County-Level Determinants of COVID-19 Testing and Cases: Are There Racial/Ethnic Disparities in Texas?

Abstract

Reports suggest that COVID-19 affects people of color disproportionately. Texas ranks second in the count of confirmed US cases. This study examined the relationship between county-level racial/ethnic composition and COVID-19 testing/cases in Texas, adjusting for population-level demographic characteristics, health factors, and health care access measures. County-level testing and case data, obtained from the Texas Department of State Health Services, were combined with the 2020 Robert Wood Johnson Foundation County Health Rankings data. Outcome variables were tests per 100,000 population and cases per 100,000 population. The independent variable of interest was percent of racial and ethnic composition. Multivariable linear regression analyses were used. There was a statistically significant increase in COVID-19 testing/100,000 population with every 1% increase in the proportion of African Americans/Blacks (β = 2065.4; P = 0.009), Asians (β = 2056.2; P = 0.015), and Hispanics (β = 1641.1; P = 764.7). After controlling for county characteristics and cases/100,000 population, these relationships were no longer significant. However, primary care physician rate was significantly associated with testing/100,000 population (β = 64.0; P = 0.027), as was the percent of uninsured (β = -469.9; P = 0.024). An analysis of case data showed that African Americans had the largest number of cases/100,000 (β = 432.2; P = 0.001), followed by Hispanics (β = 422.8; P < 0.001) and Asians (β = 415.4; P = 0.004). As in other parts of the United States, African Americans and Hispanics are most affected by COVID-19 in Texas. Community-based strategies to improve access to testing or reduce community spread outside clinical settings should target counties with low primary care physician rates or a high proportion of uninsured residents.

Introduction

The COVID-19 pandemic constitutes an unprecedented life experience. The United States has been severely affected more than any other country.¹ As of mid-October 2020, 13 million infections and 200,000 deaths have been reported in the United States.¹ Worse still, numerous states continue to report increases in the number of confirmed cases.¹ COVID-19 rates vary significantly across states, from California with more than 860,000 cases to Maine with just under 6000 confirmed cases as of October 14, 2020.¹ Texas ranks second in the counts of confirmed cases, with more than 840,000 infections and more than 17,000 fatalities.¹ Current reports suggest that Texan racial/ethnic minorities might be disproportionately affected by COVID-19.^2,3 Of the total confirmed Texas cases, 40% were Hispanic, 30% were White, 17% were Black, and 2% were Asian.⁴ COVID-19-associated deaths were pointedly higher in Harris, Hidalgo, and Bexar counties, home to Houston, the Rio Grande Valley, and San Antonio – major population centers in Texas.²

Unfortunately, Texas residents already face significant primary care challenges. The state ranks 47^th in the primary care physician (PCP) to population ratio, with only 67 PCPs per 100,000 Texas residents.⁵ In 2019, Texas ranked last in health care access and affordability, and 45^th in the availability of prevention and treatment options.⁶ Although access to care has increased in recent years through the contributions of the Texas Medicaid 1115 waiver program and considerable investment in physician recruitment, training, and education, there still is great concern for equitable and efficient access to care for most Texas residents,⁷ especially given that Texas ranks first in the percent of uninsured residents.⁸ This suboptimal access is worrisome because numerous studies have shown that limited access to PCPs and lack of health insurance are strongly correlated with suboptimal health outcomes.^9,10 Further, early reports suggest that individuals living with underlying medical conditions might be at higher risk for COVID-19-related hospitalizations.¹¹ These disparities have implications for COVID-19 outcomes among racial/ethnic minorities.

Evidence has shown that the COVID-19 outbreak has varying effects on different population subgroups.^12,13 Early racial/ethnic reports on COVID-19 morbidity and mortality suggest that minorities, including African American, Hispanic, American Indian, Alaska Native, and Pacific Islander populations bear a disproportionate burden of disease.^13

–17 As efforts to develop a vaccine and/or cure for the virus continue, strategies aimed at curtailing the virus have focused on preventive measures, including stay-at-home mandates, handwashing, social distancing, and respiratory hygiene, to slow the spread of the virus. Variations in the outbreak across subpopulations highlight a need for population-level studies to identify determinants of COVID-19 cases. This could guide prevention strategies and inform community-based strategies to improve access to testing at the state level.

