Abstract
The objective of this study was to estimate disparities in linkage to human immunodeficiency virus (HIV) care among Latinos by country/region of birth, HIV testing site, and neighborhood characteristics. A retrospective study was conducted using Florida HIV surveillance records of Latinos/Hispanics aged ≥13 diagnosed during 2014–2015. Linkage to HIV care was defined as a laboratory test (HIV viral load or CD4) within 3 months of HIV diagnosis. Multi-level Poisson regression models were used to estimate adjusted prevalence ratios (aPR) for nonlinkage to care. Of 2659 Latinos, 18.8% were not linked to care within 3 months. Compared with Latinos born in mainland United States, those born in Cuba [aPR 0.60, 95% confidence interval (CI) 0.47–0.76] and Puerto Rico (aPR 0.61, 95% CI 0.41–0.90) had a decreased prevalence of nonlinkage. Latinos diagnosed at blood banks (aPR 2.34, 95% CI 1.75–3.12), HIV case management and screening facilities (aPR 1.76, 95% CI 1.46–2.14), and hospitals (aPR 1.42, 95% CI 1.03–1.96) had an increased prevalence of nonlinkage compared with outpatient general, infectious disease, and tuberculosis/sexually transmitted diseases/family planning clinics. Latinos who resided in the lowest (aPR 1.57, 95% CI 1.19–2.07) and third lowest (aPR 1.33, 95% CI 1.01–1.76) quartiles of neighborhood socioeconomic status compared with the highest quartile were at increased prevalence. Latinos who resided in neighborhoods with <25% Latinos also had increased prevalence of nonlinkage (aPR 1.23, 95% CI 1.01–1.51). Testing site at diagnosis may be an important determinant of HIV care linkage among Latinos due to neighborhood or individual-level resources that determine location of HIV testing.
Keywords: : human immunodeficiency virus, linkage to care, Latinos, Hispanics, testing site, neighborhood
Introduction
Linkage to care is a critical step in the human immunodeficiency virus (HIV) continuum of care, with early linkage and enrolment in HIV care resulting in life expectancies similar to that of the general population for people living with HIV.1,2 The National HIV/AIDS Strategy (NHAS) stated that by 2020 reduction in HIV infections and improvement in outcomes for people living with HIV should come by linking 85% of HIV-infected persons to care within 1 month of HIV diagnosis.3 The current NHAS goals for screening, linkage, and retention in care could substantially reduce loss of life caused by HIV/AIDS in the United States (US).4
Hispanic/Latino people (hereinafter referred to as “Latinos”) are diagnosed with HIV at approximately three times the rate of their non-Latino white counterparts.5,6 Currently, 15% of Latinos diagnosed with HIV are not linked to care within 90 days of diagnosis.7 In addition, linkage to care is lower among the overall population of Latinos living with HIV, as well as among Latino youth and Latino men who have sex with men (MSM), when compared with their non-Latino white counterparts.7–9 Reported individual-level barriers to care for Latinos living with HIV have included documentation status and perceived risk of deportation, the complex nature of the US healthcare system, including provider–patient language barriers, and stigma from family and friends.10–12
Contextual characteristics about the neighborhood of residence at time of HIV diagnosis, such as the proportion of the population whose income is under the poverty level, the racial/ethnic composition, and rural/urban status, have also been associated with HIV outcomes among Latinos and other minorities, but their role in prompt linkage to HIV care among Latinos has not been explored.13–18
Despite disparities identified along other steps of the HIV care continuum and survival by country/region of birth for Latinos,14,19–22 a systematic review of the literature found no studies that compared linkage to care within Latino populations in the United States by country/region of birth.23 Further, different HIV testing sites, both in the healthcare and nonhealthcare setting, show a range in utilization by race and ethnicity.24,25 Thus, for Latinos, it is important to consider both country/region of birth and specific testing site when examining and addressing linkage to HIV care at time of diagnosis. Therefore, the objective of this study was to estimate the magnitude of disparities in linkage to HIV care among Latinos by country/region of birth, HIV testing site, and neighborhood characteristics.
Methods
Datasets
Deidentified HIV surveillance records of individuals reported as Hispanic or Latino ethnicity were obtained from the Florida Department of Health (FDOH) enhanced HIV/AIDS reporting system (eHARS). All cases reported in eHARS must meet the Center for Disease Control and Prevention (CDC) HIV case definition.26 Latino individuals were included in this study if they were diagnosed during 2014 or 2015 and were aged 13 or older. Individuals who died within 3 months of HIV diagnosis or who had missing or invalid data for ZIP code at time of diagnosis, a reported ZIP code with a population of zero based on American Community Survey (ACS), or cases diagnosed in a correctional facility were excluded. The 2009–2013 ACS was used to obtain characteristics of the neighborhood at time of diagnosis using ZIP code tabulation areas (ZCTAs).27 ZCTAs are used by the US Census Bureau to tabulate summary statistics and approximate US Postal Service ZIP codes.28
Variables
Outcome
Linkage to HIV care was defined as documentation of an HIV laboratory test (HIV viral load or CD4 cells/μL) within 3 months of HIV diagnosis.
Predictors
Individual-level variables
The following data for individuals were extracted from eHARS: ethnicity, race, year of HIV diagnosis, sex at birth, age at HIV diagnosis, county of birth, mode of HIV transmission, HIV testing site, time between HIV diagnosis and death in months, time between HIV and AIDS diagnosis (if case progressed to AIDS), time between HIV diagnosis and first viral load test, time between HIV diagnosis and first CD4 test, residential ZIP code at time of diagnosis, and whether the person was diagnosed at a correctional facility.
Country/region of birth was categorized into the following mutually exclusive areas: the United States (excluding Puerto Rico), Puerto Rico, Cuba, Mexico, Central America, South America, and other/unknown. The US Census Bureau Hispanic origin classification was used to select the countries included in the Central and South America categories.29 Mode of HIV transmission was self-reported during HIV testing, reported by a healthcare provider, or extracted from medical chart reviews. Due to the small number of cases, males with the risk factors of both injection drug use and MSM (n = 36) were included in the injection drug use category.
