Sociation Today ® 
The Official 
Journal of 
The North 
Carolina 
Sociological 
Association: A 
Peer-Reviewed
Refereed Web-Based 
Publication 
ISSN 1542-6300
Editorial Board:
Editor:
George H. Conklin,
 North Carolina
 Central University

Board:
Bob Davis,
 North Carolina
 Agricultural and
 Technical State
 University

Richard Dixon,
 UNC-Wilmington

Ken Land,
 Duke University

Miles Simpson,
 North Carolina
 Central University

Ron Wimberley,
 N.C. State University

Robert Wortham,
 North Carolina
 Central University


Editorial Assistant

John W.M. Russell,
 Technical
 Consultant

Submission
Guidelines
for Authors


Cumulative
Searchable Index
of
Sociation Today
from the
Directory of 
Open Access
Journals (DOAJ)


Sociation Today
is abstracted in
Sociological Abstracts
and a member
of the EBSCO
Publishing Group


The North
Carolina
Sociological
Association
would like
to thank
North Carolina
Central University
for its
sponsorship of
Sociation
Today


*® 

Volume 6, Number 2
Fall 2008

Healthcare Services Used by Appalachians in
Southwest Virginia*

by
Marion R. Manton
Christopher Newport University

and

Gwendolyn B. Thornton
Virginia State University

Introduction

    Although access to healthcare has improved nationally, healthcare services remain a critical issue for people living in rural areas (National Center for Health Statistics 2006).  Serious concerns exist about the health status of rural poor and vulnerable populations (Broyles, McAuley, and Baird-Holmes 1999; Ormond, Zuckerman and Lhila 2000).  Consequently, high quality healthcare services have been identified as a top priority by leaders of state agencies and local public health agencies (Gamm, Hutchison and Bellamy 2002). 

    Past studies indicate that people living in rural communities experience less than optimal preventative care and treatment regimes.  Rural communities have less contact with physicians, lower levels of prenatal care, and lower levels of disease screenings (Harris and Leininger 1993).  They also have minimal access to clinical trials (Sateren et al. 2003). 

    Research on sources of healthcare for rural populations has been increasing.  However, very limited empirical research addresses the healthcare choices of the people of Appalachia (Huttlinger, Schaller-Ayers and Lawson 2004).  This work will add to the body of existing literature by examining the primary healthcare providers used by Appalachians in southwest Virginia: a rural and remote area with a population of approximately 400,000 people situated in the heart of the Appalachia Mountains.  West Virginia is considered one of the most economically disadvantaged regions in the state with unemployment and poverty rates significantly above the national median and is reported to have a high incidence of preventable chronic health problems (Graduate Medical Education Consortium 2002; National Center for Health Statistics 2006; West Virginia Bureau of Public Health 1994). 

Theoretical Background

    This research is guided by the Aday and Anderson (1974) model of health care service utilization.  The model proposes that consumer participation in health care may be explained by three sets of factors: 

  1. predisposing, 
  2. enabling, 
  3. needs. 
Predisposing factors generally consist of demographic characteristics, enabling factors are concerned with the resources available that allow one to receive care, and needs factors refer to illnesses and injuries experienced by the individual.  Anderson (1995) broadens his model to explain healthcare use as equitable or inequitable.  He broadly defines equitable access as occurring when demographic and needs factors drive use of health care services.  Inequitable access occurs when social structure and enabling resources determine healthcare. 

Predisposing Factors

    The most common predisposing factors related to healthcare use are gender and race.  Bernstein et al. (2003) found that women generally tend to have greater contact with physicians, whereas men are more likely to use emergency room facilities.  Studies have also shown that black people tend to use hospital outpatient departments while whites are more likely to use a physician's office as their primary source of care (Shi 1999; Smedley et al.).  In this study, we also include marital status as a predisposing factor since family support may encourage sick family members to seek medical help.

Enabling Factors

    The most critical enabling factor is the ability to pay for healthcare (Anderson and Aday 1978; Anderson 1995).  Higher incomes facilitate access to medical facilities because they allow individuals to pay for health insurance coverage or for health services.  Low income families must prioritize: basic necessities are purchased before health insurance (Budetti et al. 1999; Rice et al. 2002).

    Research consistently shows the importance of health insurance as a predictor of health care utilization. (Blumberg and Liska 1996; Faulkner and Schauffler 1997: Hangdrup et al. 1997).  As a rule, services that are covered by insurance are more likely to be used than services that must be paid for by consumers (Bernstein et al. 2003). 

