Adults Aged ≥50 Years Reached by the Centers for Disease Control and Prevention’s HIV Testing Programs in the United States and US Dependent Areas, 2019

Abstract

Objectives:

Older adults represent nearly half of people living with HIV in the United States. The objective of this analysis was to describe HIV testing, demographic characteristics, and risks of adults aged ≥50 years (older adults) reached in 2019 by HIV testing programs funded by the Centers for Disease Control and Prevention (CDC).

Methods:

We collected data from 101 CDC-funded community-based organizations and 61 health departments. All funding recipients submitted deidentified program service data for 2019 through a secure online CDC-supported system. We used multivariable robust Poisson regression to assess the association between demographic and risk characteristics and the proportion of tests that resulted in a new diagnosis. We also assessed the proportion of people who received a positive test result, were linked to HIV medical care, and were interviewed for partner services.

Results:

During 2019, among 2 452 507 CDC-funded HIV tests provided in the United States, 412 164 (16.8%) were provided to older adults. Among the 1059 (0.26% positivity) older adults with newly diagnosed HIV infection for whom we had data, 582 (68.4%) were linked to HIV medical care within 30 days of diagnosis and 494 (72.1%) were interviewed for partner services. Among the 2858 older adults with previously diagnosed HIV infection, 1321 (46.2%) reported not being in HIV medical care at the time of the test; of those with linkage data, 425 (49.9%) were linked to HIV medical care within 30 days of testing HIV positive.

Conclusions:

More rapid disease progression and higher morbidity and mortality rates among older adults suggest that services are needed to ensure early diagnosis, rapid linkage, and interview for partner services.

Older adults, defined here as aged ≥50 years, accounted for nearly half of all people living with HIV in the United States and for 17% (6363) of the 37 968 new HIV diagnoses in 2018. ¹ Black or African American (hereinafter, Black) older adults constituted approximately 40% of all new HIV diagnoses in this age group. Among older adults, nearly half (48%) of new HIV diagnoses in 2018 were among gay, bisexual, and other men who have sex with men (hereinafter, MSM). ¹

Knowledge of HIV status is the first step for timely access to essential prevention and treatment services, including linkage to HIV medical care. Importantly, knowledge of HIV-positive status predicts a reduction in risk behaviors. ² More than half of adults aged 57-64 years report sexual activity; however, older adults report low levels of condom use and a low-risk perception for HIV.^3-5 Older adults are also more likely than people aged <55 years to have late-stage HIV infection when they are diagnosed, with a median gap of 4.5 years between acquiring HIV and receiving a diagnosis. ⁶ This gap is longer than for any other age group and highlights a disparity in HIV testing between older and younger adults. Because disease progression is faster in older adults than in younger adults, diagnosis as early as possible after infection is especially important for reducing risks of transmission and addressing disparities. ⁷ Identifying people with HIV, linking them to medical care, and reducing health disparities are important national goals. ⁸

The Centers for Disease Control and Prevention (CDC) directly funds community-based organizations (CBOs) and health departments to provide HIV prevention services, including HIV testing, linkage to HIV medical care, and partner services. Administered by local health departments, partner services is the process through which people with HIV are interviewed by disease intervention specialists to elicit information about their partners, who can then be confidentially notified of their possible exposure or potential risk and offered services that can protect their health and prevent HIV acquisition or transmission. The purpose of partner services is to increase the number of people brought to treatment and reduce transmission through sexual networks; partner services has been shown to be a high-yield strategy for identifying undiagnosed HIV infections. ⁹

From 2010 through 2017, CDC’s HIV testing program identified an estimated one-third of all HIV diagnoses in the United States and 24% of all diagnoses among older adults. ¹⁰ CDC routinely collects evaluation data on its testing program, but details of HIV testing outcomes among older adults have not been reported. The objectives of this analysis were to (1) describe the demographic characteristics and risks of older adults reached by CDC’s HIV testing programs and among older adults with diagnosed HIV and (2) assess how effectively these adults are linked to HIV medical care and interviewed for partner services.