This paper uses county-level Texas COVID-19 testing and case data to explore (1) the relationship between county-level racial/ethnic composition and COVID-19 tests and cases in the state of Texas, and (2) whether county-level health factors, health care access measures, and other demographic characteristics explain this relationship. Research of this nature will provide population-level insight into how this pandemic is affecting Texans and potentially guide implementation of prevention and testing strategies at the county level.

Methods

This study used publicly available county-level COVID-19 testing and case data obtained from the Texas Department of State Health Services, which comprised tests and cases reported from March 4, 2020-October 13, 2020. The data were merged with the 2020 Robert Wood Johnson Foundation (RWJF) County Health Rankings data. RWJF County Health Rankings data contain information on county-level demographic characteristics such as population size, racial/ethnic composition, unemployment rates, as well as county-level health-related data such as percent who smoke, percent who have diabees, and PCP rate.

Measures

There were 2 outcomes of interest: (1) cumulative COVID-19 tests per 100,000 population and (2) cumulative COVID-19 cases per 100,000 population. The independent variables of interest were percent of racial/ethnic groups – percent non-Hispanic White, percent Black, percent Asian, and percent Hispanic. The researchers controlled for county-level health factors (percent smokers, percent of adults with obesity, percent of adults with diabetes, percent of adults with fair or poor health), health care access measures (percent uninsured and PCP rate), and other county-level demographics (high school graduation rate, percent unemployed, black-white segregation index, and percent rural).

Statistical analyses

Summary statistics were computed for the independent variables. Multivariable linear regression was used to examine the relationship between county-level racia/ethnic composition and the outcomes of interest, controlling for health factors, health care access measures, and other county-level demographic characteristics. Three models were estimated for the test/100,000 analyses: (1) proportion of racial/ethnic groups alone, (2) proportion of racial/ethnic groups controlling for county-level health factors, health care access measures, and other demographic characteristics, and (3) proportion of racial/ethnic groups controlling for county-level factors and number of cases/100,000.

Two models were estimated for the case/100,000 analysis: (1) proportion of racial/ethnic groups alone, and (2) proportion of racial/ethnic groups controlling for county-level health factors, health care access measures, and other county demographic characteristics. Sensitivity analyses were performed using robust linear regression. All statistical tests were 2-sided, and findings were considered statistically significant at P < .05. All analyses were conducted using Stata Statistical Software, Release 15.1 (StataCorp LLC, College Station, TX). Robust regression results are reported.

Results

Table 1 shows summary statistics of the independent variables for the percentages of Asians, African Americans/Blacks, non-Hispanic Whites, and Hispanics, among other variables.

Table 1.

Summary Statistics of County-Level Characteristics

Variable	Minimum	Maximum	Mean	Std. Dev.
Percent Asian	0.00	20.79	1.33	2.08
Percent Black	0.00	33.14	6.26	6.40
Percent Non-Hispanic White	2.69	89.40	55.39	21.03
Percent Hispanic	3.72	96.36	35.35	22.99
Percent Smokers	10.64	19.87	14.94	1.56
Percent Adults with Obesity	21.60	47.30	31.25	5.26
Percent Adults with Diabetes	3.50	29.30	11.36	4.62
Percent Fair or Poor Health	12.29	40.99	20.60	4.94
Percent Uninsured	11.40	33.75	21.21	4.05
Primary Care Physician Rate	0.00	141.50	40.50	25.85
High School Graduation Rate	61.80	100.00	93.65	5.67
Percent Unemployed	1.83	10.10	3.87	1.26
Black White Segregation Index	1.67	76.78	39.98	14.03
Percent Rural	0.69	100.00	55.52	31.90

Std. Dev., standard deviation.

