommendations for CABG. Each category, or block of variables, is described subsequently.

One hypothesis regarding the reasons for racial/ethnic disparities in physician recommendations is that non-White patients see different physicians than White patients, and these physicians are less likely to recommend CABG. The physician characteristics we examined included age, race/ethnicity, specialty (cardiologist vs other), gender, and whether the physician was an attending physician or a trainee (resident or fellow). All physician characteristics were assessed through self-reports provided on the self-administered physician survey.

Information on clinical factors shown to influence receipt of CABG independent of appropriateness for CABG was obtained through abstraction of medical record data. These factors included patient age as well as the presence of disease of the left main coronary artery (left main disease) and of disease in 3 coronary arteries (3-vessel disease).

Data on patients’ health insurance coverage were abstracted from medical records. Insurance status was transformed into a dichotomous variable in which patients who had Medicare or private health insurance coverage were compared with those who did not (i.e., were solely insured by Medicaid or had no insurance).

In terms of physicians’ perceptions of patients’ social and behavioral characteristics, the following explanation was used to ask physicians to rate patients on a set of 24 characteristics:

Many studies have found certain patient characteristics to be associated with compliance with treatment regimens and following medical advice. As part of this study, we will be attempting to predict compliance and treatment outcomes on the basis of patient characteristics. We would like to ask you a few questions about your impressions of this patient. Although it is sometimes hard to get to know patients well in a short time, we have found that even the first or general impressions that physicians provide us with can be very helpful in predicting compliance. Please consider how this patient behaved in your interaction(s) with him/her.

Physicians’ perceptions of patients’ personality characteristics were assessed through ratings they made on semantic differential measures including (1) intelligent–unintelligent, (2) self-controlled–lacking self-control, (3) pleasant–unpleasant, (4) educated–uneducated, (5) rational–irrational, (6) independent–dependent, and (7) responsible–irresponsible. For each item, ratings ranged from 1 (e.g., intelligent) to 7 (e.g., unintelligent).

Physicians’ perceptions of patients’ probable behavior and social role were assessed through asking physicians their opinion on how likely the patient was to lack social support; overreport (exaggerate) discomfort; fail to comply with medical advice; abuse drugs, including alcohol; strongly desire a very physically active lifestyle; participate in cardiac rehabilitation (if prescribed); attempt to manipulate physicians; initiate a malpractice suit; have major responsibility for the care of a family member (or family members); and have significant career demands/responsibilities. Response options ranged from not at all likely to extremely likely on a 5-point scale. Because most of these items yielded heavily skewed response distributions, we transformed each into a dichotomous variable using the median as the cut point.

Analysis Plan

The goal of the analyses described here was to examine the role of 4 classes of potential mediators of the observed relationship between patient race/ethnicity and physician recommendations for CABG among patients who were appropriate candidates for the procedure: (1) physician characteristics, (2) clinical characteristics shown to influence physician decisionmaking even in the case of patients classified as appropriate candidates for CABG, (3) patients’ insurance status, and (4) physicians’ perceptions of patients’ social and behavioral characteristics. We followed the steps described by Baron and Kenny45 and Judd and Kenny46 as necessary for establishing mediation. First, the overall effect of the independent variable on the dependent variable must be statistically significant. Second, the independent variable must have a significant effect on the mediator (or mediators). Third, the mediator (or mediators) must have a significant association with the dependent variable. Finally, to establish complete (vs partial) mediation, the entry of the mediator (or mediators) into the model must eliminate the impact of the independent variable on the dependent variable.

Our first step in assessing mediation was to examine the association between patient race/ethnicity and physician recommendations among patients who were classified as appropriate candidates for CABG according to the RAND criteria and to test for interaction effects with patient gender and non-White socially constructed race/ethnicity categories. Our second step was to examine each potential mediator for a significant association with patient race/ethnicity. In this series of bivariate analyses, we used the χ2 test of association in cases in which the dependent variable was nominal or ordinal, and we used comparison of mean scores and F tests for differences in means in the case of interval-level variables.

In our third step, we examined the relationship between the potential mediators and physician recommendations for CABG. Next, we explored the possibility that the observed effects of patient race/ethnicity on physician recommendations were due to the hypothesized mediator by examining the impact of the mediator on the association between patient race/ethnicity and physician recommendations.