Methods

Data Source

In 2019, CDC funded 61 health departments and 101 CBOs to provide HIV testing and related services in the United States. Health departments use those funds to conduct HIV testing throughout their jurisdictions, typically implemented in 1 of 2 ways: (1) routine opt-out HIV testing, typically performed by medical professionals in health care settings (eg, emergency departments, primary care clinics), and (2) targeted HIV testing for priority populations (eg, MSM), typically performed by local organizations in non–health care settings (eg, nonclinical HIV counseling and testing sites, bars and clubs, retail establishments). Overall, most health department testing is conducted in health care settings. In addition to these health departments, CDC directly funded CBOs in 2019 that served populations with the greatest need for HIV prevention services, ensuring that they had the resources necessary to reach populations at increased risk for HIV infection. These populations include Black people, Hispanic/Latino people, MSM, people who inject drugs (PWID), and transgender people. All funding recipients submitted deidentified program service data for 2019 through a secure online CDC-supported system. We retrieved the data for this analysis from this secure data system, generating a National HIV Prevention Program Monitoring and Evaluation dataset (CDC, unpublished data, March 2021). Data collection was determined to be routine program monitoring and, therefore, did not require institutional review board approval; data collection was approved by the Office of Management and Budget.

HIV Tests and Testing Outcomes

Our analysis focused on 3 outcomes from 2019 CDC HIV testing data: the proportion of people with a newly identified HIV infection, the proportion of people with a newly or previously identified HIV infection linked to medical care within 30 days of diagnosis, and the proportion of people with a newly or previously identified HIV infection interviewed for partner services. People with a previous diagnosis were those who reported a previous diagnosis themselves or those for whom a previous diagnosis was reported in surveillance data. Other positive tests represented new diagnoses. An HIV test is defined as the performance of ≥1 HIV test to determine a person’s HIV infection status. A person might be tested once (eg, 1 rapid test or 1 conventional test) or multiple times (eg, 1 rapid test followed by 1 conventional test to confirm a preliminary HIV-positive test result). Linkage to HIV medical care within 30 days of diagnosis is defined as confirmation that the person attended their first HIV medical care appointment within 30 days of their HIV test date. Interview for partner services is defined as confirmation that the person was interviewed by staff trained to conduct partner services.

Demographic Characteristics, Region, Test Setting, and Population Groups

We restricted analyses to HIV tests provided to people who reported their age as 50-90 years. We stratified data by the following characteristics: age group, sex/gender, US Census region, and population group in non–health care settings. Age group categories were 50-54, 55-59, 60-64, 65-69, and ≥70 years. Sex/gender categories were male, female, and transgender. US Census regions were Northeast, Midwest, South, West, and US dependent areas (Puerto Rico and US Virgin Islands). Mutually exclusive race and ethnicity categories were Hispanic/Latino of any race, Asian, Black, “Other” (defined as people who self-identified as American Indian/Alaska Native, Native Hawaiian or Other Pacific Islander, multiple race, and other races), and White. Population group categories were MSM and MSM/PWID, PWID, heterosexual people, people with other risk (eg, women who have sex with women, having sex with a transgender person, or another behavior not listed), and people with no risk reported, meaning that the person did not report any sexual activity or injection drug use. HIV-related risk information is required for all tests administered in non–health care settings. Non–health care settings are those where HIV testing is usually performed by using a targeted testing strategy rather than a routine screening strategy. CDC-funded HIV tests conducted in health care settings (eg, inpatient clinics) are required to collect HIV-related risk information only from people who receive a positive test result for HIV.

Analysis

We used multivariable robust Poisson regression to assess the association between demographic characteristics and newly diagnosed HIV infections, linkage to HIV medical care, and interview for partner services, producing adjusted prevalence ratios (aPRs). Robust Poisson regression provides correct estimates for prevalence ratios that are more interpretable than logistic models. ¹¹ Models included the following independent variables: age group, sex/gender, race and ethnicity, and US Census region. To analyze associations with population group, which was routinely collected only in non–health care settings, we conducted separate models for tests in non–health care settings. Independent variables in these models included those used in other models (age group, sex/gender, race and ethnicity, and US Census region) plus population group. We tabulated aPRs for population group (adjusted for other covariates). For linkage to HIV medical care and interview for partner services, we separately analyzed data on newly diagnosed and previously diagnosed people, and we excluded all tests with missing data on the outcome from percentage calculations and analyses. We conducted all analyses in SAS version 9.4 (SAS Institute Inc).