Testing site variable
HIV testing site for each individual was extracted from eHARS and categorized into the following groups: Blood bank, HIV facility (HIV case management agencies and HIV counseling and testing sites), hospital (including emergency room departments), outpatient (general, infectious disease, tuberculosis/sexually transmitted disease/family planning clinic), and other/unknown (drug treatment center, laboratory, “other,” and “missing”).
Neighborhood-level variables
Neighborhoods were defined by ZCTA. Based on previous work by Niyonsenga et al.,30 13 socioeconomic status (SES) indicators were extracted from the ACS by ZCTA to develop an SES index of Florida neighborhoods: percent of households without access to a car, percent of households with ≥1 person per room, percent of population living below the poverty line, percent of owner-occupied homes worth ≥$300,000, median household income in 2013, percent of households with annual income <$15,000, percent of households with annual income ≥$150,000, income disparity (derived from percent of households with annual income <$10,000 and percent of households with annual income ≥$50,000), percent of population aged ≥25 with less than a 12th grade education, percent of population aged ≥25 with a graduate professional degree, percent of households living in rented housing, percent of population aged ≥16 who were unemployed, and percent of population aged ≥16 employed in high working-class occupation (ACS occupation group: “managerial, business, science, and arts occupations”).
Income disparity was calculated as the logarithm of 100 times the percent of households with annual income <$10,000 divided by the percent of households with annual income ≥$50,000. In addition, we extracted the proportion of the population in each ZCTA that was Hispanic/Latino and categorized it into three categories: <25%, 25–49%, and ≥50%.31,32 To categorize ZCTAs into rural or urban, we used Categorization C of Version 2.0 of the rural–urban commuting area codes, developed by the University of Washington WWAMI Rural Research Center.33
Statistical analyses
Development of neighborhood SES index
To calculate the SES index, we coded all neighborhood-level variables from the ACS so that higher scores corresponded with lower SES (higher disadvantage); they were then standardized. Then we conducted a reliability analysis. The Cronbach's alpha for all 13 indicators was 0.93. We selected seven indicators based on the correlation of the indicator with the total index (high correlation), and the Cronbach's alpha if the item was deleted (low alpha). The seven indicators selected were percent below poverty, median household income, percent of households with annual income <$15,000, percent of households with annual income ≥$150,000, income disparity, percent of population age ≥25 with less than a 12th grade education, and high-class work. The resulting Cronbach's alpha increased (0.94).
We then conducted a principal component analysis (PCA) with and without varimax rotation. PCA revealed one component, which accounted for 73.49% of the variability in the indicators. Because all the original variables were highly correlated with the component (factor loadings between 0.80 and 0.93), we retained all seven indicators. Finally, we added the standardized scores for the seven variables and categorized the scores into quartiles.
Imputation of race
A total of 88 Latinos in our study were missing data on race. We used SAS software, version 9.4 (SAS Institute, Cary, NC 2002), to impute data for race. We created 10 datasets with imputed data for race using the Multiple Imputation (MI) Procedure in SAS. We used year of HIV diagnosis, sex, age, country of birth, and HIV transmission mode to predict race in a fully conditional specification model.34 All subsequent analyses were fitted by imputation by constructing pooled estimates.
Bivariate and multi-variate analyses
Individual- and neighborhood-level data were merged by matching the current ZIP code of each case with the ZIP code's corresponding ZCTA. First, we conducted descriptive analyses comparing individual- (including testing site) and neighborhood-level characteristics by country/region of birth for Latinos. Second, we used SAS (SAS software, version 9.4, SAS Institute, Cary, NC 2002) GENMOD Procedure to estimate crude and adjusted prevalence ratios (aPRs) and 95% confidence intervals (CI) for nonlinkage to care. Models attempted using the binomial distribution and logarithm link function (log-binomial regression) did not converge. Therefore, prevalence ratios and 95% CI were estimated using Poisson regression models with robust error variance provided by the generalized estimating equation approach.35 The repeated statement with “subject = ZCTA” was used to account for correlation among cases living in the same neighborhood (ZCTA).
We estimated prevalence ratios using three models: crude model (Model 1), a model adjusting for country of birth, race, sex at birth, age at diagnosis, year of diagnosis, mode of HIV transmission, HIV diagnosis facility, and AIDS within 3 months of diagnosis (Model 2), and a model adjusting for all variables in Model 2 plus ZCTA-level SES, rural/urban status, and percentage of the population that was Latino (Model 3). The Florida International University Institutional Review Board approved this study, and the FDOH Institutional Review Board designated this study to be nonhuman subjects research.
Results
Characteristics of population
Of 2777 Latinos diagnosed with HIV in Florida during 2014–2015, four were <13 years of age, 46 died within 3 months of HIV diagnosis, 39 had missing or invalid data for ZIP code at time of diagnosis, none had a reported ZIP code with a population of zero, and 21 were diagnosed in a correctional facility. After imputation of race, 75 individuals were of a race other than black or white. Due to the small number of individuals and heterogeneity of this group, they were excluded from further analysis.
Of the remaining 2659 Latinos, the largest proportions were born in Cuba (27.3%), mainland United States (26.5%), or South America (13.8%), and the majority were of white race (95.5%), male (88.3%), and between the ages of 25 and 49 (68.0%) years (Table 1). Seventy-three percent of Latinos reported MSM as a mode of HIV transmission, although the proportion varied by country/region of birth with most of those born in South America (84.2%) and Cuba (81.1%) but only 50.9% of those born in Central America reporting MSM. Of Latinos testing positive for HIV, 46.2% were tested at an outpatient clinical facility, and 31.4% were tested at an HIV counseling and testing or case management site.
Table 1.
