Abstract
Black/African American (Black) versus White persons are unequally burdened by human immunodeficiency virus (HIV) in the United States. Structural factors can influence social determinants of health, key components in reducing HIV-related health inequality by race. This analysis examined HIV care outcomes among Black and White persons with diagnosed HIV (PWDH) in relation to three structural factors: racial redlining, Medicaid expansion, and Ryan White HIV/AIDS Program (RWHAP) use. Using National HIV Surveillance System, U.S. Census, and Home Mortgage Disclosure Act data, we examined linkage to HIV care and viral suppression (i.e., viral load < 200 copies/mL) in relation to the structural factors among 12,996 Black and White PWDH with HIV diagnosed in 2017/alive at year-end 2018, aged ≥ 18 years, and residing in 38 U.S. jurisdictions with complete laboratory data, geocoding, and census tract-level redlining indexes. Compared to White PWDH, a lower proportion of Black PWDH were linked to HIV care within 1 month after diagnosis and were virally suppressed in 2018. Redlining was not associated with the HIV care outcomes. A higher prevalence of PWDH residing (v. not residing) in states with Medicaid expansion were linked to HIV care ≤ 1 month after diagnosis. A higher prevalence of those residing (v. not residing) in states with > 50% of PWDH in RWHAP had viral suppression. Direct exposure to redlining was not associated with poor HIV care outcomes. Structural factors that reduce the financial burden of HIV care and improve care access like Medicaid expansion and RWHAP might improve HIV care outcomes of PWDH.
Keywords: HIV, AIDS, Social determinants of health, Redlining, Medicaid expansion, Ryan white program
Introduction
Black or African American (hereafter referred as Black) persons are disproportionately burdened by incidence and prevalence of human immunodeficiency virus (HIV) in the United States. In 2018, the HIV incidence rate per 100,000 for Black persons was nearly nine times as high as White persons (45.4 versus 5.2) among persons aged ≥ 13 years [1]. Additionally, the estimated prevalence of persons living with HIV between Black versus White persons aged ≥ 13 years was 1434.3 versus 198.7 (per 100,000) [1]. Despite this incidence and prevalence gap, a lower proportion of Black compared to White persons with diagnosed HIV (PWDH) have been virally suppressed (< 200 copies/mL) [2, 3]. Reducing racial inequities in HIV-related care will be critical to the success of the Ending the HIV Epidemic: A Plan for America (EHE) [4].
Addressing social determinants of health (SDH) is a key component in reducing HIV-related health inequities [5, 6]. SDH factors associated with lack of testing, poor HIV care access, nonadherence to treatment, or not reaching viral suppression include: absence of health insurance [7]; absence of local health facilities [7]; stigma, discrimination in healthcare settings, or mistrust in providers [8, 9]; lack of peer social support [10]; residing in poor neighborhoods [11, 12]; receiving an annual income below the poverty level [13]; having low educational attainment [11] or lack of health knowledge [14]; housing instability [15]; homelessness [16, 17]; and being unemployed [18]. However, more research is needed to examine HIV care outcomes as a function of structural factors that influence SDH, specifically those that widen the gap of wealth and financial instability as well as those that reduce the financial burden of care and increase access to HIV medical services.
Three structural factors that may influence HIV care inequities are worthy of further exploration. One factor that impedes acquisition of wealth and financial stability has been examined in other public health domains (e.g., breast cancer, preterm birth, pregnancy health) but not in HIV is racial mortgage redlining (hereafter referred as “redlining”) [19–21]. Redlining is a discriminatory practice by mortgage companies that prevents minority populations from acquiring home loans [20]. Redlining can contribute to the concentration of those impacted by it into poor neighborhoods as well as deprive them of acquiring financial stability and attaining wealth through homeownership [22]. The widening of the racial wealth gap may be contributing to inequities in general health and wellbeing [23]. Mendez et al. used mortgage application data provided by the Home Mortgage Disclosure Act (HMDA) to calculate a redlining index, an index that compared mortgage rejection odds between Black and White applicants after adjusting for loan amount, income, and sex of the applicant [21]. This study estimated a redlining index of 2.0 for Philadelphia County, Pennsylvania, which indicated that the odds of mortgage rejection for Black loan applicants was twice the odds of rejection for White applicants. A recent study showed that the redlining index was positively associated with preterm birth rates among Black women [20]. No analysis has examined whether redlining is associated with poor HIV care outcomes in PWDH.
Two other structural factors intended to reduce the financial burden of HIV care and increase access to HIV medical services include the state adoption of the Medicaid expansion policy and the use of the Ryan White HIV/AIDS Program (RWHAP). Medicaid is the largest insurer for adult PWDH, covering 40% of PWDH in 2018 [24]. One study has shown that Medicaid insurance coverage increased for PWDH following the implementation of Medicaid expansion policy [25]. Lack of insurance has been a risk factor for poor access to HIV-related care [7]. The Medicaid expansion policy has been positively associated with increased HIV testing and use of antiretroviral treatments [26]. This policy has also showed promise with helping PWDH reach viral suppression [26]; however, more rigorous research of this association accounting for other structural factors and social determinants of health is needed. The RWHAP is a program designed to help provide HIV medical care services and ancillary services for low-income uninsured or underinsured PWDH [27]. Every year, more than half of all PWDH in the United States receive services through RWHAP [27]. Among PWDH enrolled in the program, the difference in prevalence of viral suppression between Black and White enrollees was 13.0 percentage points in favor of White enrollees in 2010 but it nearly halved to 8.1 percentage points in 2016 [5]. It is plausible that PWDH who reside in states with Medicaid expansion or states where RWHAP is serving the majority of PWDH would more likely be linked to HIV care expeditiously following diagnosis and have viral suppression versus their counterparts who reside outside of these states. Insurance assistance and ancillary services through these programs are more available and accessible to PWDH to reduce the financial burden of HIV care and address structural barriers associated HIV care access (e.g., transportation).
Expanding on the existing SDH and HIV literature, we examined associations of census tract-level characteristics of redlining and state-level characteristics of Medicaid expansion and RWHAP use in relation to two HIV care outcomes, linkage to HIV care within 1 month after diagnosis and attaining viral suppression, for Black and White PWDH who were initially diagnosed with HIV in 2017 to help inform HIV prevention efforts toward the EHE goals [4].
