Benefits of Frequent HIV Testing in the THRIVE Demonstration Project: United States, 2015–2020

Abstract

Objectives. To describe HIV testing among clients in the Targeted Highly Effective Interventions to Reverse the HIV Epidemic (THRIVE) demonstration project and evaluate testing frequency.

Methods. We identified factors associated with an average testing frequency of 180 days or less compared with more than 180 days using adjusted Poisson regression models. We performed the Kaplan–Meier survival analysis to compare time to diagnosis by testing frequency.

Results. Among 5710 clients with 2 or more tests and no preexposure prophylaxis (PrEP) prescription, 42.4% were tested frequently. Black/African American clients were 21% less likely and Hispanic/Latino clients were 18% less likely to be tested frequently than were White clients. Among 71 Black/African American and Hispanic/Latino cisgender men who have sex with men and transgender women with HIV diagnoses, those with frequent testing had a median time to diagnosis of 137 days, with a diagnostic testing yield of 1.5% compared with those tested less frequently, with 559 days and 0.8% yield.

Conclusions. HIV testing at least every 6 months resulted in earlier HIV diagnosis and was efficient. Persons in communities with high rates of HIV who are not on PrEP can benefit from frequent testing, and collaborative community approaches may help reduce disparities. (Am J Public Health. 2023;113(9):1019–1027. https://doi.org/10.2105/AJPH.2023.307341)

In 2019, there were more than 36 000 persons diagnosed with HIV infection in the United States. About 70% of these diagnoses were among gay, bisexual, and other men who have sex with men (MSM) and transgender women (TGW).¹ Black/African American and Hispanic/Latino persons were disproportionately affected. An estimated 13% of US persons with HIV infection were undiagnosed in 2019.² One of the 4 pillars of the Ending the HIV Epidemic in the US initiative, launched by the US Department of Health and Human Services in 2019, is to diagnose all persons with HIV as early as possible.³

HIV testing is a key intervention throughout both the HIV prevention continuum and the HIV care continuum. It allows persons with HIV to be diagnosed, initiate antiretroviral treatment, and become virally suppressed, which have the individual health benefit of preserving immune function and the public health benefit of reduced HIV transmission.^4,5 HIV testing also identifies persons who do not have HIV but might benefit from preexposure prophylaxis (PrEP) to prevent HIV acquisition.⁶

Since 2006, the Centers for Disease Control and Prevention (CDC) has recommended HIV testing at least once for all persons aged 13 to 64 years and at least annually for persons whose behaviors, circumstances, or sexual networks may be associated with increased likelihood of HIV acquisition; these persons include sexually active MSM and TGW, sex partners of persons with HIV infection, persons who exchange sex for money or drugs, and persons who inject drugs and their sex partners.⁷ The CDC recommends that providers consider more frequent testing, every 3 to 6 months, for sexually active MSM.⁴ Additionally, the CDC recommends HIV testing every 2 to 3 months for persons using long-acting injectable or oral PrEP.⁶ Persons who might benefit from frequent HIV testing may not seek or receive it because of lack of awareness of HIV risk and structural and social barriers, including stigma, fear, and lack of access.⁸

We describe HIV testing of clients in the Targeted Highly Effective Interventions to Reverse the HIV Epidemic (THRIVE) demonstration project from 2015 to 2020 in the United States. We identified factors associated with receipt of frequent HIV testing and calculated the incidence of HIV infection, time to diagnosis, and diagnostic testing yield per test to evaluate the effectiveness and efficiency of frequent testing in communities with high rates of HIV diagnoses.

METHODS

In the THRIVE demonstration project, the CDC funded 7 US health departments to develop and lead community collaboratives to provide comprehensive, culturally sensitive HIV prevention and care services for cisgender MSM and TGW in communities with high numbers of Black/African American or Hispanic/Latino MSM with HIV infection.^1,9,10 The 7 THRIVE recipients were health departments in Birmingham, Alabama; Baltimore, Maryland; New Orleans, Louisiana; New York, New York; Philadelphia, Pennsylvania; Hampton Roads, Virginia; and the District of Columbia. Recipients collected longitudinal client-level data, including HIV testing and PrEP service provision, and reported de-identified data to the CDC for evaluation of the effectiveness of THRIVE. We considered anyone who received a THRIVE-funded service, including HIV and sexual health services and essential support services, a THRIVE client.