Latinos Aged 13 and Older Diagnosed with HIV Infection by Birth Country/Region and Selected Characteristics, Florida, 2014–2015
| All Latinosa | United States-born Latinos | Puerto Rico | Mexico | Cuba | Central Americab | South Americac | Other | |
|---|---|---|---|---|---|---|---|---|
| Total (n) | 2659 | 704 (26.5) | 193 (7.3) | 125 (4.7) | 726 (27.3) | 159 (6.0) | 367 (13.8) | 386 (14.5) |
| Individual-level variables | ||||||||
| Raced | ||||||||
| Black | 121 (4.6) | 53 (7.5) | 9 (4.7) | 1 (0.8) | 27 (3.7) | 4 (2.5) | 4 (1.1) | 23 (6.0) |
| White | 2538 (95.5) | 651 (92.5) | 184 (95.3) | 124 (99.2) | 699 (96.3) | 155 (97.5) | 363 (98.9) | 362 (94.0) |
| Sex at birth | ||||||||
| Female | 310 (11.7) | 94 (13.4) | 36 (18.7) | 8 (6.4) | 54 (7.4) | 29 (18.2) | 26 (7.1) | 63 (16.4) |
| Male | 2349 (88.3) | 610 (86.7) | 157 (81.4) | 117 (93.6) | 672 (92.6) | 130 (81.8) | 341 (92.9) | 322 (83.7) |
| Age at diagnosis | ||||||||
| 13–24 | 440 (16.6) | 199 (28.3) | 23 (11.9) | 18 (14.4) | 92 (12.7) | 18 (11.3) | 41 (11.2) | 49 (12.7) |
| 25–49 | 1808 (68.0) | 431 (61.2) | 124 (64.3) | 94 (75.2) | 498 (68.6) | 127 (79.9) | 276 (75.2) | 258 (67.0) |
| 50 or older | 411 (15.4) | 74 (10.5) | 46 (23.8) | 13 (10.4) | 136 (18.8) | 14 (8.8) | 50 (13.6) | 78 (20.3) |
| Year of HIV diagnosis | ||||||||
| 2014 | 1215 (45.7) | 303 (43.0) | 94 (48.7) | 61 (48.8) | 339 (46.7) | 81 (50.9) | 162 (44.1) | 175 (45.5) |
| 2015 | 1444 (54.3) | 401 (57.0) | 99 (51.3) | 64 (51.2) | 387 (53.3) | 78 (49.1) | 205 (55.9) | 210 (54.6) |
| Mode of transmission | ||||||||
| Heterosexual | 480 (18.1) | 102 (14.5) | 53 (27.5) | 29 (23.2) | 107 (14.7) | 63 (39.6) | 42 (11.4) | 84 (21.8) |
| IDU | 94 (3.5) | 38 (5.4) | 26 (13.5) | 2 (1.6) | 10 (1.4) | 2 (1.3) | 2 (0.5) | 14 (3.6) |
| MSM | 1941 (73.0) | 520 (73.9) | 107 (55.4) | 87 (69.6) | 589 (81.1) | 81 (50.9) | 309 (84.2) | 249 (64.5) |
| Other/unknown | 144 (5.4) | 44 (6.3) | 7 (3.6) | 7 (5.6) | 20 (2.8) | 13 (8.2) | 14 (3.8) | 39 (10.1) |
| HIV diagnosis facility | ||||||||
| Blood bank | 82 (3.1) | 30 (4.3) | 3 (1.6) | 2 (1.6) | 12 (1.7) | 8 (5.0) | 11 (3.0) | 16 (4.2) |
| HIV counseling and testing or case management site | 836 (31.4) | 173 (24.6) | 37 (19.2) | 28 (22.4) | 311 (42.8) | 43 (27.0) | 158 (43.1) | 86 (22.3) |
| Hospital | 264 (9.9) | 67 (9.5) | 29 (15.0) | 24 (19.2) | 47 (6.5) | 25 (15.7) | 20 (5.7) | 51 (13.3) |
| Outpatiente | 1229 (46.2) | 357 (50.7) | 110 (57.0) | 56 (44.8) | 294 (40.5) | 62 (39.0) | 149 (40.6) | 201 (52.2) |
| Other/unknownf | 248 (9.3) | 77 (10.9) | 14 (7.3) | 15 (12.0) | 62 (8.5) | 21 (13.2) | 28 (7.6) | 31 (8.1) |
| AIDS in 3 months | ||||||||
| Yes | 477 (17.9) | 100 (14.2) | 42 (21.8) | 43 (34.4) | 118 (16.3) | 48 (30.2) | 56 (15.3) | 70 (18.2) |
| No | 2182 (82.1) | 604 (85.8) | 151 (78.2) | 82 (65.6) | 608 (83.8) | 111 (69.8) | 311 (84.7) | 315 (81.8) |
| Linkage to HIV care | ||||||||
| No | 500 (18.8) | 166 (23.6) | 28 (14.5) | 20 (16.0) | 102 (14.1) | 28 (17.6) | 82 (22.3) | 74 (19.2) |
| Yes | 2159 (81.2) | 538 (76.4) | 165 (85.5) | 105 (84.0) | 624 (86.0) | 131 (82.4) | 285 (77.7) | 311 (80.8) |
| ZCTA-level variables | ||||||||
| SES index, quartiles | ||||||||
| 1 (lowest) | 1078 (40.5) | 276 (39.2) | 79 (40.9) | 45 (36.0) | 368 (50.7) | 80 (50.3) | 113 (30.8) | 117 (30.4) |
| 2 | 597 (22.5) | 150 (21.3) | 54 (28.0) | 39 (31.2) | 146 (20.1) | 29 (18.2) | 82 (22.3) | 97 (25.2) |
| 3 | 642 (24.1) | 185 (26.3) | 43 (22.3) | 32 (25.6) | 158 (21.8) | 32 (20.1) | 98 (26.7) | 94 (24.4) |
| 4 (highest) | 342 (12.9) | 93 (13.2) | 17 (8.8) | 9 (7.2) | 54 (7.4) | 18 (11.3) | 74 (20.2) | 77 (20.0) |
| RUCA classification | ||||||||
| Urban | 2630 (98.9) | 695 (98.7) | 189 (97.9) | 118 (94.4) | 724 (99.7) | 158 (99.4) | 364 (99.2) | 382 (99.2) |
| Rural | 29 (1.1) | 9 (1.3) | 4 (2.1) | 7 (5.6) | 2 (0.3) | 1 (0.6) | 3 (0.8) | 3 (0.8) |
| Percentage of population Hispanic/Latino | ||||||||
| ≥50% | 1147 (43.1) | 209 (29.7) | 45 (23.3) | 30 (24.0) | 529 (72.9) | 57 (35.9) | 154 (42.0) | 123 (32.0) |
| 25–49% | 812 (30.5) | 230 (32.7) | 81 (42.0) | 42 (33.6) | 138 (19.0) | 55 (34.6) | 133 (36.2) | 133 (34.6) |
| <25 | 700 (26.3) | 265 (37.6) | 67 (34.7) | 53 (42.4) | 59 (8.1) | 47 (29.6) | 80 (21.8) | 129 (33.5) |
Percentage may not add up to 100 due to rounding.