Methods
Study Design and Population
Using data from Centers for Disease Control and Prevention’s National HIV Surveillance System (NHSS), the U.S. Census Bureau’s American Community Survey (ACS), and the Home Mortgage Disclosure Act (HMDA), we conducted a cross-sectional analysis to examine HIV care outcomes in relation to structural and SDH factors among Black and White PWDH aged ≥ 18 years (i.e., adults), with HIV diagnosed in 2017, and alive at year-end 2018 in 38 jurisdictions that had complete laboratory reporting, geocoding, and census tract-level redlining indexes (i.e., Alabama, Alaska, California, Colorado, Delaware, Florida, Georgia, Hawaii, Illinois, Indiana, Iowa, Louisiana, Maryland, Michigan, Minnesota, Mississippi, Missouri, Montana, Nebraska, Nevada, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Utah, Virginia, Washington, West Virginia, Wisconsin, and Wyoming). The NHSS data were used to distinguish between Black and White PWDH. The NHSS collects information on race/ethnicity from existing documents (e.g., medical records, records from partner services). Details on how the standard NHSS race/ethnicity categories are generated can be reviewed in the technical notes at https://www.cdc.gov/hiv/library/reports/hiv-surveillance/vol-32/content/technical-notes.html. Figure 1 provides details on the selection process for the analytic sample.
Fig. 1.
Study population selection steps
Datasets and Linkage
Datasets for this analysis included:
The Geocoding and Data Linkage (GDL) data contained HIV surveillance data for person with HIV initially diagnosed in the period of January 1, 2017 to December 31, 2017. The addresses of residence at the time of diagnosis were geocoded to the census tract-level and then linked, at the census tract-level, to SDH factors (i.e., poverty, education, income, and health insurance coverage) from the ACS 2013–2017 5-year estimates.
NHSS data contained age, sex at birth, transmission category, and HIV care outcomes for cases that could be geo-linked and the corresponding SDH variable pulled from ACS.
Home Mortgage Disclosure Act-ACS linked data (HDMA-ACS) contained data used to generate the census tract redlining index.
We used a unique person identifier to link the GDL and NHSS data, to retrieve the demographic information and HIV care outcomes on PWDH. The NHSS database available as of December 31, 2019 was used, which allowed 12 months to report the 2018 viral suppression status in the NHSS for those diagnosed with HIV in 2017. We used the census tract variable to link the GDL dataset to the HDMA-ACS dataset to link PWDH to their respective redlining index. The unit of analysis was the individual record in the NHSS data.
HIV Care Outcomes
Similar to a previous report, we included two HIV care outcomes [28]:
Linked to HIV medical care within 1 month after diagnosis: measured as having documentation of at least one CD4 or viral load test performed ≤ 1 month after diagnosis
Viral suppression in 2018: measured as having a documented viral load result of < 200 copies/mL at the most recent viral load test during 2018
Structural Factors
Redlining, a census tract-level index incorporated accumulated mortgage application data from 2014 to 2017 (see below)
Residing in a state with Medicaid expansion implemented during 2014–2016 (see data source at https://www.kff.org/medicaid/issue-brief/status-of-state-medicaid-expansion-decisions-interactive-map/)
Residing in a state with > 50% PWDH receiving RWHAP services based on the 2017 data from the Health Resources and Services Administration (This threshold marked the median percent of PWDH using RWHAP services across 50 U.S. states and Washington D.C.).
SDH Factors
Five census tract variables from the ACS 2013–2017 5-year estimates:
Racial segregation (% of residents in the census tract that are of Black race)
Poverty (% of residents with income below the poverty level)
Education (% of residents with less than a high school education)
Median household income (median income for a household)
Insurance coverage (% of residents without health insurance)
For the five SDH factors, we created categories by using empirically derived quartiles, with each quartile cut-point rounded to the nearest integer and determined based on data from all census tracts in the United States and Puerto Rico.
Individual-Level Factors
Additionally, three individual-level factors included in the analysis were selected based on the World Health Organization’s (WHO’s) Conceptual Framework for Action on the Social Determinants of Health [29]:
Age at diagnosis
Sex at birth: male and female
Transmission category: male-to-male sexual contact, injection drug use (IDU); male-to-male sexual contact and IDU; and heterosexual contact. Individuals are assigned to one category using hierarchical rules of probability to identify the one risk factor most likely responsible for transmission. A multiple imputation method was used to assign individuals with missing risk factor information (see the technical notes on assigning transmission categories located at https://www.cdc.gov/hiv/library/reports/hiv-surveillance/vol-32/content/technical-notes.html
Redlining Index Generation
We used the HMDA datasets to derive the racial mortgage redlining index at the census tract-level in the United States. To examine the Black-versus-White disparity in loan disposition, we limited the analytic sample to applicants who identified themselves as Black or White on their loan applications. Using the method developed by Mendez et al., we created the redlining index to compare the odds of loan rejection among Black applicants to the odds of loan rejection among White applicants after controlling for the applicant’s sex, gross annual income, and loan amount [21]. Random effects models were used. Any significant odds ratio > 1.0 indicated the odds of rejection for Black applicants was higher than that of White applicants. Among the census tracts in our study of which the redlining index values were available, the mean of the redlining index was 2.0; the median was 1.8, and the standard deviation was 1.1; the influence of outliers should be limited as suggested by the closeness of mean and median. Mendez also estimated the mean redlining index to be 2.0 (ranging 0.01 to 7.0) across the census tracts in Philadelphia County, Pennsylvania [21]. We compared those residing in tracts with scores higher than 2.0 in our analysis with those in tracts at or below 2.0. GLIMMIX Procedure in SAS® version 9.4 was used to create the redlining indices.
Statistical Analysis
We first described Black and White PWDH by individual-level characteristics, structural and SDH factors, and HIV care outcomes. Bivariate statistics for indicating differences in characteristics between the two race groups were not provided, because we used surveillance (population) data and not sampled data in the analysis. To estimate the independent association between each covariate and each outcome adjusting for all other factors ascertained, we used Poisson regression with robust variance estimates to provide adjusted prevalence ratios (aPRs). All aPRs accounted for all structural, SDH, and individual-level factors. The outcomes were too rare for the regression models to examine the differential associations between the covariates and the outcomes by race via interaction terms. Therefore, we only examined each population independently; all analyses were stratified by race. Because two outcome measures were analyzed for each race group, we included 99% confidence intervals (99%CIs) to reduce the overall probability of having a type 1 error. Analyses were conducted using the SAS 9.4 software. When aPRs were calculated, variance inflation factor (VIF) and condition index (C) were used to diagnose and then correct redundancy/collinearity among the structural and SDH factors. Thresholds of VIF ≥ 10 or C ≥ 10 were used to determine if redundancy/collinearity existed among factors.