Initially, we tested the mediating role of provider characteristics. We then examined the possibility that clinical characteristics were distributed differently according to patient race/ethnicity, along with the possibility that these clinical characteristics mediated the effects of race/ethnicity on provider recommendations. The importance of these analyses is highlighted by an earlier report showing that the presence of left main disease, 3-vessel disease, or both, increased patients’ likelihood of undergoing CABG, whether or not they were rated as either appropriate candidates for CABG or in need of the procedure.30 In other words, patients who were appropriate candidates for CABG but did not have left main artery disease or 3-vessel disease were less likely to undergo the appropriate treatment.

Next, we examined the effects of insurance coverage on the relationship between patient race/ethnicity and physician recommendation. Finally, we tested the degree to which physicians’ perceptions of patients mediated the effects of patient race/ethnicity on physician recommendations, independent of physician characteristics, clinical characteristics, and insurance status.

Each of the potential mediating variables that exhibited a significant bivariate association with both the independent variable (patient race/ethnicity) and the dependent variable (physician recommendations) was entered into a multivariate model in blocks corresponding to the 4 categories listed earlier: physician characteristics, clinical characteristics, insurance status, and physicians’ perceptions of patients. Initially, patient race/ethnicity was entered into a logistic regression equation, followed by physician characteristics, clinical variables, insurance status, and, finally, the social and behavioral factors. This strategy allowed us to examine whether a given block of variables attenuated the relationship between race/ethnicity and provider recommendations and thus, fulfilled the statistical requirements for mediation.

Individual variables that did not achieve statistical significance at a particular step were dropped from the equation at the subsequent step. Physicians’ perceptions of patients’ social and behavioral characteristics were entered in the final step to provide the most conservative test of these potential mediators.

Our preliminary analyses of the effects of patient race/ethnicity on physicians’ recommendations for CABG among patients who were appropriate candidates revealed a statistically significant interaction between race/ethnicity and gender of patient in predicting physician recommendations. An examination of results unadjusted for any other covariates showed that 21% of Black men received a recommendation for CABG, in comparison with 40% of White and Hispanic men (P < .05). In contrast, 40% of Black women received a recommendation for CABG (similar to White men), and there were no race/ethnicity differences among women. Supporting this latter result, the interaction between race/ethnicity and gender remained statistically significant (P < .002) in a multivariate model including clinical factors, insurance status, and physician characteristics. Because there were no statistically significant race/ethnicity effects among women or between Hispanics and Whites, the next set of analyses focused on identifying mediators of the effects of race/ethnicity on physician recommendations among Black and White men (n=199).

Types of Providers Seen

Table 1 presents the results of tests assessing bivariate relationships among all of the potential mediators. Patients in all race/ethnicity categories were equally likely to be seen by an attending physician (vs a resident or fellow), a cardiologist, and a male physician; however, White patients saw, on average, older physicians than Blacks (among Blacks: mean = 43.13, SD= 7.18; among Whites: mean = 45.9, SD= 7.86; P < .01) and were less likely to see a non-White physician than their Black counterparts (Black patients, 30%; White patients, 9%; χ21 = 28.26, P < .01). The fact that Blacks were more likely to see non-White physicians did not reflect race/ethnicity congruence among non-Whites, as only 1% of the physicians in the sample identified themselves as Black. The remainder were largely of ethnic backgrounds originating in Asia (India and bordering nations).

There was no significant bivariate association between physicians’ race/ethnicity or age and their treatment recommendations. Independent effects of physician characteristics were also examined in multivariate analyses, as described subsequently.

Distribution of Clinical Characteristics

There were no significant racial/ethnic differences in rates of 3-vessel disease in this sample of male patients who were appropriate candidates for CABG according to the RAND criteria. Black patients were significantly younger than Whites (mean = 43 years vs mean = 46 years; P < .01) and had lower rates of left main disease (10% vs 19%; P < .06).

Role of insurance status as a mediator.

As can be seen in Table 1, there were significant race/ethnicity differences in terms of insurance coverage. Only 3% of White patients’ medical records indicated that they had Medicaid coverage or no coverage, while 21% of Black patients had Medicaid or no coverage (P < .00).

Influence of insurance status on physician recommendations.

Insurance status exhibited significant bivariate associations with physician recommendations. Only 26% of patients with Medicaid coverage or no insurance coverage received a recommendation for CABG, as opposed to 43% of patients with Medicare or private coverage (P < .01).

Mediating effect of race/ethnicity on provider recommendations for CABG.

As shown in Table 1, physicians rated Black patients significantly more negatively in terms of their likelihood of having sufficient social support, failing to comply with medical advice, abusing drugs, having significant career demands, and desiring a physically active lifestyle. Blacks were also rated as less likely than Whites to comply with medical advice or participate in cardiac rehabilitation if prescribed, although these differences were of only borderline statistical significance. Few Black patients were rated as more educated or intelligent than their White counterparts.