Results

During 2019, among the 2 452 507 CDC-funded HIV tests provided in the United States, 412 164 (16.8%) were provided to adults aged 50-90 years. Among these adults, the highest percentages of tests were provided, by age, to adults aged 50-54 years (n = 133 837; 32.5%); by sex/gender, to males (n = 243 369; 59.0%); by race and ethnicity, to Black people (n = 163 268; 39.6%); and by US Census region, to people living in the South (n = 268 435; 65.1%) (Table 1). Of the 3917 older adults who received a positive test result for HIV infection in 2019, 1059 (27.0%) received a new diagnosis (Table 2) (0.26% new positivity; 1059 of 412 164), and 2858 (73.0%) had previously received a diagnosis of HIV infection (Table 3).

Table 1.

HIV tests and newly diagnosed HIV infection among adults aged ≥50 years receiving CDC-funded HIV tests, by selected characteristics, United States and US dependent areas, 2019 ^a

		Newly diagnosed HIV infection
Characteristic	HIV tests, no. (column %)	No. (column %)	Positivity, %	aPR (95% CI) [P value] b
Age group, y
50-54	133 837 (32.5)	423 (39.9)	0.32	1 [Reference]
55-59	119 177 (28.9)	337 (31.8)	0.28	0.89 (0.77-1.03) [.11]
60-64	83 834 (20.3)	184 (17.4)	0.22	0.71 (0.60-0.85) [<.001]
65-69	40 295 (9.8)	67 (6.3)	0.17	0.53 (0.41-0.69) [<.001]
≥70	35 021 (8.5)	48 (4.5)	0.14	0.47 (0.35-0.64) [<.001]
Sex/gender
Male	243 369 (59.0)	757 (71.5)	0.31	1 [Reference]
Female	165 027 (40.0)	286 (27.0)	0.17	0.57 (0.49-0.65) [<.001]
Transgender	1139 (0.3)	14 (1.3)	1.23	3.04 (1.77-5.22) [<.001]
Race and ethnicity
Asian, non-Hispanic	9849 (2.4)	17 (1.6)	0.17	0.55 (0.33-0.91) [.02]
Black, non-Hispanic	163 268 (39.6)	455 (43.0)	0.28	1 [Reference]
Hispanic/Latino	93 136 (22.6)	288 (27.2)	0.31	0.99 (0.84-1.16) [.88]
Other race, non-Hispanic c	5256 (1.3)	16 (1.5)	0.30	0.91 (0.55-1.49) [.71]
White, non-Hispanic	119 941 (29.1)	256 (24.2)	0.21	0.72 (0.62-0.84) [<.001]
US Census region
Northeast	50 033 (12.1)	162 (15.3)	0.32	1 [Reference]
Midwest	36 033 (8.7)	68 (6.4)	0.19	0.60 (0.45-0.81) [<.001]
South	268 435 (65.1)	584 (55.1)	0.22	0.74 (0.62-0.89) [.001]
West	53 687 (13.0)	204 (19.3)	0.38	1.18 (0.96-1.46) [.12]
US dependent areas	3976 (1.0)	41 (3.9)	1.03	3.02 (2.12-4.31) [<.001]
Total	412 164 (100.0)	1059 (100.0)	0.26	—
Population group: non–health care settings d
MSM + MSM who inject drugs	19 955 (20.6)	145 (41.8)	0.73	1 [Reference]
People who inject drugs	7054 (7.3)	21 (6.1)	0.30	0.35 (0.22-0.58) [<.001]
Heterosexual	47 301 (48.9)	139 (40.1)	0.29	0.31 (0.23-0.41) [<.001]
Other risk	1141 (1.2)	1 (0.3)	0.09	0.08 (0.01-0.56) [.01]
No reported risk	10 508 (10.9)	29 (8.4)	0.28	0.28 (0.18-0.44) [<.001]

Abbreviations: —, not applicable; aPR, adjusted prevalence ratio; CDC, Centers for Disease Control and Prevention; MSM, men who have sex with men.