Excludes the following: 46 who died within 3 months of HIV diagnosis, 39 who had missing or invalid data for ZIP code at time of diagnosis, and 21 who were diagnosed in a correctional facility.
Includes Latinos born in Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama.
Includes Latinos born in Argentina, Bolivia, Chile, Colombia, Ecuador, Paraguay, Peru, Uruguay, and Venezuela.
Race was missing for 88 Latinos who met our inclusion/exclusion criteria. After multiple imputation, 75 individuals were from a race other than black or white (American Indian/Alaskan Native, Asian, Native Hawaiian/Pacific Islander) and were excluded.
Includes general, infectious disease, tuberculosis/sexually transmitted disease/family planning clinic.
Includes drug treatment center, laboratory, “other,” and “missing.”
AIDS, acquired immune deficiency syndrome; HIV, human immunodeficiency virus; IDU, injection drug use; MSM, male to male sexual contact; RUCA, rural urban commuting area; SES, socioeconomic status; ZCTA, ZIP code tabulation area.
In terms of neighborhood characteristics, 40.5% of Latinos resided in the lowest quartile of neighborhood SES, 98.9% in urban areas, and 43.1% in areas where greater than 50% of the population was Latino. While half of those born in Cuba and Central America lived in the lowest SES neighborhoods, only 30% of those born in South America did. Nearly three-quarters (72.9%) of those born in Cuba lived in a predominantly Latino neighborhood.
Linkage to HIV care
Of 2659 Latinos, 18.8% were not linked to care within 3 months of HIV diagnosis (Table 1). Compared with Latinos born in mainland United States, those born in Cuba (aPR 0.60, 95% CI 0.47–0.76) and Puerto Rico (aPR 0.61, 95% CI 0.41–0.90) had a decreased prevalence of nonlinkage to care (Table 2). Those diagnosed in 2014 compared with 2015 had an increased prevalence of nonlinkage (aPR 1.29, 95% CI 1.10–1.51), and those who had an AIDS diagnosis within 3 months of the HIV diagnosis had a decreased prevalence of nonlinkage (aPR 0.02, 95% CI 0.01–0.07).
Table 2.
Nonlinkage to HIV Care by Individual and Neighborhood Characteristics Among Latinos Aged 13 Years and Older Diagnosed with HIV 2014–2015 in Florida
| Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|
| Total, n | Not linked to care, n (%) | Crude PR for nonlinkage to care (95% CI) | Adjusted PR for nonlinkage to care (95% CI) | Adjusted PR for nonlinkage to care (95% CI) | |
| Individual-level variables | |||||
| Country of birth | |||||
| US-born Latino | 704 | 166 (23.6) | Referent | Referent | Referent |
| Central America | 159 | 28 (17.6) | 0.75 (0.50–1.11) | 0.79 (0.52–1.20) | 0.78 (0.55–1.10) |
| Cuba | 726 | 102 (14.1) | 0.60 (0.47–0.76) | 0.59 (0.45–0.77) | 0.60 (0.47–0.76) |
| Mexico | 125 | 20 (16.0) | 0.68 (0.43–1.08) | 0.95 (0.58–1.53) | 0.91 (0.62–1.36) |
| Other Latino | 385 | 74 (19.2) | 0.82 (0.62–1.07) | 0.78 (0.58–1.04) | 0.82 (0.64–1.03) |
| Puerto Rico | 193 | 28 (14.5) | 0.62 (0.41–0.92) | 0.61 (0.40–0.94) | 0.61 (0.41–0.90) |
| South America | 367 | 82 (22.3) | 0.95 (0.73–1.23) | 0.92 (0.70–1.22) | 0.97 (0.78–1.20) |
| Race | |||||
| Black | 121 | 22 (18.2) | 0.97 (0.63–1.48) | 0.96 (0.61–1.49) | 0.90 (0.60–1.35) |
| White | 2538 | 478 (18.8) | Referent | Referent | Referent |
| Sex at birth | |||||
| Female | 310 | 57 (18.4) | Referent | Referent | Referent |
| Male | 2349 | 443 (18.9) | 1.03 (0.78–1.35) | 1.21 (0.85–1.73) | 1.22 (0.91–1.64) |
| Age at diagnosis | |||||
| 13–24 | 440 | 107 (24.3) | Referent | Referent | Referent |
| 25–49 | 1808 | 323 (17.9) | 0.73 (0.59–0.91) | 0.89 (0.71–1.13) | 0.91 (0.74–1.12) |
| 50 or older | 411 | 70 (17.0) | 0.70 (0.52–0.95) | 1.02 (0.74–1.42) | 1.03 (0.76–1.42) |
| Year of HIV diagnosis | |||||
| 2014 | 1215 | 262 (21.6) | 1.31 (1.10–1.56) | 1.29 (1.07–1.54) | 1.29 (1.10–1.51) |
| 2015 | 1444 | 238 (16.5) | Referent | Referent | Referent |
| Mode of HIV transmission | |||||
| Heterosexual | 480 | 75 (15.6) | Referent | Referent | Referent |
| IDU | 94 | 24 (25.5) | 1.63 (1.03–2.59) | 1.39 (0.83–2.34) | 1.31 (0.82–2.10) |
| MSM | 1941 | 343 (17.7) | 1.13 (0.88–1.45) | 0.80 (0.58–1.12) | 0.80 (0.61–1.07) |
| Other/unknown | 144 | 58 (40.3) | 2.58 (1.83–3.63) | 1.85 (1.27–2.70) | 1.87 (1.41–2.49) |
| HIV diagnosis facility | |||||
| Outpatient | 1229 | 180 (14.7) | Referent | Referent | Referent |
| Blood bank | 82 | 43 (52.4) | 3.58 (2.57–4.99) | 2.34 (1.63–3.35) | 2.34 (1.75–3.12) |
| HIV counseling and testing or case management site | 836 | 213 (25.5) | 1.74 (1.43–2.12) | 1.74 (1.41–2.14) | 1.76 (1.46–2.14) |
| Hospital | 264 | 31 (11.7) | 0.80 (0.55–1.17) | 1.42 (0.97–2.10) | 1.42 (1.03–1.96) |
| Other/unknown | 248 | 33 (13.3) | 0.91 (0.63–1.32) | 0.96 (0.66–1.39) | 0.96 (0.66–1.42) |
| AIDS in 3 months | |||||
| No | 2182 | 498 (22.8) | Referent | Referent | Referent |
| Yes (AIDS) | 477 | 2 (0.4) | 0.02 (0.01–0.07) | 0.02 (0.005–0.08) | 0.02 (0.01–0.07) |
| ZCTA-level variables | |||||
| SES index, quartiles | |||||
| 1 (lowest) | 1078 | 225 (20.9) | 1.33 (1.00–1.77) | 1.57 (1.19–2.07) | |
| 2 | 597 | 105 (17.6) | 1.15 (0.85–1.56) | 1.33 (1.00–1.78) | |
