[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
January 2007

Disparities in Pediatric Preventive Care in the United States, 1993-2002

Author Affiliations

Author Affiliations: Denver Community Health Services, Denver, Colo (Drs Hambidge and Federico); Departments of Pediatrics (Drs Hambidge and Federico), Family Medicine (Ms Emsermann), and Internal Medicine (Dr Steiner), and Colorado Health Outcomes Program (Drs Hambidge and Steiner), University of Colorado School of Medicine, Denver.

Arch Pediatr Adolesc Med. 2007;161(1):30-36. doi:10.1001/archpedi.161.1.30

Objectives  To quantify physician-reported preventive counseling and screening during well-child visits (WCVs) and to examine racial and ethnic disparities in these activities.

Design  Cross-sectional study using the National Ambulatory Medical Care Survey, January 1993 through December 2002.

Setting  Office-based physician practices.

Participants  Children from birth to 18 years old who were seen by a physician for a WCV.

Main Outcome Measure  Preventive counseling and screening.

Results  Well-child visits were shorter for Latino children than for white or black children. At WCVs, white children were more likely to receive preventive counseling than were black or Latino children (72% vs 61% vs 61%, respectively; P = .01) but not more likely to receive screening for elevated blood pressure, anemia, vision and hearing acuity, or lead toxicity. There were no differences in secondary diagnoses made at WCVs for white, black, or Latino children (15% vs 17% vs 14%, respectively; P = .65). The children who received the least counseling were Latino children in the public sector non–health maintenance organization setting (counseled at 39% of visits) and Latino children who self-paid for the visits (counseled at 26% of visits). After adjusting for possible confounders, including medications prescribed at the visit, black and Latino children were less likely to receive counseling than were white children (odds ratios, 0.68 and 0.63; 95% confidence interval, 0.48-0.97 and 0.44-0.90, respectively), and black children were less likely to receive preventive screening services (odds ratios, 0.65; 95% confidence interval, 0.45-0.93).

Conclusions  By physician report in a nationally representative sample, black and Latino children received less counseling at WCVs than did white children. These disparities were unexplained by the competing demands of other secondary diagnoses or medications prescribed or dispensed.

A large body of work has delineated racial and ethnic disparities in medicine, as highlighted in a report from the Institute of Medicine in 2003 that stated, “Racial and ethnic disparities in health care exist and, because they are associated with worse outcomes in many cases, are unacceptable.”1(p6) However, research on disparities in health care in children is sparse, especially about the quality of preventive services. Given the centrality of the preventive visit to quality pediatric health care, the purposes of this study were to quantify physician-reported preventive counseling and screening during well-child visits (WCVs) from a nationally representative sample and to examine racial and ethnic disparities in the receipt of counseling and screening services at those visits.

Study design and data source

We conducted a cross-sectional study using data from the National Ambulatory Medical Care Survey (NAMCS) for the 10 years from January 1993 through December 2002. The NAMCS, conducted annually by the National Center for Health Statistics, is a national survey designed to provide reliable information about the use of ambulatory medical care services in the United States.2

To derive representative national estimates, the NAMCS uses a multistage probability sample design. A sample of approximately 2500 non–federally employed physicians, identified from the master files of the American Medical Association and the American Osteopathic Association, is screened annually; approximately 700 physicians are ineligible because of retirement or employment primarily in teaching, research, or administration. Of eligible physicians, approximately 69% participate annually.2 Participating physicians complete a patient record form for a systematic random sample of office visits during a randomly assigned 1-week reporting period. Before this reporting period, specially trained NAMCS field representatives meet with participating physicians and their staffs to explain proper completion of the survey instrument. The final database contains information from approximately 24 000 patient visits annually.

Data extraction and variable definition

We defined a WCV using the following criteria: age 18 years or younger; primary physician diagnosis code V20.2, routine infant or child health check (International Classification of Diseases, Ninth Revision, Clinical Modification); major reason for visit, nonillness care; visit to a primary care physician; and the child was seen by a physician during the visit, to exclude nurse-only visits in which vaccines were administered but no other preventive health care was delivered. A primary care physician was defined as a pediatrician, family physician, or general internist; all visits to pediatric specialists were excluded. We analyzed age in 4 clinically relevant categories: infants and toddlers (age, birth to 2 years), preschool-aged children (3-5 years), school-aged children (6-12 years), and adolescents (13-18 years).