Results
In 2017, 21,546 Black or White persons, aged ≥ 18 years, were initially diagnosed with HIV, were alive at year-end 2018, and lived in states with complete laboratory reporting (12,846 census tracts). However, a total of 12,996 (60.3%) PWDH residing in 8,769 (68.3%) census tracts met the final inclusion criteria for this analysis (Fig. 1). Redlining was widespread with higher index values in the southeast census tracts. Figure 2 shows the included census tracts with > 2.0 indexes.
Fig. 2.
Redline index by census tracta. aThe redlining index was generated from the Home Mortgage Disclosure Act data source. The index compares the odds of mortgage rejection among Black applicants versus that of White applicants after accounting for applicant’s sex, gross annual income, and loan amount. The mean index across census tracts examined in this analysis was 2.0, indicating that the odds of loan denial for Black people was twice that for White people
Of the 12,996 PWDH included in the analysis, 8,487 (65.3%) were of Black race and 4509 (34.7%) were of White race (Table 1). Many demographic characteristics were similar for both race groups, such as high proportions were aged 18–34 years, male, and with HIV infection attributed to male-to-male sexual contact. Although, the proportion of White PWDH with infection attributed to injection drug use was more than twice that of Black PWDH (8.6% versus 3.7%) (Table 1).
Table 1.
Characteristics of Black and White adults with HIV diagnosed in 2017 and alive at year-end 2018, 38 Jurisdictions
Black No. (%) | White No. (%) | |
---|---|---|
| ||
Total | 8487 (100.0) | 4509 (100.0) |
Age at diagnosis (years) | ||
18–24 years | 2290 (27.0) | 671 (14.9) |
25–34 years | 3092 (36.4) | 1524 (33.8) |
35–44 years | 1374 (16.2) | 882 (19.6) |
45–54 years | 970 (11.4) | 871 (19.3) |
55 years and older | 761 (9.0) | 561 (12.4) |
Sex at birth | ||
Male | 6336 (74.7) | 3890 (86.3) |
Female | 2151 (25.3) | 619 (13.7) |
Transmission categorya | ||
Male-to-male sexual contact | 5153 (60.7) | 3201 (71.0) |
Injection drug use (IDU)b | 317 (3.7) | 386 (8.6) |
Male-to-male sexual contact/IDU | 138 (1.6) | 274 (6.1) |
Heterosexual contactb,c | 2862 (33.7) | 642 (14.2) |
Otherb | 17 (0.2) | 6 (0.1) |
% in census tract Black/African American | ||
< 24.99% | 2591 (30.5) | 3636 (80.6) |
25–74.99% | 3544 (41.8) | 799 (17.7) |
≥ 75% | 2352 (27.7) | 74 (1.6) |
% in census tract below federal poverty level | ||
< 7% | 907 (10.7) | 979 (21.7) |
7.00–10.99% | 976 (11.5) | 958 (21.2) |
11.00–18.99% | 2291 (27.0) | 1360 (30.2) |
≥ 19% | 4313 (50.8) | 1212 (26.9) |
% in census tract with less than high school diploma | ||
< 6% | 981 (11.6) | 1152 (25.5) |
6.00–10.99% | 1642 (19.3) | 1243 (27.6) |
11.00–17.99% | 2465 (29.0) | 1074 (23.8) |
≥ 18% | 3399 (40.0) | 1040 (23.1) |
Median household income in census tract, $ | ||
< 40,000 | 4332 (51.0) | 1028 (22.8) |
40,000–53,999 | 1948 (23.0) | 1266 (28.1) |
54,000–74,999 | 1484 (17.5) | 1293 (28.7) |
≥ 75,000 | 715 (8.4) | 920 (20.4) |
Unknown | 8 (0.1) | 2 (0.0) |
% in census tract without health insurance | ||
< 6 | 531 (6.3) | 789 (17.5) |
6.00–9.99 | 1226 (14.4) | 949 (21.0) |
10.00–15.99 | 2319 (27.3) | 1287 (28.5) |
≥ 16 | 4411 (52.0) | 1484 (32.9) |
Census tract redlining indexd | ||
< 2.0 | 3183 (37.5) | 2130 (47.2) |
≥ 2.0 | 5304 (62.5) | 2379 (52.8) |
Resides in a state with expanded Medicaid policy | ||
No | 5238 (61.7) | 2418 (53.6) |
Yes | 3249 (38.3) | 2091 (46.4) |
Resides in a state with > 50% PLWH receiving Ryan White Programs | ||
No | 4642 (54.7) | 2617 (58.0) |
Yes | 3845 (45.3) | 1892 (42.0) |
Linked to HIV care within one month after diagnosis | ||
No | 2054 (24.2) | 801 (17.7) |
Yes | 6433 (75.8) | 3708 (82.3) |
Viral suppression in 2018 | ||
No | 2734 (32.2) | 1024 (22.7) |
Yes | 5753 (67.8) | 3485 (77.3) |
Data have been statistically adjusted to account for missing transmission category
Categories include both males and females
Heterosexual contact with a person known to have or to be at high risk for, HIV infection
The redlining index compares the mortgage rejections of Black/African American applicants to White applicants after accounting for loan amount, gross annual income, and sex of applicant (higher index values represent greater rejection inequities)
Regarding the structural and SDH factors, some similarities were observed for both race groups. There were high proportions of PWDH residing in states without Medicaid expansion (Black = 61.7%; White = 53.6%) and in states without high proportions (> 50%) of PWDH receiving RWHAP services (Black = 54.7%; White = 58.0%) (Table 1). For the remaining factors, there were stark differences between the two groups. When comparing Black to White PWDH, 50.8% vs. 26.9%, respectively, resided in tracts where ≥ 19% of residents lived below the poverty level; 40.0% vs. 23.1%, respectively, resided in tracts where ≥ 18% of residents had less than a high school education; 51.0% vs. 22.8%, respectively, resided in tracts where the median household income was < $40,000; and 52.0% vs. 32.9%, respectively, resided in tracts where ≥ 16% of the residents did not have health insurance. Further, 62.5% of Black PWDH vs. 52.8% of White PWDH resided in tracts where the redlining index was > 2.0. An examination of these factors by race among those who were linked to HIV care and among those with viral suppression are provided in the appendix (See Appendix in the online materials for details).