Inclusion Criteria

To describe HIV testing patterns in THRIVE, we included all clients who were aged 18 years or older, received an HIV test, and had a negative result on their first HIV test conducted in THRIVE. To estimate the frequency of testing, HIV incidence rates, and time to diagnosis, we included only clients who received 2 or more tests and excluded those who were prescribed PrEP in THRIVE, because persons using PrEP have a decreased risk of acquiring HIV and, per the CDC, should be tested for HIV every 2 to 3 months, depending on the type of PrEP.⁶

Definitions

We assessed the race/ethnicity, gender identity, and sexual behaviors of clients because these characteristics are associated with increased rates of HIV diagnoses, understanding that race/ethnicity may serve as a proxy for underlying social, environmental, and structural factors, including systemic racism.^1,11 We categorized age groups by clients’ age at enrollment as 18 to 24, 25 to 34, 35 to 44, 45 to 54, and 55 or older years. We determined racial/ethnic group by client self-reported race and ethnicity and categorized it as Hispanic/Latino (including persons of any race), Black/African American, White, other race (including persons who identified as American Indian/Alaska Native, Asian, Native Hawaiian/other Pacific Islander, multiracial, or another race), and unknown. We determined clients’ gender identity and sexual behavior group by client self-report and categorized it as cisgender MSM, TGW, transgender men, cisgender heterosexual men, cisgender women, and other (including clients who could not be assigned to 1 of the categories because of missing data). We categorized cisgender men as MSM or heterosexual based on whether they reported having sex with men. We did not stratify other gender identity groups by sexual behavior because of the small numbers of persons in these groups.

Estimating Testing Frequency

We included all types of HIV tests and identified unique testing events with a minimum gap of 20 days between 2 tests, accounting for the HIV infection window period.¹² We categorized clients with 2 tests performed less than 20 days apart as having 1 HIV test. We defined frequent testing as an average interval of 180 days or less between tests and less frequent testing as an interval of more than 180 days.

Statistical Analyses

We calculated the percentage increase in the number of HIV antigen/antibody tests performed, including lab-based and point-of-care tests. We described characteristics of THRIVE clients with an initial negative HIV test result stratified by receipt of only 1 HIV test and 2 or more tests. We also described characteristics of clients with 2 or more HIV tests who were not prescribed PrEP, stratified by frequency of HIV testing. We estimated adjusted risk ratios (ARRs) and 95% confidence intervals (CIs) using Poisson regression models for the association between client characteristics and frequency of testing. We adjusted for THRIVE site, age group, race/ethnicity group, and gender identity and sexual behavior group. We included THRIVE site as a covariate because of differences in the size and capacity of each program, as well as different social, cultural, and structural contexts in the communities served at each site.

To estimate the incidence rate of HIV diagnosis, we identified THRIVE clients who were not prescribed PrEP who had an initial negative HIV test result followed by a positive test result with at least 20 days between tests. We described the number of clients with HIV diagnoses, cumulative follow-up time for each person with 2 or more tests, incidence rate per 100 person-years of follow-up time, median number of tests per client, median interval between tests, and diagnostic testing yield per test stratified by race/ethnicity group and gender identity and sexual behavior group. We calculated the cumulative follow-up time for clients diagnosed with HIV with the time interval between the first date of testing and the first positive test date. For clients with no positive tests, we calculated the cumulative follow-up time as the gap between the first and last date of testing plus 90 days of potential continued enrollment, as clients would not be expected to receive another HIV test for at least 3 months.

We calculated the median number of tests for clients who had only negative tests using all testing data collected and for clients diagnosed with HIV infection using all testing data collected up until the positive test. We calculated the incidence rate per 100 person-years by dividing the number of new diagnoses by the cumulative follow-up time per person-years. We calculated the testing yield per HIV test by dividing the number of new diagnoses by the number of tests performed among clients.