Children's race and ethnicity were defined using the terms on the NAMCS survey: white, black, or Latino. Other racial categories were coded together as “other” because of low visit numbers and were excluded from the main analysis. Physicians and their staffs who complete the NAMCS form are instructed to report race and ethnicity based on “observation or the physician's of the patient”2(p12) We analyzed 4 mutually exclusive racial and ethnic groups: non-Latino white, non-Latino black, Latino, and other non-Latino. The public NAMCS data files contain no data on language preference of the patient, or race or ethnicity of the physician.

Sources of health insurance categories included the following: private, public (predominantly Medicaid), uninsured (self-pay or no charge), and unknown (other, unknown, and blank). health maintenance organization (HMO) status was determined directly from question 10 on the survey: “Does patient belong to an HMO?” In addition, we used the NAMCS office setting variable characterized as follows: private office (either private solo or group practice), clinic (freestanding clinic, federally qualified health center, non–federal government clinic, family planning clinic, or non–hospital faculty practice plan), HMO, and other (other, unknown, and blank). We determined whether a practice was urban or nonurban using the metropolitan statistical area, based on the standard definition of the US Census Bureau.

The provision of counseling or patient education services was determined from the “therapeutic and preventive services” item on the survey form. This item instructs the physician to “Check ALL appropriate boxes for any of the following types of counseling, advice, education, instructions, or recommendations to the patient that were provided or ordered during the visit.”2 The following topics have individual boxes to check: diet/nutrition, exercise, human immunodeficiency virus/sexually transmitted disease transmission, family planning and contraception, prenatal instructions, breast self-examination, tobacco use or exposure, growth and development, mental health, stress management, skin cancer prevention, and injury prevention. There is also a text box to record counseling in any area not included in these topics. We considered counseling to have occurred if any of these boxes were checked. In addition, we specifically analyzed 5 topic areas that are especially important in pediatrics, as follows: growth and development, injury prevention, exercise, diet, and tobacco exposure. For screening services, we analyzed the 3 tests with the most clinical relevance to children of all ages, that is, blood pressure, and vision and hearing acuity, as well as 2 screening tests that are more age-specific, hematocrit and blood lead levels.

We hypothesized that racial and ethnic inequities in receipt of counseling or screening, if they occurred, could be accounted for by competing demands at the WCV. For example, a child who has an acute illness or a chronic condition at a WCV may require more physician time, leaving less time for preventive counseling. To examine possible competing demands using the NAMCS, we derived data on medications given or prescribed at the WCV from the “medications/injections” box on the patient record form. In addition, we used the “physician's diagnosis for this visit” box to derive any secondary diagnoses made by the physician in addition to the primary diagnosis of well-child care. The NAMCS does not contain any additional data fields that facilitate analysis of severity of illness or comorbidity.

Nonresponse rates were less than 5% for most items on the survey form. The National Center for Health Statistics imputes data from patients with similar characteristics for the following items2: race (18% nonresponse) and time spent with physician (16%). The ethnicity box had a 17% nonresponse rate in our sample, and because of the centrality of ethnicity to our study we excluded all of these visits from our analysis. The following variables were not included in the NAMCS in certain years and were, therefore, excluded from the analysis in those years: injury prevention (missing 2001-2002), diet and nutrition (1993-1994), tobacco cessation (1993-1994), visual acuity (1993-1994 and 2001-2002), hearing acuity (1993-1996), anemia (1993-1996), lead toxicity (1993-1994 and 2001-2002), saw physician at visit (1993-1994), and office setting (missing 1993-1996). Of note, the nonresponse rate for the HMO item was 13%.