    Other important considerations include education and access to transportation.  Individuals with lower levels of education may have a limited understanding of the necessity of medical intervention (LeClere et al. 1994), and those who have limited access to public transportation, or who must depend on others for transportation, find it difficult both to schedule healthcare visits and to attend their appointments. 

Needs Factors 

    Finally, the Aday and Anderson (1974) model of healthcare providers proposes that individuals with the greatest healthcare needs are more likely to seek out healthcare services.  This might include conditions diagnosed by a health professional, or health issues recognized by the individual.

Research Methods and Data

    This exploratory study uses a convenience sample of 139 men and women  who attended a Remote Area Medical (RAM) Health Expedition held in Wise County, Virginia, July, 2005.  The annual RAM event provides free medical, dental and eye care.  The 55 item questionnaire used for the study was developed in consultation with the RAM Interdisciplinary Planning Team.  Data was collected using face-to-face interviews.  Each interview lasted approximately 40 minutes.  Despite the limitations associated with a convenience sample, we feel that the insights drawn from this work will prove useful to the development of future research.

Measures

Health Care Utilization

    In our study we distinguish among three types of healthcare use:

  1. emergency room,
  2. physician, and 
  3. other health care services. 
To measure healthcare utilization respondents were asked the question "What services do you most frequently use as your primary provider for medical care?"   Responses included emergency room, physician, free healthcare services, and "other" sources of healthcare.  The "other" category includes religious leaders such as a minister or pastor, and various other members of the community who offer cultural, or herbal remedies.   All responses other than emergency room and physician were categorized as "other healthcare services."

   We created three sets of control variables in order to incorporate the three sets of factors predicted by Aday and Anderson to impact healthcare use: 

  1. predisposing factors, 
  2. enabling factors, and 
  3. needs factors
Predisposing Factors

    This study uses three measures of predisposing factors: 

  1. gender, 
  2. race, and 
  3. marital status. 
Gender.  The reference category for this variable is male.

Race.  Since 91.4% of the sample is white, the remaining racial categories were combined to form "other race."  The reference category for this variable is "other race."

Marital Status.  As measures of marital status we create three dummy variables: currently married, previously married, and never married.  Currently married is the reference category.

Enabling Factors

    We include three measures of enabling factors: 

  1. health insurance, 
  2. education, and 
  3. transportation. 
Health Insurance. We use three dummy variables for health insurance coverage: public health insurance, private health insurance and no health insurance.  Individuals were classified as having public health insurance if they reported using Medicare, Medicaid, State Children's Health Insurance Program (SCHIP), state-sponsored or other government-sponsored health plan, or military plan.  People were classified as having private health insurance if they self-report participation in a private health insurance plan.  A person was defined as uninsured if he or she did not have any private health insurance or public health insurance.  Private insurance is the reference category.

Education.  We use three dummy variables to measure respondent's education: less than high school, high school, and at least some college.  At least some college is the reference category.

Transportation is measured 1 if the respondent indicated that he or she had problems with transportation, else=0.

Needs Factors

    We include three measures of needs factors: 

  1. diagnosed medical condition, 
  2. alcohol problem, and 
  3. drug problem. 
Diagnosed medical condition is measured with the question "Do you currently have a diagnosed physical/emotional/medical condition?"  "Yes" response coded 1, "no" response coded 0.

Alcohol problem is measured with the question "Are you struggling with an alcohol problem?" "Yes" response coded 1, "no" response coded 0.

Drug problem is measured with the question "Are you struggling with a drug problem?"  "Yes" response coded 1, "no" response coded 0.

Analytic Approach

    We begin by presenting means and percentages for the social characteristics of respondents in our sample.  We move on to present a bivariate analyses showing the relationship between the three types of healthcare utilized (emergency room, physician, and other less traditional healthcare services) and the three sets of factors predicted by Aday and Anderson to impact healthcare use (predisposing factors, enabling factors, and needs factors).

    Finally, multivariate models were employed to determine whether the bivariate relationships remain after simultaneously controlling for all of the factors.  We present three sets of logistic regression models.  Model 1 presents log-odds ratios for emergency room care. Model 2 presents log-odds ratios for physician care.  Model 3 presents log-odds for less traditional forms of healthcare. 