^aData source: CDC, National HIV Prevention Program Monitoring and Evaluation Data, unpublished data, March 2021.

^baPRs, 95% CIs, and P values resulted from multivariable robust Poisson regression that included age group, sex/gender, race and ethnicity, and US Census region as independent variables. Separate models for non–health care settings included population group adjusted for age group, sex/gender, race and ethnicity, and US Census region. P < .05 considered significant.

^cOther race includes non-Hispanic American Indian/Alaska Native, Native Hawaiian or Other Pacific Islander, those who identified as multiple races or ethnicities, and those who reported another race or ethnicity.

^dPopulation group was routinely reported only for tests conducted in non–health care settings.

Table 2.

Linkage to HIV medical care and interviews for partner services among newly diagnosed HIV-positive adults aged ≥50 years receiving CDC-funded HIV tests, by selected characteristics, United States and US dependent areas, 2019 ^a

		Linked to HIV medical care within 30 days of diagnosis			Interviewed for HIV partner services
Characteristic	Newly diagnosed, no. (column %)	No. b (%)	Missing data, no. (%)	aPR (95% CI) [P value] c	No. b (%)	Missing data, no. (%)	aPR (95% CI) [P value] c
Age group, y
50-54	423 (39.9)	223/324 (68.8)	99/423 (23.4)	1 [Reference]	197/268 (73.5)	155/423 (36.6)	1 [Reference]
55-59	337 (31.8)	189/279 (67.7)	58/337 (17.2)	1.00 (0.89-1.11) [.96]	161/221 (72.9)	116/337 (34.4)	1.00 (0.90-1.12) [.95]
60-64	184 (17.4)	90/151 (59.6)	33/184 (17.9)	0.88 (0.76-1.02) [.10]	90/122 (73.8)	62/184 (33.7)	1.05 (0.93-1.18) [.42]
65-69	67 (6.3)	47/56 (83.9)	11/67 (16.4)	1.17 (1.02-1.35) [.02]	27/44 (61.4)	23/67 (34.3)	0.82 (0.64-1.05) [.12]
≥70	48 (4.5)	33/41 (80.5)	7/48 (14.6)	1.14 (0.97-1.35) [.11]	19/30 (63.3)	18/48 (37.5)	0.88 (0.67-1.16) [.37]
Sex/gender
Male	757 (71.5)	417/606 (68.8)	151/757 (19.9)	1 [Reference]	362/500 (72.4)	257/757 (33.9)	1 [Reference]
Female	286 (27.0)	160/233 (68.7)	53/286 (18.5)	1.00 (0.90-1.10) [.95]	129/181 (71.3)	105/286 (36.7)	1.00 (0.90-1.12) [.96]
