The 2013 HIV Continuum of Care in Tennessee

Abstract

Objectives:

We measured patient engagement in the human immunodeficiency virus (HIV) continuum of care in Tennessee after implementation of enhanced surveillance activities to assess progress toward 2015 statewide goals. We also examined subgroup disparities to identify groups at risk for poor outcomes.

Methods:

We estimated linkage to care, retention in care, and viral suppression among HIV-infected people in Tennessee in 2013, overall and by subgroup, after implementation of enhanced laboratory reporting, address verification, and death-matching procedures.

Results:

Of 792 people newly diagnosed with HIV infection in 2013, 632 (79.8%) were linked to care, close to the 2015 goal of ≥80%. Of 15 473 people living and diagnosed with HIV infection before 2013, 8458 (54.7%) were retained in care, approaching the 2015 goal of ≥64.0%. A total of 8640 (55.8%) were virally suppressed, surpassing the 2015 goal of ≥51.0%. Compared with people living and diagnosed with HIV infection before 2013, newly diagnosed people were more likely to be younger, male, non-Hispanic black, and men who have sex with men (MSM). For linkage to care, retention in care, and viral suppression, younger and non-Hispanic black people fared worse, whereas females and those enrolled in the Ryan White program fared better. For retention in care and viral suppression, Hispanic people, injection drug users, and East Tennessee residents fared worse than those in Memphis, whereas MSM fared better. Nashville residents fared worse in retention in care than Memphis residents.

Conclusion:

Tennessee’s HIV continuum of care in 2013 showed progress toward 2015 goals. Future efforts to improve the HIV continuum of care should be directed toward vulnerable groups and regions, particularly young, non-Hispanic black, and Hispanic people; injection drug users; and residents of the East Tennessee and Nashville regions.

The Centers for Disease Control and Prevention (CDC) estimates that approximately 900 000 people aged ≥13 years in the United States are living with diagnosed human immunodeficiency virus (HIV) infection, and about 44 000 new cases are reported annually. Given improved survival rates with antiretroviral therapy (ART), the number of people living with HIV infection is rising, and the number of new infections is stable.¹ However, new HIV infections disproportionately affect black people and men who have sex with men (MSM),^1,2 especially black MSM.³

The success of HIV prevention efforts depends on early testing and initiation of ART.⁴ The importance of treatment as prevention is highlighted in the US National HIV/AIDS Strategy.⁵ The HIV continuum of care, introduced in 2005 by Giordano et al,⁶ refers to a series of steps required for a patient to benefit from ART and suppression of HIV viral replication (ie, viral suppression). These steps include learning of one’s HIV-positive status, entering health care services (ie, linkage to care), being prescribed ART, and adhering to medical appointments (ie, retention in care) and medications. Assessment of the continuum of care can help policymakers monitor the engagement of various populations in HIV medical care.^7–10 The CDC encourages states and localities to conduct HIV continuum-of-care analyses to tailor local programs and HIV-related services.^11–13

The southern United States faces substantial challenges in addressing its HIV epidemic^14–18 and was named a key geographic area to focus HIV care and prevention efforts by the National HIV/AIDS Strategy.⁵ The southern United States has the highest rate of HIV diagnoses compared with other regions of the country. Tennessee ranks seventeenth highest among the 50 US states and the District of Columbia in the rate of HIV diagnoses (14.7 per 100 000 population) and twenty-first highest in HIV prevalence (297.2 per 100 000 population) among adults and adolescents ≥13 years of age.^1,19 Poverty, low levels of access to care in rural areas, racial/ethnic discrimination, and the disproportionate incarceration rates of black men that alter sexual networks in at-risk communities are notable determinants of health in the South. A lack of comprehensive sexual education in schools and syringe exchange programs may also increase HIV transmission risk. Southern states typically have restrictive eligibility requirements for Medicaid and comparatively limited contributions to Acquired Immunodeficiency Syndrome (AIDS) Drug Assistance Programs.²⁰

To address goals of the National HIV/AIDS Strategy, the Tennessee Department of Health (TDH) conducted a baseline HIV continuum-of-care surveillance study in 2010, which found that fewer people with newly diagnosed HIV infection were linked to care in Tennessee than nationally (64% vs 80%).²¹ Furthermore, fewer people living with diagnosed HIV infection in Tennessee compared with nationally were retained in care (29% vs 51%) and virally suppressed (35% vs 43%).²² Based on these results, TDH set goals for 2015: (1) ≥80% of people newly diagnosed with HIV infection will be linked to medical care within 3 months of diagnosis, (2) ≥64% of people living with diagnosed HIV infection will be retained in care, and (3) ≥51% of people living with diagnosed HIV infection will have achieved viral suppression. Consistent with the National HIV/AIDS Strategy, TDH also established goals to achieve ≥20% improvement in HIV viral suppression rates among disproportionately affected populations, including MSM, non-Hispanic black and Hispanic people, and people <35 years of age (Table 1).²³

Table 1.

