• Health services;
  • pediatrics;
  • process assessment (health care);
  • renal transplantation


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Transplantation is the treatment goal for youth with kidney failure. To assess the effects of compliance, parental education and race on nephrologists' recommendations for transplantation in children, we surveyed a national random sample of adult and pediatric nephrologists. We elicited transplant recommendations for case vignettes created from random combinations of patient age, gender, race, cause of renal failure, family structure, parental education and compliance.

Of 519 eligible physicians, 316 (61%) responded. Nephrologists were more likely to recommend transplantation for children of college-educated parents than children of parents who did not finish high school, despite identical clinical and demographic characteristics (adjusted OR 1.48, 95% CI 1.18, 1.86). Patient noncompliance negatively influenced transplant recommendations (adjusted OR 0.1, 95% CI 0.08, 0.13). Additionally, compliance had a different effect on transplant recommendations for white compared with black patients. The adjusted OR of a white, compliant patient being referred for transplantation were twice that of a black compliant patient (OR 2.06, 95% CI 1.17, 3.6).

Education and compliance with therapy independently influence nephrologists' recommendations for transplantation in youth with kidney failure. Among the most compliant candidates, referral for transplantation may vary with patient race.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Despite the extension of Medicare health insurance to virtually all patients with end stage renal disease (ESRD) (1), differential access to kidney transplantation apparently still exists along racial and socioeconomic lines. This holds true for patients of all ages, even after blood and histocompatibility differences between the kidney donor pool and patients on the cadaveric transplant waiting list are taken into account (2–11).

Multiple steps are involved in activating a dialysis patient for kidney transplantation (4). Small differences between racial and social or economic groups within each of these steps may count towards the persistent large differences in access that have been repeatedly documented. Patients of different racial, social and economic backgrounds (1) may make different treatment choices regarding transplantation (2), may have different burdens of comorbid conditions, or (3) may be subject to different treatment recommendations by the nephrologists based on race, social class or economic status.

Recent studies have directly examined the first two points, and indirectly examined the latter. Patient preferences for transplantation (3), as well as other invasive procedures (12), do differ according to race, but this only partly explains the magnitude of apparent racial and socioeconomic differences in access to transplantation. Comorbidities that are absolute or relative contraindications to transplantation may be more frequent among black dialysis patients, however, even among ‘appropriate’ transplant candidates, blacks are less likely to be referred for transplant evaluation than are whites (13). In sum, these reports suggest the possibility that ‘race-based barriers to the receipt of appropriate care’ exist, perhaps at the level of the referring physician (13–15).

Race is a marker for skin color, but can also be a proxy for a multitude of factors that correlate with race: socioeconomic status, discrimination, cultural factors, or biologic differences between racial groups (16). We used an experimental design to determine whether patient race, or social factors that may be subjectively associated with race, such as parental education and compliance with therapy, independently influence nephrologists' transplant recommendations for children and adolescents with ESRD. We chose to study children and adolescents because kidney transplantation is universally regarded as the treatment of choice for ESRD in this age group, and significant clinical comorbidities rarely exist as contraindications to kidney transplantation.


In a national study, we randomly selected a sample of adult and pediatric nephrologists to receive clinically representative case vignettes for which we elicited transplant recommendations. The methods for this study design have been previously described (17). We purposely over-sampled pediatric nephrologists as our vignettes all dealt with young kidney disease patients, aged < 19 years.