Transgender	14 (1.3)	5/10 (50.0)	4/14 (28.6)	0.77 (0.43-1.40) [.40]	3/3 (100.0)	11/14 (78.6)	1.83 (1.33-2.52) [<.001]
Race and ethnicity
Asian, non-Hispanic	17 (1.6)	12/15 (80.0)	2/17 (11.8)	1.35 (1.04-1.75) [.02]	5/9 (55.6)	8/17 (47.1)	1.00 (0.58-1.71) [>.99]
Black, non-Hispanic	455 (43.0)	241/366 (65.8)	89/455 (19.6)	1 [Reference]	201/298 (67.4)	157/455 (34.5)	1 [Reference]
Hispanic/Latino	288 (27.2)	168/222 (75.7)	66/288 (22.9)	1.15 (1.02-1.28) [.02]	126/178 (70.8)	110/288 (38.2)	1.02 (0.89-1.18) [.75]
Other race, non-Hispanic d	16 (1.5)	8/12 (66.7)	4/16 (25.0)	1.08 (0.70-1.67) [.72]	4/8 (50.0)	8/16 (50.0)	0.75 (0.39-1.44) [.39]
White, non-Hispanic	256 (24.2)	140/214 (65.4)	42/256 (16.4)	1.03 (0.91-1.16) [.67]	152/182 (83.5)	74/256 (28.9)	1.28 (1.15-1.42) [<.001]
US Census region
Northeast	162 (15.3)	126/150 (84.0)	12/162 (7.4)	1 [Reference]	82/102 (80.4)	60/162 (37.0)	1 [Reference]
Midwest	68 (6.4)	36/52 (69.2)	16/68 (23.5)	0.85 (0.70-1.04) [.11]	48/57 (84.2)	11/68 (16.2)	0.97 (0.83-1.13) [.65]
South	584 (55.1)	281/434 (64.7)	150/584 (25.7)	0.78 (0.70-0.86) [<.001]	286/404 (70.8)	180/584 (30.8)	0.87 (0.77-0.98) [.02]
West	204 (19.3)	105/175 (60.0)	29/204 (14.2)	0.71 (0.61-0.82) [<.001]	48/88 (54.5)	116/204 (56.9)	0.65 (0.53-0.81) [<.001]
US dependent areas	41 (3.9)	34/40 (85.0)	1/41 (2.4)	0.90 (0.77-1.06) [.20]	30/34 (88.2)	7/41 (17.1)	1.18 (0.98-1.41) [.08]
Total	1059 (100.0)	582/851 (68.4)	208/1059 (19.6)	—	494/685 (72.1)	374/1059 (35.3)	—
Population group: non–health care settings e
MSM + MSM who inject drugs	145 (41.8)	80/108 (74.1)	37/145 (25.5)	1 [Reference]	63/99 (63.6)	46/145 (31.7)	1 [Reference]
People who inject drugs	21 (6.1)	8/17 (47.1)	4/21 (19.0)	0.56 (0.34-0.94) [.03]	6/16 (37.5)	5/21 (23.8)	0.58 (0.30-1.12) [.11]
Heterosexual	139 (40.1)	62/100 (62.0)	39/139 (28.1)	0.70 (0.53-0.92) [.01]	56/80 (70.0)	59/139 (42.4)	1.17 (0.87-1.57) [.31]
Other risk	1 (0.3)	—	1/1 (100.0)	—	—	1/1 (100.0)	—
No reported risk	29 (8.4)	13/22 (59.1)	7/29 (24.1)	0.69 (0.46-1.04) [.08]	6/16 (37.5)	13/29 (44.8)	0.54 (0.27-1.10) [.09]