2015 HIV prevention and care goals and progress in Tennessee (2010 and 2013) compared with national progress (2012-2013)^a

	Tennessee performanceb		National
Tennessee 2015 goalsb	2010	2013	performancec (y)
Goals for increasing access to care and improving health outcome among people living with HIV
Linkage to HIV medical care within 3 months of diagnosis is ≥80%	63.7	79.8	82.1 (2013)
Retention in HIV medical care is ≥64%	29.2	54.7	53.8 (2012)
Viral suppression is ≥51%	35.0	55.8	50.1 (2012)
Goals for reducing HIV-related disparities
Increase viral suppression among MSM by 20% compared with 2010 performance baseline (≥47%)	39.1	59.5	53.2 (2012)
Increase viral suppression among black people by 20% compared with 2010 performance baseline (≥51%)	31.1	52.7	44.2 (2012)
Increase viral suppression among Hispanic people by 20% compared with 2010 performance  baseline (≥36%)	29.7	45.9	50.1 (2012)
Increase viral suppression among 25- to 34-year-olds by 20% from 2010 performance baseline (≥34%)	27.8	48.6	43.5 (2012)

Abbreviations: HIV, human immunodeficiency virus; MSM, men who have sex with men. ^aFigures are percentages.

^bData source: Tennessee Department of Health, HIV/STD/Viral Hepatitis Program. 2010 HIV continuum of care: Tennessee. https://tn.gov/assets/entities/health/attachments/2010.TN_HIV_Cont_of_Care.White_Paper.Final.03March14.pdf. Accessed March 16, 2016.

^cData source: Centers for Disease Control and Prevention (US). Monitoring selected national HIV prevention and care objectives by using surveillance data—United States and 6 dependent areas—2013. HIV Surveillance Supplemental Report. 2015;20:1-70. http://www.cdc.gov/hiv/library/reports/surveillance. Accessed January 28, 2016.

Tennessee changed its HIV surveillance and registry linkage practices after the 2010 baseline analysis to more accurately enumerate people living with diagnosed HIV and measure their engagement in HIV care. First, all laboratories performing CD4+ cell counts (CD4) and viral load testing on patients in Tennessee were required to report these test results to TDH effective January 1, 2012.²⁴ Second, TDH used commercial person-searching software (LexisNexis Accurint) to verify identities and update address information for people living with diagnosed HIV infection.²⁵ Third, TDH matched state HIV surveillance data with Tennessee death certificate and National Death Index data.²⁶ The TDH also received CDC funding for the Care and Prevention in the United States Demonstration Project (CAPUS) in 2012.²⁷ The objective of this study was to describe TDH’s most recent statewide HIV continuum-of-care data to gauge progress and highlight disparities in outcomes.

Methods

Study Population and Data Collection

The TDH conducted a cross-sectional study among people living with diagnosed HIV infection in Tennessee in 2013, a diagnosis-based rather than prevalence-based HIV continuum of care.²⁸ Eligible patients included all those in the Enhanced HIV/AIDS Reporting System (eHARS), CDC’s National HIV Surveillance System.²⁹

We used laboratory and demographic data from eHARS, including data on age (<15, 15-24, 25-34, 35-44, 45-54, and ≥55 years), sex at birth (male or female, regardless of current gender identity), race/ethnicity (Hispanic, non-Hispanic white [white], non-Hispanic black [black], and non-Hispanic other), HIV risk factors, residential address, and date of death. Self-reported HIV transmission risk factors were MSM, injection drug use (IDU), both MSM and IDU (MSM/IDU), heterosexual contact, other (including perinatal infection), and unknown. We defined 4 geographic regions by site of patient residence: Davidson County Ryan White Transitional Grant Area (TGA) (Nashville TGA), Shelby County Ryan White TGA (Memphis TGA), other West (the western portion of Tennessee excluding the Memphis TGA), and other East (the eastern portion of the state including Knoxville but excluding the Nashville TGA). A TGA is defined by the Health Resources and Services Administration as an area with 1000 to 1999 AIDS cases in the most recent 5 years and a population of at least 50 000.³⁰ People were defined as enrollees in the Ryan White HIV/AIDS program if they were enrolled at any time during 2013 for services including assistance with medications, case management, insurance, and/or nutrition.