Results

Sample Characteristics

    Table 1 indicates that around 32 percent of people report using the emergency room as their primary source of healthcare, around 42 percent use a physician service, and 27 percent turn to other healthcare services.

Table 1
Selected Characteristics of Sample

 
N
%
Mean (sd)
Emergency Room Primary
38
31.7
 
Physician Care Primary
50
41.7
 
Other Health Care Primary
32
26.7
 
Male
53
38.1
 
Female
86
61.9
 
White
127
91.4
 
Other
12
8.6
 
Age
139
 
41.35 (13.30)
Education (Overall)
139
 
11.58 (2.42)
Less than HS
41
29.2
 
High School
60
43.8
 
Some College
36
26.3
 
Married
31
22.3
 
Div/Sep/Widowed
44
31.7
 
Never Married
64
46.0
 
Has Public Insurance
53
43.8
 
Has Private Insurance
19
15.7
 
Has No Insurance
49
40.5
 
Has Diagnosed Medical Condition
77
55.4
 
Has Problem with Alcohol
18
12.9
 
Has Problem with Drug Abuse
21
15.1 
 

The sample is predominantly female (62 percent) and white (91 percent).  The average age of these respondents is 41 years.  A large portion of people did not complete high school (30 percent), but a surprisingly high portion had attended college (26 percent).  The mean years of education is 12 years.  The majority of the sample has never married (46 percent).  22% are currently married, and around 32 percent are either widowed, separated, or divorced.  Very few people had private health insurance (16 percent).  Most had public health coverage (44 percent) or no health insurance (41 percent).  Self-reported health status indicates 55 percent have been diagnosed with a physical or mental health condition, 13 percent report having problems with alcohol abuse, and 15 percent report having problems with drug abuse.

    Table 2 compares medical services to the three sets of factors proposed by Aday and Anderson (1974) to impact healthcare utilization: predisposing factors, enabling factors, and needs factors. 

Table 2
Percentage Distribution of Health Care Services by Predisposing, Enabling and Needs Factors

 Factors
Emer-
gency
Room
%
Phy-
sician
%
Other
Health-
Care
Pro-
vider
%
Chi
Square
Predisposing
Factors
       
Male
46.8
36.2
17.0
10.462**
Female
21.9
42.2
32.9
 
White
30.9
41.8
27.3
.428
Other
40.0
40.0
20.0
 
Married
33.3
40.7
25.9
10.942*
Div/Sep/Widowed
41.0
38.5
20.5
 
Never Married
24.1
44.4
31.5
 
Enabling
Factors
       
Has Public Insurance
19.2
59.6
21.2
27.807***
Has Private Insurance
22.2
667
11.1
 
Has No Insurance
52.2
13.0
34.8
 
Under $10,000 Income
31.4
49.0
19.6
3.392
$10,000-19,999
34.5
34.5
31.0
 
$20,000 + Income
21.8
43.5
34.8
 
Problems with Transportation
34.3
45.7
20.0
1.125
Education less than HS
34.3
34.3
31.4
7.606*
High School
38.0
36.0
26.0
 
Some College
15.2
60.6
24.2
 
Needs
Factors
       
Has Diagnosed Medical Condition
23.9
53.5
22.2
10.252***
Has Problem with Alcohol
64.3
21.4
14.3
 
Has Drug Problems
62.5
25.0
12.5
 
*p<.05 **p<.01 ***p<.001

Predisposing Factors

    As expected, results for the gender show males to be twice as likely (46.8 percent) to report using the emergency room as their primary healthcare service than females (21.9 percent).    On the other hand, women are more likely to use the services of a physician (42 percent) compared to males (36 percent) and to use other types of healthcare (33 percent) compared to males (17 percent). 

    Unlike earlier research on healthcare, we did not find a statistically significant difference in healthcare services for the race variable.  This is not surprising since there is very little racial diversity in this community: The population of Wise County, Virginia is more than 92 percent white. 

    Divorced, separated and widowed people are much more likely to use an emergency room as their primary healthcare provider.  41% of these respondents report using the emergency room compared to only 33% percent of married respondents and 24% of never married respondents.  However, marital status has little impact in terms of the tendency to use the services of a physician.  There is only a six percentage point difference between the marital categories: Around 41% of married people compared to 39% of divorced, separated, or widowed people and 44% of never married people.  Our survey data indicate that never married respondents are more likely to report using the other less traditional forms of healthcare: around 32% compared to 26% of married people and around 21% of divorced, separated, and widowed respondents. 