Table 2 provides the results of the multivariate analyses predicting provider recommendations for CABG among men who were rated as appropriate candidates for the procedure (n=199). Variables were entered in a series of 4 blocks to allow identification of the degree to which the 4 categories of various sets of predictors mediated the effects of race/ethnicity on provider recommendations (unadjusted Black–White odds ratio [OR]=0.40; P<.01).

First, none of the physician characteristics that significantly differed according to patient race/ethnicity had a significant association with physician recommendation. Second, of the clinical variables that had a bivariate association with patient race, only left main disease had a significant association with physician recommendations (OR = 4.01; P < .00). In addition, the amount of variance in provider recommendations accounted for by patient race decreased, indicating that some of the effect of patient race on physicians’ recommendations may be mediated by lower rates of left main disease among African Americans.

Third, the effect of insurance status on physician recommendations had, at best, a very small influence on the relationship between race and physician recommendations for CABG. Finally, of the measures of physicians’ perceptions of patients’ social and behavioral characteristics, only physician perception that the patient desired a physically active lifestyle and was educated had significant associations with recommendations. Most notably, the addition of these variables eliminated the effect of patient race/ethnicity on physician recommendations for CABG, providing support for the hypothesis that physicians’ perceptions of patients mediate the effects of patient race on their treatment recommendations.

The results of this study have several implications for future research, intervention, and policy on racial/ethnic disparities in health care. First, the observed interaction between patient race and gender associated with disparities in treatment suggests the possibility that researchers may want to reexamine (if they have not already done so) their data sets to determine whether such an interaction is present. Surprisingly, our finding contradicts the results of an earlier study in which primary care physicians were asked to provide opinions in response to videotaped vignettes. In this study, there were race/ethnicity-related differences in providers’ decisionmaking in the case of women but not men.

The complex relationship between race/ethnicity and gender reinforces the need for researchers to focus on the complex and interconnected nature of the influence of socio-demographic variables on provider behavior. This idea is consistent with a great deal of social cognitive research showing that people tend to categorize others at the intermediate level of subtype (e.g., Black woman, elderly White man) rather than in terms of overarching categories of race and gender, the reason being that subtypes are more descriptive and consequently allow more precision and cognitive efficiency.47,48

Our study also provided support for the primary hypotheses regarding the central importance of the way patient race (as well as other nonclinical characteristics) influences providers’ conscious and unconscious beliefs about patients and how these beliefs in turn influence their decisionmaking. Specifically, disparities in coronary artery disease treatment were shown to be at least partially mediated by the ways in which patient race or ethnicity influenced physicians’ perceptions of patients’ social and behavioral characteristics. This finding adds to a small but growing body of evidence8,34,4759 on how physicians often use patients’ demographic characteristics (e.g., race/ethnicity) as decisionmaking heuristics, including a very interesting line of research focusing on providers’ use of “base rates” (e.g., population statistics on prevalence of a characteristic in a subgroup) to inform their decisionmaking.55,60,61

The present findings are supported by the results of a number of other studies indicating that social and behavioral factors influence providers’ clinical decisionmaking, both explicitly (consciously) and implicitly (unconsciously). There is considerable empirical evidence that patients’ gender, age, socioeconomic status, diagnosis, marital status, sexual orientation, symptom severity, and race/ethnicity can influence providers’ beliefs about and expectations of patients independent of other factors.33,39,6273 Furthermore, both randomized vignette studies and examinations of actual provider practices indicate that patients’ demographic characteristics influence clinical decisionmaking.8,34,5159

Moreover, the fact that physicians’ perceptions of the degree to which patients were educated (P < .001) and desired a physically active lifestyle (P < .07) were independent predictors of physicians’ recommendations for CABG has important implications for understanding provider decisionmaking in general. Physicians’ exclusion of candidates for CABG on the basis of factors unrelated to accepted clinical guidelines may disproportionately exclude non-White patients. Although it is impossible to ascertain the accuracy of physicians’ ratings of patients’ desire for a physically active lifestyle, it is clear from a previous study39 that physicians underestimate Black patients’ education levels.