Calculation of rates

National estimates of WCV rates were obtained by using the assigned patient visit weight in the NAMCS database to calculate the numerators. The weights are derived from the probability of being sampled and are adjusted for nonresponse.2 Denominators were calculated using national population estimates from the National Health Interview Survey (NHIS).3 Because NHIS data were not available in the same format for the years 1993 to 1996, our calculated WCV rates do not include those years. Population rates using NHIS and NAMCS data were restricted to the years 1997 to 2002 and ages birth to 18 years when race and ethnicity of the respondent were white non-Latino, black, or Latino. Because the NHIS definition of race or ethnicity differs from that of the NAMCS in how “other” race is defined, we did not calculate a rate for this variable. In general, the NHIS and NAMCS variables for age, sex, race or ethnicity, metropolitan area, region, and survey year corresponded. However, NHIS defines insurance in separate categories, whereas NAMCS has just 1 insurance category. To create corresponding insurance categories between the 2 surveys, we determined NHIS categories using a hierarchical approach: children who reported private insurance were classified as such, while children with State Children’s Comprehensive Health Insurance Program, Medicare, or other public insurance and no private coverage were classified as publicly insured. Children who reported no coverage were classified as uninsured. Because of the discrepancies in reporting between the NAMCS and NHIS surveys, the “unknown” insurance category could not be assessed. Because rates were estimated using 2 separate databases, standard errors for each rate were calculated in accordance with the formulas for the variance of a product and quotient.4

Statistical analysis

To measure the unadjusted association between independent variables (race and ethnicity stratified by insurance type and HMO status) and outcome variables (receipt of counseling, screening, or medications), we performed χ2 tests. To determine differences between the mean duration of a visit by patient race or ethnicity, pairwise comparisons were performed. To determine general adjusted associations by race or ethnicity and HMO status, we performed the Cochran-Mantel-Haenszel χ2 test. We determined associations for all analyses at an α level of .05 (2-sided).

For multivariate analysis, since NAMCS survey questions pertaining to HMO status were asked differently during the 1993-2002 period of our study, we determined whether a child was in an HMO (yes or no) by the following data fields for each survey year: insurance field “HMO” (1993-1994); insurance field “typepay, 5” (1995-1996); HMO field, “Does patient belong to an HMO?” (1997-2000); and the office setting field “retypoff, 7” (2001-2002). For bivariate, stratified bivariate, and multivariate analyses, we used SUDAAN statistical software (Survey Data Analysis, version 9.1; Research Triangle Institute, Research Triangle Park, NC) to account for the multistage weighting scheme of the NAMCS.

Independence of associations was tested using multivariate logistic regression and is presented as adjusted odds ratios (ORs) with 95% confidence intervals (CIs), adjusting for the following covariates: sex, age, metropolitan statistical area, region of the United States, health insurance, physician specialty, medications prescribed at visit, and office setting. In addition, the NAMCS sample year was included as a fixed effect to account for possible changes with time, such as modifications of preventive service guidelines. Covariates were included in the final model only if hypothesized a priori to be potential confounders or, as in the case of medications, if they had a significant bivariate association with race or ethnicity (P<.20). Interactions between covariates that were of potential clinical interest were examined; none were statistically significant and, therefore, none were included in the final multivariate models. There was some evidence of collinearity between HMO and insurance status, but because the HMO β coefficients and standard errors did not change substantially in models both with and without insurance, we included both HMO status and health insurance in the final model. Because this study analysis used preexisting national data sets, it was classified as exempt from institutional board review.


Children in the United States aged from birth to 18 years made more than 248 million WCVs from January 1993 through December 2002. More than 234 million of these visits (94%) were for children identified as white, black, or Latino. Reflecting the age-specific WCV recommendations of the American Academy of Pediatrics and the Maternal and Child Health Bureau,5,6 young children had the greatest rate of preventive health care visits (Table 1). Children identified as white had a WCV rate of 30 per 100 children per year, compared with 25 and 34 WCVs per 100 children per year for children identified as black or Latino, respectively.

Table 1. 
Characteristics of Children in NAMCS Well-Child Care Sample, 1997-2002*
Characteristics of Children in NAMCS Well-Child Care Sample, 1997-2002*

At bivariate analysis, receipt of preventive counseling differed by race and ethnicity, with children identified as black or Latino receiving counseling at 61% of visits, compared with 72% for white children (P = .01; Table 2). Most of this difference was accounted for by less counseling for minority children in growth and development (46%-47% of visits vs 62% for white children; P<.001). There were no significant differences in screening services at bivariate analysis. Latino children had WCVs that were shorter, on average (15.9 minutes), than for non-Latino white or black children (17.3 and 17.6 minutes, respectively; P<.05 for both comparisons).