Table 2 shows multivariable regression analyses of total COVID-19 tests per 100,000 population. Model 1 shows a statistically significant increase in COVID-19 testing/100,000 population for every 1% increase in the proportion of African Americans/Blacks (P = 0.01), Asians (P = 0.02), Hispanics (P = 0.03), and non-Hispanic Whites (P = 0.05). Controlling for health factors, health care access measures, and other county demographic characteristics, the positive relationship with county proportion of African Americans, Asians, and Hispanics remained significant while the percent of non-Hispanics Whites was no longer statistically significant. Additionally, increasing PCP ratio was associated with increasing number of tests/100,000 (β = 65.4; P = 0.03), while increases in the proportion of uninsured were negatively associated with tests/100,000 (β = -568.4; P = 0.01). The final model shows that after adjusting for cases/100,000, county racial/ethnic composition was no longer statistically significant while increasing PCP rates and percent uninsured remained significantly associated with higher number of tests/100,000 (Table 2).

Table 2.

Multivariable Regression Analyses of Total COVID-19 Tests in Texas

	Unadjusted model		Adjusted Model 1		Adjusted Model 2
	β (std. err)	P	β (std. err)	P	β (std. err)	P
Percent Asian	2056.2 (840.3)	0.02	2260.6 (1069.2)	0.04	771.5 (1070.6)	0.47
Percent Black	2065.4 (784.1)	0.01	2366.1 (989.0)	0.02	1012.0 (999.4)	0.31
Percent Non-Hispanic White	1549.0 (781.6)	0.05	1950.2 (1010.8)	0.06	577.8 (1016.0)	0.57
Percent Hispanic	1641.1 (764.7)	0.03	2004.8 (977.3)	0.04	674.8 (986.3)	0.50
Percent Smokers			587.2 (976.0)	0.55	1099.5 (938.8)	0.24
Percent Adults with Obesity			-133.4 (124.4)	0.29	-171.2 (119.8)	0.16
Percent Adults with Diabetes			25.7 (132.9)	0.85	-36.3 (128.4)	0.78
Percent Fair or Poor Health			433.4 (654.9)	0.51	-219.2 (632.8)	0.73
Percent Uninsured			-568.4 (210.3)	0.01	-469.9 (206.4)	0.02
Primary Care Physician Rate			65.4 (29.8)	0.03	64.0 (28.7)	0.03
High School Graduation Rate			-18.0 (122.6)	0.88	-27.2 (117.8)	0.82
Percent Unemployed			275.1 (630.5)	0.66	504.1 (606.9)	0.41
Black White Segregation Index			30.5 (41.6)	0.46	24.2 (40.0)	0.55
Percent Rural			-4.8 (34.4)	0.89	6.2 (33.1)	0.85
Percent ages 65 and older			213.6 (240.0)	0.38	333.0 (232.1)	0.15
Cases per 100,000					-71.1 (69.8)	<0.01

Bold = P < 0.05.

Table 3 shows multivariable regression analyses of total COVID-19 cases per 100,000 population. Model 1 shows a statistically significant increase in COVID-19 cases/100,000 population for every 1% increase in the proportion of African Americans/Blacks (P < 0.01), Asians (P < 0.01), Hispanics (P < 0.01), and non-Hispanic Whites (P < 0.01). Controlling for county health factors, health care access measures, and other demographics (model 2), the association of county racial/ethnic composition and cases/100,000 remained statistically significant for all 4 groups. However, African Americans had the largest number of cases/100,000 (β = 432.2; P < 0.01), followed by Hispanics, Asians, and non-Hispanic Whites.

Table 3.

Multivariable Regression Analyses of Total COVID-19 Cases per 100,000 Population in Texas

	Unadjusted Model 1		Adjusted Model 2
	β (std. err)	P	β (std. err)	P
Percent Asian	311.6 (97.3)	<0.01	415.4 (142.2)	<0.01
Percent Black	330.0 (90.8)	<0.01	432.2 (131.5)	<0.01
Percent Non-Hispanic White	281.5 (90.5)	<0.01	408.5 (134.4)	<0.01
Percent Hispanic	308.0 (88.6)	<0.01	422.8 (120.0)	<0.01
Percent Smokers			116.9 (129.8)	0.37
Percent Adults with Obesity			-22.1 (16.5)	0.18
Percent Adults with Diabetes			7.34 (17.7)	0.68
Percent Fair or Poor Health			80.2 (87.1)	0.36
Percent Uninsured			-61.2 (28.0)	0.03
Primary Care Physician Rate			-4.6 (4.0)	0.24
High School Graduation Rate			7.2 (16.3)	0.66
Percent Unemployed			-28.5 (83.9)	0.74
Black White Segregation Index			5.3 (5.5)	0.34
Percent Rural			-4.2 (4.6)	0.36
Percent ages 65 and older			-32.6 (31.9)	0.31

Bold = P < 0.05.