Enabling Factors

    Health insurance status and education were the only variables found to be statistically significant among the enabling factors.  The data showed large differences in healthcare use depending on the availability and type of health insurance coverage.  People without health insurance were more than twice as likely to report the emergency room as their primary source of medical care (52 percent) compared to 22% of people with private insurance and 19 percent of people with public health insurance.  In contrast, only 13% of people with no insurance reported using a physician as their primary source of healthcare compared to around 60% of people who have access to public health coverage, and 67% of those who have access to private health insurance.  Those with no health insurance were also more likely to use non-traditional forms of healthcare: around 35% compared to 21% of those with public health coverage, and 11% of those with private insurance. 

    Individuals with at least some education past high school were the least likely to cite the emergency room as their primary source of care and the most likely to report using a physician.  Only 15% of respondents with at least some college education used the emergency room as their primary care facility compared to 34% of people with less than a high school degree and 38% of those with a high school diploma.  Conversely,  around 61% of respondents with some college education reported using the services of a physician compared to around 34% of those with less than a high school education, and 36% of those with a high school degree.

    In terms of the needs factors, we found that people with a diagnosed medical condition are more likely to use a physician (54%) compared to 24% who indicated they were more likely to go to an emergency room, and roughly 23% who chose alternative medical care.  On the other hand, people who self-report having problems with alcohol or drugs are much more likely to use the emergency room.  Of those respondents reporting problems with alcohol, around 64% choose the emergency room, 21% choose physician care, and 14% choose other non-traditional services.  Similarly, 63% of respondents indicating drug related problems choose the emergency room as their primary source of care, 25% choose physician care, and 13% choose other services. 

    Results from the bivariate analysis identify seven key factors that influence healthcare service use: gender, marital status, health insurance coverage, education, having a diagnosed medical condition, problems with alcohol abuse, and problems with drug abuse.  These key factors are included as control variables in the multivariate analysis reported below.

    Table 3 presents separate logistic regression models for each of the primary healthcare services: emergency room, physician's office, and other less traditional health providers. 

Table 3
Logistic Regression for Healthcare Services

Factors
Model
1
Emer-
gency
Room 
Log 
Odds
Model
 2
Phy-
sician
Log 
Odds
Model
3
Other
Health
Services
Log 
Odds
Predisposing
Factors
     
Female
.208**
1.732
1.237*
Div/Sep/Widowed^
.243
2.897
1.305
Never Married^
.406*
1.999
.811
Enabling
Factors
     
Public Health Insurance^^
2.002
.532
2.196
No Insurance^^
9.194**
.029***
5.261
Education Less Than High School
.089
.211*
1.134
High School
.174
.573
.798
Needs
Factors
     
Medical Condition
.189
4.639***
.471
Alcohol Problem
.624
.068
.000
Drug Problem
.838
2.149
.000
Constant      
-2 log likelihood
.281
.758*
.089
Omnibus Chi-Square
108.766
109.079
115.357
Test of Fitness
37.124***
43.577***
16490*
*p<.05 **p<.01 ***p<.001
^Reference category = currently married.
^^Reference catgory = personal insurance.

Emergency Room Use

    Model 1 tests the effects of each of the factors on the propensity to use a hospital emergency room as the primary source of healthcare.  Regression results indicate the overall model is statistically reliable (-2 Log Likelihood=108.766; chi square (9)=37.124, p<.001).  The model correctly classified 89.2 percent of the cases.

    Consistent with our descriptive results, we see that both predisposing factors (gender and marital status) in model 1 significantly predict the likelihood of using emergency room facilities.  Being female reduces the odds of emergency room use.  Females are only about 1/5th as likely as men to use an emergency room as their primary source of healthcare.  We also see the influence of marital status.  While divorced, separated, and widowed people are not statistically different from currently married couples in their use of emergency rooms, never married people are around 1/4th as likely to use the emergency room. 

    Turning to the enabling factors, health insurance status is a strong predictor of emergency room use.  The odds ratio for having no insurance is 9.19.  This suggests that those who without any form of health insurance (personal or public) are nine times more likely to use emergency room services than their counterparts who have access to personal or public insurance.

    None of the needs factors variables was useful in the analysis of emergency room use.  Since the needs factors variables are insignificantly different from zero, interpretation of their magnitude has little meaning.