Table 2. 
Percentage of Well-Child Visits With Counseling or Screening, and Visit Duration*
Percentage of Well-Child Visits With Counseling or Screening, and Visit Duration*

The differences in counseling among children of different racial and ethnic groups were more pronounced in children with public insurance compared with private insurance and in children not enrolled in HMOs compared with those enrolled in HMOs (Table 3). Receipt of preventive counseling did not differ among children of different racial and ethnic groups for children enrolled in HMOs, whether the child had private or public insurance. The children who received the least counseling were Latino children in the public sector non-HMO setting (counseled at 39% of visits) and Latino children whose parents self-paid for WCV (counseled at 26% of visits). For all children, the racial and ethnic differences in counseling occurred regardless of whether the child had been seen in the practice before (P = .03, data not shown). Black and Latino children were more likely to receive a medication or injection at WCVs than white children were (Table 4). However, the number of secondary diagnoses did not differ by race or ethnicity.

Table 3. 
Percentage of Well-Child Visits With Any Counseling, Stratified by Insurance Type and HMO Status
Percentage of Well-Child Visits With Any Counseling, Stratified by Insurance Type and HMO Status
Table 4. 
Percentage of 196 Million Well-Child Visits at Which a Medication or Injection Was Ordered or a Secondary Diagnosis Was Made*
Percentage of 196 Million Well-Child Visits at Which a Medication or Injection Was Ordered or a Secondary Diagnosis Was Made*

At multivariate analysis, black and Latino children were less likely to receive counseling and black children were less likely to undergo screening compared with white children (Table 5). Differences in counseling or screening were noted for different age groups, in different regions of the United States, with different physician specialties, and when medications were dispensed.

Table 5. 
Multivariate Analysis of Factors Associated With Counseling and Screening During Well-Child Visits by US Children
Multivariate Analysis of Factors Associated With Counseling and Screening During Well-Child Visits by US Children

In this analysis of data for 10 years from a nationally representative sample of primary care office visits, we did not find that WCV rates differed by race or ethnicity. However, disparity in access to care was not the focus of our study and has been previously shown in other work.7-11 The duration of a WCV was 10% shorter for Latino children than for white or black children; the shorter time for Latino children may represent a language barrier because families with limited English proficiency experience some of the worst disparities in primary care.12-18 Furthermore, Latino parents have been shown to be more likely to report that pediatric providers did not spend enough time with their child during the last WCV.14

Physicians reported, on average, counseling at 11% to 15% fewer visits for minority children than for white children. These differences were greater for children with public health insurance, with black children receiving 22% less and Latino children receiving 26% less counseling than non-Latino white children. Half of all publicly insured minority children received no preventive counseling. Children who are most vulnerable are, by physician report, receiving less counseling at WCVs. This finding suggests that interventions to address this disparity be directed first to those children of lowest socioeconomic status.

The disparity in counseling for children of different racial and ethnic groups was not seen for Latino children who received care in an HMO setting. This finding was true for children with public as well as private insurance, suggesting that it was not simply due to examining children of better socioeconomic status. There is little published data on the effect of managed care on preventive counseling at pediatric visits. In adults, racial or ethnic disparities in counseling are ameliorated in the managed care setting,19,20 although this finding is not universal.21 Attributes of managed care organizations that might reduce disparities include integrated medical information systems (including electronic prompts and reminders) and systemwide standards of practice. It is also possible that this finding may reflect HMO enrollment of children from families with higher educational achievement or other attributes that result in a different physician-family dynamic.

At multivariate analysis, black and Latino children were 32% to 37% less likely to receive counseling than were white children, and black children were 35% less likely to receive screening services. There are many possible reasons for these differences. First, the racial or ethnic differences may reflect differences in socioeconomic status. Although we could account for health insurance status, the NAMCS has no direct measure of socioeconomic factors such as family income, place of residence, and parental educational achievement and employment status. Second, it is possible that the population of physicians who care for many minority children is different from those who do not. For example, among Medicare beneficiaries, most primary care visits by black adults were to a group of physicians who did not see many white patients.22 Those physicians were less likely to be board certified and more likely to report barriers to providing appropriate care than their counterparts who took care of predominantly white patients. Although the race or ethnicity of the physician may have a role, there is evidence that in pediatrics, in contrast to adult medicine, patient-provider race or ethnicity concordance is not associated with parental perceptions of the quality of children's primary care experiences.23,24 Third, the observed differences in counseling at WCVs could be the result of competing demands25 for other services during the WCV. Minority children were somewhat more likely to receive a medication or injection at the WCV, but there were no significant racial or ethnic differences in receipt of antimicrobial agents or in the number of visits with a secondary diagnosis. Furthermore, in multivariate models that included receipt of medications as a covariate, the disparities in counseling and screening services persisted. Therefore, this competing demand did not explain the differential receipt of preventive counseling. Perhaps the most troubling explanation for the disparities observed in our study is the possibility, as raised in the Institute of Medicine report, of “health provider bias—conscious or unconscious, individual or institutional.”26(p418)