There are a number of initiatives aimed at ameliorating treatment disparities, including a variety of programs focused on increasing providers’ cultural competence. Evidence regarding the effectiveness of these approaches is not yet available, and, because most cultural competence programs focus on conscious beliefs rather than unconsciously activated stereotypes, it is unclear whether these programs will have an impact on the disparities reflected in these findings. However, research on unconscious bias and stereotyping suggests that current cultural competence programs that focus on improving interpersonal and communication skills, ability to elicit patient self-disclosure, active listening skills, and self-awareness are promising.

There is research evidence that, under certain conditions, individuals can consciously replace automatically activated stereotypes. These studies suggest that a provider is less likely to be influenced by a group stereotype if the provider (1) is aware of the potential for judgments, emotions, and behaviors to be biased; (2) is aware of a stereotype being activated; (3) is highly motivated to replace stereotypes with individual information; and (4) has sufficient cognitive resources (time and cognitive capacity) available for all of these cognitive tasks.

In conclusion, the present findings point to the importance of research and intervention strategies addressing the ways in which providers’ conscious and unconscious beliefs about patients mediate racial/ethnic disparities in treatment. It will be difficult to develop effective intervention strategies without a deeper understanding of how the context in which providers train and practice influences these processes. There is a great need for innovative studies examining the degree to which changes in reimbursement methodologies, organizational factors, provider education, and professional norms—including adherence to clinical practice guidelines—influence disparities in care.

Our findings also highlight the urgent need for discourse and consensus development on the role of social factors in clinical decisionmaking. What factors should matter? When should they matter? What valid and reliable methods can clinicians use to assess patients’ status on these factors and thus avoid reliance on “base rates”60,61 or stereotypes? What changes in organization, delivery, and reimbursement methodologies are needed to support such strategies? Failure to attend to these issues will in all likelihood condemn us to continuing inequities in—and thus, variable quality of—health care and health outcomes.

TABLE 1— Associations Between Patient Race and Potential Mediators of the Effects of Patient Race on Physician Recommendations for Coronary Artery Bypass Graft Surgery (n=199)
TABLE 1— Associations Between Patient Race and Potential Mediators of the Effects of Patient Race on Physician Recommendations for Coronary Artery Bypass Graft Surgery (n=199)
Potential MediatorBlack (n = 88)White (n = 111)P
Physician characteristics, %
Clinical characteristics, %
    Left main disease1019.06
    3-vessel disease5645NS
    Mean age, y4346.01
Medicaid coverage or uninsured, %213.00
Physician perceptions of patient social and behavioral characteristics, %
    Lacks social support5328.01
    Likely to overreport discomfort4531.03
    Has significant responsibility for family member3038NS
    Likely to fail to comply with medical advice5347.02
    Likely to have significant career demands2846.01
    Likely to abuse drugs3518.03
    Likely to initiate malpractice suit3940NS
    Desires physically active lifestyle1736.01
    Likely to participate in cardiac rehabilitation3753.02
    Likely to try to manipulate physician3736NS

Note. NS = nonsignificant. Data refer to men who were appropriate candidates for CABG but not for percutaneous transluminal coronary angioplasty.

TABLE 2— Factors Associated With Physician Recommendation for Coronary Artery Bypass Graft Surgery (CABG) Among Men Who Were Appropriate Candidates for CABG (n = 199)
TABLE 2— Factors Associated With Physician Recommendation for Coronary Artery Bypass Graft Surgery (CABG) Among Men Who Were Appropriate Candidates for CABG (n = 199)
Block of VariablesPROdds Ratio (95% Confidence Interval)
Block 1
    Race/ethnicity (reference group: White).01. . .. . .
    Black (vs White) patient.01−0.200.40 (0.23, 0.69)
Block 2
    Black patient.01−0.200.45 (0.24, 0.83)
    White physicianNS. . .. . .
    Age of physicianNS. . .. . .
Block 3  . . .
    Black patient.02−0.230.45 (0.23, 0.90)
    Left main disease.000.185.99 (2.06, 17.39)
    Patient age (interval level).160.000.99 (0.99, 1.01)
Block 4a
    Black patient.01−0.160.43 (0.23, 0.79)
    Left main disease.010.205.16 (1.80, 14.76)
    Medicaid coverage or uninsured.07−0.080.29 (0.07, 1.17)
Block 5b
    Black patient.170.000.58 (0.27, 1.25)
    Left main disease.010.216.10 (1.99, 18.64)
    Medicaid coverage or uninsured.07−0.070.24 (0.10, 0.58)
    Patient desires physically active lifestyle.070.082.20 (0.91, 5.29)
    Patient educated.0010.143.28 (1.24, 8.73)

aInsurance and reimbursement added.

bPhysicians’ perceptions of patients’ social and behavioral characteristics added.