Continuum-of-care Outcomes

Consistent with CDC surveillance definitions,²⁸ we used laboratory data (CD4 counts and viral load tests) to ascertain linkage to care, retention in care, and viral suppression. We defined people diagnosed and living with HIV infection as people diagnosed through December 31, 2012, and alive and living in Tennessee as of December 31, 2013. We defined people with newly diagnosed HIV infection as people newly diagnosed with HIV infection in Tennessee during 2013. People newly diagnosed with HIV infection during 2013 were defined as linked to HIV care within 3 months of diagnosis (up to March 31, 2014, at the latest), as evidenced in eHARS by ≥1 CD4 count or viral load test within 3 months of diagnosis. We defined people diagnosed and living with HIV infection (excluding newly diagnosed) as retained in care if they had ≥2 HIV care visits in 2013 at least 3 months apart, as evidenced in eHARS by ≥2 CD4 counts and/or viral load tests ≥3 months apart. We defined people diagnosed and living with HIV infection (excluding newly diagnosed) as virally suppressed if they achieved viral suppression by the end of 2013, as evidenced in eHARS by ≥1 viral load test in 2013 and the most recent viral load test in 2013 measuring ≤200 copies per milliliter.

Because national estimates of the number of people living with undiagnosed HIV infection differ by subpopulation and therefore may differ from local jurisdictional estimates, Tennessee’s HIV continuum of care begins with people living with diagnosed HIV infection.¹⁹ We did not include data on receipt of ART because patient-level data on this indicator were not available.

We compared HIV continuum-of-care outcomes for Tennessee during 2013 with 2010 results to assess progress toward 2015 statewide goals. We also compared 2013 Tennessee HIV continuum-of-care outcomes with national results for 2012.³¹

Statistical Analysis

We conducted statistical comparisons across characteristics by using Pearson’s χ² tests of differences in proportions. We used modified Poisson regression analysis to estimate prevalence ratios and 95% confidence intervals (CIs) for associations.³² We included all factors in the adjusted models. Values for all variables included in the regression models were available for every person diagnosed and living with HIV infection in Tennessee during the study period. All tests were 2-tailed. We conducted statistical analyses using Stata version 12.1.³³ We performed a preplanned sensitivity analysis only among MSM and MSM <35 years of age to identify any racial/ethnic disparities masked in the overall MSM population.

Results

The TDH eHARS database documented a cumulative total of 21 995 people diagnosed with HIV infection in Tennessee in 2013. Of these, 8609 (39.1%) had a confirmed Tennessee address: 6988 (31.8%) had a new Tennessee address that could be updated from the existing Tennessee address in the database, 3311 (15.2%) were living out of state, and 3087 (14.0%) had no address match. After matching HIV surveillance data with Tennessee death certificate data and National Death Index data, we determined that 15 473 people diagnosed with HIV infection by the end of 2012 were alive and residing in Tennessee at the end of 2013; 792 were newly diagnosed as HIV positive in 2013 (Table 2). Among these 15 473 people, the mean time since HIV diagnosis was 10.7 years (median = 10, interquartile range, 5-16).

Table 2.

Estimated number and characteristics of people with HIV infection engaged in the HIV continuum of care, Tennessee, 2013