For Black/African American and Hispanic/Latino cisgender MSM and TGW, the groups with the highest incidence of HIV in THRIVE, we performed the Kaplan–Meier survival analysis and used log-rank testing to compare the time to diagnosis between clients who received frequent testing and those who received less frequent testing. We assessed time to diagnosis as the time interval between the first negative test and the first positive test among persons diagnosed with HIV infection. We determined the median time to diagnosis among Black/African American and Hispanic/Latino cisgender MSM and TGW diagnosed with HIV infection and compared persons who received frequent testing with those who received less frequent testing. We performed all analyses using SAS version 9.4 (SAS Institute, Cary, NC) and prepared figures using R version 4.1.1 (RStudio, Boston, MA).

RESULTS

Overall, THRIVE provided HIV testing for 29 687 clients with an initial negative HIV test (Table 1). The number of lab-based antigen/antibody HIV tests performed in THRIVE increased 253%, from 2256 in 2016 to 7953 in 2019, and the number of point-of-care antigen/antibody tests increased 449%, from 1171 to 6427. Most clients (71.7%) received only 1 HIV test, and this did not vary by age or race/ethnicity. Almost half of cisgender MSM (4653; 44.2%) and more than half of TGW (429; 51.9%) received 2 or more tests.

TABLE 1—

Characteristics of Clients Enrolled in THRIVE by the Number of HIV Tests: United States, 2015–2020

	Clients, No. (%)
	Total	1 HIV Testa	≥ 2 HIV Testsb
Total	29 687	21 289	8 398
Age group, y
18–24	8 223	6 069 (28.5)	2 154 (25.6)
25–34	11 086	7 458 (35.0)	3 628 (43.2)
35–44	5 081	3 590 (16.9)	1 491 (17.8)
45–54	2 868	2 192 (10.3)	676 (8.0)
≥ 55	2 398	1 959 (9.2)	439 (5.2)
Unknown/missing	31	21 (0.1)	10 (0.1)
Race/ethnicity
White	4 341	3 259 (15.3)	1 082 (12.9)
Black/African American	18 341	13 537 (63.6)	4 804 (57.2)
Hispanic/Latinoc	4 603	2 916 (13.7)	1 687 (20.1)
Otherd	1 709	1 070 (5.0)	639 (7.6)
Unknown	693	507 (2.4)	186 (2.2)
Gender identity/sexual behavior
Cisgender MSM	10 526	5 873 (27.6)	4 653 (55.4)
TGW	827	398 (1.9)	429 (5.1)
TGM	287	169 (0.8)	118 (1.4)
Cisgender women	8 729	7 285 (34.2)	1 444 (17.2)
Cisgender heterosexual men	9 099	7 393 (34.7)	1 706 (20.3)
Othere	219	171 (0.8)	48 (0.6)
Prescribed PrEP
Yes	3 887	1 199 (5.6)	2 688 (32.0)
No	25 800	20 090 (94.4)	5 710 (68.0)

Note. MSM = gay, bisexual, and other men who have sex with men; PrEP = preexposure prophylaxis; TGM = transgender men; TGW = transgender women; THRIVE = Targeted Highly Effective Interventions to Reverse the HIV Epidemic demonstration project.

^a Includes clients with only 1 HIV test with a negative result. Only tests > 20 d apart were considered to be independent testing events.

^b Includes clients with ≥ 2 HIV tests among those with an initial negative HIV test and > 20 d between tests.

^c Hispanic/Latino clients can be of any race.

^d Other includes clients who self-identified as American Indian/Alaska Native, Asian, Native Hawaiian/other Pacific Islander, and other.

^e Includes clients with missing gender identity or sexual behavior variables.