The original project on which this study was based was supported by the New York Department of Health’s Quality Assurance Project Fund and the University of Albany’s Faculty Grant Program. Portions of the development of this article and associated analyses were supported by the Center for Chronic Disease Outcomes Research, which is funded by the VA Health Services Research and Development Service.

We wish to acknowledge Edward D. Hannan for his efforts in conceptualizing and directing the medical record abstraction portions of the original study; Jane Burke for her excellent project management of the original study; Hanna Bloomfield, director of the Center for Chronic Disease Outcomes Research, for her leadership in finding protected time for writing; and H. Jack Geiger for his helpful comments on earlier presentations of these results.

Human Participant Protection This study was approved by the University of Albany’s institutional review board. Participants provided informed consent to take part in the study.


1. Fiscella K, Franks P, Doescher MP, Saver BG. Disparities in health care by race, ethnicity, and language among the insured: findings from a national sample. Med Care. 2002;40:52–59. Crossref, MedlineGoogle Scholar
2. Smedley BD, Stith AY, Nelson AR, eds. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC: National Academy Press; 2002. Google Scholar
3. Etchason J, Armour B, Ofili E, et al. Racial and ethnic disparities in health care. JAMA. 2001;285:883. Crossref, MedlineGoogle Scholar
4. Anderson GM, Brown AD. Appropriateness ratings: overuse, underuse, or misuse? Lancet. 2001;358:1475–1476. Crossref, MedlineGoogle Scholar
5. Bell PD, Hudson S. Equity in the diagnosis of chest pain: race and gender. Am J Health Behav. 2001; 25:60–71. Crossref, MedlineGoogle Scholar
6. Collins T, Johnson M, Daley J, Henderson W, Khuri S. Risk factors for lower extremity nontraumatic amputation in peripheral arterial disease: is race/ethnicity an independent factor? Paper presented at: 19th Annual VA Health Services Research Conference, February 2001, Washington, DC. Google Scholar
7. Einbinder LC, Schulman KA. The effect of race on the referral process for invasive cardiac procedures. Med Care Res Rev. 2000;57(suppl 1):162–180. Crossref, MedlineGoogle Scholar
8. Halm EA, Atlas SJ, Borowsky LH, et al. Understanding physician adherence with a pneumonia practice guideline: effects of patient, system, and physician factors. Arch Intern Med. 2000;160:98–104. Crossref, MedlineGoogle Scholar
9. Maynard C, Fisher LD, Passamani ER, Pullum T. Blacks in the Coronary Artery Surgery Study (CASS): race and clinical decision making. Am J Public Health. 1986;76:1446–1448. LinkGoogle Scholar
10. Kressin NR, Petersen LA. Racial differences in the use of invasive cardiovascular procedures: review of the literature and prescription for future research. Ann Intern Med. 2001;135:352–366. Crossref, MedlineGoogle Scholar
11. Petersen LA, Normand SL, Druss BG, Rosenheck RA. Process of care and outcome after acute myocardial infarction for patients with mental illness in the VA health care system: are there disparities? Health Serv Res. 2003;38:41–63. Crossref, MedlineGoogle Scholar
12. Petersen LA, Wright SM, Peterson ED, Daley J. Impact of race on cardiac care and outcomes in veterans with acute myocardial infarction. Med Care. 2002; 40(suppl 1):I86–I96. MedlineGoogle Scholar
13. Peterson ED, Wright SM, Daley J, Thibault GE. Racial variation in cardiac procedure use and survival following acute myocardial infarction in the Department of Veterans Affairs. JAMA. 1994;271:1175–1180. Crossref, MedlineGoogle Scholar
14. Ferguson JA, Weinberger M, Westmoreland GR, et al. Racial disparity in cardiac decision making: results from patient focus groups. Arch Intern Med. 1998; 158:1450–1453. Crossref, MedlineGoogle Scholar
15. Leape LL, Hilborne LH, Bell R, Kamberg C, Brook RH. Underuse of cardiac procedures: do women, ethnic minorities, and the uninsured fail to receive needed revascularization? Ann Intern Med. 1999;130:183–192. Crossref, MedlineGoogle Scholar