Discussion

This study examined county-level determinants of COVID-19 testing and cases in Texas, adjusting for county-level health factors, health care access measures, and other county demographic characteristics. Lack of insurance and PCP rates, both important measures of access to health care, were found to be significantly associated with COVID-19 testing and cases in Texas. This study also found that cases/100,000 were significantly higher among non-Hispanic Blacks and Hispanics.

The finding of decreasing tests/100,000 with increasing percent of uninsured individuals in a given county has implications for detection of COVID-19 among residents of Texas given that low health insurance coverage rates is a particularly pressing problem in Texas. Although Texas participates in the Medicaid 1115 waiver program through the Texas Healthcare Transformation and Quality Improvement Program¹⁸ allowing for greater safety net services, the state ranks first in the proportion of uninsured persons. Others have attributed this to the state's refusal to expand Medicaid under the Affordable Care Act.^19,20 The state also has a large proportion of undocumented immigrants, most of whom are Hispanic and ineligible to receive employer-sponsored or publicly available health insurance benefits; 61% of the uninsured in Texas are Hispanic.²¹ Unfortunately, non-expansion states, such as Texas, not only lead nationally in uninsured rates but also have worse health outcomes.²² The setting of COVID-19 has only made the lack of insurance more dire. Before the onset of the COVID-19 pandemic, more than 4 million adults ages 18–65 years were uninsured in Texas.²³ As the pandemic continued to destroy jobs, 650,000 more adults between ages 18 and 65 lost health insurance. Currently 29% of Texas adults are uninsured compared to a nationwide uninsured rate of 16%.²³ These ominous statistics suggest the need for policies to address access to care during this pandemic. These study findings also highlight the need for targeted interventions to provide access to testing for communities with a high proportion of uninsured individuals.

The positive association between higher PCP ratios and testing availability cannot be ignored, especially for a state like Texas that already has one of the lowest PCP to population ratios. A 2015 survey reported that 35 Texas counties have no physicians, and 80 counties have 5 or fewer physicians (115 of 254 total counties in Texas).²⁴ As a result of this PCP shortage, many Texans living in these underserved communities may be forced to travel to neighboring counties to receive health care. Worse still, counties with lower health rankings and limited access to primary care services²⁵ struggle to find healthy foods because they live in food deserts, increasing the risk for poor health outcomes.^26,27 The present study finding of increasing tests/100,000 with increasing PCP rates is an indication that limiting COVID-19 testing to only clinical settings could exacerbate COVID-19-related disparities and outcomes in communities with inadequate PCP availability. Thus, there is a need for coordinated strategies to provide access to testing in community settings.

In the present study, African Americans had the greatest number of cases/100,000, followed by Hispanics, Asians, and non-Hispanic Whites. This finding aligns with earlier studies that suggest that African Americans and Hispanics in the United States bear a disproportionate burden of COVID-19 cases.^13,28
–30 The researchers proffer that the higher impacts in African Americans and Hispanics may be related to differential exposure, as these populations comprise greater percentages of frontline workers, putting them at greater risk for infection than other Americans who work from home.³¹ In addition, living conditions that may be crowded and/or include multigenerational family members may increase the risk of spreading the pathogen.³² Thus, this study highlights the importance of targeted community-level strategies to increase awareness of preventive measures and access to personal protective equipment in counties with a high proportion of individuals of minority race/ethnicity.

This study is not without limitations. First, the analysis is at the county level, using county-level aggregates of health factors, health care access, and other county-level demographic characteristics. Chowkwanyun and Reed argue that although reporting racial disparities in COVID-19 cases could highlight the impact of the endemic disparities experienced by marginalized subgroups, they caution that reporting such data without context or adequate analyses could perpetuate myths that could adversely affect strategies aimed at reducing health disparities for these groups.³³ Therefore, individual-level analyses could provide additional insights beyond this index study to allow researchers to accurately capture the impact of these social factors on COVID-19 testing and cases at the individual level. Second, study data were limited to Texans and thus the findings might not be generalizable to the entire US population.