Questionnaire format

Nephrologists were surveyed by a mailed questionnaire containing 10 randomly generated case vignettes. The survey was sent with a cover letter stating that this was a study of clinical decision making to gain a better understanding of the most important factors determining treatment recommendations for dialysis and transplant. Characteristics included in the case vignettes included: patient age (5–19 years), race (black or white), gender (male or female), cause of ESRD (‘acquired glomerular’ or ‘congenital urologic’ disease), level of parental education (< high school, some high school, or some college), distance from a dialysis facility (< 1 h drive, 1–2 h or > 2 h), compliance with therapy (‘compliant’, ‘not compliant’, ‘questionably compliant’), and number of parents in the home (one or two). We asked the nephrologists to assume that the patient in the vignette had never undergone a transplant, had no other comorbid conditions and their underlying renal disease was unlikely to re-occur after transplantation. Ten case vignettes were randomly selected and assigned to 100 different questionnaires. The distributions of clinical characteristics in the assigned case vignettes were analyzed to assure that all possible combinations of patient characteristics were present. We asked nephrologists whether or not they recommended transplantation for each of these hypothetical cases.

Sampling of nephrologists

The survey samples were drawn from a national database of adult and pediatric nephrologists. Questionnaires were mailed to a random sample of 600 office- and hospital-based nephrologists selected after stratification by 10 geographic regions.

Survey follow up

Questionnaires that were returned as undeliverable for incorrect addresses were replaced by random selection of another pediatric or adult nephrologist from the same geographic region. Non-respondents after the second mailing were phoned to verify that the selected nephrologists remained active in clinical practice at that address and had received the survey. A duplicate questionnaire was faxed to those participants whose address was verified, and telephone follow up was carried out within 2 weeks.

Statistical analysis

Preliminary analyses explored the geographic distribution, practice characteristics and training of respondents compared with nonrespondents to assess the possibility of nonresponse bias. We then examined how clinical ( age, race, gender, and cause of ESRD) and ‘nonclinical’ (number of parents in the home, compliance, and education) factors affected these recommendations using multivariate logistic regression analysis. Possible correlations among the nephrologists' responses to the 10 separate case vignettes included in each questionnaire were accounted for using techniques for handling clustered data (the generalized estimation equation) (18). Additional stratified analyses were performed according to race (black or white), level of parental education, or compliance. Race-compliance and race-education interaction terms were also included in the multivariate models.

Because the outcome of interest (referral for transplantation) was a common occurrence in the survey, and the adjusted OR therefore overestimated the prevalence ratio, we converted the adjusted ORs to prevalence ratios and adjusted probabilities using reported methods (19).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Response rate

Of 600 nephrologists selected in our national random sample, 81 were excluded because of undeliverable addresses after the first replacement (n = 21), telephone follow up demonstrated they had moved out of the region (n = 29), they were no longer in practice (n = 18) or the physicians described themselves as never seeing patients in the pediatric or adolescent age group and declined to make treatment recommendations for young patients in the case vignettes (n = 13). Of the 519 remaining nephrologists, 316 (61%) responded. Fifty-four percent (191/351) of adult nephrologists responded compared with 74% (125/168) of pediatric nephrologists.

Characteristics of responding nephrologists

The demographic, training and practice characteristics available on all nephrologists surveyed and by response category are presented in Table 1. There were no statistically significant differences between respondents and nonrespondents bygeographic region. There were significant differences in the response rates between pediatric and adult nephrologists, (74% vs. 54%) and among physicians practicing in university compared with community settings. University-based nephrologists, and those with fewer years in practice were more likely to respond to the survey (Table 1).

Table 1.  Charactersitics of survey respondents compared with nonrespondents
 Respondents n=316 n(%)Nonrespondents n=203 n(%)p-value
  1. Data in parentheses are percentages.

Gender  0.03
Male251 (79.0)177 (87.2) 
Female 64 (20.3) 26 (12.8) 
Unknown  1 (0.7) – 
Years in practice  < 0.006
< 10102 (32.3) 47 (23.2) 
11–20127 (40.2) 79 (38.9) 
> 20 87 (27.5) 76 (37.4) 
Practice setting  < 0.005
Independent center145 (45.9)133 (65.5) 
University hospital133 (42.1) 37 (18.2) 
Other 38 (12.0) 33 (16.3) 

Transplant recommendations according to patient clinical and ‘nonclinical’ characteristics

Responding nephrologists recommended transplantation in 80% of the 3160 case vignettes presented to them. In overall analyses, there were no differences in transplant recommendations according to patient age, race, sex, cause of ESRD, or number of parents in the home (Table 2). Additionally, there were no differences in transplant recommendations to these vignettes for adult compared with pediatric nephrologists (82% vs. 78%). However, transplantation was recommended in a significantly higher percentage of case vignettes describing college-educated parents, and higher levels of patient compliance.