| 3 | 642 | 116 (18.1) | 1.19 (0.88–1.60) | 1.33 (1.01–1.76) | |
| 4 (highest) | 342 | 54 (15.8) | Referent | Referent | |
| RUCA classification | |||||
| Urban | 2631 | 496 (18.9) | Referent | Referent | |
| Rural | 29 | 5 (17.2) | 0.96 (0.48–1.92) | 1.24 (0.67–2.29) | |
| Percentage of population Hispanic/Latino | |||||
| ≥50% | 1147 | 208 (18.1) | Referent | Referent | |
| 25–49% | 812 | 145 (17.9) | 0.91 (0.73–1.12) | 0.90 (0.74–1.10) | |
| <25 | 700 | 147 (21.0) | 1.16 (0.95–1.41) | 1.23 (1.01–1.51) | |
Model 1: Crude rates (each variable in a separate model).
Model 2: Controlling for individual-level variables (country of birth, race, sex at birth, age at diagnosis, year of diagnosis, mode of HIV transmission, HIV diagnosis facility, and AIDS within 3 months of diagnosis).
Model 3: Controlling for individual-level variables and neighborhood-level variables (ZCTA-level SES, rural/urban status, and percentage of the population who was Hispanic/Latino).
AIDS, acquired immune deficiency syndrome; CI, confidence interval; HIV, human immunodeficiency virus; IDU, injection drug use; MSM, male to male sexual contact; PR, prevalence ratio; RUCA, rural urban commuting area; SES, socioeconomic status; US, United States; ZCTA, ZIP code tabulation area.
Latinos diagnosed at blood banks (aPR 2.34, 95% CI 1.75–3.12), HIV case management and counseling and testing facilities (aPR 1.76, 95% CI 1.46–2.14), and hospitals (aPR 1.42, 95% CI 1.03–1.96) had an increased prevalence of nonlinkage to care compared with outpatient facilities. Latinos who resided in the lowest (aPR 1.10, 95% CI 1.04–1.17) and third lowest (aPR 1.33, 95% CI 1.01–1.76) quartiles of neighborhood SES compared with the highest quartile also had an increased prevalence of nonlinkage to care. Latinos who resided in neighborhoods with <25% Latinos also had increased prevalence of nonlinkage to care (aPR 1.23, 95% CI 1.01–1.51).
In post hoc analyses, we attempted to examine two-way interactions between country/region of birth, testing site, and neighborhood poverty. However, models with interactions between country/region of birth and testing site, and country/region of birth and neighborhood poverty did not converge due to small cell sizes.
Discussion
Between 2014 and 2015 in the state of Florida, linkage to care varied by country/region of birth, with nonlinkage to care within 3 months of HIV diagnosis more likely among mainland US-born Latinos than those born in Cuba and Puerto Rico. Nonlinkage was also more likely among Latinos who tested at a blood bank, HIV, and hospital facilities compared with Latinos who tested at outpatient facilities. Differences in linkage to care existed at the neighborhood-level such that nonlinkage was more likely among Latinos who resided in neighborhoods of lower SES and those in neighborhoods with low Latino population density.
The proportion of Latinos who were not linked to care within 3 months of an HIV diagnosis in our Florida study (18.8%) was similar to findings from two studies of Latinos with HIV in 19 US jurisdictions using National HIV Surveillance System (NHSS) data from 2010 (19.7%) and 2011 (18%).6,36 Similar to our study, these two previous studies used laboratory data to define linkage to HIV care services. Our finding was not consistent with a study of Latinos in the United States, Puerto Rico, and US Virgin Islands that examined individuals diagnosed in 2014 using National HIV Prevention Program Monitoring and Evaluation (NHM&E) data and found a 3-month nonlinkage to care rate of 38.9%.5
Methodological differences may explain these discrepant outcomes. The study conducted by Rao et al.5 was of publicly funded testing events, and thus, its results would be most comparable with cases in our study tested at HIV counseling and testing or case management sites where nonlinkage was higher (25.5%) than the overall percentage. In addition, Rao et al. defined linkage to care by attendance to first medical appointment.5 Our study used the date of the first laboratory test. It is possible, that in some settings a laboratory test is ordered before the first medical appointment in light of a positive HIV test.37 Rao et al. also stated that their findings represent the minimum percentage of Latinos linked to care and that their calculations likely underestimate the actual percentage of individuals linked by including records with invalid data in the denominator.5
Our findings suggest that Latinos born in Cuba and Puerto Rico are more likely to be linked to care within 3 months of HIV diagnosis than mainland US-born Latinos. In a recent study by our group, Latinos living with HIV who were born in Cuba were also more likely to be retained in HIV care and be virally suppressed than US-born Latinos.22 Presumably our findings are not due to differences in access to care given that Latinos born in Cuba (due to their refugee status) and Puerto Rico (due to birth in a US territory) have similar legal provisions for accessing care than US-born Latinos.38 However, differences in SES between these groups that affect eligibility for HIV services such as the Ryan White HIV/AIDS Program may play a role in these results.