16. Schneider EC, Leape LL, Weissman JS, Piana RN, Gatsonis C, Epstein AM. Racial differences in cardiac revascularization rates: does “overuse” explain higher rates among white patients? Ann Intern Med. 2001; 135:328–337. Crossref, MedlineGoogle Scholar
17. Schulman KA, Berlin JA, Harless W, et al. The effect of race and sex on physicians’ recommendations for cardiac catheterization. N Engl J Med. 1999;340:618–626. Crossref, MedlineGoogle Scholar
18. Sedlis SP, Fisher VJ, Tice D, Esposito R, Madmon L, Steinberg EH. Racial differences in performance of invasive cardiac procedures in a Department of Veterans Affairs medical center. J Clin Epidemiol. 1997;50:899–901. Crossref, MedlineGoogle Scholar
19. Sheifer SE, Escarce JJ, Schulman KA. Race and sex differences in the management of coronary artery disease. Am Heart J. 2000;139:848–857. Crossref, MedlineGoogle Scholar
20. Whittle J, Conigliaro J, Good CB, Lofgren RP. Racial differences in the use of invasive cardiovascular procedures in the Department of Veterans Affairs medical system. N Engl J Med. 1993;329:621–627. Crossref, MedlineGoogle Scholar
21. LaVeist TA, Arthur M, Morgan A, et al. The Cardiac Access Longitudinal Study: a study of access to invasive cardiology among African American and white patients. J Am Coll Cardiol. 2003;41:1159–1166. Crossref, MedlineGoogle Scholar
22. O’Connell L, Brown SL. Do nonprofit HMOs eliminate racial disparities in cardiac care? J Health Care Finance. 2003;30:84–94. MedlineGoogle Scholar
23. LaVeist TA, Arthur M, Morgan A, Plantholt S, Rubinstein M. Explaining racial differences in receipt of coronary angiography: the role of physician referral and physician specialty. Med Care Res Rev. 2003;60:453–467. Crossref, MedlineGoogle Scholar
24. LaVeist TA, Morgan A, Arthur M, Plantholt S, Rubinstein M. Physician referral patterns and race differences in receipt of coronary angiography. Health Serv Res. 2002;37:949–962. Crossref, MedlineGoogle Scholar
25. Ayanian JZ, Cleary PD, Weissman JS, Epstein AM. The effect of patients’ preferences on racial differences in access to renal transplantation. N Engl J Med. 1999; 341:1661–1669. Crossref, MedlineGoogle Scholar
26. Conigliaro J, Whittle J, Good CB, Skanderson M, Kelley M, Goldberg K. Delay in presentation for cardiac care by race, age, and site of care. Med Care. 2002;40(suppl 1):97–105. Crossref, MedlineGoogle Scholar
27. Kressin NR, Clark JA, Whittle J, et al. Racial differences in health-related beliefs, attitudes, and experiences of VA cardiac patients: scale development and application. Med Care. 2002;40(suppl 1):I72–I85. Crossref, MedlineGoogle Scholar
28. Whittle J, Conigliaro J, Good CB, Joswiak M. Do patient preferences contribute to racial differences in cardiovascular procedure use? J Gen Intern Med. 1997; 12:267–273. Crossref, MedlineGoogle Scholar
29. Valenstein M, Kales H, Mellow A, et al. Psychiatric diagnosis and intervention in older and younger patients in a primary care clinic: effect of a screening and diagnostic instrument. J Am Geriatr Soc. 1998;46:1499–1505. Crossref, MedlineGoogle Scholar
30. Hannan EL, van Ryn M, Burke J, et al. Access to coronary artery bypass surgery by race/ethnicity and gender among patients who are appropriate for surgery. Med Care. 1999;37:68–77. Crossref, MedlineGoogle Scholar
31. Naumburg EH, Franks P, Bell B, Gold M, Engerman J. Racial differentials in the identification of hyper-cholesterolemia. J Fam Pract. 1993;36:425–430. MedlineGoogle Scholar
32. Ortega AN, Gergen PJ, Paltiel AD, Bauchner H, Belanger KD, Leaderer BP. Impact of site of care, race, and Hispanic ethnicity on medication use for childhood asthma. Pediatrics. 2002;109:E1. Crossref, MedlineGoogle Scholar
33. Krupat E, Irish JT, Kasten LE, et al. Patient assertiveness and physician decision-making among older breast cancer patients. Soc Sci Med. 1999;49:449–457. Crossref, MedlineGoogle Scholar