Among 5710 clients who received 2 or more HIV tests and were not prescribed PrEP (Figure A, available as a supplement to the online version of this article at http://www.ajph.org), 42.4% received frequent HIV testing (Table 2). Clients aged 18 to 24 years were more likely to receive frequent testing than were those aged 35 to 44 years, with an ARR of 1.15 (95% CI = 1.01, 1.31), after adjusting for THRIVE site, race/ethnicity group, and gender identity and sexual behavior group. Compared with White clients, Black/African American clients were 21% less likely to receive frequent testing (ARR = 0.79; 95% CI = 0.70, 0.90) and Hispanic/Latino clients were 18% less likely (ARR = 0.82; 95% CI = 0.70, 0.96). Compared with cisgender heterosexual men, MSM were 25% more likely to receive frequent testing (ARR = 1.25; 95% CI = 1.11, 1.41), TGW were 24% more likely (ARR = 1.24; 95% CI = 0.99, 1.54), and cisgender women were 15% more likely (ARR = 1.15; 95% CI = 1.01, 1.31).

TABLE 2—

Characteristics Associated With Frequent Testing Among THRIVE Clients With ≥ 2 HIV Tests and Not Prescribed PrEP: United States, 2015–2020

	Clients, No. (%)			ARR (95% CI)d
	Totala	Testing Frequency of > 180 db	Testing Frequency of ≤ 180 dc
Total	5 710	3 289 (57.6)	2 421 (42.4)
Age group at enrollment, y
18–24	1 747	982 (56.2)	765 (43.8)	1.15 (1.01, 1.31)
25–34	2 287	1 288 (56.3)	999 (43.7)	1.10 (0.97, 1.24)
35–44	907	543 (59.9)	364 (40.1)	1 (Ref)
45–54	433	261 (60.3)	172 (39.7)	1.04 (0.86, 1.24)
≥ 55	334	214 (64.1)	120 (35.9)	0.97 (0.79, 1.19)
Unknown/missing	2	1 (50.0)	1 (50.0)	1.27 (0.18, 9.05)
Race/ethnicity
White	730	286 (39.2)	444 (60.8)	1 (Ref)
Black/African American	3 633	2 293 (63.1)	1 340 (36.9)	0.79 (0.70, 0.90)
Hispanic/Latinoe	918	486 (52.9)	432 (47.1)	0.82 (0.70, 0.96)
Otherf	329	170 (51.7)	159 (48.3)	0.85 (0.70, 1.04)
Unknown	100	54 (54.0)	46 (46.0)	0.78 (0.57, 1.07)
Gender identity/sexual behavior
Cisgender MSM	2 465	1 190 (48.3)	1 275 (51.7)	1.25 (1.11, 1.41)
TGW	225	116 (51.6)	109 (48.4)	1.24 (0.99, 1.54)
TGM	93	55 (59.1)	38 (40.9)	1.00 (0.71, 1.41)
Cisgender women	1 328	825 (62.1)	503 (37.9)	1.15 (1.01, 1.31)
Cisgender heterosexual men	1 566	1 085 (69.3)	481 (30.7)	1 (Ref)
Otherg	33	18 (54.5)	15 (45.5)	1.07 (0.64, 1.80)

Note. ARR = adjusted risk ratio; CI = confidence interval; MSM = gay, bisexual, and other men who have sex with men; PrEP = preexposure prophylaxis; TGM = transgender men; TGW = transgender women; THRIVE = Targeted Highly Effective Interventions to Reverse the HIV Epidemic demonstration project.

^a Includes clients with ≥ 2 HIV tests > 20 d apart among those with an initial negative HIV test result and not prescribed PrEP in THRIVE.

^b Less frequent testing was defined as an average interval of > 180 d between tests.

^c Frequent testing was defined as an average interval of ≤ 180 d between tests.

^d Adjusted for THRIVE site, age group, racial/ethnic group, and gender identity/sexual behavior group.

^e Hispanic/Latino clients can be of any race.

^f Other includes clients who self-identified as American Indian/Alaska Native, Asian, Native Hawaiian/other Pacific Islander, and other.

^g Includes clients with missing gender identity/sexual behavior variables.