Table 2.  Patient characteristics and odds of recommending transplantation
 % Transplant recommendedp-valueOdds ratio Unadjusted (95% Cl)Adjusted* (95% Cl)
  • *

    Adjusted for variables stated, pediatric vs. adult nephrologist, and geographic region.

Parental education 0.0008  
< High school77% 1.0 (reference)1.0 (reference)
High school81% 1.29 (1.20, 1.84)1.38 (1.10, 1.73)
College83% 1.5 (1.20, 1.84)1.48 (1.18, 1.86)
Compliance < 0.0001  
Compliant94% 1.0 (reference)1.0 (reference)
Questionably compliant85% 0.35 (0.26, 0.40)0.35 (0.25, 0.47)
Not compliant62% 0.1 (0.08,0.14)0.1 (0.08, 0.13)
Parents in home NS  
One parent45% 1.0 (reference)1.0 (reference)
Two parents55% 0.95 (0.08, 1.13)0.99 (0.82, 1.20)
Age (years) NS  
5–983% 1.0 (reference)1.0 (reference)
10–1479% 0.82 (0.6, 1.12)0.81 (0.58,1.14)
15–1980% 0.86 (0.63, 1.17)0.82 (0.59, 1.14)
Race NS  
Black81% 1.0 (reference)1.0 (reference)
White80% 0.93 (0.78, 1.11)0.91 (0.76, 1.10)
Gender NS  
Female81% 1.0 (reference)1.0 (reference)
Male79% 0.87 (0.73, 1.04)0.87 (0.72, 1.05)
Disease NS  
Acquired disease81% 1.0 (reference)1.0 (reference)
Congenital urologic80% 1.07 (0.9, 1.27)0.98 (0.81, 1.18)

Descriptions of patient compliance had the strongest influence on the recommendation for transplantation. Patients described as ‘not compliant’ (62%) or ‘questionably compliant’ (85%) were less likely than those described as compliant (94%) to be recommended for transplantation. Multivariate analyses, controlling for patient clinical and nonclinical factors outlined in Table 2 confirmed this gradedassociation. Compared with compliant patients, the adjusted odds of being recommended for transplantation was 0.1 (95% CI 0.08, 0.13) and 0.35 (95% CI 0.25, 0.47) for ‘not compliant’ and ‘questionably compliant’ patients, respectively (Table 2).

The level of parents' education also influenced the nephrologists' recommendation for transplantation. In univariate analyses, 77% of children whose parents did not complete high school were recommended for kidney transplant, compared with 83% of children whose parents had some college education. In multivariate logistic regression analysis, the odds of nephrologists recommending transplantation were almost 50% greater for children of college-educated parents than those with less than a high-school education, even when compliance, patient age, race, gender, cause of ESRD, and number of parents in the home were held constant (OR 1.48, 95% CI 1.18, 1.86) (Table 2). When adjustment was made for the fact that each nephrologist surveyed made recommendations for several case vignettes, using the generalized estimation equation, the results were robust (OR 1.43, 95% CI 1.23, 1.66).

The adjusted probabilities for referral for transplantation for children of college-educated parents (83%, 95% CI 80, 86%) and for noncompliant patients (61%, 95% CI 56, 67%) were virtually identical to the actual percentages reported in Table 2.