In spite of these limitations, this study has several strengths. First, by combining multiple data sources, the researchers were able to control for contextual factors that could affect county-level racial/ethnic disparities in COVID-19 testing and cases. Further, this study provides important insights that could guide community-based COVID-19 prevention or mitigation measures among residents of Texas or states with similar characteristics.

Conclusion

In conclusion, the pervasive health and health care disparities experienced by individuals of minority race/ethnicity in the United States has been exacerbated by the COVID-19 pandemic. These disparities could be worse for residents of communities with a high proportion of individuals with limited health care access. Community-based interventions that target counties with a high proportion of minority populations, uninsured individuals, or limited access to PCPs have the potential to reduce COVID-19-related disparities and should be explored.

Footnotes

Author Contribution Statement

Drs. Adepoju and Ojinnaka both contributed to study conception and design, statistical analysis, data interpretation, and the drafting and revising of this manuscript.

Author Disclosure Statement

The authors declare that there are no conflicts of interest.

Funding Information

No funding was received for this article.

References

Center for Systems Science and Engineering (CSSE) at Johns Hopkins University. COVID-19 dashboard. https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6 Accessed September 21, 2020 .

Texas Department of State Health Services. Texas case counts: COVID-19. https://txdshs.maps.arcgis.com/apps/opsdashboard/index.html#/ed483ecd702b4298ab01e8b9cafc8b83 Accessed June 11, 2020 .

Kaiser Family Foundation. COVID-19 cases by race/ethnicity. 2020. https://www.kff.org/other/state-indicator/covid-19-cases-by-race-ethnicity/?currentTimeframe=0&sortModel=%7B%22colId%22:%22Location%22,%22sort%22:%22asc%22%7D Accessed October 11, 2020 .

Texas Department of State Health Services. Texas case counts COVID-19. https://txdshs.maps.arcgis.com/apps/opsdashboard/index.html#/ed483ecd702b4298ab01e8b9cafc8b83 Accessed May 25, 2020 .

Association of American Medical Colleges. 2019 Texas Physician Workforce Profile. 2019. https://www.aamc.org/system/files/2019–12/state-physician-Texas-2019%5B1%5D.pdf Accessed May 27, 2020 .

The Commonwealth Fund. Scorecard on state health system performance. https://scorecard.commonwealthfund.org/state/texas/ Accessed May 27, 2020 .

Gunja

, Collins

, Doty

, Beutel

. Insurance coverage, access to care, and medical debt since the ACA: a look at California, Florida, New York, and Texas. Issue Brief (Commonw Fund), 2017; 7:1–12.

Kaiser Family Foundation. Health insurance coverage of the total population. 2020. https://www.kff.org/other/state-indicator/total-population/?currentTimeframe=0&sortModel=%7B%22colId%22:%22Uninsured%22,%22sort%22:%22desc%22%7D Accessed October 18, 2020 .

Kilbourne

AM.

Care without coverage: too little, too late. J Natl Med Assoc, 2005; 97:1578.

10.

Hossain

, Laditka

. Using hospitalization for ambulatory care sensitive conditions to measure access to primary health care: an application of spatial structural equation modeling. Int J Health Geogr, 2009; 8:51–51.

11.

Centers for Disease Control and Prevention. Coronavirus disease 2019: At risk for severe illness. 2020. https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/groups-at-higher-risk.html Accessed June 8, 2020 .

12.

Price-Haywood

, Burton

, Fort

, Seoane

. Hospitalization and mortality among black patients and white patients with Covid-19. N Engl J Med, 2020; 382:2534–2543.

13.

Webb Hooper

, Nápoles

, Pérez-Stable

. COVID-19 and racial/ethnic disparities. JAMA, 2020; 323:2466–2467.

14.

COVID-19 Statistics. http://dph.illinois.gov/covid19/covid19-statistics Accessed June 10, 2020 .

15.

Thebault

, Tran

, Williams

. The coronavirus is infecting and killing black

Americans at an alarmingly high rate

. Washington

Post

. https://www.washingtonpost.com/world/2020/04/07/coronavirus-latest-news/ Accessed August 15, 2020 .