Our study raises 2 key questions: Why do these disparities exist? and What can be done to eliminate them? The possible causes of disparities in these pediatric preventive services include communication barriers, lack of physician cultural competence and difficulties in cross-racial or ethnic physician-patient communication,27 biased attitudes, socioeconomic issues, or some other factor. Addressing the disparities found in this study will require an increased recognition of the problem, and possibly training modules for residents who will care for children and continuing medical education modules for practicing physicians.

Except for Latino children, the duration of a WCV was similar to that reported by parents on the 2000 National Survey of Early Childhood Health, which reported a mean visit duration of 17.7 minutes (95% CI, 17.1-18.4).28 Other studies of the effect of race and ethnicity on receipt of pediatric preventive services have found either no difference in basic preventive services29 or evidence for increased counseling in some topics among minority families.14 The latter study found increased counseling of minority families in areas of household violence and substance abuse, which may represent racial or ethnic stereotyping. Both of these studies relied on data from the National Survey of Early Childhood Health, which uses parental report from a nationwide telephone survey. Physicians in our study reported any counseling at 66% and any screening at 59% of WCVs, rates that are lower than in studies using direct parental and physician surveys.28-30 Differences between the NAMCS and National Survey of Early Childhood Health likely reflect the different designs of the 2 surveys: the NAMCS relies on a survey form that contains boxes to check for individual counseling items and is filled out by the participating physician or nurse, whereas the National Survey of Early Childhood Health is a telephone survey of households performed in 2000. Both surveys are subject to recall or selection bias.

While it is possible that physicians participating in the NAMCS underreported counseling and screening activities overall, underreporting should be nonselective by race and ethnicity. Therefore underreporting should not account for counseling at fewer visits for minority children than for white children, a difference driven by racial and ethnic disparities in counseling on growth and development. The 2004 National Healthcare Disparities Report did not find many differences in childhood screening and counseling by race and ethnicity, based on data from the 2001 Medical Expenditure Panel Survey.31 That survey, however, does not include a question on growth and development counseling and, thus, would not elicit this disparity.

The biggest limitation in this study is the lack of any direct measure of socioeconomic measure in the NAMCS. Although the disparities we found were seen regardless of health insurance status, we could not assess other important unmeasured socioeconomic determinants.32 Recent evidence suggests that social disadvantage, as determined by poverty, low parental educational achievement, and single-parent households, may have a more profound effect on child health than race or ethnicity per se.33 In addition to poverty, the presence of an underlying chronic health condition can have a major effect on use of health services. However, the NAMCS does not contain information on chronic health conditions other than diagnoses made at the visit. Additional limitations include the lack of data on primary family language and race of the physician. In addition, the NAMCS samples office-based practices but does not include hospital-based practices, which include most of the child health clinics found at large academic children's hospitals in the United States and which serve a disproportionate number of poor and minority children.

In summary, we found evidence in this nationally representative sample of racial and ethnic disparities in the quality of well-child care. Equitable care is one dimension of quality health care,34 since disparities in health care are, in essence, disparities in quality.35-37 Efforts not only to improve access to care for all children but to improve the quality of care for all children once they have access will go a long way toward eliminating disparities in pediatric primary care.

Back to top
Article Information

Correspondence: Simon J. Hambidge, MD, PhD, Mailcode 0132, Denver Health Medical Center, 777 Bannock St, Denver, CO 80204 (simon.hambidge@uchsc.edu).

Accepted for Publication: July 26, 2006.

Author Contributions: Dr Hambidge had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Hambidge and Steiner. Acquisition of data: Hambidge. Analysis and interpretation of data: Hambidge, Emsermann, Federico, and Steiner. Drafting of the manuscript: Hambidge, Federico, and Steiner. Critical revision of the manuscript for important intellectual content: Hambidge, Emsermann, Federico, and Steiner. Statistical analysis: Hambidge, Emsermann, and Steiner. Obtained funding: Hambidge, Emsermann, Federico, and Steiner. Administrative, technical, and material support: Federico. Study supervision: Hambidge.