Studies of healthcare utilization among Latino groups have reported mixed findings. For example, a study using data from the National Health Interview Survey found that Cubans were more likely to delay or forgo care than other Latino heritage groups.39 However, they also found that Puerto Ricans were less likely than any other group to delay or forgo care. A second study using a national sample of Latinas found that Puerto Ricans had the highest rate of mental health service utilization and Cubans the highest rate of specialized care utilization compared with other Latino heritage groups.40 It is important to note that our study compared Latinos by country/region of birth and not heritage. Our data do not allow us to identify the heritage of US-born Latinos; thus, our cultural discussion of this group is limited. It is also important to highlight that Cubans are the predominant Latino ethnic origin group in Florida and that increased exposure to Cuban social networks may provide a protective effect.
Our study did not find differences in linkage to care by race, sex, or age. In addition, we did not find differences by mode of HIV transmission except for the “other/unknown” category. We were unable to identify any study that examined disparities in linkage to HIV care by race among Latinos. Our finding that men, women, and individuals in varying age groups have similar linkage to care is consistent with a study of Latinos in 19 US jurisdictions.6
Testing at a blood bank, HIV counseling and testing site or case management site, or a hospital compared with an outpatient facility was a significant predictor of nonlinkage to care among Latinos in our study. Previous studies have suggested that if the person who provided the patient with a positive HIV diagnosis scheduled the follow-up appointment, the patient was more likely to be linked to care.24,41 It is possible that this level of assistance linking individuals to HIV care is not possible at blood banks. It is also possible that individuals testing at blood banks are different in demographics, risk perception, access to care, HIV disclosure preferences, or neighborhood factors from those testing in other settings. For example, in our study, 26.8% of blood bank cases had no identified or reported risk factor compared with 5.0% of outpatient facilities cases. In our study, HIV facilities included HIV screening sites.
These sites may have included health departments, community testing sites, or mobile testing units. A CDC study that examined linkage to care by HIV testing site reported that the proportion of newly diagnosed persons linking to care within 3 months was lower for nonhealthcare community HIV testing sites (67.7%) compared with primary care clinic (89.2%) and STD clinic testing sites (85.3%).25 Similar findings were observed in two other studies.42,43
Our study found a marginal difference in the prevalence of nonlinkage to care between hospitals and outpatient facilities. Due to the way cases are reported in Florida, our group of Latinos diagnosed at hospitals/inpatient facilities included Latinos diagnosed at emergency room settings. Therefore, the increased prevalence of nonlinkage to care among those diagnosed in hospital settings may be a reflection of emergency room testing that does not lead to hospitalization. Linkage to care among individuals diagnosed in emergency departments (86.8%) has been reported to be lower than among individuals diagnosed in primary care settings (89.2%) and inpatient facilities (94.4%).25
Latinos who resided in the lowest and third lowest quartiles of neighborhood SES were at increased prevalence of nonlinkage to care in our study. Neighborhood poverty has been associated with HIV outcomes, particularly with survival after HIV diagnosis, in other studies.13–16,18 Possible mechanism for these associations could be increased psychosocial stress among individuals who live in disadvantaged neighborhoods, decreased positive social networks, and fewer health and social support services.44–46 Notably, in a previous study by our group, we were unable to find an association between neighborhood SES status and retention in HIV care or viral load suppression among Latinos, suggesting that the mechanisms for the effect are different for each step along the HIV care continuum.47
Our study also found a marginal increased prevalence among Latinos residing in areas where <25% of the population identified as Latino/Hispanic compared with areas where ≥50% of the population was Latino. In a previous study that examined delayed HIV diagnosis among Latinos, residing in an area with <25% Latino population density was also a risk factor compared with areas with ≥50% Latinos.48 These findings may reflect a shortage of culturally targeted support services in nonpredominant Latino communities. Future studies are needed to explore characteristics of low and high Latino density areas, such as geographically clustered social capital or social support networks, that may explain the relationship between Latino population density and continuum of care outcomes.49 Our study did not find a significant association between rural/urban status and linkage to care.
Our study has four main limitations. First, limited research suggests that foreign-born Latinos may be seeking care in their home countries. Therefore, we may underestimate linkage to care by examining care in the United States only.50,51 Second, laboratory tests may be ordered before a medical visit. Thus, defining linkage to care by laboratory test only may overestimate the proportion of Latinos linked to HIV care.37 Third, HIV surveillance systems do not collect information on heritage, limiting our discussion of cultural factors affecting US-born Latinos. Finally, testing facility type for HIV tests performed at emergency room departments in Florida hospitals are documented as inpatient/hospital facility. Therefore, it was not possible for us to determine the associated risk of nonlinkage to care among Latinos tested in emergency settings.
Blood banks and HIV testing facilities appear to be two important junctures for intervention where Latinos who are likely to link to care late get tested and diagnosed for HIV. Further, Latinos residing in areas of low SES and in areas of low Latino population density may require additional help linking to HIV care after diagnosis. The protective associations observed among Latinos born in Cuba and Puerto Rico should be explored further to determine the extent that culture, heritage, and social capital influence linkage to care.
Acknowledgments
Research reported in this publication was supported by the National Institute on Minority Health & Health Disparities (NIMHD) under Award Number 5R01MD004002 and 5S21MD010683, and the National Institute of Mental Health (NIMH) under Award Number K23-100978-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of NIMHD, NIMH, or the National Institutes of Health. The authors acknowledge Karalee Poschman, MPH for her work in linking the HIV/AIDS Reporting System data with data from Florida databases of HIV-related services.