Evaluation for interaction between compliance, parental education and race

Results of stratified analyses according to race are presented in Table 3. Compared with children whose parents had less than a high-school education, parental college education had a stronger positive effect on recommendations for transplantation among black patients (Adj. OR 1.65, 95% CI 1.18, 2.31) than among white patients (Adj. OR 1.32, 95% CI 0.96, 1.82). Conversely, explicit descriptions of noncompliance had a greater negative impact on transplant referral for white patients. Additionally, stratified analyses limited to ‘compliant’ patients showed the adjusted odds of compliant white patients being referred for transplantation were twice that of black patients (OR 2.06, 95% CI 1.17, 3.6). An analysis with a formal interaction term in the model showed that interaction terms for race and compliance (p < 0.003) were statistically significant. This effect was not seen in those patients labeled as not compliant (adj. OR 0.83, 95% CI 0.64–1.06) or questionably compliant (adj. OR 0.75, 0.53–1.06). In analyses grouping race, education and compliance characteristics together, we found that 97% of compliant, white children of college-educated parents were recommended for transplantation, and 91% of compliant black children of college-educated parents were recommended for transplantation (p < 0.05).

Table 3.  Association between patient characteristics and physicians' recommendations for transplantation for white and black children with end stage renal disease
 White child n=1617 (51%)Black child n=1543 (49%)
Characteristic% Transplant recommendedUnadjusted odds ratio (95% Cl)Adjusted odds ratio (95% Cl)% Transplant recommendedUnadjusted odds ratio (95% Cl)Adjusted odds ratio (95% Cl)
Age (years)
 5–980.91.0 (reference)1.0 (reference)84.21.0 (reference)1.0 (reference)
 10–1478.50.86 (0.55, 1.35)0.75 (0.46, 1.21)80.40.77 (0.49, 1.23)0.82 (0.50, 1.35)
 15–1980.20.96 (0.61, 1.48)0.80 (0.50, 1.29)80.20.76 (0.48, 1.20)0.77 (0.48, 1.25)
 Female81.11.0 (reference)1.0 (reference)81.41.0 (reference)1.0 (reference)
 Male78.00.82 (0.65, 1.05)0.82 (0.63, 1.07)80.10.92 (0.72, 1.19)0.91 (0.69, 1.19)
 Acquired disease78.21.0 (reference)1.0 (reference)81.11.0 (reference)1.0 (reference)
 Congenital urol81.01.19 (0.93, 1.52)1.02 (0.79, 1.33)80.30.95 (0.74, 1.22)0.92 (0.71, 1.21)
Parents in home
 One parent79.51.0 (reference)1.0 (reference)79.81.0 (reference)1.0 (reference)
 Two parents79.70.99 (0.78, 1.26)1.01 (0.77, 1.31)81.40.91 (0.70, 1.17)0.98 (0.75, 1.28)
Parental education
 < High school77.01.0 (reference)1.0 (reference)76.21.0 (reference)1.0 (reference)
 High school79.91.19 (0.89, 1.59)1.32 (0.97, 1.82)81.81.41 (1.04, 1.90)1.43 (1.04, 1.97)
 College81.91.36 (1.01, 1.82)1.33 (0.96, 1.82)84.21.66 (1.22, 2.27)1.65 (1.18, 2.31)
 Not compliant59.80.06 (0.04, 0.10)0.06 (0.04, 0.1)64.20.15 (0.10, 0.22)0.15 (0.10, 0.21)
 Questionable83.20.20 (0.12, 0.33)0.20 (0.12, 0.32)86.60.54 (.35, .82)0.53 (0.35, 0.81)
 Compliant96.21.0 (reference)1.0 (reference)92.31.0 (reference)1.0 (reference)


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

To directly examine whether social factors affect transplant recommendations, we surveyed a national random sample of adult and pediatric nephrologists. Random combinations of patient characteristics generated hypothetical case vignettes of children and adolescents with kidney failure, describing patient age, race, number of parents in the home, parents' educational level, and compliance. Physicians were asked whether or not they recommended transplantation for each of these case vignettes. With a 61% response rate, we found that race did not primarily affect pediatric renal transplant referral, however, the physicians' transplant recommendations varied in a graded fashion, with explicit labels describing parents' education and level of compliance. Children of more educated parents were more likely to be referred for transplantation. Patients described as ‘noncompliant’ were much less likely to be referred for transplantation than those labeled ‘compliant’. Furthermore, our analysis suggested a label of noncompliance had a stronger negative effect on transplant referral for white compared with black patients, while among patients labeled as compliant, white patients were statistically significantly more likely than black patients to be referred for transplantation. How do these findings shed light on the tangle of biologic, social and economic factors that contribute to ‘race-based barriers’ to kidney transplantation?