Use of Physicians

Model 2 shifts attention to physicians as primary caregivers.  Regression results indicate the overall model is statistically reliable (-2 Log Likelihood = 110.079; chi square (9)=51.165, p<.0001).  The model correctly classified 78.6 percent of the cases.

    According to the Wald statistic, neither of the predisposing factors (gender and marital status) are statistically significant.  This finding is not surprising.  Initial bivariate analysis indicated a small difference between males and females (6%) and similarly small differences (2.2% to 6%) between marital status categories.

    Conversely, both of the enabling factors contribute to our understanding of physician's office as the primary choice of healthcare.  As expected, health insurance status is a strong predictor of the likelihood of choosing a physician.  People with no insurance coverage were less than 3 percent as likely to use a physician's office as their counterparts who have access to some form of medical insurance.  The type of health care coverage does not impact the decision to choose a physician's office: public coverage is not significantly different from personal insurance coverage.

    Education serves as a predictor of physician care.  The odds ratio for having less than a high school education is .211 indicating that those who did not earn a high school degree are only 1/5th as likely to visit a physician's office as those who have a high school education or higher.  There is no significant difference in the likelihood of physician's visits for those with a high school degree and those who have attended college.

    In terms of the needs factors, being diagnosed with a physical, emotional or medical condition is an important predictor.  Those diagnosed with one or more of these conditions are 4.6 times more likely to visit a physician.  However, once all of the factors were controlled for, there is no evidence that experiencing alcohol problems or drug problems has any effect on the use of physician services.

Use of Other Health Care Services

    Model 3 tests the effects of each of the factors on the propensity to use other, less traditional forms of healthcare service as the primary source of healthcare.  The model is not particularly useful.  Regression results suggest a poor-fitting model, -2 Log Likelihood=155.357; chi square (9)=16.806, p>.05).  The model correctly classified only 24 percent of the cases.  Wald statistics indicated only gender to be statistically significant.  Community belief systems may be more useful predictors of other health care services.  Individuals commented that they sought health care from members of the community who offer cultural or herbal remedies.   Unfortunately, we did not collect systematic information on this type of health service.

Summary and Conclusions

    The results of this analysis provide several insights that might prove useful to future research.  Overall, our results underscore the critical role of insurance in health care.  Individuals with no health care coverage are eight times more likely to use the emergency room as their primary source of care.  This is an important issue since individuals without insurance coverage may delay seeking treatment until their health problems reach a crisis stage, and late stage diagnosis can result in poorer outcomes.

    Our results demonstrate very little difference in type of healthcare service among individuals with private health insurance and those with public health insurance.  Both groups are more likely to use a physician as their primary provider.   This outcome suggests that efforts made by the State of Virginia to reduce the number of Medicaid patients who use the emergency room for routine care is probably succeeding. 

    41% of the people in this study were without health insurance.  Health insurance is a critical factor in influencing health outcomes.  Consequently, policy makers need to develop programs addressing the needs of the uninsured.  Since small businesses are less likely to offer health insurance, one solution might be to develop health insurance cooperatives for small businesses.  Another promising approach might be to develop a program similar to that of CHOICE Regional Health Network.  Located in Olympia, Washington, this program specifically addresses the problem of access to insurance.  CHOICE provides guidance to the uninsured and underinsured by helping clients to choose an affordable health plan, connecting people to social services, and providing help in completing the necessary paperwork.  In 2002, CHOICE assisted more than 3,000 people to access health care services.  Additional information on this organization can be found at http://www.choicenet.org.

    Our research also suggests that future studies should investigate the impact of local cultural beliefs and traditions on healthcare decisions.  A number of individuals in our study cited members of the community who offer cultural or herbal remedies as their primary source of healthcare.  It was not within the scope of this research to pursue this area of investigation.

    We recognize several limitations of this study.  First, the characteristics of the sample prevent generalizing from the results of this study.  The sample included only individuals who attended the annual Remote Area Medical (RAM) Health Expedition.  Second, omitted variables may cause bias between the relationships studied: No data were collected regarding cultural beliefs and behavior, but it became apparent that local culture might be an important part of understanding healthcare choices in this community.  Third, since this study took place in one time period it is appropriate to consider associations between variables rather than causation.

    Nonetheless, the Appalachians of southwest Virginia are a unique community, and we feel that the insights drawn from this work will prove useful to the development of future research.