Disparities in HIV care outcomes by race were also observed (Table 1). Respectively, 82.3% of White PWDH vs. 75.8% of Black PWDH were linked to HIV care within 1 month after diagnosis; 77.3% of White PWDH vs. 67.8% of Black PWDH were virally suppressed in 2018.
Among Black PWDH, a higher prevalence of being linked to HIV care within 1 month after HIV diagnosis was observed for those who were female (vs. male) (aPR: 1.05; 99%CI 1.01–1.11, p < 0.01), residing in tracts with 6.0–9.99% without health insurance (vs. residing in tracts ≥ 16% without health insurance) (aPR: 1.06; 99%CI 1.00–1.11, p < 0.01) and residing in states with Medicaid expansion (vs. not residing in such states) (aPR: 1.06; 99%CI 1.02–1.10, p < 0.01) (Table 2).
Table 2.
Factors associated with linked to care and viral suppression for Black and White adults with HIV diagnosed in 2017 and alive at year-end 2018, 38 jurisdictions
Factors | Linked to HIV care within one month after diagnosis |
Viral suppression in 2018 |
||||||||
---|---|---|---|---|---|---|---|---|---|---|
Black |
White |
Black |
White |
Black | White | |||||
No. (%)a | aPR (99%CI) | No. (%) | aPR (99%CI) | No. (%) | aPR (99%CI) | No. (%) | aPR (99%CI) | Total | Total | |
Age at diagnosis (years) | ||||||||||
18–24 years | 1672 (73.0) | 0.98 (0.94, 1.02) | 541 (80.6) | 1.01 (0.96, 1.08) | 1472 (64.3) | 0.95 (0.91, 1.00) | 510 (76.0) | 0.98 (0.92, 1.05) | 2290 | 671 |
25–34 years | 2324 (75.2) | Reference | 1207 (79.2) | Reference | 2088 (67.5) | Reference | 1160 (76.1) | Reference | 3092 | 1524 |
35–44 years | 1056 (76.9) | 1.01 (0.96, 1.06) | 738 (83.7) | 1.06 (1.00, l.ll)*f | 950 (69.1) | 1.03 (0.97, 1.09) | 681 (77.2) | 1.02 (0.96, 1.08) | 1374 | 882 |
45–54 years | 771 (79.5) | 1.04 (0.99, 1.10) | 741 (85.1) | 1.07 (1.02, 1.12)* | 703 (72.5) | 1.07(1.01, 1.14)* | 688 (79.0) | 1.03 (0.97, 1.09) | 970 | 871 |
55 years and older | 610 (80.2) | 1.04 (0.98, 1.10) | 481 (85.7) | 1.08 (1.02, 1.14)* | 540 (71.0) | 1.05 (0.98, 1.13) | 446 (79.5) | 1.03 (0.96, 1.10) | 761 | 561 |
Sex at birth | ||||||||||
Male | 4712 (74.4) | Reference | 3209 (82.5) | Reference | 4245 (67.0) | Reference | 3040(78.1) | Reference | 6336 | 3890 |
Female | 1721 (80.0) | 1.05 (1.01, 1.11)* | 499 (80.6) | 1.00 (0.92, 1.08) | 1508 (70.1) | 1.02 (0.96, 1.08) | 445 (71.9) | 0.94 (0.85, 1.04) | 2151 | 619 |
Transmission categoryb | ||||||||||
Male-to-male sexual contact | 3831 (74.4) | Reference | 2674 (83.5) | Reference | 3460 (67.1) | Reference | 2540 (79.3) | Reference | 5153 | 3201 |
Injection drug use (IDU)c | 240 (75.7) | 1.00 (0.95, 1.05) | 312(80.8) | 0.98 (0.91, 1.05) | 195 (61.4) | 0.93 (0.86, 1.00)*f | 253 (65.5) | 0.87 (0.79, 0.96)* | 317 | 386 |
Male-to-male sexual contact/IDU | 96 (69.7) | 0.99 (0.89, 1.10) | 200 (72.9) | 0.89 (0.81, 0.98)* | 83 (60.1) | 0.96 (0.85, 1.08) | 199 (72.6) | 0.91 (0.83, 1.00) | 138 | 274 |
Heterosexual contactc,d | 2250 (78.6) | 1.01 (0.97, 1.05) | 517 (80.6) | 0.98 (0.91, 1.05) | 2004 (70.0) | 1.02 (0.97, 1.07) | 490 (76.3) | 1.01 (0.93, 1.10) | 2862 | 642 |
Otherc | 16 (94.1) | - | 5 (83.3) | - | 12 (70.6) | - | 4 (66.7) | - | 17 | 6 |
% in census tract Black/African American | ||||||||||
<24.99% | 1959 (75.6) | 0.99 (0.95, 1.03) | 2987 (82.2) | 0.97 (0.92, 1.02) | 1795 (69.3) | 1.02 (0.97, 1.07) | 2832 (77.9) | 1.00 (0.94, 1.06) | 2591 | 3636 |
25–74.99% | 2658 (75.0) | Reference | 663 (83.0) | Reference | 2410 (68.0) | Reference | 609 (76.2) | Reference | 3544 | 799 |
≥75% | 1816 (77.2) | 1.03 (0.99, 1.08) | 58 (78.4) | 0.97 (0.83, 1.14) | 1548 (65.8) | 0.97 (0.92, 1.02) | 44 (59.5) | 0.80 (0.62, 1.02) | 2352 | 74 |
% in census tract below federal poverty level | ||||||||||
<7% | 706 (77.8) | 1.00 (0.92, 1.08) | 802 (81.9) | 0.95 (0.88, 1.02) | 621 (68.5) | 1.02 (0.92, 1.13) | 788 (80.5) | 1.01 (0.92, 1.11) | 907 | 979 |
7.00–10.99% | 763 (78.2) | 1.02 (0.95, 1.09) | 788 (82.3) | 0.96 (0.90, 1.03) | 698 (71.5) | 1.08 (0.99, 1.17) | 738 (77.0) | 0.98 (0.90, 1.07) | 976 | 958 |
11.00–18.99% | 1717 (74.9) | 0.98 (0.93, 1.03) | 1136 (83.5) | 1.00 (0.94, 1.06) | 1562 (68.2) | 1.02 (0.96, 1.08) | 1055 (77.6) | 1.00 (0.93, 1.07) | 2291 | 1360 |
≥19% | 3247 (75.3) | Reference | 982 (81.0) | Reference | 2872 (66.6) | Reference | 904 (74.6) | Reference | 4313 | 1212 |
% in census tract with less than high school diploma | ||||||||||
<6% | 750 (76.5) | 0.99 (0.93, 1.06) | 973 (84.5) | 1.05 (0.98, 1.13) | 692 (70.5) | 1.05 (0.97, 1.13) | 924 (80.2) | 1.11 (1.02, 1.20)* | 981 | 1152 |