A number of authors have asserted that ‘sociocultural’ status, including educational attainment, influences medical decision making by health professionals (20,21). A patients' educational level may be known or subjectively assessed. As parents are responsible for carrying out the medical plan for pediatric transplant recipients, we included the parents' educational status in our case vignettes. Our study shows that when the level of parental education is made explicit, higher education is associated with increasing likelihood of referral for transplantation. Although little data exists directly measuring parental education in pediatric renal disease, among adult patients with ESRD substantial differences in the levels of education exist between black and white dialysis patients. In one large recent study, over 25% of black dialysis patients had not completed high school while only 17% of whites had less than a high school education (3). In that study, when the probability of referral for transplantation among blacks and white patients was adjusted for sociodemographic factors and type of dialysis facility, black–white differences became substantially smaller, suggesting that a large part of the observed racial difference could be attributed to differences in sociodemographic factors.

While labels of educational level were made explicit in our study, in clinical practice, patients' educational level is often subjectively assessed by their physician. Recent reports have shown that a physician's perception of a patients' education can be associated with patient race (22). Blacks are frequently perceived to be less educated than whites. Therefore, physician perceptions of level of parental education may correlate with race, and an increased referral rate for transplantation associated with higher levels of parental education may partly explain the previously described racial differences in referral for transplantation.


In our overall analysis, a label of noncompliance was the most important factor influencing transplant recommendations. Compliance, or adherence, to the immunosuppressive regimen is essential for long-term graft survival. The important consideration given to descriptions of compliance may therefore be appropriate in these case vignettes. In organ transplant recipients, noncompliance rates range between 20 and 50%, and at least in one study, 91% of patients who were noncompliant with medications suffered either graft rejection or death compared with 18% of compliant patients (20).

In our vignettes, cases were presented with labels of ‘compliant’, ‘questionably compliant’ or ‘not compliant’. In practice, subjective assessments on the part of the physicians would substitute for our explicit labels. Several studies have demonstrated that subjective physician assessments of compliance may be associated with race. In a study of the effect of race on a physician's recommendation for cardiac catheterization, physicians were asked to predict the likelihood that a patient would comply with therapy, and to judge the characteristics of patients believed to be predictors of patient compliance. Actors reading identical scripted symptoms played patients. In this study, black actor patients were deemed less likely to comply with therapy than whites (14). Another recent study of the effect of patient race and socioeconomic status on physicians' perceptions of patients, also showed that compared with white patients, black patients were deemed less likely to comply with medical advice, and less likely to participate in physician-prescribed rehabilitation (22). If physicians' perceptions of post-transplant compliance are not only important determinants of transplant referral, but also are associated with patient race, subtle differences in perceptions of compliance could contribute to systematic racial differences in access to transplantation. In our stratified analyses, when explicit descriptions of compliance were held constant, differences in transplant recommendations according to race existed among patients labeled as ‘compliant’. In this group, white patients were more likely to be recommended for transplantation than were black patients. This finding could be interpreted to show that among patients in whom physicians are confident of compliance, there is a bias against transplant recommendation for black compared with white patients. However, this association did not persist among those patients labeled as ‘questionably’ or ‘not’ compliant. It must also be cautioned that the multiple comparisons performed in our analysis may have raised to the level of significance a potentially random association between race, compliance and recommendation for transplantation.