Author Disclosure Statement
No competing financial interests exist.
References
- 1.McManus H, O'Connor CC, Boyd M, et al. Long-term survival in HIV positive patients with up to 15 Years of antiretroviral therapy. PLoS One 2012;7:e48839. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nakagawa F, Lodwick RK, Smith CJ, et al. Projected life expectancy of people with HIV according to timing of diagnosis. AIDS 2012;26:335–343 [DOI] [PubMed] [Google Scholar]
- 3.White House Office of National AIDS Policy. National HIV/AIDS Strategy for the United States. 2015. Available at: https://files.hiv.gov/s3fs-public/nhas-update.pdf (Last accessed April1, 2016)
- 4.Shah M, Perry A, Risher K, et al. Effect of the US National HIV/AIDS Strategy targets for improved HIV care engagement: A modelling study. Lancet HIV 2016;3:e140–e146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rao S, Seth P, Walker T, et al. HIV testing and outcomes among Hispanics/Latinos—United States, Puerto Rico, and U.S. Virgin Islands, 2014. MMWR Morb Mortal Wkly Rep 2016;65:1099–1103 [DOI] [PubMed] [Google Scholar]
- 6.Gant Z, Bradley H, Hu X, et al. Hispanics or Latinos living with diagnosed HIV: Progress along the continuum of HIV care—United States, 2010. MMWR Morb Mortal Wkly 2014;63:886–890 [PMC free article] [PubMed] [Google Scholar]
- 7.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, 2015. HIV Surveillance Supplemental Report 22(No. 2). 2017. Available at: www.cdc.gov/hiv/library/reports/hivsurveillance.html (Last accessed January30, 2018) [Google Scholar]
- 8.Philbin MM, Tanner AE, DuVal A, et al. Factors affecting linkage to care and engagement in care for newly diagnosed HIV-positive adolescents within fifteen adolescent medicine clinics in the United States. AIDS Behav 2014;18:1501–1510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Singh S, Bradley H, Hu X, Skarbinski J, Hall HI, Lansky A. Men living with diagnosed HIV who have sex with men: Progress along the continuum of HIV care—United States, 2010. MMWR Morb Mortal Wkly Rep 2014;63:829–833 [PMC free article] [PubMed] [Google Scholar]
- 10.Dang BN, Giordano TP, Kim JH. Sociocultural and structural barriers to care among undocumented Latino immigrants with HIV infection. J Immigr Minor Health 2012;14:124–131 [DOI] [PubMed] [Google Scholar]
- 11.Bowden WP, Rhodes SD, Wilkin AM, Jolly CP. Sociocultural determinants of HIV/AIDS risk and service use among immigrant Latinos in North Carolina. Hisp J Behav Sci 2006;28:546–562 [Google Scholar]
- 12.Morales-Aleman MM, Sutton MY. Hispanics/Latinos and the HIV continuum of care in the Southern USA: A qualitative review of the literature, 2002–2013. AIDS Care 2014;26:1592–1604 [DOI] [PubMed] [Google Scholar]
- 13.Arnold M, Hsu L, Pipkin S, McFarland W, Rutherford GW. Race, place and AIDS: The role of socioeconomic context on racial disparities in treatment and survival in San Francisco. Soc Sci Med 2009;69:121–128 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hanna DB, Pfeiffer MR, Torian LV, Sackoff JE. Concurrent HIV/AIDS diagnosis increases the risk of short-term HIV-related death among persons newly diagnosed with AIDS, 2002–2005. AIDS Patient Care STDS 2008;22:17–28 [DOI] [PubMed] [Google Scholar]
- 15.Joy R, Druyts EF, Brandson EK, et al. Impact of neighborhood-level socioeconomic status on HIV disease progression in a universal health care setting. J Acquir Immune Defic Syndr 2008;47:500–505 [DOI] [PubMed] [Google Scholar]
- 16.McFarland W, Chen S, Hsu L, Schwarcz S, Katz M. Low socioeconomic status is associated with a higher rate of death in the era of highly active antiretroviral therapy, San Francisco. J Acquir Immune Defic Syndr 2003;33:96–103 [DOI] [PubMed] [Google Scholar]
- 17.Sheehan DM, Trepka MJ, Fennie KP, Prado G, Cano MÁ, Maddox LM. Black-white Latino racial disparities in HIV survival, Florida, 2000–2011. Int J Environ Res Public Health 2015;13:ijerph13010009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Trepka MJ, Fennie KP, Sheehan DM, Lutfi K, Maddox L, Lieb S. Late HIV diagnosis: Differences by rural/urban residence, Florida, 2007–2011. AIDS Patient Care STDS 2014;28:188–197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Espinoza L, Hall HI, Selik RM, Hu X. Characteristics of HIV infection among Hispanics, United States 2003–2006. J Acquir Immune Defic Syndr 2008;49:94–101 [DOI] [PubMed] [Google Scholar]
- 20.Wohl AR, Tejero J, Frye DM. Factors associated with late HIV testing for Latinos diagnosed with AIDS in Los Angeles. AIDS Care 2009;21:1203–1210 [DOI] [PubMed] [Google Scholar]
- 21.Espinoza L, Hall HI, Hu X. Diagnoses of HIV infection among Hispanics/Latinos in 40 states and Puerto Rico, 2006–2009. J Acquir Immune Defic Syndr 2012;60:205–213 [DOI] [PubMed] [Google Scholar]
- 22.Sheehan DM, Mauck DE, Fennie KP, et al. Black-white and country of birth disparities in retention in HIV care and viral suppression among Latinos with HIV in Florida, 2015. Int J Environ Res Public Health 2017;14:pii: [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sheehan DM, Trepka MJ, Dillon FR. Latinos in the United States on the HIV/AIDS care continuum by birth country/region: A systematic review of the literature. Int J STD AIDS 2015;26:1–12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pollini RA, Blanco E, Crump C, Zúñiga ML. A community-based study of barriers to HIV care initiation. AIDS Patient Care STDS 2011;25:601–609 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Seth P, Wang G, Collins NT, Belcher L. Identifying new positives and linkage to HIV medical care—23 testing site types, United States, 2013. MMWR Morb Mortal Wkly Rep 2015;64:663–667 [PMC free article] [PubMed] [Google Scholar]