References

Aday, L.A., and R. Anderson 1974. "A Framework for the Study of Access to Medical Care.” Health Services Research 9:208-220.

Anderson, R., and L. A. Aday 1978. “Access to Medical Care in the U.S.: Realized and Potential.” Medical Care 16(7):533-546.

Anderson, R. 1995. "Revisiting the Behavioral Model and Access to Medical Care: Does it Matter?" Journal of Health and Social Behavior 36:1-10.

Bernstein, A.B., E. Hing, A. J. Moss, K. F. Allen, A.B. Siller, and R. G. Tiggle 2003. Health Care in America: Trends in Utilization. Hyattsville, Maryland: National Center for Health Statistics.

Blumberg, L., and D. W. Liska 1996. The Uninsured in the United States: A Status Report. Washington DC: The Urban Institute.

Broyles, R.W., W.J. McAuley and D. Baird-Holmes 1999. "The Medically Vulnerable: Their Health Risks, Health Status, and Use of Physician Care." Journal of Health Care for the Poor and Underserved 10(2):186-200.

Budetti, J., L. Duchon, C. Schoen et al. 1999. Can't Afford to Get Sick: A Reality for Millions of Working Americans. Commonwealth Fund Report.

Cohen, R. A., and M. E. Martinez 2006. Health Insurance Coverage: Estimates from the National Health Interview Survey, 2005. Available at: http://www.cdc.gov/nchs/nhis.htm. Accessed June 2006.

Faulkner, L.A., and H.H. Schauffler 1997. "The Effect of Health Insurance Coverage on the Appropriate Use of Recommended Clinical Preventive Services. American Journal of Preventive Medicine 13(6):453-458.

Gamm, L., L. Hutchison, G. Bellamy, B. Dabney, and A. Dorsey., eds. 2003. Rural Healthy People 2010: A Companion Document to Healthy People 2010. Volume 2. College Station, Texas: The Texas A&M University System Health Science Center, School of Rural Public Health, Southwest Rural Health Research Center.

Graduate Medical Education Consortium. 2002. Report to the Board. August. Wise, VA: Graduate Medical Education Consortium.

Hangdrup, N.A., E. J. Simoes, and R. C. Brownso 1997. "Health Care Coverage: Traditional and Preventive Measures and Associations with Chronic Disease Risk Factors." Journal of Community Health 22:387-399.

Harris R., and L. Leininger 1993. "Preventative Care in Rural Primary Practice." Cancer 72(suppl 3):1113-1118.

Huttlinger, K., J. Schaller-Ayers, and T. Lawson 2004. "Healthcare in Appalachia: A Population-based Approach." Public Health Nursing 21(2):103-11.

LeClere, F., L. Jensen, and A. Biddlecom 1994 "Health Care Utilization, Family Context and Adaptation among Immigrants to the U.S." Journal of Health and Social Behavior 35:370-384.

National Center for Health Statistics 2006. Health in the United States: With Chartbook on Trends in the Health of Americans. U.S. Department of Health and Human Services, Washington D.C.: U.S. Government Printing Office.

Ormond, B., S. Zuckerman, and A. Lhila 2000.  Rural/urban Differences in Health Care not Uniform across States. Assessing the New Federalism. Series B, No. B-11. The Urban Institute.

Rice, T., J. Gabel, L. Levitt, et al. 2002.  "Workers and their health plans: Free to choose?" Health Affairs 21(1):182-184 

Sateren, W.B., E.L. Trimble, and  J. Abrams 2002.  "How Sociodemographics, Presence of Oncology Specialists, and Hospital Cancer Programs Affect Accrual to Cancer Treatment Trials." Journal of Clinical Oncology 20:2109-2117.

Shi, L. 1999. "Experience of Primary Care by Racial and Ethnic Groups in the United States." Medical Care 37(10):1068-1077.

Smedley B. D., A. Y. Stith, and and A. R. Nelson 2002. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care.  Institute of Medicine. Washington, DC: National Academy Press.

West Virginia Bureau of Public Health 1994. The 1992 West Virginia Behavioral Risk Factor Survey. Charleston, WV: Health Statistics Center.

Authors' Note

*We would like to thank the anonymous reviewers for their constructive and insightful comments.
 
 

 Return to Sociation Today Fall 2008

©2008 by the North Carolina Sociological Association