Among all 5710 clients, 92 were diagnosed with HIV infection while enrolled in THRIVE during 8432.2 observed person-years, with an incidence rate of 1.1 per 100 person-years (Table 3). Among 92 clients diagnosed with HIV infection, 72 were cisgender MSM, 6 were TGW, 6 were cisgender women, and 8 were cisgender heterosexual men (Table 3; Table A, available as a supplement to the online version of this article at http://www.ajph.org). The highest incidence rates were among cisgender MSM (2.0 per 100 person-years) and TGW (1.7 per 100 person-years). By known race/ethnicity group, the highest incidence rates were observed among Black/African American TGW (3.7 per 100 person-years), Black/African American MSM (3.1 per 100 person-years), and Hispanic/Latino MSM (1.9 per 100 person-years). The diagnostic testing yield per HIV test was highest among Black/African American TGW (1.6%), Black/African American MSM (1.4%), and Hispanic/Latino MSM (0.8%).

TABLE 3—

Clients Diagnosed With HIV Infection, Incidence Rate, and Testing Yield in THRIVE: United States, 2015–2020

Gender Identity, Sexual Behavior, and Race/Ethnicity Group	No. Clients With ≥ 2 HIV Testsa	No. Clients With New HIV Diagnoses	Follow-Up Time, Person-Years	Incidence Rate per 100 Person-Years	No. HIV Tests	Median No. of Tests (IQR)	Median No. of Days Between Tests (IQR)	Diagnostic Testing Yield per Test,b %
Total	5 710	92	8 432.2	1.1	17 401	2 (2–3)	150 (91–288)	0.53
Cisgender MSM	2 465	72	3 544.5	2.0	8 357	3 (2–4)	119 (80–233)	0.86
White	562	2	758.6	0.3	2 038	3 (2–4)	98 (77–181)	0.10
Black/African American	1 069	49	1 596.3	3.1	3 532	2 (2–4)	137 (83–257)	1.39
Hispanic/Latinoc	568	16	829.0	1.9	1 912	2 (2–4)	124 (82–246)	0.84
Other raced	218	2	308.7	0.6	752	3 (2–4)	112 (75–222)	0.27
Unknown	48	3	52.0	5.8	123	2 (2–3)	175 (98–310)	2.44
TGW	225	6	358.3	1.7	828	3 (2–4)	138 (88–231)	0.72
White	15	0	18.5	0.0	49	2 (2–5)	113 (70–205)	0.00
Black/African American	87	5	134.5	3.7	304	3 (2–4)	147 (96–227)	1.64
Hispanic/Latinoc	95	1	157.6	0.6	362	3 (2–5)	134 (87–235)	0.28
Otherd	15	0	26.9	0.0	51	3 (2–4)	183 (113–340)	0.00
Unknown	13	0	20.8	0.0	62	4 (2–5)	92 (70–142)	0.00
TGM	93	0	145.5	0.0	298	3 (2–4)	156 (98–298)	0.00
White	6	0	6.6	0.0	15	2 (2–3)	154 (119–342)	0.00
Black/African American	46	0	79.0	0.0	149	3 (2–4)	188 (119–308)	0.00
Hispanic/Latinoc	24	0	37.8	0.0	89	2.5 (2–5)	105 (76–247)	0.00
Otherd	7	0	9.3	0.0	14	2 (2–2)	378 (168–595)	0.00
Unknown	10	0	12.8	0.0	31	3 (2–4)	140 (98–252)	0.00
Cisgender women	1 328	6	1 882.5	0.3	3 577	2 (2–3)	183 (104–341)	0.17
White	74	1	78.0	1.3	168	2 (2–2)	178 (103–304)	0.60
Black/African American	1 100	5	1 604.2	0.3	3 026	2 (2–3)	182 (104–337)	0.17
Hispanic/Latinoc	102	0	138.0	0.0	260	2 (2–3)	206 (102–364)	0.00
Otherd	37	0	45.7	0.0	85	2 (2–2)	196 (141–404)	0.00
Unknown	15	0	16.6	0.0	38	2 (2–3)	182 (88–241)	0.00
Cisgender heterosexual men	1 566	8	2 453.4	0.3	4 232	2 (2–3)	210 (113–370)	0.19
White	64	0	70.2	0.0	147	2 (2–2)	197 (96–312)	0.00
Black/African American	1 320	7	2 109.1	0.3	3 559	2 (2–3)	217 (119–379)	0.20
Hispanic/Latinoc	121	1	189.1	0.5	357	2 (2–3)	165 (91–329)	0.28
Otherd	52	0	72.2	0.0	147	2 (2–3)	139 (88–308)	0.00
Unknown	9	0	12.8	0.0	22	2 (2–3)	206 (127–307)	0.00
Black/African American and Hispanic/Latino cisgender MSM and TGW	1 819	71	2 717.4	2.6	6 110	3 (2–4)	133 (83, 251)	1.20
Frequent testinge	860	51	1 025.3	5.0	3 496	3 (2–5)	98 (66, 143)	1.50
Less frequent testingf	959	20	1 692.1	1.2	2 614	2 (2–3)	278 (181, 432)	0.80