6.00–10.99% | 1243 (75.7) | 0.99 (0.94, 1.04) | 1022 (82.2) | 1.02 (0.96, 1.09) | 1136 (69.2) | 1.03 (0.96, 1.09) | 977 (78.6) | 1.09(1.01, 1.18)* | 1642 | 1243 |
11.00–17.99% | 1899 (77.0) | 1.01 (0.97, 1.06) | 887 (82.6) | 1.03 (0.97, 1.09) | 1672 (67.8) | 1.01 (0.96, 1.07) | 848 (79.0) | 1.10(1.03, 1.18)* | 2465 | 1074 |
≥18% | 2541 (74.8) | Reference | 826 (79.4) | Reference | 2253 (66.3) | Reference | 736 (70.8) | Reference | 3399 | 1040 |
Median household income in census tract, $ | ||||||||||
<40,000 | 3227 (74.5) | Reference | 820 (79.8) | Reference | 2896 (66.9) | Reference | 760 (73.9) | Reference | 4332 | 1028 |
40,000–53,999 | 1496 (76.8) | 1.04 (0.99, 1.10) | 1045 (82.5) | 1.03 (0.97, 1.10) | 1342 (68.9) | 0.99 (0.93, 1.05) | 978 (77.3) | 1.00 (0.93, 1.07) | 1948 | 1266 |
54,000–74,999 | 1135 (76.5) | 1.03 (0.96, 1.10) | 1080 (83.5) | 1.05 (0.98, 1.13) | 1009 (68.0) | 0.95 (0.87, 1.03) | 1004 (77.6) | 0.99 (0.91, 1.08) | 1484 | 1293 |
≥75,000 | 567 (79.3) | 1.05 (0.97, 1.15) | 762 (82.8) | 1.05 (0.96, 1.14) | 499 (69.8) | 0.97 (0.86, 1.08) | 742 (80.7) | 1.02 (0.92, 1.13) | 715 | 920 |
Unknown | 8 (100) | - | 1 (50.0) | - | 7 (87.5) | - | 1 (50.0) | - | 8 | 2 |
% in census tract without health insurance | ||||||||||
<6 | 413 (77.8) | 1.02 (0.95, 1.10) | 673 (85.3) | 1.03 (0.95, 1.10) | 367 (69.1) | 0.99 (0.90, 1.09) | 634 (80.4) | 0.98 (0.90, 1.07) | 531 | 789 |
6.00–9.99 | 977 (79.7) | 1.06 (1.00, l.ll)*f | 785 (82.7) | 1.00 (0.94, 1.07) | 856 (69.8) | 1.02 (0.95, 1.09) | 742 (78.2) | 0.98 (0.91, 1.05) | 1226 | 949 |
10.00–15.99 | 1803 (77.7) | 1.04(1.00, 1.08) | 1067 (82.9) | 1.01 (0.96, 1.06) | 1606 (69.3) | 1.02 (0.97, 1.08) | 1007 (78.2) | 1.00 (0.94, 1.06) | 2319 | 1287 |
≥16 | 3240 (73.5) | Reference | 1183 (79.7) | Reference | 2924 (66.3) | Reference | 1102(74.3) | Reference | 4411 | 1484 |
Census tract redlining indexe | ||||||||||
<2.0 | 2441 (76.7) | Reference | 1775 (83.3) | Reference | 2149 (67.5) | Reference | 1650 (77.5) | Reference | 3183 | 2130 |
≥2.0 | 3992 (75.3) | 1.00 (0.96, 1.02) | 1933 (81.3) | 0.97 (0.94, 1.01) | 3604 (67.9) | 1.01 (0.97, 1.06) | 1835 (77.1) | 1.00 (0.96, 1.04) | 5304 | 2379 |
Resides in a state with expanded Medicaid policy | ||||||||||
No | 3843 (73.4) | Reference | 1927 (79.7) | Reference | 3520 (67.2) | Reference | 1833 (75.8) | Reference | 5238 | 2418 |
Yes | 2590 (79.7) | 1.06 (1.02, 1.10)* | 1781 (85.2) | 1.06(1.01, 1.10)* | 2233 (68.7) | 1.01 (0.96, 1.05) | 1652 (79.0) | 1.03 (0.99, 1.08) | 3249 | 2091 |
Resides in a state with > 50% PLWH receiving Ryan White Programs | ||||||||||
No | 3481 (75.0) | Reference | 2101 (80.3) | Reference | 3067 (66.1) | Reference | 1978 (75.6) | Reference | 4642 | 2617 |
Yes | 2952 (76.8) | 1.01 (0.98, 1.05) | 1607 (84.9) | 1.05 (1.01, 1.09)* | 2686 (69.9) | 1.07 (1.02, 1.11)* | 1507 (79.7) | 1.05 (1.00, 1.10)*f | 3845 | 1892 |
% is a row percent to show the prevalence of the outcome for each stratified group
Data have been statistically adjusted to account for missing transmission category
Categories include both males and females
Heterosexual contact with a person known to have or to be at high risk for HIV infection
The redlining index compares the mortgage rejections of Black/African American applicants to White applicants after accounting for loan amount, gross annual income, and sex of applicant (higher index values represent greater rejection inequities)
aPR was significant at the 0.01 level. The lower or upper bound of 99%CI presents 1.00 because of rounding
P < 0.01
Among White PWDH, compared to those aged 25–34 years, a higher prevalence of being linked to HIV care within 1 month after HIV diagnosis was found for those aged 35–44 years (aPR: 1.06; 99%CI 1.00–1.11, p < 0.01), 45–54 years (aPR: 1.07; 99%CI 1.02–1.12, p < 0.01), and ≥ 55 years (aPR: 1.08; 99%CI 1.02–1.14, p < 0.01) (Table 2). A higher prevalence of being linked to HIV care within 1 month after HIV diagnosis was also observed for those residing in states with Medicaid expansion (vs. not residing in such states) (aPR: 1.06; 99%CI 1.01–1.10, p < 0.01) and residing in states with > 50% PWDH receiving RWHAP (vs. not residing in such states) (aPR: 1.05; 99%CI 1.01–1.09, p < 0.01). White PWDH with HIV possibly attributed to either male-to-male sexual contact or injection drug use (vs. male-to-male sexual contact alone) had a lower prevalence of being linked to HIV care within 1 month after diagnosis (aPR: 0.89; 99%CI 0.81–0.98, p < 0.01).