16.

Laurencin

, McClinton

. The COVID-19 pandemic: a call to action to identify and address racial and ethnic disparities. J Racial Ethn Health Disparities, 2020; 7:398–402.

17.

Azar

KMJ

, Shen

, Romanelli

, et al. Disparities in outcomes among COVID-19 patients in a large health care system in California. Health Aff (Millwood), 2020; 39:1253–1262.

18.

Kaiser Family Foundation. Medicaid waiver tracker: approved and pending section 1115 waivers by state. 2020. https://www.kff.org/medicaid/issue-brief/medicaid-waiver-tracker-approved-and-pending-section-1115-waivers-by-state/ Accessed October 16, 2020 .

19.

Revere

, Kavarthapu

, Hall

, Begley

. Achieving triple aim outcomes: an evaluation of the Texas Medicaid waiver. Inquiry, 2020; 57:0046958020923547.

20.

Chen

Lucy

, Madubuonwu

, Pecos-Duarte

, Sommers

. Medicaid expansion in Texas: potential economic and employment implications. 2019. https://www.commonwealthfund.org/publications/issue-briefs/2019/aug/medicaid-expansion-texas-potential-economic-employment-implications Accessed October 16, 2020 .

21.

Texas Medical

Association

. Quick Statistics on the Uninsured in Texas and the

U.S

. 2020. https://www.texmed.org/template.aspx?id=5519 Accessed September 19, 2020 .

22.

Han

, Nguyen

, Drope

, Jemal

. Health-related outcomes among the poor: Medicaid expansion vs. non-expansion states. PLoS One, 2015; 10:e0144429.

23.

National Center for Coverage Innovation. The COVID-19 pandemic and resulting economic crash have caused the greatest health insurance losses in American history. 2020. https://www.familiesusa.org/wp-content/uploads/2020/07/COV-254_Coverage-Loss_Report_7-17-20.pdf Accessed on August 30, 2020 .

24.

Commins

. 35 Texas counties have zero physicians. HealthLeaders 2015. https://www.healthleadersmedia.com/strategy/35-texas-counties-have-zero-physicians Accessed on August 30, 2020 .

25.

Office of Disease Prevention and Health Promotion. Healthy people 2020. https://www.healthypeople.gov/ Accessed on August 30, 2020 .

26.

Fehr

, Kates

, Cox

, Michaud

. COVID-19 in Rural America - Is There Cause for Concern? . https://www.kff.org/other/issue-brief/covid-19-in-rural-america-is-there-cause-for-concern/# Accessed June 10, 2020 .

27.

Raifman

, Raifman

. Disparities in the population at risk of severe illness from COVID-19 by race/ethnicity and income. Am J Prev Med, 2020; 59:137–139.

28.

Kaufman

, Niles

, Nash

. Disparities in SARS-CoV-2 positivity rates: associations with race and ethnicity. Popul Health Manag 2020. DOI: 10.1089/pop.2020.0163

29.

Mahajan

, Larkins-Pettigrew

Racial demographics and COVID-19 confirmed cases and deaths: a correlational analysis of 2886 US counties. J Public Health (Oxf) 2020. DOI:10.1093/pubmed/fdaa070

30.

Moore

, Ricaldi

, Rose

, et al. Disparities in incidence of COVID-19 among underrepresented racial/ethnic groups in counties identified as hotspots during june 5–18, 2020—22 states, February–June 2020. Morb Mortal Wkly Rep, 2020; 69:1122.

31.

Rho

, Brown

, Fremstad

. A Basic Demographic Profile of Workers in Frontline

Industries

. 2020. https://cepr.net/a-basic-demographic-profile-of-workers-in-frontline-industries/ Accessed July 31, 2020 .

32.

Almagro

, Coven

, Gupta

, Orane-Hutchinson

Racial disparities in frontline workers and housing crowding during COVID-19: evidence from geolocation data. Available at SSRN https://ssrn.com/abstract=3695249 or https://dx-doi-org.web.bisu.edu.cn/10.2139/ssrn.3695249 Accessed September 27, 2020 .

33.

Chowkwanyun

, Reed

Jr . Racial health disparities and Covid-19—caution and context. N Engl J Med, 2020; 383:201–203.