	Total newly diagnosed with HIV infectiona	Linked to care within 3 months of diagnosisb		Living and diagnosed with HIV infectionc	Retained in cared		Virally suppressede
Characteristic	No.	No. (%)	P valuef	No.	No. (%)	P valuef	No. (%)	P valuef
Total	792	632 (79.8)		15 473	8458 (54.7)		8640 (55.8)
Age, yg			.04			<.001		<.001
<15	4	4 (100.0)		110	61 (55.5)		55 (50.0)
15-24	202	151 (74.8)		629	367 (58.3)		300 (47.7)
25-34	244	183 (75.0)		2567	1280 (49.9)		1247 (48.6)
35-44	138	121 (87.7)		3954	2006 (50.7)		2043 (51.7)
45-54	130	107 (82.3)		5168	2944 (57.0)		3104 (60.1)
≥55	74	66 (89.2)		3045	1800 (59.1)		1891 (62.1)
Sex			.025			.341		.027
Male	633	495 (78.2)		11 357	6182 (54.4)		6402 (56.4)
Female	159	137 (86.2)		4116	2276 (55.3)		2238 (54.4)
Race/ethnicity			.08			<.001		<.001
Non-Hispanic white	247	209 (84.6)		5765	3136 (54.4)		3521 (61.1)
Non-Hispanic black	477	361 (75.7)		8667	4747 (54.8)		4567 (52.7)
Hispanic	38	34 (89.5)		688	320 (46.5)		316 (45.9)
Otherh	30	28 (93.3)		353	255 (72.2)		236 (66.9)
HIV risk category			.176			<.001		<.001
Heterosexual	166	140 (84.3)		3593	2002 (55.7)		1974 (54.9)
MSM	465	372 (80.0)		7204	4057 (56.3)		4284 (59.5)
IDU	8	7 (87.5)		1097	545 (49.7)		557 (50.8)
MSM/IDU	8	6 (75.0)		464	234 (50.4)		232 (50.0)
Other	66	52 (78.8)		249	129 (51.8)		123 (49.4)
Unknown (no identified risk)	79	55 (69.6)		2866	1491 (52.0)		1470 (51.3)
Region of residence			.768			<.001		<.001
Memphis TGAi	329	258 (78.4)		5866	3314 (56.5)		3183 (54.3)
Nashville TGAi	223	183 (82.1)		4705	2695 (57.3)		2784 (59.2)
West otherj	81	65 (80.2)		1424	836 (58.7)		842 (59.1)
East otherj	159	126 (79.2)		3478	1613 (46.4)		1831 (52.6)
Enrollment in Ryan White services			<.001			<.001		<.001
No	475	352 (74.1)		9724	4474 (46.0)		4608 (47.4)
Yes	317	280 (88.3)		5749	3984 (69.3)		4032 (70.1)

Abbreviations: HIV, human immunodeficiency virus; IDU, injection drug use; MSM, men who have sex with men; TGA, transitional grant area. ^aNewly diagnosed with HIV infection refers to people newly diagnosed with HIV infection in Tennessee during 2013.

^bLinked to care refers to people with ≥1 CD4 count or viral load test reported within 3 months of diagnosis date among people newly diagnosed with HIV infection in 2013 in Tennessee.

^cDiagnosed and living with HIV refers to people diagnosed and living with HIV infection through December 31, 2012, and alive and living in Tennessee as of December 31, 2013.

^dRetained in care refers to people with ≥2 CD4 count and/or viral load tests reported ≥3 months apart in 2013 among people diagnosed with HIV infection on or before December 31, 2012, and alive and residing in Tennessee as of December 31, 2013.

^eVirally suppressed refers to people with ≥1 viral load test in 2013 and whose most recent test result in 2013 indicated ≤200 copies per milliliter among all people diagnosed with HIV infection on or before December 31, 2012, and alive and residing in Tennessee as of December 31, 2013.

^fP value from Pearson’s χ² test for difference in proportions by categorical variable (comparing yes with no proportions within demographic categories).

^gAge was calculated as of December 31, 2013.

^hOther race includes Native Hawaiian/other Pacific Islander, Asian, American Indian/Alaska Native, and multiple races.

ⁱA TGA is an area with 1000 to 1999 cases of acquired immunodeficiency syndrome in the most recent 5 years and a population of at least 50 000.

^jWest other refers to the western portion of Tennessee, excluding the Memphis TGA; East other refers to the eastern portion of Tennessee, excluding the Nashville TGA.

Compared with the 15 473 cumulatively diagnosed people, the 792 newly diagnosed people were younger (56.8% vs 21.4% <35 years of age), male (79.9% vs 73.4%), black (60.2% vs 56.0%), and MSM (58.7% vs 46.6%). Both groups were similar to each other in their geographic distribution and enrollment in the Ryan White program (Table 2).