Regarding viral suppression among Black PWDH, a higher prevalence of having viral suppression in 2018 was found for those who were aged 45–54 years (vs. 25–34 years) (aPR: 1.07; 99%CI 1.01–1.14, p < 0.01) and residing in states with > 50% PWDH receiving RWHAP (vs. not residing in such states) (aPR: 1.07; 99%CI 1.02–1.11, p < 0.01) (Table 2). Further, among Black PWDH, those with HIV infection attributed to injection drug use (vs. male-to-male sexual contact) had a lower prevalence of being virally suppressed in 2018 (aPR: 0.93; 99%CI 0.86–1.00, p < 0.01).
Among White PWDH, compared to those residing in tracts with ≥ 18% of residents having less than a high school diploma, a higher prevalence of viral suppression in 2018 was found for those in tracts where the percent of residents with less than a high school diploma was < 6% (aPR: 1.11; 99%CI 1.02–1.20, p < 0.01), 6.00–10.99% (aPR: 1.09; 99%CI 1.01–1.18, p < 0.01), and 11.00–17.99% (aPR: 1.10; 99%CI 1.03–1.18, p < 0.01) (Table 2). The prevalence of viral suppression in 2018 was also higher for those residing in states with > 50% PWDH receiving RWHAP (vs. not residing in such states) (aPR: 1.05; 99%CI 1.00–1.10, p < 0.01). Last, those with HIV infection attributed to injection drug use (vs. male-to-male sexual contact) also had a lower prevalence of being virally suppressed in 2018 (aPR: 0.87; 99%CI 0.79–0.96, p < 0.01).
For both populations, none of the outcomes were associated with the redlining index.
Discussion
To our knowledge, this is the first analysis that simultaneously examined redlining, Medicaid expansion, and RWHAP use in relation to HIV care outcomes for Black and White PWDH in the United States. Black PWDH in our analysis fared poorly in each SDH category compared to White PWDH. Direct exposure to areas with redlining was not associated with poor HIV care outcomes. However, factors that reduced the financial burden of HIV care and improved care access such as Medicaid expansion and RWHAP were positively associated with HIV care outcomes.
Expansion of state Medicaid programs has been linked to increases in coverage, service use, and quality of care [30]. For both race groups, our analysis found that PWDH who resided in a state with expanded Medicaid coverage had a higher prevalence of being linked to care within 1 month after diagnosis compared to those in states without such benefits, which built upon previous findings that Medicaid expansion was associated with increased HIV testing and diagnoses [26, 31]. While our findings were encouraging, the relationship between Medicaid expansion and viral suppression warrants further research. What is covered for PWDH under Medicaid expansion can differ by state and the variation might have reduced the probability to detect a positive association with viral suppression in this analysis. Multi-level analyses using census tracts and states as levels or the inclusion of random effects in the statistical modelling might help mitigate some of this problem. Although, a recent study found that among PWDH in Washington D.C., those privately insured had greater durable viral suppression than those publicly insured, pointing out the need for more research to further examine HIV quality of care in association with Medicaid expansion [32]. Longitudinal analyses examining HIV viral suppression in relation to the timing of Medicaid expansion implementation across states may provide additional evidence for the impact of this policy on achieving the EHE goals.
Residing in states with higher levels of enrollment in RWHAP also showed promising associations for both Black (viral suppression) and White PWDH (linkage to care and viral suppression). Given there was no available census tract-level data on RWHAP use for this analysis, we used the overall median (> 50% PWDH receiving RWHAP services across all 50 U.S. states and Washington D.C.) as a proxy indicator of accessibility to HIV primary medical care and essential support services among low income PWDH for each jurisdiction. RWHAP improves insurance coverage for HIV care among populations experiencing financial disadvantages and provides ancillary services that address structural barriers (e.g., housing, transportation, mental health and substance abuse services). States with higher RWHAP usage might also have better systems in place to help clients and providers navigate various programs provided and funded under Ryan White to meet client’s needs. As more states expand their Medicaid program, eligible enrollees for RWHAP may decide to receive financial support from Medicaid coverage alone [25]; however, one study showed that prevalence of viral suppression is much higher among those who have RWHAP coverage in combination with public insurance coverage compared to those with public insurance coverage alone [24]. Therefore, a deeper examination of the joint impact of both programs on reducing HIV care inequities is warranted.
Our analysis examined the redlining index across census tracts spanning multiple states making it one of the largest analyses to examine this index in relation to health outcomes. It is also the first to examine redlining in relation to HIV care outcome inequities. Our estimated average census tract-level index of 2.0 was similar to the estimated index provided by Mendez and colleagues for Philadelphia County [21]. However, we did not find any associations between exposure to redlining and the HIV care outcomes for both Black and White PDWH after controlling for Medicaid expansion and RWHAP service use, programs/policies that reduced financial barriers and improved access to HIV care. The inequities in mortgaging practices in no doubt contribute to the racial wealth gap [22]. Redlining is an upstream cause that may contribute to other factors that result in negative outcomes of HIV care for those subpopulations disproportionately affected by it. Our analysis is only an initial exploration into mortgaging inequities and HIV care outcomes; more work will hopefully build on our findings to inform efforts for reducing HIV care inequities related to housing [33].
For Black and White PWDH, older persons had a higher prevalence of being linked to care within 1 month after diagnosis or having viral suppression, while those whose infection attributed to injection drug use had a lower prevalence of viral suppression. Among White PWDH, those living in areas with more educated residents also had a higher prevalence of viral suppression than their counterparts who resided in other areas, a finding similar to that of a previous study on the education status and HIV care outcomes [34]. More research is needed to explore the underpinnings of these associations but at least suggest a need to heighten efforts that improve linkage to care and viral suppression among younger adults, persons who inject drugs, and persons with lower education.