The strengths of this study include the use of randomized, hypothetical case vignettes in which patient race, social and economic factors are randomly combined. In the study of actual dialysis patients, these factors are frequently closely linked, and their effects on referral for transplantation are difficult to separate. The experimental design of this study allowed us to isolate the effects of these characteristics on the nephrologists' treatment recommendations, as we used only vignettes of medically ‘appropriate’ transplant candidates with no contradictions for this procedure, and no stated patient preference. In analyses of observational patient data, it is difficult to tease out the relative contributions of patient preferences, physician recommendations and the nature of the interaction between the physician and patient, which cumulatively impact on treatment decisions in clinical practice.

Our conclusions are also strengthened by the national representation of the nephrologists in our sample. Stratified sampling by geographic location ensured that we sampled nephrologists from both urban and rural areas as geography may certainly influence treatment recommendations.

Our study has several limitations. Our response rate of 61% leaves open the possibility that respondents to our survey may differ in significant ways to nonrespondents and that our results will not be generalizable. We cannot judge the effects of the differential response of the pediatric compared with the adult nephrologists. However, transplant recommendations by the adult and pediatric nephrologists (82% vs. 78%) were similar. Additionally, our overall response rate surpasses the mean response rate of 54% for published physician surveys (23). Furthermore, the analyses comparing responders to nonresponders demonstrated no differences in response rate by geographic region. Although we saw that pediatric nephrologists were more likely than adult nephrologists to respond to our survey, this is likely related to the fact that the case vignettes included only patients up to 19 years of age. Although response rates were lower, the evaluation of the adult nephrologists' transplant recommendations in our survey was important to the generalizability of our results, as previous studies of USRDS data suggest that almost one-third of chronic pediatric ESRD patients are cared for in facilities that predominantly serve adults (24).

Although the use of case vignettes does not allow the clinician the wealth of clinical information gleaned in a genuine physician–patient interaction, and the response to the patient description on the printed page may vary from actual practice, the use of case vignettes is also a unique strength of our study. Case vignettes have been shown to approximate the gold standard of standardized patient interviews in studies focusing on the process of care provided in actual clinical practice (25).