Abbreviations: —, not applicable; aPR, adjusted prevalence ratio; CDC, Centers for Disease Control and Prevention; MSM, men who have sex with men.

^aData source: CDC, National HIV Prevention Program Monitoring and Evaluation Data, unpublished data, March 2021.

^bThose with missing data were excluded from the denominator and analysis.

^caPRs, 95% CIs, and P values resulted from multivariable robust Poisson regression that included age group, sex/gender, race and ethnicity, and US Census region as independent variables. Separate models for non–health care settings included population group adjusted for age group, sex/gender, race and ethnicity, and US Census region. P < .05 considered significant. Because of small cell sizes and model nonconvergence, data for Asian people were combined into data in the “other” race and ethnicity category in the analysis of data on people interviewed for partner services in non–health care settings.

^dOther race includes American Indian/Alaska Native, Native Hawaiian or Other Pacific Islander, those who identified as multiple races or ethnicities, and those who reported another race or ethnicity.

^ePopulation group was routinely reported only for tests conducted in non–health care settings.

Table 3.

Linkage to HIV medical care and interviews for partner services among previously diagnosed HIV-positive adults aged ≥50 years receiving CDC-funded HIV tests, by selected characteristics, United States and US dependent areas, 2019 ^a

		Previously diagnosed and linked to HIV medical care
Characteristic	Previously diagnosed, no. (column %)	Not in medical care at time of test, no. (row %)	Missing data, no. (%)	Linked, no. b (%)	aPR (95% CI) [P value] c
Age group, y
50-54	1059 (37.1)	507 (47.9)	161/507 (31.8)	173/346 (50.0)	1 [Reference]
55-59	984 (34.4)	452 (45.9)	172/452 (38.1)	132/280 (47.1)	0.93 (0.79-1.09) [.36]
60-64	555 (19.4)	259 (46.7)	102/259 (39.4)	84/157 (53.5)	1.04 (0.87-1.25) [.67]
65-69	166 (5.8)	62 (37.3)	21/62 (33.9)	22/41 (53.7)	1.03 (0.75-1.40) [.87]
≥70	94 (3.3)	41 (43.6)	13/41 (31.7)	14/28 (50.0)	1.15 (0.80-1.64) [.45]
Sex/gender
Male	2120 (74.2)	989 (46.7)	336/989 (34.0)	314/653 (48.1)	1 [Reference]
Female	711 (24.9)	320 (45.0)	131/320 (40.9)	106/189 (56.1)	1.15 (0.99-1.34) [.07]
Transgender	20 (0.7)	10 (50.0)	2/10 (20.0)	3/8 (37.5)	0.76 (0.32-1.80) [.53]
Race and ethnicity
Asian, non-Hispanic	19 (0.7)	11 (57.9)	2/11 (18.2)	4/9 (44.4)	0.82 (0.39-1.71) [.60]
Black, non-Hispanic	1660 (58.1)	767 (46.2)	269/767 (35.1)	255/498 (51.2)	1 [Reference]
Hispanic/Latino	437 (15.3)	188 (43.0)	65/188 (34.6)	68/123 (55.3)	1.06 (0.88-1.29) [.53]
Other race, non-Hispanic d	33 (1.2)	26 (78.8)	11/26 (42.3)	10/15 (66.7)	1.31 (0.88-1.96) [.18]
White, non-Hispanic	609 (21.3)	276 (45.3)	85/276 (30.8)	81/191 (42.4)	0.85 (0.70-1.02) [.08]
US Census region
Northeast	527 (18.4)	262 (49.7)	21/262 (8.0)	104/241 (43.2)	1 [Reference]
Midwest	79 (2.8)	38 (48.1)	12/38 (31.6)	14/26 (53.8)	1.22 (0.84-1.77) [.30]
South	1970 (68.9)	845 (42.9)	347/845 (41.1)	256/498 (51.4)	1.20 (1.01-1.41) [.04]
West	252 (8.8)	166 (65.9)	87/166 (52.4)	45/79 (57.0)	1.36 (1.06-1.74) [.02]
US dependent areas	30 (1.0)	10 (33.3)	2/10 (20.0)	6/8 (75.0)	1.59 (1.01-2.51) [.045]
Total	2858 (100.0)	1321 (46.2)	469/1321 (35.5)	425/852 (49.9)	—
Population group (non–health care settings) e
MSM + MSM who inject drugs	212 (27.8)	111 (52.4)	57/111 (51.4)	31/54 (57.4)	1 [Reference]
People who inject drugs	70 (9.2)	32 (45.7)	10/32 (31.3)	9/22 (40.9)	0.64 (0.36-1.13) [.12]
Heterosexual	288 (37.8)	136 (47.2)	71/136 (52.2)	35/65 (53.8)	0.84 (0.56-1.25) [.39]
Other risk	8 (1.0)	6 (75.0)	1/6 (16.7)	1/5 (20.0)	0.31 (0.05-1.96) [.21]
No reported risk	111 (14.6)	56 (50.5)	18/56 (32.1)	23/38 (60.5)	0.96 (0.63-1.46) [.85]

Abbreviations: —, not applicable; aPR, adjusted prevalence ratio; CDC, Centers for Disease Control and Prevention; MSM, men who have sex with men.

^aData source: CDC, National HIV Prevention Program Monitoring and Evaluation Data, unpublished data, March 2021.

^bThose with missing data were excluded from the denominator and analysis.

^caPRs, 95% CIs, and P values resulted from multivariable robust Poisson regression that included age group, sex/gender, race and ethnicity, and US Census region as independent variables. Separate models for non–health care settings included population group adjusted for age group, sex/gender, race and ethnicity, and US Census region. P < .05 considered significant.

^dOther race includes American Indian/Alaska Native, Native Hawaiian or Other Pacific Islander, those who identified as multiple races or ethnicities, and those who reported another race or ethnicity.

^ePopulation group was routinely reported only for tests conducted in non–health care settings.