Of those newly diagnosed with HIV infection in 2013, 632 (79.8%) were linked to care within 3 months of diagnosis. Of those living and diagnosed with HIV infection prior to 2013, 8458 (54.7%) were retained in care, and 8640 (55.8%) were virally suppressed (Table 2). All of these proportions represent marked increases from the 2010 Tennessee analysis (63.7%, 29.2%, and 35.0% for linkage to care, retention in care, and viral suppression, respectively). One of the 3 measures, viral suppression, exceeded the 2012 national average (55.8% in Tennessee vs 50.1% nationally; 2013 data were not yet available), and the other 2 measures were close to national averages (79.8% in Tennessee vs 82.1% nationally for linkage to care; 54.7% in Tennessee vs 53.8% nationally for retention in care) (Table 1).³¹

Levels of engagement in Tennessee’s HIV continuum of care differed by age group. People aged 25 to 34 years had lower rates of participation in the HIV continuum of care than people aged 35 to 44 years, with 75.0% linked to care (vs 87.7%), 49.9% retained in care (vs 50.7%), and 48.6% virally suppressed (vs 51.7%) in 2013. People aged ≥55 years had the highest rates of engagement in care, with 89.2% (P = .004) linked to care, 59.1% (P < .001) retained in care, and 62.1% (P < .001) virally suppressed (Table 2). This age differential in HIV continuum-of-care elements applied to each adjusted regression model (Table 3).

Table 3.

Adjusted prevalence ratios and 95% confidence intervals for engagement in the HIV continuum of care among people living and diagnosed or newly diagnosed with HIV infection, Tennessee, 2013

	Linked to carea	Retained in careb	Virally suppressedc
Characteristic	aPRd (95% CI)	aPRd (95% CI)	aPRd (95% CI)
Age, ye
<15	1.22 (0.99-1.50)	1.22 (1.00-1.49)	1.07 (0.85-1.34)
15-24	0.85* (0.76-0.95)	1.06 (0.99-1.14)	0.87* (0.79-0.95)
25-34	0.86* (0.78-0.94)	0.94* (0.89-0.98)	0.91* (0.87-0.95)
35-44	Reference	Reference	Reference
45-54	0.97 (0.88-1.07)	1.15* (1.10-1.19)	1.17* (1.13-1.21)
≥55	1.08 (0.98-1.20)	1.24* (1.19-1.29)	1.26* (1.21-1.31)
Sex
Male	Reference	Reference	Reference
Female	1.14* (1.02-1.28)	1.06* (1.02-1.11)	1.05* (1.01-1.10)
Race/ethnicity
Non-Hispanic white	Reference	Reference	Reference
Non-Hispanic black	0.90* (0.82-0.98)	0.96* (0.93-0.99)	0.86* (0.84-0.89)
Hispanic	1.10 (0.96-1.26)	0.91* (0.81-0.95)	0.79* (0.73-0.86)
Otherf	1.14 (0.99-1.31)	1.28* (1.15-1.33)	1.10* (1.02-1.19)
HIV risk category
Heterosexual	Reference	Reference	Reference
MSM	1.07 (0.94-1.21)	1.06* (1.01-1.11)	1.07* (1.02-1.13)
IDU	0.94 (0.71-1.23)	0.84* (0.83-0.95)	0.88* (0.82-0.94)
MSM/IDU	0.92 (0.61-1.37)	0.88 (0.84-1.02)	0.86* (0.78-0.95)
Other	0.95 (0.82-1.10)	0.93 (0.84-1.14)	1.04 (0.88-1.21)
Unknown	0.93 (0.79-1.09)	0.93 (0.93-1.02)	0.97 (0.93-1.02)
Region of residence
Memphis TGAg	Reference	Reference	Reference
Nashville TGAg	0.99 (0.90-1.08)	0.94* (0.91-0.98)	0.97 (0.94-1.01)
West otherh	0.97 (0.86-1.10)	0.97 (0.92-1.02)	0.98 (0.93-1.03)
East otherh	0.92 (0.82-1.04)	0.75* (0.72-0.79)	0.83* (0.79-0.86)
Enrollment in Ryan White services
No	Reference	Reference	Reference
Yes	1.21 (1.13-1.29)	1.55* (1.51-1.60)	1.52* (1.48-1.56)

Abbreviations: aPR, adjusted prevalence ratio; CI, confidence interval; HIV, human immunodeficiency virus; IDU, injection drug use; MSM, men who have sex with men; TGA, transitional grant area. ^aLinkage refers to ≥1 CD4 count or viral load test reported within <3 months of diagnosis date among people newly diagnosed with HIV infection in 2013 in Tennessee (compared with those not meeting the definition).

^bRetention refers to ≥2 CD4 count and/or viral load tests reported ≥3 months apart in 2013 among people diagnosed with HIV infection on or before December 31, 2012, and alive and residing in Tennessee as of December 31, 2013 (compared with those not meeting the definition).

