The Geography of Viral Hepatitis C in Texas, 1992-1999


This study examines the relationship between viral hepatitis C (HCV), race, gender, and population density in Texas counties. Previous studies related HCV rates to residence in urban areas, race, and gender as common risk markers. HCV morbidity rates in Texas counties from 1992 to 1999 were obtained from the Department of State Health Services (DSHS). Race, gender, and population density data were extracted from the 2000 U.S. Census. Simple linear regression, Pearson's rank correlation, and Friedman's test were used for the analysis. The results indicated that population density was not a significant indicator of HCV rates, but race and gender were significant. For example, Black males have a much higher risk than White and Hispanic groups. Thus, HCV rates in Texas counties seem to be influenced not by urbanization, but by ethnicity and gender. Previously considered as having low risk for HCV, rural areas in Texas appear to have a higher risk than urban areas. Yet, the latter have been the main focus of previous intervention efforts. New intervention efforts in Texas must focus especially on minorities in rural areas.

Table of Contents: 


    Viral hepatitis C (HCV) is an emerging health problem in the United States because it causes serious illness, affects millions, and has a close connection with HIV/AIDS. HCV is often asymptomatic and undiagnosed; this leads to extensive damage of the liver, liver failure, or hepatocellular carcinoma. HCV is common in Texas with a prevalent infection rate of 1.79%. According to the Texas Department of State Health Services, “ Texas is facing a silent epidemic of hepatitis C (HCV), which is the leading cause of liver disease” (2005, para. 1). The impact of HCV infection may explode over the next 10 to 20 years because it takes 20 to 30 years for chronic liver disease, cirrhosis, and liver cancer to develop. “[Nationwide], it is conservatively estimated that illness and deaths from HCV-related liver disease will increase two- to threefold over the next two decades. Direct medical costs may range from $6.5 to $13.6 billion, with even higher indirect and social costs" (U.S. Department of Health and Human Services, 2002, p. 3). Thus, it is essential to identify the geographical regions where prevalence is highest to better prepare Texas’ health and economic institutions.

    This study examines the geography of HCV illness rates in Texas counties as well as the factors related to gender and ethnic mix. Imperative to this endeavor is the review of factors implicated in the spread of HCV. This analysis seeks to illuminate the geographic regions in Texas counties with high prevalence rates using population density, race and ethnicity, and gender as predictors. For example, does the rate of illness vary insignificantly between the races? Does the rate of HCV increase in urban areas that have high population density or does it cluster in rural areas? Do males have higher HCV infection rates than females or does HCV spread regardless of gender?

    Literature Review

    Yalamanchili et al. (2005) established that HCV was a long-term and costly disease that clustered near major Texas cities and along the Texas-Mexico border. The study also stated that the urban population centers with the higher rates of infection would be burdened to absorb the majority of the healthcare costs related to the future of those patients. Shi and Stevens (2005) argued that race and ethnicity are common determinants of disparity and conflict. The book explains the roles of social and individual determinants in vulnerability while providing empirical evidence of disparities in health quality and outcome for vulnerable populations. Shi and Stevens’ work led me to consider a statistical analysis of HCV illness among the races in Texas. Cases of HCV illness from January 2000 through December 2003 in the United States were generated using gender as a factor and statistically reported from the National Health and Nutrition Examination Survey III (U.S. Department of Health and Human Services, 2005). The data showed HCV to have higher infection rates in males when compared to females.

    Research Hypotheses

    Three main hypotheses are examined in this research:

    Hypothesis 1: Degree of urbanization as defined by population density is a factor in viral hepatitis C. Counties that have high population densities and are within metropolitan areas will have higher rates of HCV illness. Conversely, counties that are rural and have low population density will have lower rates of illness.

    Hypothesis 2: Race is a factor of viral hepatitis C. HCV rates vary among White, Black, and Hispanic ethnic groups.

    Hypothesis 3: Gender is a factor of viral hepatitis C. The rates of HCV are significantly different between the sexes. Male HCV rates will not be equal to female HCV rates.

    Methodology and Data Sources

    Viral hepatitis C (HCV) will, in this discussion, be defined by rate as occurrence of illness per 100,000 people. Rates of HCV illness by county in Texas were extracted from the Vitalweb system contained in the online database of infectious diseases provided by the Vital Statistics Division of the Texas Department of Health. County-level data on total rate of HCV patients, gender of HCV patients by county, and ethnicity of HCV patients by county are all taken from the 1992 to 1999 year span. Population and square mile area data of Texas counties were acquired from the 2000 U.S. Census.