Note. IQR = interquartile range; MSM = gay, bisexual, and other men who have sex with men; TGM = transgender men; TGW = transgender women; THRIVE = Targeted Highly Effective Interventions to Reverse the HIV Epidemic demonstration project.

^a Includes clients with ≥ 2 HIV tests > 20 d apart among those with an initial negative HIV test result and not prescribed preexposure prophylaxis in THRIVE.

^b Testing yield per HIV test was calculated by dividing the number of new diagnoses by the number of tests among clients.

^c Hispanic/Latino clients can be of any race.

^d Other includes clients who self-identified as American Indian/Alaska Native, Asian, Native Hawaiian/other Pacific Islander, and other.

^e Frequent testing was defined as an average interval of ≤ 180 days between tests.

^f Less frequent testing was defined as an average interval of > 180 days between tests.

Among 1819 Black/African American and Hispanic/Latino cisgender MSM and TGW, 71 were diagnosed with HIV infection during an observed 2717.4 person-years, with an incidence rate of 2.6 per 100 person-years and a testing yield of 1.2% per test. Among these clients, 71.8% received frequent testing. The testing yield per test among clients with frequent testing was 1.5% compared with a testing yield of 0.8% among those with less frequent testing. Clients with frequent testing had a shorter time to diagnosis (or end of follow-up period) than did those who received less frequent testing (P < .001). Clients diagnosed with HIV infection who received frequent testing had a shorter median time to diagnosis (137 days; IQR = 83–503) than did those who received less frequent testing (559 days; IQR = 311–709; Figure 1).

FIGURE 1—

Time to Diagnosis of HIV Infection or End of Follow-Up and Frequency of Testing Among Black/African American and Hispanic/Latino Cisgender Gay, Bisexual, and Other Men Who Have Sex With Men (MSM) and Transgender Women (TGW) Who (a) Had ≥ 2 HIV Tests > 20 Days Apart, Had an Initial Negative HIV Test Result, and Were Not Prescribed Preexposure Prophylaxis; and (b) Were Diagnosed With HIV Infection: THRIVE, United States, 2015–2020

Note. THRIVE = Targeted Highly Effective Interventions to Reverse the HIV Epidemic demonstration project. The sample size in part a was n = 1819; we used a log-rank test to compare the time to diagnosis between those with frequent and those with less frequent testing (P < .001). The sample size in part b was n = 71. Dashed lines represent median time to diagnosis. The median time to diagnosis among persons with frequent testing was 137 days (interquartile range [IQR] = 83–503) and among persons with less frequent testing was 559 days (IQR = 311–709).

DISCUSSION

THRIVE successfully implemented sensitive antigen/antibody HIV testing for the priority populations of cisgender MSM and TGW. Black/African American and Hispanic/Latino MSM and TGW had the highest HIV incidence rates in THRIVE, mirroring the epidemiology of HIV in the United States.¹ Among this group of THRIVE clients with high incidence rates, we found that clients with frequent testing had shorter time to diagnosis than did those who tested less frequently. Frequent testing was also efficient, with a high diagnostic testing yield per test among tests performed in this group. These findings can inform HIV testing recommendations for persons not using PrEP and can serve as inputs in modeling studies of HIV transmission and cost-effectiveness.