It is important to remember that in this study of social factors affecting a nephrologist's transplant recommendations, the nephrologists surveyed recommended transplantation in an overwhelming majority (80%) of the cases presented to them. Our study suggests that nephrologists' perceptions of patient compliance and the level of parental education may impact transplant recommendations. Race did not directly affect transplant recommendations in this survey. However, if a physician's perception of patient compliance and education are linked with race, these factors could contribute to ‘race-based barriers’ to referral for transplantation. In disentangling the effects of race, parental education and compliance on nephrologists' recommendations for transplantation in hypothetical, clinically ‘appropriate’ transplant candidates, our study points to further research in the direction of targeted interventions to erase differential access to transplantation. These interventions may include efforts to eradicate education biases, and the development of standardized, objective measures of compliance with care as part of the transplant evaluation (26).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Presented in part at the Pediatric Academic Societies' 2000 Annual Meeting, Boston, MA, May 12–15, 2000, and the American Society of Transplantation 19th Annual Meeting, Chicago, IL, May 15–19, 2000. Dr Furth is supported by grant #K08 DK02586–01A1 from the National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, and the Johns Hopkins Children's Center Clinical Care Outcomes Research Project. Dr Powe is supported by grant #K24 DK02643 from the National Institute of Diabetes and Digestive and Kidney Diseases.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References
  • 1
    Evans RW, Blagg CR, Bryan FA Jr. Implications for health care policy. A social and demographic profile of hemodialysis patients in the United States. JAMA 1981; 245: 48791.
  • 2
    Eggers RW. Racial differences in access to kidney transplantation. Health Care Financ Rev 1995; 17: 89103.
  • 3
    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: 16611669.
  • 4
    Alexander GC, Sehgal AR. Barriers to cadaveric renal transplantation among blacks, women, and the poor. JAMA 1998; 280: 11481152.
  • 5
    Hunsicker LG, Wolfe RA, Kahan BD et al. The impact of comorbid and sociodemographic factors on access to renal transplantation. JAMA 1993; 269: 603608.
  • 6
    Held PJ, Pauly MV, Bovbjerg RR, Newmann J, Salvatierra O Jr. Access to kidney transplantation. Has the United States eliminated income and racial differences? Arch Intern Medical 1988; 148: 25942600.
  • 7
    Kjellstrand CM. Age, sex and race inequality in renal transplantation. Arch Intern Med 1988; 148: 13051309.
  • 8
    Sanfilippo FP, Vaughn WK, Peters TG et al. Factors affecting the waiting time of cadaveric kidney transplant candidates in the United States. JAMA 1992; 267: 247252.
  • 9
    Soucie JM, Neylan JF, McClellan W. Race and sex differences in the identification of candidates for renal transplantation. Am J Kidney Dis 1992; 19: 414419.
  • 10
    Furth SL, Garg PP, Neu AM, Hwang W, Fivush BA, Powe NR. Racial differences in access to the kidney transplant waiting list for children and adolescents with end-stage renal disease. Pediatrics 2000; 106: 756761.
  • 11
    Garg PP, Diener-West M, Powe NR. Reducing racial disparities in transplant activation. whom should we target? Am J Kidney Dis 2001; 37: 921931.
  • 12
    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: 267273.
  • 13
    Noonan SJ, Armistead N, Cleary PD et al. Racial disparities in access to renal transplantation – clinically appropriate or due to underuse or overuse? N Engl J Med 2000; 343: 15371544.
  • 14
    Kerner JF, Sistrunk S, Gersh BJ et al. The effect of race and sex on physicians' recommendations for cardiac catheterization. N Engl J Med 1999; 340: 618626.
  • 15
    Vu HT, Powe NR, Nelson C et al. Race, gender, and partnership in the patient-physician relationship. JAMA 1999; 282: 583589.
  • 16
    LaVeist TA. Beyond dummy variables and sample selection: what health services researchers ought to know about race as a variable. Health Serv Res 1994; 29: 116.
  • 17
    Furth SL, Hwang W, Yang C, Neu AM, Fivush BA, Powe NR. Relation between pediatric experience and treatment recommendations for children and adolescents with kidney failure. JAMA 2001; 285: 10271033.
  • 18
    Liang KY, Zeger SL. Regression analysis for correlated data. Annu Rev Public Health 1993; 14: 4368.
  • 19
    Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998; 280: 16901691.
  • 20
    Laederach-Hofmann K, Bunzel B. Noncompliance in organ transplant recipients: a literature review. Gen Hosp Psychiatry 2000; 22: 412424.
  • 21
    Eisenberg JM. Sociologic influences on decision-making by clinicians. Ann Intern Med 1979; 90: 957964.
  • 22
    Van Ryn M, Burke J. The effect of patient race and socio-economic status on physicians' perceptions of patients. Soc Sci Med 2000; 50: 813828.
  • 23
    Asch DA, Jedriziewski MK, Christakis NA. Response rates to mail surveys published in medical journals. J Clin Epidemiol 1997; 50: 11291136.
  • 24
    Furth SL, Powe NR, Hwang W, Neu AM, Fivush BA. Does pediatric specialization influence treatment choice in chronic disease Management? Dialysis Modality Choice for children with ESRD. Arch Pediatr Adolesc Med 1997; 151: 545550.
  • 25
    Peabody JW, Luck J, Glassman P, Dresselhaus TR, Lee M. Comparison of vignettes, standardized patients, and chart abstraction. a prospective validation study of 3 methods for measuring quality. JAMA 2000; 283: 17151722.
  • 26
    Bunzel B, Laederach-Hofmann K. Solid organ transplantation: are there predictors for posttransplant noncompliance? A literature overview. Transplantation 2000; 70: 711716.