The number of new HIV diagnoses was highest, by age, among people aged 50-54 years (n = 423; 39.9%); by sex/gender, among males (n = 757; 71.5%); by race and ethnicity, among Black people (n = 455; 43.0%); by US Census region, among people in the South (n = 584; 55.1%); and by population group, among MSM (n = 145; 41.8%) (Table 2). New positivity was lower among people aged ≥60 years than among people aged 50-54 years (60-64 years: aPR = 0.71; 65-69 years: aPR = 0.53; ≥70 years: aPR = 0.47; all P < .001) (Table 1). When compared with new positivity among males, new positivity was 3 times higher among transgender people (aPR = 3.04; P < .001) and lower among females (aPR = 0.57; P < .001). New positivity was lower among White (aPR = 0.72; P < .001) and Asian (aPR = 0.55; P = .02) people than among Black people. When compared with the rate in the Northeast, new positivity was higher in the US dependent areas (aPR = 3.02; P < .001) and lower in the Midwest (aPR = 0.60; P < .001) and the South (aPR = 0.74; P = .001). In non–health care settings, as compared with the rate among MSM, new positivity was lower among PWID (aPR = 0.35; P < .001), heterosexual people (aPR = 0.31; P < .001), people with other risk (aPR = .08; P = .01), and people with no reported risk (aPR = 0.28; P < .001).

Among older adults with newly diagnosed HIV infection, of 851 people with linkage data, 582 (68.4%) were linked to HIV medical care within 30 days of diagnosis, and of 685 people with partner services data, 494 (72.1%) were interviewed for partner services (Table 2). Linkage to care was higher among people aged 65-69 years (aPR = 1.17; P = .02) than among people aged 50-54 years. Linkage was also higher among Hispanic/Latino (aPR = 1.15; P = .02) and Asian (aPR = 1.35; P = .02) people than among Black people. By US Census region, linkage was lower in the South (aPR = 0.78; P < .001) and the West (aPR = 0.71; P < .001) than in the Northeast. In non–health care settings, linkage was lower among PWID (aPR = 0.56; P = .03) and heterosexual people (aPR = 0.70; P = .01) than among MSM. Interview for partner services was higher among transgender people (aPR = 1.83; P < .001) than among males and higher among White people (aPR = 1.28; P < .001) than among Black people. Interview for partner services was lower in the South (aPR = 0.87; P = .02) and West (aPR = 0.65; P < .001) than in the Northeast.

Of the 2858 older adults in this analysis with a previously diagnosed HIV infection, 1321 (46.2%) were not in HIV medical care at the time of the test. Of these 1321 people, 852 (64.5%) had linkage information, and of these 852 people, 425 (49.9%) were linked to HIV medical care within 30 days of receiving a positive test result for HIV (Table 3). We found no significant differences in linkage by age, sex/gender, race and ethnicity, or population groups. Linkage was higher among older adults tested in the South (aPR = 1.20; P = .04), West (aPR = 1.36; P = .02), and US dependent areas (aPR = 1.59; P = .045) than in the Northeast.

Practice Implications

Studies on sexuality among older adults report that condom use is low when compared with use among younger adults, as are discussions about sexual health during health care visits and health care providers’ perception of risk for sexually transmitted diseases and HIV among older patients.^12,13 HIV testing among this population is important because older adults who know their HIV-positive status are less likely to engage in risk behaviors than those who do not know their HIV status. ² The proportionally higher number of health care visits among older adults ¹⁴ presents an opportunity for health care providers to discuss HIV-related risks and prevention, such as HIV testing and preexposure prophylaxis medication. Specifically, older MSM and transgender people especially would appear to benefit from frequent HIV testing. Current CDC HIV testing guidelines recommend that people aged 13-64 years be screened for HIV at least once and that people at higher risk for HIV, including MSM, be tested more frequently. ¹⁵ Our results suggest that because people in the United States are living longer and healthier lives than they have previously, CDC might consider reexamining these guidelines to include screening people aged ≥65 years for HIV.

Our findings show that nearly three-quarters of older adults who received a positive test result for HIV through CDC’s HIV testing program were previously diagnosed with HIV. Previous analyses of CDC’s HIV testing program show that, generally, approximately half of all HIV-positive test results were among previously diagnosed HIV-positive people. ¹⁶ The higher proportion in our study may be attributable to several factors, such as older people interacting more than younger people with health care providers or needing to be relinked to care when changing health care providers. ¹⁷ CDC does not collect information on why a person is tested for HIV. Each health department and CBO approaches HIV testing differently, so we cannot know precisely why a high percentage of people previously diagnosed with HIV receive a CDC-funded HIV test. For example, some policies may require that clients or patients receive a test regardless of whether they disclose their HIV status. Regardless of why a person is retested, the interaction presents a unique opportunity to ensure that people with HIV are promptly linked to HIV medical care.