^cViral suppression refers to ≥1 viral load test in 2013 and those whose most recent test in 2013 indicated ≤200 copies per milliliter among all people diagnosed with HIV infection on or before December 31, 2012, and alive and residing in Tennessee as of December 31, 2013 (compared with those not meeting the definition).

^dEstimates derived from adjusted modified Poisson regression model including all factors in table.

^eAge was calculated as of December 31, 2013.

^fOther race includes Native Hawaiian/other Pacific Islander, Asian, American Indian/Alaska Native, and multiple races.

^gA TGA is an area with 1000 to 1999 cases of acquired immunodeficiency syndrome in the most recent 5 years and a population of at least 50 000.

^hWest other refers to the western portion of Tennessee, excluding the Memphis TGA; East other refers to the eastern portion of Tennessee, excluding the Nashville TGA.

*P < .05.

Females were more likely than males to be linked to care (86.2% vs 78.2%) but had similar rates of retention in care (Table 2). In adjusted analyses, females had significantly better outcomes than males for linkage to care (adjusted prevalence ratio [aPR] = 1.14, 95% CI, 1.02-1.28), retention in care (aPR = 1.06, 95% CI, 1.02-1.11), and viral suppression (aPR = 1.05, 95% CI, 1.01-1.10) (Table 3).

Black people had a larger burden of HIV infection and were less likely to be engaged in care than white people. Among those newly diagnosed with HIV infection, 84.6% of white people, compared with 75.7% of black people and 89.5% of Hispanic people (P = .08), were linked to care. Racial/ethnic disparities persisted at later stages, with 54.4% of white people, 54.8% of black people, and 46.5% of Hispanic people retained in care (P < .001). Black (52.7%) and Hispanic (45.9%) people had lower rates of viral suppression than white people (61.1%) (P < .001) (Table 2). These disparities persisted after adjusting for all other demographic characteristics and enrollment in Ryan White services, with black people faring worse than white people in linkage to care, retention in care, and viral suppression (Table 3).

We also found differences in engagement by transmission category. Among those living with an HIV diagnosis in 2013, MSM (56.3%) and those with heterosexual risk (55.7%) were more likely than IDU (49.7%) and MSM/IDU (50.4%) to be retained in care (P < .001). Viral suppression was higher among MSM (59.5%) than among heterosexuals (54.9%), IDU (50.8%), or MSM/IDU (50.0%) (P < .001) (Table 2). After accounting for all other demographic characteristics and enrollment in Ryan White services, MSM fared better than heterosexuals and IDU fared worse than heterosexuals in retention in care and viral suppression (Table 3).

We also identified differences in HIV continuum-of-care outcomes by region and enrollment in the Ryan White program. In the adjusted model, those residing in the Nashville TGA (aPR = 0.94, 95% CI, 0.91-0.98) and the east (aPR = 0.75, 95% CI, 0.72-0.79) fared worse than those in the Memphis TGA for retention in care, yet only the east (aPR = 0.83, 95% CI, 0.79-0.86) lagged in viral suppression behind the Memphis TGA (Table 3). People enrolled in the Ryan White program, compared with those not enrolled, had higher rates of linkage to care (88.3% vs 74.1%), retention in care (69.3% vs 46.0%), and viral suppression (70.1% vs 47.4%) (Table 2).

In an adjusted sensitivity analysis among MSM, black MSM were less likely than white MSM to be retained in care (aPR = 0.94, 95% CI, 0.89-0.98), and both black MSM (aPR = 0.82, 95% CI, 0.79-0.86) and Hispanic MSM (aPR = 0.78, 95% CI, 0.69-0.87) were less likely than white MSM to be virally suppressed. The MSM aged 15 to 24 and 25 to 34 years were less likely than MSM aged 35-44 years to be virally suppressed (aPR = 0.80, 95% CI, 0.71-0.91 vs aPR = 0.88, 95% CI, 0.82-0.94, respectively). The MSM aged 25 to 34 years were less likely than MSM aged 35 to 44 years to be retained in care and virally suppressed (Table 4). In a sensitivity analysis among MSM <35 years of age, black MSM were less likely than white MSM to be virally suppressed (aPR = 0.79, 95% CI, 0.70-0.89).

Table 4.