- 26.Centers for Disease Control and Prevention. Revised surveillance case definition for HIV infection—United States, 2014. MMWR Recomm Rep 2014;63:1–10 [PubMed] [Google Scholar]
- 27.American Community Survey. 2009–2013. American Community Survey 5-Year Estimates. 2015. Available at: http://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk (Last accessed April1, 2016)
- 28.United States Census Bureau. Geography: ZIP Code Tabulation Areas (ZCTAs). n.d.. Available at: www.census.gov/geo/reference/zctas.html (Last accessed April1, 2016)
- 29.American Community Survey. American Community Survey and Puerto Rico Community Survey 2011 Code List. 2011. Available at: www.census.gov/acs/www/Downloads/data_documentation/CodeLists/2011_ACS_Code_Lists.pdf (Last accessed April1, 2016)
- 30.Niyonsenga T, Trepka MJ, Lieb S, Maddox LM. Measuring socioeconomic inequality in the incidence of AIDS: Rural-urban considerations. AIDS Behav 2013;17:700–709 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Alvarez KJ, Levy BR. Health advantages of ethnic density for African American and Mexican American elderly individuals. Am J Public Health 2012;102:2240–2242 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shaw RJ, Pickett KE, Wilkinson RG. Ethnic density effects on birth outcomes and maternal smoking during pregnancy in the US linked birth and infant death data set. Am J Public Health 2010;100:707–713 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.WWAMI Rural Health Research Center. Rural Urban Commuting Areas (RUCA). 2013. Available at: http://depts.washington.edu/uwruca/ruca-uses.php (Last accessed April27, 2015)
- 34.Van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification. Stat Methods Med Res 2007;16:219–242 [DOI] [PubMed] [Google Scholar]
- 35.Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004;159:702–706 [DOI] [PubMed] [Google Scholar]
- 36.Bradley H, Hall HI, Wolitski RJ, et al. Vital Signs: HIV diagnosis, care, and treatment among persons living with HIV—United States, 2011. MMWR Morb Mortal Wkly Rep 2014;63:1113–1117 [PMC free article] [PubMed] [Google Scholar]
- 37.Bertolli J, Shouse RL, Beer L, et al. Using HIV surveillance data to monitor missed opportunities for linkage and engagement in HIV medical care. Open AIDS J 2012;6:131–141 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Livingston G. DHS: Hispanics, Health Insurance, and Health Care Access. 2009. Available at: www.pewhispanic.org/2009/09/25/hispanics-health-insurance-and-health-care-access (Last accessed January15, 2017)
- 39.Alcalá HE, Chen J, Langellier BA, Roby DH, Ortega AN. Impact of the Affordable Care Act on health care access and utilization among Latinos. J Am Board Fam Med 2017;30:52–62 [DOI] [PubMed] [Google Scholar]
- 40.Ai AL, Appel HB, Huang B, Lee K. Overall health and healthcare utilization among Latino American women in the United States. J Womens Health 2012;21:878–885 [DOI] [PubMed] [Google Scholar]
- 41.Hightow-Weidman LB, Jones K, Wohl AR, et al. Early linkage and retention in care: Findings from the outreach, linkage, and retention in care initiative among young men of color who have sex with men. AIDS Patient Care STDS 2011;25:S31–S38 [DOI] [PubMed] [Google Scholar]
- 42.Tripathi A, Gardner LI, Ogbuanu I, et al. Predictors of time to enter medical care after a new HIV diagnosis: A statewide population-based study. AIDS Care 2011;23:1366–1373 [DOI] [PubMed] [Google Scholar]
- 43.Reddy EA, Agala CB, Maro VP, et al. Test site predicts HIV care linkage and antiretroviral therapy initiation: A prospective 3.5 year cohort study of HIV-positive testers in northern Tanzania. BMC Infect Dis 2016;16:497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Boardman JD, Finch BK, Ellison CG, Williams DR, Jackson JS. Neighborhood disadvantage, stress, and drug use among adults. J Health Soc Behav 2001;42:151–165 [PubMed] [Google Scholar]
- 45.Galea S, Ahern J, Vlahov D. Contextual determinants of drug use risk behavior: A theoretic framework. J Urban Health 2003;80:iii50–iii58 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Kirby JB, Kaneda T. Neighborhood socioeconomic disadvantage and access to health care. J Health Soc Behav 2005;46:15–31 [DOI] [PubMed] [Google Scholar]
- 47.Sheehan DM, Fennie KP, Mauck DE, Maddox LM, Lieb S, Trepka MJ. Retention in HIV care and viral suppression: Individual- and neighborhood-level predictors of racial/ethnic differences, Florida, 2015. AIDS Patient Care STDS 2017;31:167–175 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Sheehan DM, Trepka MJ, Fennie KP, et al. Individual and neighborhood determinants of late HIV diagnosis among Latinos, Florida, 2007–2011. J Immigr Minor Health 2017;19:825–834 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Ransome Y, Dean LT, Crawford ND, Metzger DS, Blank MB, Nunn AS. How do social capital and HIV/AIDS outcomes geographically cluster and which sociocontextual mechanisms predict differences across clusters? J Acquir Immune Defic Syndr 2017;76:13–22 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Bergmark R, Barr D, Garcia R. Mexican immigrants in the US living far from the border may return to Mexico for health services. J Immigr Minor Health 2010;12:610–614 [DOI] [PubMed] [Google Scholar]
- 51.Zúñiga ML, Organista KC, Scolari R, Olshefsky AM, Schulhof R, Colón M. Exploring care access issues for Mexican-origin Latinos living with HIV in the San Diego/Tijuana border region. J HIV AIDS Soc Serv 2006;5:37–54 [Google Scholar]