    A simple linear regression and Pearson’s correlation coefficient were then utilized to examine the relationship between urbanization and total HCV rates. Pearson’s correlation measures the strength and direction of the relationship between the two variables, but this alone is not enough to establish a relationship between two tested variables. The simple linear regression evaluates the magnitude of change in total HCV rate that is explained by population density. All statistical methods used were tested with a .01 significance level.

    A Friedman test was utilized to analyze the variance for three or more independent variables. This nonparametric test is equivalent to one-way repeated measures ANOVA. It is used when there are more than two measurements from related subjects. Only three races were used for comparison because the rates of the other ethnic groups were null or improperly skewed due to the minute rates of HCV. Parameters of the Friedman statistical test are usually strict in that only counties with rates above five are accepted, which did not fit the data used. Three tables of statistical information containing means, standard deviations, rank of the means, chi-square, and significance values are calculated.

    A paired-samples t test was used to examine the hypothesis that rates of HCV are significantly different between the sexes. The results are presented in three tables. A scattergram was also plotted to help illustrate the correlation of male HCV rates and female HCV rates. Mode of transmission was not used as a predictor due to the lack of collected data for the counties of Texas.

    Results and Findings

    To test the first hypothesis of urbanization as a predictor of total rate of HCV, a linear regression model and Pearson’s correlation coefficient were used. A simple linear regression, presented in Table 1, was used to describe the relationship between total rate and population density of HCV. The R-square is equal to zero, meaning 0.0% of total HCV rates are explained by population density, and the slope of the linear regression model is 0.000005377. Extrapolating the data from the linear regression model did not prove the hypothesis true. This is illustrated by the scatterplot in Figure 1 fitted with the linear regression line. The almost completely horizontal line substantiates how poorly the variables are related.

    Using Pearson’s correlation coefficients to measure the relationship between population density and total rate of HCV, population density did not prove to be an important factor of HCV for this study (see Table 2). Density was not statistically significant when compared to total rates of HCV, p=. 992. Pearson’s correlation coefficient was calculated as r = +. 001, which indicates absence of correlation. High HCV rates did not cluster around counties with large cities (Figure 2 and Figure 3), which suggests that urbanization does not correspond with higher HCV rates.

    A Friedman test was utilized to analyze the second hypothesis of whether HCV rates are equal among the races (see Table 3). The Black rate (m=2.08, sd=12.82) ranked first with 1.88, the Hispanic rate (m=1.58, sd=3.40) had the second highest mean rank of 1.96 and the White rate (m=1.417, sd=2.04) scored the lowest mean rank of 2.16. A significant difference was found among the Black, Hispanic, and White ethnic groups (χ 2 (2) = 17.36, p=. 000). This statistical analysis suggests that the Black ethnic group has the greatest prevalence of HCV and is most at risk. Race appears to have an impact on the rate of HCV.

    The correlation between male patient rates and female patient rates was determined at r = +. 307 (see Table 4). This indicates that there is a positive correlation; that is, an increase in male patient rates is related to an increase in female patient rates. A scattergram (Figure 4) was plotted to examine the nature of the functional relationship between the two variables of male rate and female rate. This diagram illustrates the positive correlation between the two variables.

    A paired-samples t test was calculated to compare the mean of HCV male illness rates to female illness rates (see Table 5). A significant difference was found between the means of the two patient groups t (253) = 6.225, p=. 000. The mean of male HCV rates (m=2.46, sd=3.45) is significantly higher than the mean of female HCV rates (m=1.15, sd=1.75).


    Although the rates of HCV illness in Texas has declined for the past decade, the proportion of cases progressing to cirrhosis is "expected to increase" (CDC, 2002). Thus, HCV is common in Texas and will result in an increase in complications of cirrhosis in coming years.

    The degree of urbanization as defined by population density did not have the predicted effect on HCV rates despite their supposed importance in rates of HCV illness. Counties in Texas with higher population densities, which should produce higher rates according to the Baylor Medical Center proceedings, are actually lower in total rates. This suggests that factors other than high population density rates are controlling the rates of HCV infection within each county. Urban location was not found to be a factor in HCV.