For people living with HIV to be virally suppressed, it is essential that they be tested for HIV and promptly linked to and retained in HIV medical care.^18,19 The national goal is to link 95% of people with newly diagnosed HIV to HIV medical care within 30 days of diagnosis. ²⁰ We observed low linkage rates among newly and previously diagnosed older adults (68% and 50%, respectively). Linkage to HIV medical care among all people tested through CDC’s HIV testing program was slightly higher than it was among older adults, in general, at 71% among newly diagnosed people. ²¹ Other recent publications on priority populations, including MSM and transgender people, report low rates of linkage to HIV medical care,^21,22 suggesting the need for enhanced linkage-to-care efforts in general and not just among older adults. One core pillar of the US Department of Health and Human Services initiative, Ending the HIV Epidemic in the United States, is treating HIV infection. ²³ Through this initiative, CDC is funding jurisdictions to scale up linkage and treatment, with >60% of jurisdictions expanding telemedicine services, especially in rural areas. ²³ This scale-up is also especially important for older adults because transportation and mobility may be challenges to receiving care.

Provision of partner services is an important strategy for identifying people with HIV. Because older adults may have a low perception of risk, they may not be getting tested for HIV as often as they should. ⁵ Partner services can help bridge this gap by reaching out to potentially affected partners. CDC recommends that all people with newly diagnosed HIV receive partner services to help them identify sex or needle-sharing partners who may also be infected or may be at very high risk for becoming infected. ⁹ Our analysis found that 72.1% of newly diagnosed older adults were interviewed for partner services. This percentage is lower than the national average for all age groups (80% in 2019), suggesting room for improvement among this population. ²¹ Partner services has proven to be an effective strategy for identifying priority populations for HIV testing, ⁹ and older adults are in need of enhanced partner services. Enhancement may include embedding more disease intervention staff at organizations that provide HIV-related services and when conducting partner notification, paying increased attention to older adults.

Our study had at least 3 limitations. First, findings described CDC-funded HIV tests only and are not generalizable to HIV testing rates among all older adults in the United States. Although these results are not generalizable, our analysis helps to establish CDC’s HIV prevention efforts among older adults. Health departments are funded to provide HIV prevention services to their entire jurisdictions, and CBOs provide targeted testing, for example, to young MSM of color and young transgender people of color. CDC does not have notices of funding opportunities that directly prioritize older adults; many funding opportunities focus on disproportionately affected younger populations. The results from our analysis support the need for enhanced linkage to care among older people, especially at health departments. Second, linkage to HIV medical care and interview for partner services data excluded records with missing or invalid data in the denominator and may therefore have overestimated the percentage of people linked to care or interviewed for partner services. Missing data in national-scale data collection have been reported, and CDC continues to work closely with health departments and CBOs to ensure high-quality data and completeness. Third, when surveillance data were unavailable to verify previous HIV status, misclassification of new or previous diagnoses may have occurred if people inaccurately reported their HIV testing history. A recently published analysis found that the proportion of tests confirmed in surveillance was increasing over time, with 81% of CDC-funded new diagnoses confirmed in surveillance and not based on self-report in 2017 (the latest year analyzed). ¹⁰

Conclusion

Older adults represent nearly half of all people living with HIV in the United States. ¹ More rapid disease progression and higher morbidity and mortality rates among older adults, as compared with younger people, suggest that early diagnosis and rapid linkage to care in this population are critical.^7,18 Our results show, however, that linkage to HIV medical care can be improved, with linkage among newly diagnosed older adults at 68.4% and previously diagnosed older adults at 49.9%. Similarly, the rate of interview for partner services among older adults was low (72.1%) and can be improved. In sum, testing among older adults continues to be an important part of HIV prevention in the United States, and future efforts should focus on improving linkage to care in this population.