Adjusted prevalence ratios and 95% confidence intervals for engagement in the HIV continuum of care among only MSM living and diagnosed or newly diagnosed with HIV infection, Tennessee, 2013

	Total newly diagnosed with HIVa	Linked to careb			Total diagnosed and living with HIVc	Retained in cared			Virally suppressede
Characteristic	No.	No. (%)	P value	aPRf (95% CI)	N	No. (%)	P value	aPRf (95% CI)	No. (%)	P value	aPRf (95% CI)
Total	465	372 (80.0)			7204	4057 (56.3)			4284 (59.5)
Age, yg			.083				<.001			<.001
15-24	156	118 (75.6)		0.87 (0.76-1.00)	301	168 (55.8)		0.95 (0.85-1.06)	143 (47.5)		0.80* (0.71, 0.91)
25-34	168	131 (78.0)		0.92 (0.81-1.04)	1323	673 (50.9)		0.91* (0.85-0.97)	670 (50.6)		0.88* (0.82-0.94)
35-44	73	63 (86.3)		Reference	1647	875 (53.1)		Reference	933 (56.6)		Reference
45-54	56	48 (85.7)		1.03 (0.89-1.18)	2586	1520 (58.8)		1.13* (1.07-1.20)	1648 (63.7)		1.13* (1.07-1.19)
≥55	12	12 (100.0)		1.17* (1.04-1.31)	1347	821 (61.0)		1.21* (1.14-1.29)	890 (66.1)		1.20* (1.13-1.27)
Race/ethnicity			.072				<.001			<.001
Non-Hispanic white	175	147 (84.0)		Reference	3657	2069 (56.6)		Reference	2372 (64.9)		Reference
Non-Hispanic black	251	190 (75.7)		0.92 (0.81-1.03)	3130	1748 (55.8)		0.94* (0.89-0.98)	1678 (53.6)		0.82* (0.79-0.86)
Hispanic	17	15 (88.2)		1.06 (0.88-1.29)	274	137 (50.0)		0.89 (0.79-1.00)	134 (48.9)		0.78* (0.69-0.87)
Otherh	22	20 (90.9)		1.13 (0.93-1.37)	143	103 (72.0)		1.20* (1.08-1.33)	100 (69.9)		1.07 (0.96-1.20)
Enrollment in Ryan White services			.008				<.001			<.001
No	253	191 (75.5)		Reference	4320	2090 (48.4)		Reference	2240 (51.9)		Reference
Yes	212	181 (85.4)		1.16* (1.06-1.27)	2884	1,967 (68.2)		1.48* (1.42-1.54)	2044 (70.9)		1.44* (1.38-1.49)

Abbreviations: aPR, adjusted prevalence ratio; CI, confidence interval; HIV, human immunodeficiency virus; MSM, men who have sex with men. ^aNewly diagnosed with HIV refers to people newly diagnosed with HIV infection in Tennessee during 2013.

^bLinked to care refers to people with ≥1 CD4 count or viral load test reported within 3 months of diagnosis date among people newly diagnosed with HIV infection in 2013 in Tennessee.

^cDiagnosed and living with HIV refers to people diagnosed and living with HIV infection through December 31, 2012, and alive and living in Tennessee as of December 31, 2013.

^dRetained in care refers to people with ≥2 CD4 count and/or viral load tests reported ≥3 months apart in 2013 among people diagnosed with HIV infection on or before December 31, 2012, and alive and residing in Tennessee as of December 31, 2013.

^eVirally suppressed refers to people with ≥1 viral load test in 2013 and whose most recent test result in 2013 indicated ≤200 copies per milliliter among all people diagnosed with HIV infection on or before December 31, 2012, and alive and residing in Tennessee as of December 31, 2013.

^fEstimates derived from adjusted modified Poisson regression model including all factors and region of residence.

^gAge was calculated as of December 31, 2013.

^hOther race includes Native Hawaiian/other Pacific Islander, Asian, American Indian/Alaska Native, and multiple races.

*P < .05.

Discussion

In 2012, TDH staff members met with partners throughout the state (ie, consumers, public health professionals, and HIV health care providers) to establish state goals for 2015 that were bold and achievable.²³ Results from the baseline 2010 HIV continuum-of-care surveillance study were considered in setting the 2015 goals. For linkage to care and viral suppression, Tennessee achieved its 2015 goals in 2013. For retention in care, Tennessee fell short of its 2015 goal in 2013, but substantial progress was made toward achieving it.