Financial Disclosure: None reported.

Funding/Support: Drs Hambidge and Federico were recipients of Primary Care Research Faculty fellowships, grants D14HP00153 and D55HP05157 from the US Department of Health and Human Services, Health Resources and Services Administration, during part of this project. Dr Hambidge was supported, in part, by a Generalist Physician Faculty Scholars award from the Robert Wood Johnson Foundation. Ms Emsermann and Dr Steiner were supported, in part, by grant D54HP00054 from the US Department of Health and Human Services, Health Resources and Services Administration, for a Collaborative Primary Care Research Unit.

Role of the Sponsors: Sponsors and funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript.

Previous Presentation: Preliminary results from this study were presented as a platform session at the annual meeting of the Pediatric Academic Societies; May 7, 2002; Baltimore, Md.

Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care; Institute of Medicine,Smedley  BDedStith  AYedNelson  ARed Summary.  Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC National Academies Press2003;1- 28Google Scholar
Ambulatory Care Statistics Branch; National Center for Health Statistics, NAMCS Micro-data file documentation. 2000;ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Dataset_Documentation/NAMCS/doc00.pdfAccessed October 15, 2006
Botman  SLMoore  TFMoriarity  CLParsons  VL Design and estimation for the National Health Interview Survey, 1995-2004.  Vital Health Stat 2000;21- 31Google Scholar
Mood  AMedGraybill  FAedBoes  DCed Introduction to the Theory of Statistics.  New York, NY McGraw-Hill Co1974;180- 181
American Academy of Pediatrics, Committee on Practice and Ambulatory Medicine, Recommendations for preventative pediatric health care (policy statement).  Pediatrics 2000;105645- 646Google ScholarCrossref
Green  MedPalfray  JSed Bright Futures: Guidelines for Health Supervision of Infants, Children, and Adolescents. 2nd ed. Arlington, Va National Center for Education in Maternal and Child Health2002;
Flores  GAbreu  MOlivar  MAKastner  B Access barriers to health care for Latino children.  Arch Pediatr Adolesc Med 1998;1521119- 1125PubMedGoogle ScholarCrossref
Lieu  TANewacheck  PWMcManus  MA Race, ethnicity, and access to care among US adolescents.  Am J Public Health 1993;83960- 965PubMedGoogle ScholarCrossref
Newacheck  PWHughes  DCStoddard  JJ Children's access to primary care: differences by race, income, and insurance status.  Pediatrics 1996;9726- 32PubMedGoogle Scholar
Holl  JLSzilagyi  PGRodewald  LEByrd  RSWeitzman  ML Profile of uninsured children in the United States.  Arch Pediatr Adolesc Med 1995;149398- 406[published correction appears in Arch Pediatr Adolesc Med. 1995;149:763]PubMedGoogle ScholarCrossref
Brousseau  DCHoffmann  RGYauck  JNattinger  ABFlores  G Disparities for Latino children in the timely receipt of medical care.  Ambul Pediatr 2005;5319- 325PubMedGoogle ScholarCrossref
Flores  G Culture, ethnicity, and linguistic issues in pediatric care: urgent priorities and unanswered questions.  Ambul Pediatr 2004;4276- 282PubMedGoogle ScholarCrossref
Ku  LFlores  G Pay now or pay later: providing interpreter services in health care.  Health Aff (Millwood) 2005;24435- 444PubMedGoogle ScholarCrossref
Flores  GOlson  LTomany-Korman  SC Racial and ethnic disparities in early childhood health and health care.  Pediatrics 2005;115e183- e193PubMedGoogle ScholarCrossref
Weech-Maldonado  RMorales  LSSpritzer  KElliott  MHays  RD Racial and ethnic differences in parents' assessments of pediatric care in Medicaid managed care.  Health Serv Res 2001;36575- 594PubMedGoogle Scholar
Morales  LSElliott  MWeech-Maldonado  RHays  RD The impact of interpreters on parents' experiences with ambulatory care for their children.  Med Care Res Rev 2006;63110- 128PubMedGoogle ScholarCrossref
Flores  GAbreu  MTomany-Korman  SC Limited English proficiency, primary language at home, and disparities in children's health care: how language barriers are measured matters.  Public Health Rep 2005;120418- 430PubMedGoogle Scholar