Some limitations with the data sources warrant consideration. Data were limited to persons whose residential addresses were complete and could be geocoded, who lived in jurisdictions with complete laboratory reporting, and who resided in tracts with generated HMDA redlining indexes; therefore, results did not reflect the entire population of PWDH in the United States. However, based on individual-level demographic factors (i.e., race, age, sex, transmission risk) and residential census tract characteristics (e.g., % Black/African American, % below the federal poverty level, % less than a high school diploma, median household income, and % without health insurance), the included and excluded samples were very similar suggesting that a bias related to these factors might be minimal. Another data limitation was that the NHSS data geocoded used the residential address at the time of diagnosis. It was possible that some of the individuals moved during the 2018 assessment period but were still counted in their residential census tract at time of diagnosis. We also did not have access to information detailing whether individual PWDH were receiving RWHAP; only state-level data were available (i.e., > 50% PLWH receiving RWHAP) as a proxy. Similarly, our analysis examined whether the individuals resided in an area with evidence of redlining rather than examine whether they were personally rejected on a mortgage loan.
Further, because our analysis used secondary resources, some limitations to the analysis must also be considered. The process for categorizing Black and White persons in the HMDA database might be different from the process of categorizing race in the NHSS. Therefore, we could only approximate the association between the racial redlining index and the outcomes for Black and White PWDH. Data on antiretroviral treatment were not well captured in the NHSS and thus analysis was not conducted to account for access to these treatments, which might have impacted some of the associations with viral suppression. Another analytic limitation was that there was insufficient power to examine differential associations (i.e., interaction terms) between the structural and SDH factors and the outcomes by race. We were only able to conduct stratified analyses by race and one should not make direct comparisons in these associations (e.g., even if a covariate was statistically significant with an outcome for one race group but not the other race group, one still cannot infer the two race groups significantly differed by these associations). Finally, our analysis was a cross-sectional examination of associations and therefore does not imply causality.
Conclusions
Our analysis of Black and White adults with diagnosed HIV provides insight into the impact of structural and SDH factors that affect wealth acquisition, through homeownership, and medical care access on HIV care outcomes. Programs and policies that decrease financial burden of care and increase access to HIV treatments such as Medicaid expansion and the RWHAP might improve health outcomes for persons with HIV. Although redlining was not directly associated with these HIV care outcomes, more work is needed to examine if redlining perpetuates circumstances such as poverty for those disproportionately affected, which in turn might elevate risk of disease infection. Also, while this study focused on HIV care disparities between Black versus White PWDH, more research is needed to examine factors in relation to HIV care outcome inequities and disparities between White persons and other minority populations that historically have been disadvantaged with respect to acquiring community health care resources. This initial analysis focused on White and Black PWDH because these racial groups provided sufficient samples to examine the care outcomes of interest at the census tract-level. A future study is planned to examine the association between the redlining index and disparities in HIV diagnoses across multiple race/ethnic groups by U.S. county. The WHO’s Conceptual Framework for Action on the Social Determinants of Health can be an instrument to guide such research [29]. To reach the goals set forth by EHE and other federal initiatives, more attention is needed to improve access to care for those subgroups and areas negatively affected by structural and SDH factors as well as improve programs to achieve the goals of promoting health equity and reduce health disparities among persons with HIV [35].
Supplementary Material
Funding
Centers for Disease Control and Prevention.
Footnotes
Conflict of interest None of the authors have a conflict of interest or competing interest.
Disclaimer The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Code Availability Not applicable.
Ethical Approval Not applicable.
Consent to Participate Not applicable.
Consent for Publication Not applicable.
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10461-022-03641-5.
Data Availability
All datasets used in this study were public data sources.
References
- 1.Centers for Disease Control and Prevention. Estimated HIV incidence and prevalence in the United States, 2014–2018. HIV Surveillance Supplemental Report 2020;25(No. 1). https://www.cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-supplemental-report-vol-25-1.pdf. Accessed 23 Sept 2020. [Google Scholar]
- 2.Nance RM, Delaney JAC, Simoni JM, et al. HIV Viral suppression trends over time among HIV-infected patients receiving care in the United States, 1997 to 2015: a cohort study. Ann Intern Med. 2018;169(6):376–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Centers for Disease Control and Prevention. Monitoring selected national HIV prevention and care objectives by using HIV surveillance data—United States and 6 dependent areas, 2018. HIV Surveillance Supplemental Report 2020;25(No. 2). https://www.cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-supplemental-report-vol-25-2.pdf. Accessed 23 Sept 2020. [Google Scholar]
- 4.U.S. Department of Health and Human Services. Ending the HIV Epidemic: A Plan for America. https://www.hiv.gov/federal-response/ending-the-hiv-epidemic/overview. Accessed 23 Sept 2020.