Increases in the percentage of people linked to care, retained in care, and virally suppressed may reflect true improvements in HIV continuum-of-care outcomes, enhanced surveillance activities, or both. The state improved its surveillance activities by requiring laboratory reporting of all CD4 and viral load test results to TDH starting in 2013. Tennessee added other surveillance strategies, such as matching state death data with National Death Index data and updating residential status using Accurint software. These strategies helped to provide more accurate information on the number of people living with diagnosed HIV infection in Tennessee. Jurisdictions that do not use these methods might inaccurately construct their HIV continuum-of-care data because they may be missing critical laboratory data or including people in analyses who are no longer living or residing in their jurisdiction.

In addition to enhanced surveillance, Tennessee also improved its HIV continuum-of-care activities. New activities initiated between 2010 and 2013 included the 2012 CAPUS award from CDC.²⁷ The CAPUS seeks to increase the proportion of people with HIV infection from racial/ethnic minority groups who know their status by expanding testing capacity and optimizing linkage to, retention in, and reengagement in care and prevention services. In Tennessee, CAPUS funds supported (1) transition from third-generation (ie, HIV antibody only) to fourth-generation (ie, combined HIV antibody and antigen) HIV testing in 3 state laboratories, (2) implementation of a social networking strategy to engage hard-to-reach black MSM in HIV testing, (3) use of “corrections navigators” (ie, TDH staff members) to facilitate linkage of inmates to HIV care upon release, and (4) use of surveillance data to help disease intervention specialists identify and locate people living with HIV infection who are not engaged in care and to inform HIV health care providers of clients who are in care but are not virally suppressed.

Our findings reinforce concerns about newly infected people, many of whom are young, male, black, and MSM.^2,17 These findings may be the result of true increases in HIV incidence or increases in HIV testing among these groups. People at risk of new HIV infections are typically the same subgroups that perform poorly along the HIV continuum of care. Consistent with other studies, we found that younger, black, and Hispanic people had less favorable continuum-of-care outcomes than older and white people.^19,22 Although MSM fared better in retention and viral suppression than groups with other HIV transmission risk factors, this finding of improved performance among MSM masks concerns for black MSM, whose access to and engagement in care have been limited in other jurisdictions.^3,34,35 Our sensitivity analysis among MSM alone showed that black MSM performed significantly worse than white MSM in retention in care and viral suppression.

Differences in the HIV continuum of care by region persisted after adjusting for demographic factors and enrollment in the Ryan White program. Regions may differ in the quality and proximity of health care services and the prevalence of social and behavioral determinants of health, such as stigma, discrimination, mental illness, and poverty. Those enrolled in the Ryan White program had higher rates of linkage to care, retention in care, and viral suppression than those not enrolled. This finding supports the results of a similar national study, which highlighted the success of the Ryan White program in improving HIV continuum-of-care outcomes.³⁶

The TDH applies data obtained from local jurisdictional analyses to interventions to reduce locally identified health disparities in HIV infection. Line-item data about people identified as not retained in care are provided to CAPUS-funded disease intervention specialists who locate and navigate people for reengagement in HIV care. By repeated continuum analyses, we will assess the effects of changes in surveillance practices and CAPUS or other programs over time.

Limitations

Our surveillance data analysis had several limitations. Although laboratory values were used as a proxy for clinic visits, they may not accurately reflect clinic visits. Dates or names can be misapplied to given laboratory results. Although Tennessee met CDC criteria for designation as a complete CD4 and viral load laboratory-reporting jurisdiction during 2013, noncompliance with reporting of laboratory results may have reduced the completeness of the data set.³²

Conclusion

The TDH progress toward 2015 state goals and the goals of the National HIV/AIDS Strategy has been a collaborative process with stakeholders, including public health officials, patient/activist groups, community-based organizations, and care and social service providers. The TDH 2015 goals were ambitious, but 2 of 3 goals (linkage to care and viral suppression) were achieved 2 years early. More work is needed to achieve the 2015 goal for retention in care. The 2013 data provide a fresh baseline on which TDH can set new goals for those already achieved. An academic partnership among Vanderbilt University, Meharry Medical College, and TDH culminated in the 2015 Tennessee Center for AIDS Research (TN-CFAR) award from the National Institutes of Health. Using TN-CFAR academic expertise in data management and epidemiology/biostatistics, TDH is improving data analyses and dissemination of results and facilitating future work to incorporate data on social and behavioral determinants of health. Studies of interventions aimed at improving HIV continuum-of-care outcomes will be targeted to vulnerable groups identified using local jurisdictional data.