Despite implementation of culturally sensitive programs, Black/African American and Hispanic/Latino THRIVE clients were less likely to receive frequent testing than were White clients, underscoring ongoing disparities in HIV prevention that were not eliminated during the project period. Persons in racial and ethnic minority groups and sexual and gender minority groups who are disproportionately affected by HIV, and who may benefit most from frequent HIV testing, often encounter social and structural barriers that adversely affect their health.¹³ These social and structural factors, including systemic racism, poverty, stigma, and discrimination, prevent equity in receipt of HIV prevention services.^8,14 Collaborative, culturally sensitive, community-tailored programs like THRIVE can increase the use of HIV testing and PrEP and can help to reduce disparities,¹⁵ but additional large-scale and long-term interventions may be needed to overcome the long-standing, deeply rooted barriers that drive disparities.

THRIVE successfully engaged a large number of MSM and TGW in HIV testing, the first step in receipt of HIV prevention and care services. Frequent testing resulted in earlier diagnosis of HIV infection, which has both individual and public health benefits. Early initiation of antiretroviral therapy supports viral suppression, leading to reduced HIV-related morbidity and mortality and prevention of community transmission.^4,5 A negative HIV test is an opportunity to discuss HIV prevention options and, when appropriate, to start PrEP.⁶ Although many MSM and TGW were tested frequently in THRIVE, large proportions received only 1 HIV test or were tested less frequently.

Interventions are needed to increase HIV testing coverage and frequency of testing of persons in communities with a high prevalence of HIV or with behaviors associated with HIV acquisition. Among MSM and TGW, about 20% self-reported that they had not received an HIV test in the previous 12 months.¹⁶ A nationally representative study of visits to ambulatory health care providers found that HIV testing must increase at least threefold among US men to provide universal testing coverage.¹⁷ Increasing routine HIV testing coverage and increasing the frequency of testing among persons with risk factors will require interventions in both clinical and public health settings. Innovative strategies include self-test kit distribution and clinical decision support tools that are programmed in electronic health record systems to generate automated test orders.^13,18,19 Community-based programs that are tailored for the local context may also help to increase frequent HIV testing among persons who may benefit.

Limitations

Although this analysis has many strengths, including an analysis of longitudinal data from a large cohort of Black/African American and Hispanic/Latino MSM and TGW, it has at least 4 limitations. First, we were unable to evaluate clients’ behaviors that might be associated with acquisition of HIV, such as frequency of sexual and injection drug use behaviors, because these data were not collected in THRIVE. If a client disclosed behaviors to their provider, they might have been tested more frequently. We were also unable to assess clients’ intent or motivation for testing and whether the clients’ health care providers recommended HIV testing. Given the high rates of persons diagnosed with HIV infection in the priority populations served by THRIVE, we assumed that most clients were at substantial risk of acquiring HIV and would probably have benefited from frequent testing.

Second, in this analysis, we could not assess whether a client received an HIV test or an HIV diagnosis outside the THRIVE collaborative, as data were collected only on the services provided by members of the collaborative, so we might have underestimated the proportion of clients who were frequently tested or diagnosed with HIV. Third, if a client’s gender identity and sexual behavior group was misclassified, we might have underestimated the number of Black/African American and Hispanic/Latino cisgender MSM and TGW who were tested. If clients experienced differential levels of stigma, certain gender identities and sexual behaviors might have been underreported. Fourth, HIV testing patterns observed in THRIVE may not be generalizable to other populations, communities, or geographic locations.

Public Health Implications

HIV testing is the first step to receipt of HIV prevention and care services. Frequent HIV testing, at least every 6 months, identified HIV infections earlier, with shorter time to diagnosis, and was efficient, with high testing yield for persons in the priority populations who are disproportionately affected by HIV. Although interventions were implemented in THRIVE jurisdictions to increase community collaboration and to provide culturally sensitive and community-tailored HIV prevention services, we observed disparities in the frequency of HIV testing. Additional long-term, multifaceted programs may be needed to increase HIV testing among persons in priority populations and to overcome the social and structural barriers that drive disparities in HIV incidence.