- 5.Mandsager P, Marier A, Cohen S, Fanning M, Hauck H, Cheever LW. Reducing HIV-related health disparities in the Health Resources and Services Administration’s Ryan White HIV/AIDS program. Am J Public Health. 2018;108(S4):S246–s250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sullivan PS, Rosenberg ES, Sanchez TH, et al. Explaining racial disparities in HIV incidence in Black and White men who have sex with men in Atlanta, GA: a prospective observational cohort study. Ann Epidemiol. 2015;25(6):445–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Napravnik S, Eron JJ Jr., McKaig RG, Heine AD, Menezes P, Quinlivan E . Factors associated with fewer visits for HIV primary care at a tertiary care center in the Southeastern U.S. AIDS Care. 2006;18(Suppl 1):S45–50. [DOI] [PubMed] [Google Scholar]
- 8.Earnshaw VA, Bogart LM, Dovidio JF, Williams DR. Stigma and racial/ethnic HIV disparities: moving toward resilience. Am Psychol. 2013;68(4):225–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Brincks AM, Shiu-Yee K, Metsch LR, et al. Physician mistrust, medical system mistrust, and perceived discrimination: associations with HIV care engagement and viral load. AIDS Behav. 2019;23(10):2859–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Scott HM, Pollack L, Rebchook GM, Huebner DM, Peterson J, Kegeles SM. Peer social support is associated with recent HIV testing among young Black men who have sex with men. AIDS Behav. 2014;18(5):913–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Beer L, Mattson CL, Bradley H, Skarbinski J. Understanding cross-sectional racial, ethnic, and gender disparities in antiretroviral use and viral suppression among HIV patients in the United States. Medicine. 2016;95(13):e3171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Rebeiro PF, Howe CJ, Rogers WB, et al. The relationship between adverse neighborhood socioeconomic context and HIV continuum of care outcomes in a diverse HIV clinic cohort in the Southern United States. AIDS Care. 2018;30(11):1426–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Neaigus A, Reilly KH, Jenness SM, Wendel T, Marshall DM, Hagan H. Multilevel risk factors for greater HIV infection of Black men who have sex with men in New York City. Sex Transm Dis. 2014;41(7):433–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rebeiro PF, McPherson TD, Goggins KM, et al. Health literacy and demographic disparities in HIV care continuum outcomes. AIDS Behav. 2018;22(8):2604–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wilson EC, Turner C, Arayasirikul S, et al. Housing and income effects on HIV-related health outcomes in the San Francisco Bay Area—findings from the SPNS transwomen of color initiative. AIDS Care. 2018;30(11):1356–9. [DOI] [PubMed] [Google Scholar]
- 16.Liu Y, Silenzio VMB, Nash R, et al. Suboptimal recent and regular HIV testing among Black men who have sex with men in the United States: implications from a meta-analysis. J Acquir Immune Defic Syndr. 2019;81(2):125–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wohl AR, Benbow N, Tejero J, et al. Antiretroviral prescription and viral suppression in a representative sample of HIV-infected persons in care in 4 large metropolitan areas of the United States, Medical Monitoring Project, 2011–2013. J Acquir Immune Defic Syndr. 2017;76(2):158–70. [DOI] [PubMed] [Google Scholar]
- 18.Maulsby CH, Ratnayake A, Hesson D, Mugavero MJ, Latkin CA. A scoping review of employment and HIV. AIDS Behav. 2020;24(10):2942–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Beyer KMM, Zhou Y, Matthews K, Bemanian A, Laud PW, Nattinger AB. New spatially continuous indices of redlining and racial bias in mortgage lending: links to survival after breast cancer diagnosis and implications for health disparities research. Health Place. 2016;40:34–43. [DOI] [PubMed] [Google Scholar]
- 20.Matoba N, Suprenant S, Rankin K, Yu H, Collins JW. Mortgage discrimination and preterm birth among African American women: an exploratory study. Health Place. 2019;59:102193. [DOI] [PubMed] [Google Scholar]
- 21.Mendez DD, Hogan VK, Culhane J. Institutional racism and pregnancy health: using Home Mortgage Disclosure Act data to develop an index for mortgage discrimination at the community level. Public Health Rep. 2011;126(Suppl. 3):102–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Herring C, Henderson L. Wealth inequality in Black and White: cultural and structural sources of the racial wealth gap. Race Soc Probl. 2016;8(1):4–17. [Google Scholar]
- 23.Deaton A Policy implications of the gradient of health and wealth. Health Aff. 2002;21(2):13–30. [DOI] [PubMed] [Google Scholar]
- 24.Dawson L, Kates J. Insurance coverage and viral suppression among people with HIV, 2018. KFF Organization. https://www.kff.org/hivaids/issue-brief/insurance-coverage-and-viral-suppression-among-people-with-hiv-2018/. Accessed 28 Sept 2020. [Google Scholar]
- 25.Berry SA, Fleishman JA, Yehia BR, et al. Healthcare coverage for HIV provider visits before and after implementation of the Affordable Care Act. Clin Infect Dis. 2016;63(3):387–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Adamson B, Lipira L, Katz AB. The impact of ACA and Medicaid expansion on progress toward UNAIDS 90–90–90 Goals. Curr HIV/AIDS Rep. 2019;16(1):105–12. [DOI] [PubMed] [Google Scholar]
- 27.Health Resources and Services Administration: Ryan White HIV/AIDS Program. https://hab.hrsa.gov/about-ryan-white-hivaids-program/about-ryan-white-hivaids-program. Accessed 25 Sept 2020
- 28.Centers for Disease Control and Prevention. Monitoring selected national HIV prevention and care objectives by using HIV surveillance data—United States and 6 dependent areas, 2017. HIV Surveillance Supplemental Report 2019;24(No. 3). https://www.cdc.gov/hiv/pdf/library/reports/surveillance/cdc-hiv-surveillance-supplemental-report-vol-24-3.pdf. Accessed 25 Sept 2020 [Google Scholar]
- 29.Solar O, Irwin A. A conceptual framework for action on the social determinants of health. Social Determinants of Health Discussion Paper 2 (Policy and Practice). World Health Organization. https://www.who.int/sdhconference/resources/ConceptualframeworkforactiononSDH_eng.pdf. Accessed 25 Sept 2020 [Google Scholar]
- 30.Mazurenko O, Balio CP, Agarwal R, Carroll AE, Menachemi N. The effects of Medicaid expansion under the ACA: a systematic review. Health Aff. 2018;37(6):944–50. [DOI] [PubMed] [Google Scholar]
- 31.Gai Y, Marthinsen J. Medicaid expansion, HIV testing, and HIV-related risk behaviors in the United States, 2010–2017. Am J Public Health. 2019;109(10):1404–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Goldstein D, Hardy WD, Monroe A, Hou Q, Hart R, Terzian A. Despite early Medicaid expansion, decreased durable virologic suppression among publicly insured people with HIV in Washington, DC: a retrospective analysis. BMC Public Health. 2020;20(1):509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Griffin A, Dempsey A, Cousino W, et al. Addressing disparities in the health of persons with HIV attributable to unstable housing in the United States: the role of the Ryan White HIV/AIDS Program. PLoS Med. 2020;17(3):e1003057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Collazos J, Asensi V, Carton JA, Ibarra S. The influence of the patients’ educational levels on socioeconomic, clinical, immunological and virological endpoints. AIDS Care. 2009;21(4):511–9. [DOI] [PubMed] [Google Scholar]
- 35.Centers for Disease Control and Prevention. Division of HIV/AIDS Prevention Strategic Plan 2017–2020. https://www.cdc.gov/hiv/pdf/dhap/cdc-hiv-dhap-external-strategic-plan.pdf. Accessed 14 Jan 2021.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All datasets used in this study were public data sources.