34. Feldman HA, McKinlay JB, Potter DA, et al. Non-medical influences on medical decision making: an experimental technique using videotapes, factorial design, and survey sampling. Health Serv Res. 1997;32:343–366. MedlineGoogle Scholar
35. Todd KH, Samaroo N, Hoffman JR. Ethnicity as a risk factor for inadequate emergency department analgesia. JAMA. 1993;269:1537–1539. Crossref, MedlineGoogle Scholar
36. Cleeland CS, Gonin R, Baez L, Loehrer P, Pandya KJ. Pain and treatment of pain in minority patients with cancer: the Eastern Cooperative Oncology Group Minority Outpatient Pain Study. Ann Intern Med. 1997;127:813–816. Crossref, MedlineGoogle Scholar
37. Strakowski SM, Lonczak HS, Sax KW, et al. The effects of race on diagnosis and disposition from a psychiatric emergency service. J Clin Psychiatry. 1995;56:101–107. MedlineGoogle Scholar
38. van Ryn M, Fu SS. Paved with good intentions: do public health and human service providers contribute to racial/ethnic disparities in health? Am J Public Health. 2003;93:248–255. LinkGoogle Scholar
39. van Ryn M, Burke J. The effect of patient race and socio-economic status on physicians’ perceptions of patients. Soc Sci Med. 2000;50:813–828. Crossref, MedlineGoogle Scholar
40. van Ryn M. Research on the provider contribution to race/ethnicity disparities in medical care. Med Care. 2002;40(suppl 1):I140–I151. Crossref, MedlineGoogle Scholar
41. Winslow CM, Kosecoff JB, Chassin M, Kanouse DE, Brook RH. The appropriateness of performing coronary artery bypass surgery. JAMA. 1988;260:505–509. Crossref, MedlineGoogle Scholar
42. Eagle KA, Guyton RA, Davidoff R, et al. ACC/AHA guidelines for coronary artery bypass graft surgery: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1991 Guidelines for Coronary Artery Bypass Graft Surgery). J Am Coll Cardiol. 1999;34:1262–1347. Crossref, MedlineGoogle Scholar
43. Percutaneous Transluminal Coronary Angioplasty: A Literature Review and Ratings of Appropriateness and Necessity. Santa Monica, Calif: RAND Corp; 1991. Google Scholar
44. Coronary Artery Bypass Graft Surgery: A Literature Review and Ratings of Appropriateness and Necessity. Santa Monica, Calif: RAND Corp; 1991. Google Scholar
45. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51:1173–1182. Crossref, MedlineGoogle Scholar
46. Judd CM, Kenny DA. Process analysis: estimating mediation in treatment evaluations. Eval Rev. 1981;5:602–619. CrossrefGoogle Scholar
47. Taylor C. A categorization approach to stereotyping. In: Hamilton DL, ed. Cognitive Processes in Stereotyping and Intergroup Behavior. Hillsdale, NJ: Lawrence Erlbaum Associates; 1981:83–114. Google Scholar
48. Fiske S. Stereotyping, prejudice, and discrimination. In: Gilbert DT, Fiske ST, Lindzey G, eds. The Handbook of Social Psychology. New York, NY: McGraw-Hill; 1998:357–415. Google Scholar
49. Bogart LM, Catz SL, Kelly JA, Benotsch EG. Factors influencing physicians’ judgments of adherence and treatment decisions for patients with HIV disease. Med Decis Making. 2001;21:28–36. Crossref, MedlineGoogle Scholar
50. Bogart LM, Kelly JA, Catz SL, Sosman JM. Impact of medical and nonmedical factors on physician decision making for HIV/AIDS antiretroviral treatment. J Acquir Immune Defic Syndr. 2000;23:396–404. Crossref, MedlineGoogle Scholar
51. vanIneveld CH, Cook DJ, Kane SL, King D. Discrimination and abuse in internal medicine residency. J Gen Intern Med. 1996;11:401–405. Crossref, MedlineGoogle Scholar
52. Shye D, Freeborn DK, Romeo J, Eraker S. Understanding physicians’ imaging test use in low back pain care: the role of focus groups. Int J Qual Health Care. 1998;10(2):83–91. Crossref, MedlineGoogle Scholar
53. McKinlay JB, Burns RB, Durante R, et al. Patient, physician and presentational influences on clinical decision making for breast cancer: results from a factorial experiment. J Eval Clin Pract. 1997;3:23–57. Crossref, MedlineGoogle Scholar
54. McKinlay JB, Burns RB, Feldman HA, et al. Physician variability and uncertainty in the management of breast cancer: results from a factorial experiment. Med Care. 1998;36:385–396. Crossref, MedlineGoogle Scholar