    Race had a significant influence on the rates of HCV and supports Shi and Stevens’ (2005) empirical evidence, which concluded that minority races were more at risk for HCV than the White, non-Hispanic race. Blacks in this study had the highest prevalence of HCV followed by Hispanics and lastly, by Whites.

    Gender was found to be a significant factor, with rates among males being significantly higher than females. The size of the difference between the mean male HCV rates and female HCV rates were compared in relation to the amount of inherent variability in the data. The random error was not large and lead to a 100% probability that random error did not occur and that the hypothesis was proved.

    Overall, the spatial distribution of HCV rates seems to be influenced not by urbanization, but by ethnicity and gender. Also, it appears that risks of HCV infection in Texas rural areas are at least as high in urban areas. Yet, the latter have been the main focus of previous intervention efforts. The implications of this study are that new intervention efforts in Texas must focus especially on minorities in rural areas.


    • Shi, L., & Stevens, G. D. (2005). Vulnerable populations in the United States. San Francisco: Jossey-Bass.
    • Texas Department of State Health Services, Infectious Disease Control Unit. (2005, October 13a). Hepatitis C initiative and services. Retrieved December 1, 2005.
    • Texas Department of State Health Services, Infectious Disease Control Unit. (2005, October 13b). Texas Viral Hepatitis Cases (1992–1999) – Main Window. Retrieved December 1, 2005.
    • United States Department of Health and Human Services. (2005, August 23). Vital Statistics Division of the Texas Department of Health. Retrieved November 10, 2005, from
    • United States Census Bureau. Census 2000. American fact finder. Retrieved November 20, 2005, from
    • United States Department of Health and Human Services, Centers for Disease Control and Prevention. (2002, September). Viral Hepatitis and Injection Drug Users.Retrieved March 23, 2006, from
    • Yalamanchili, K., Saadeh, S., Lepe, R., & Davis, G. L. (2005). The prevalence of hepatitis C virus infection in Texas: Implications for future health care. PubMed Central, 18, 3–6.

    Table 1: Regression model summary and regression coefficients.

    Model Summaryb

    Model R R Square Adjusted R Square Std. Error of the Estimate
    1 .001a .000 -.004 2.1839

    a. Predictors: (Constant), DENSITY
    b. Dependent Variable: TRATE


      Unstandardized Coefficients Standardized Coefficients   95% Confidence Interval for B
    Model B Std. Error Beta t Sig. Lower Bound Upper Bound
    1 (Constant) 1.827 .144   12.665 .000 1.543 2.111
      DENSITY 5.377E-06 .001 .001 .010 .992 -.001 .001

    a. Dependent Variable: TRATE

    Table 2: Correlated population density and total rate of illness per 100,000.


    _DENSITY Pearson Correlation
    Sig. (2-tailed)

    Pearson Correlation
    Sig. (2-tailed)


    Table 3: White, Black, and Hispanic HCV rate statistics, ranks, and significance levels.

    Descriptive Statistics

      N Mean Std. Deviation Minimum Maximum
    White Rate
    Black Rate
    Hispanic Rate 


      Mean Rank
    White Rank
    Black Rank
    Hispanic Rank

    Test Statisticsa

    N 255
    Chi-Square 17.362
    df 2
    Asymp. Sig .000

    a. Friedman Test

    Table 4. Male and female HCV rate correlation.

    Paired Samples Correlations

      N Correlation Sig.
    Pair 1 Male Rate & Female Rate 254 .307 .000

    Table 5. Male and female HCV rate statistics.

    Paired Samples Statistics

      Mean N Std. Deviation Std. Error Mean
    Male Rate
    Female Rate

    Paired Samples Test

      Paired Differences t df Sig. (2-tailed)
    Mean St. Duration Std. Error Mean 95% Confidence Interval of the Differ
    Lower Upper
    Pair 1 Male Rate - Female Rate 1.308 3.3495 .2102 .894 1.722 6.225 253 .000

    Figure 1: Linear regression line.

    Figure 2: Viral Hepatitis C rates, 1992 to 1999.

    (Texas Department of State Health Services, 2005a)

    Figure 3: Population density 2000 U.S. Census.

    (U.S. Census Bureau, 2004)

    Figure 4. Male and female HCV rates.