Cohen  ALChristakis  DA Primary language of parent is associated with disparities in pediatric preventive care.  J Pediatr 2006;148254- 258PubMedGoogle ScholarCrossref
Fiscella  KFranks  PDoescher  MPSaver  BG Do HMOs affect educational disparities in health care?  Ann Fam Med 2003;190- 96PubMedGoogle ScholarCrossref
DeLaet  DEShea  SCarrasquillo  O Receipt of preventive services among privately insured minorities in managed care versus fee-for-service insurance plans.  J Gen Intern Med 2002;17451- 457PubMedGoogle ScholarCrossref
Fiscella  KFranks  P Is patient HMO insurance or physician HMO participation related to racial disparities in primary care?  Am J Manag Care 2005;11397- 402PubMedGoogle Scholar
Bach  PBHoangmai  HPSchrag  DTate  RCHargraves  JL Primary care physicians who treat blacks and whites.  N Engl J Med 2004;351575- 584PubMedGoogle ScholarCrossref
Stevens  GDShi  LCooper  LA Patient-provider racial and ethnic concordance and parent reports of the primary care experiences of children.  Ann Fam Med 2003;1105- 112PubMedGoogle ScholarCrossref
Stevens  GDMistry  RZuckerman  BHalfon  N The parent-provider relationship: does race/ethnicity concordance or discordance influence parent reports of the receipt of high quality basic pediatric preventive services?  J Urban Health 2005;82560- 574PubMedGoogle ScholarCrossref
Jaen  CRStange  KCNutting  PA Competing demands of primary care: a model for the delivery of clinical preventive services.  J Fam Pract 1994;38166- 171PubMedGoogle Scholar
Geiger  HJSmedley  BDedStith  AYedNelson  ARed Racial and ethnic disparities in diagnosis and treatment: a review of the evidence and a consideration of causes.  Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC National Academies Press2003;415- 454Google Scholar
Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care; Institute of Medicine,Smedley  BDedStith  AYedNelson  ARed Interventions: cross-cultural education in the health professions.  Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, DC National Academies Press2003;199- 214Google Scholar
Olson  LMInkelas  MHalfon  NSchuster  MAO’Connor  KGMistry  R Overview of the content of health supervision for young children: reports from parents and pediatricians.  Pediatrics 2004;1131907- 1916PubMedGoogle Scholar
Zuckerman  BStevens  GDInkelas  MHalfon  N Prevalence and correlates of high-quality basic pediatric preventive care.  Pediatrics 2004;1141522- 1529PubMedGoogle ScholarCrossref
Bethell  CReuland  CHPHalfon  NSchor  EL Measuring the quality of preventive and developmental services for young children: national estimates and patterns of clinician's performance.  Pediatrics 2004;1131973- 1983PubMedGoogle Scholar
Agency for Healthcare Research and Quality; US Department of Health and Human Services, 2004 National Healthcare Disparities Report. Rockville, MdAHRQ publication 05-00142004;http://www.ahrq.gov/qual/nhdr05/nhdr05.htmAccessed October 15, 2006
Braveman  PACubbin  CEgerter  S  et al.  Socioeconomic status in health research: one size does not fit all.  JAMA 2005;2942879- 2888PubMedGoogle ScholarCrossref
Bauman  LJSilver  EJStein  REK Cumulative social disadvantage and child health.  Pediatrics 2006;1171321- 1328PubMedGoogle ScholarCrossref
Committee on Quality Health Care in America; Institute of Medicine, Crossing the Quality Chasm: A New Health System for the 21st Century.  Washington, DC National Academies Press2000;http://www.nap.edu/books/0309072808/html/Accessed October 15, 2006
Beal  AC Policies to reduce racial and ethnic disparities in child health and health care.  Health Aff 2004;23171- 179Google ScholarCrossref
Andrulis  DP Moving beyond the status quo in reducing racial and ethnic disparities in children's health.  Public Health Rep 2005;120370- 377PubMedGoogle Scholar
Fiscella  KFranks  PGold  MRClancy  CM Inequality in quality: addressing socioeconomic, racial, and ethnic disparities in health care.  JAMA 2000;2832579- 2584PubMedGoogle ScholarCrossref