55. McKinlay JB, Potter DA, Feldman HA. Nonmedical influences on medical decision-making. Soc Sci Med. 1996;42:769–776. Crossref, MedlineGoogle Scholar
56. Christakis NA, Asch DA. Biases in how physicians choose to withdraw life support. Lancet. 1993;342:642–646. Crossref, MedlineGoogle Scholar
57. Bradley CP. Factors which influence the decision whether or not to prescribe: the dilemma facing general practitioners. Br J Gen Pract. 1992;42:454–458. MedlineGoogle Scholar
58. Graber MA, Bergus G, Dawson JD, Wood GB, Levy BT, Levin I. Effect of a patient’s psychiatric history on physicians’ estimation of probability of disease. J Gen Intern Med. 2000;15:204–206. Crossref, MedlineGoogle Scholar
59. Lockey AS, Hardern RD. Decision making by emergency physicians when assessing cardiac arrest patients on arrival at hospital. Resuscitation. 2001;50:51–56. Crossref, MedlineGoogle Scholar
60. Balsa AI, McGuire TG. Prejudice, clinical uncertainty and stereotyping as sources of health disparities. J Health Econ. 2003;22:89–116. Crossref, MedlineGoogle Scholar
61. Balsa AI, McGuire TG. Statistical discrimination in health care. J Health Econ. 2001;20:881–907. Crossref, MedlineGoogle Scholar
62. Hall JA, Epstein AM, DeCiantis ML, McNeil BJ. Physicians’ liking for their patients: more evidence for the role of affect in medical care. Health Psychol. 1993; 12:140–146. Crossref, MedlineGoogle Scholar
63. Gerbert B. Perceived likeability and competence of simulated patients: influence on physicians’ management plans. Soc Sci Med. 1984;18:1053–1059. Crossref, MedlineGoogle Scholar
64. Kearney PM, Griffin T. Between joy and sorrow: being a parent of a child with developmental disability. J Adv Nurs. 2001;34:582–592. Crossref, MedlineGoogle Scholar
65. Kearney N, Miller M, Paul J, Smith K. Oncology healthcare professionals’ attitudes toward elderly people. Ann Oncol. 2000;11:599–601. Crossref, MedlineGoogle Scholar
66. Lewis DO, Balla DA, Shanok SS. Some evidence of race bias in the diagnosis and treatment of the juvenile offender. Am J Orthopsychiatry. 1979;49:53–61. Crossref, MedlineGoogle Scholar
67. Revenson TA. Compassionate stereotyping of elderly patients by physicians: revising the social contact hypothesis. Psychol Aging. 1989;4:230–234. Crossref, MedlineGoogle Scholar
68. Shortt S. Venerable or vulnerable? Ageism in health care. J Health Serv Res Policy. 2001;6:1–2. Crossref, MedlineGoogle Scholar
69. Stern M, Arenson E. Childhood cancer stereotype: impact on adult perceptions of children. J Pediatr Psychol. 1989;14:593–605. Crossref, MedlineGoogle Scholar
70. Stern M, Moritzen SK, Carmel S, Olexa-Andrews M. The prematurity stereotype in Israeli health care providers. Med Educ. 2001;35:129–133. Crossref, MedlineGoogle Scholar
71. Ross M, Glisson C. Bias in social work intervention with battered women. J Soc Serv Res. 1991;14:79–105. CrossrefGoogle Scholar
72. Kelly JA, St. Lawrence JS, Smith S, Hood HV, Cook DJ. Medical students’ attitudes toward AIDS and homosexual patients. J Med Educ. 1987;62:549–556. MedlineGoogle Scholar
73. Kelly JA, St. Lawrence JS, Smith S Jr, Hood HV, Google Scholar


No related items




Michelle van Ryn, PhD, MPH, Diana Burgess, PhD, Jennifer Malat, PhD, and Joan Griffin, PhDMichelle van Ryn is with the Department of Family Practice and Community Health, University of Minnesota Medical School; the Division of Epidemiology, School of Public Health, University of Minnesota, Minneapolis; and the Center for Chronic Disease Outcomes Research, Minneapolis VA Medical Center. Diana Burgess and Joan Griffin are with the Center for Chronic Disease Outcomes Research, Minneapolis VA Medical Center, and the Department of Medicine, University of Minnesota Medical School. Jennifer Malat is with the Department of Sociology, University of Cincinnati, Cincinnati, Ohio. “Physicians’ Perceptions of Patients’ Social and Behavioral Characteristics and Race Disparities in Treatment Recommendations for Men With Coronary Artery Disease”, American Journal of Public Health 96, no. 2 (February 1, 2006): pp. 351-357.


PMID: 16380577