Vital Signs: Deaths Among Persons with Diagnosed HIV Infection, United States, 2010–2018

Background Life expectancy for persons with human immunodeficiency virus (HIV) infection who receive recommended treatment can approach that of the general population, yet HIV remains among the 10 leading causes of death among certain populations. Using surveillance data, CDC assessed progress toward reducing deaths among persons with diagnosed HIV (PWDH). Methods CDC analyzed National HIV Surveillance System data for persons aged ≥13 years to determine age-adjusted death rates per 1,000 PWDH during 2010–2018. Using the International Classification of Diseases, Tenth Revision, deaths with a nonmissing underlying cause were classified as HIV-related or non–HIV-related. Temporal changes in total deaths during 2010−2018 and deaths by cause during 2010–2017 (2018 excluded because of delays in reporting), by demographic characteristics, transmission category, and U.S. Census region of residence at time of death were calculated. Results During 2010–2018, rates of death decreased by 36.6% overall (from 19.4 to 12.3 per 1,000 PWDH). During 2010–2017, HIV-related death rates decreased 48.4% (from 9.1 to 4.7), whereas non–HIV-related death rates decreased 8.6% (from 9.3 to 8.5). Rates of HIV-related deaths during 2017 were highest by race/ethnicity among persons of multiple races (7.0) and Black/African American persons (5.6), followed by White persons (3.9) and Hispanic/Latino persons (3.9). The HIV-related death rate was highest in the South (6.0) and lowest in the Northeast (3.2). Conclusion Early diagnosis, prompt treatment, and maintaining access to high-quality care and treatment have been successful in reducing HIV-related deaths and remain necessary for continuing reductions in HIV-related deaths.


Introduction
Persons with human immunodeficiency virus (HIV) infection require lifelong treatment to reduce HIV-related morbidity and mortality; advances in HIV treatment have resulted in a life expectancy that approaches that of the general population (1,2). Deaths attributable to HIV infection are preventable, yet during 2017, HIV was still among the 10 leading causes of death among certain population groups (3).
The National HIV Surveillance System (NHSS) is the primary source of population-based information about HIV in the United States (4). A previous analysis demonstrated that, during 1990-2011, deaths among persons with stage 3 HIV infection (acquired immunodeficiency syndrome [AIDS]) decreased, with larger decreases in HIV-attributable deaths (−89%) than in non-HIV-attributable deaths (−57%) (5). On the basis of increasing evidence of the benefits of antiretroviral therapy both for persons with HIV and for preventing secondary transmission, treatment guidelines were updated in 2012 to recommend antiretroviral therapy for all persons with HIV (6). A national target for reducing the death rate among persons with diagnosed HIV (PWDH) by ≥33% during 2010-2020 was established to encourage progress toward improving health outcomes among PWDH (7). Using NHSS data, CDC assessed such progress, with an emphasis on HIV-related deaths, at the national and state levels.  Rates of HIV-related deaths during 2017 were higher among females (5.4 per 1,000 PWDH) than males (4.5) and transgender females (females assigned male sex at birth) (4.3), and highest among persons of multiple races (7.0) and Black persons (5.6), followed by White persons (3.9) and Hispanic/ Latino persons (3.9) ( Table 1). The rates of HIV-related deaths increased with age, from 1.6 among PWDH aged 13-24 years to 8.4 among persons aged ≥55 years. However, the proportion of deaths that were HIV-related decreased with increasing age from 48.6% among PWDH aged 13-24 years with a known cause of death to 30.0% among PWDH aged ≥55 years with a known cause of death because the rate of non-HIV-related death increased with age more than the rate of HIV-related death. Among males, the rate of HIV-related death was lower among those whose infection was attributed to male-to-male sexual contact (3.9) than among those whose infection was attributed to other transmission categories; among females, the rate was lower among those with infection attributed to heterosexual contact (4.6) than among those in other transmission categories. The rate of HIV-related deaths was highest in the South (6.0) and lowest in the Northeast (3.2).

Discussion
By 2018, the rate of death among PWDH in the United States had decreased by 36.6% from what it was in 2010, surpassing the 2020 national target of ≥33% (7). This decrease, which was primarily attributable to reductions in HIV-related deaths, likely reflects the increase during 2010-2018 in the proportion of persons who knew their serostatus from 82.2% to 86.2% and the implementation of updated treatment guidelines resulting in increased viral suppression among PWDH from 46.0% to 64.7% (6,8). Absolute and relative differences in HIV-related deaths among Black persons and Hispanic/ Latino persons, compared with those among White persons, also decreased during 2010-2017. This reduction likely reflects a greater relative improvement during 2012-2017 in the time from diagnosis to viral suppression among Black persons, Rates age-adjusted using the U.S. 2000 standard population. Rates presented by age at time of death are not age-adjusted. Rates and percentage change calculated on the basis of <12 deaths are considered unstable and should be interpreted with caution. § "Transgender male-to-female" includes persons who were assigned "male" sex at birth but have ever identified as "female. " "Transgender female-to-male" includes persons who were assigned "female" sex at birth but have ever identified as "male. " ¶ Additional gender identity examples include "bigender, " "gender queer, " and "two-spirit. " ** Data by cause of death should be interpreted with caution because <85% of reported deaths were reported with a known underlying cause of death. † † Hispanic/Latino persons can be of any race. § § Data have been statistically adjusted to account for missing transmission category; therefore, values might not sum to column subtotals and total. ¶ ¶ Data presented are based on sex at birth and include transgender persons. *** Heterosexual contact with a person known to have, or to be at high risk for, HIV infection. † † † Includes hemophilia, blood transfusion, perinatal, and risk factor not reported or not identified.

TABLE 2. Total deaths and human immunodeficiency virus (HIV)-related deaths among persons aged ≥13 years with diagnosed HIV infection, by area of residence at time of death, and selected race/ethnicity categories -United States and Puerto Rico,* 2017
Area of residence

TABLE 2. (Continued) Total deaths and human immunodeficiency virus (HIV)-related deaths among persons aged ≥13 years with diagnosed HIV infection, by area of residence at time of death, and selected race/ethnicity categories -United States and Puerto Rico,* 2017
Abbreviation: PWDH = persons with diagnosed HIV infection. * Other U.S. dependent areas are excluded because they do not report underlying cause of death information. † HIV-related deaths include deaths with an underlying cause with an International Classification of Diseases, Tenth Revision code of B20-B24, O98.7, or R75. Non-HIV-related deaths include all other deaths with a known underlying cause. Deaths with an unknown underlying cause are excluded. § PWDH includes persons living with HIV infection at the end of the calendar year plus the number of diagnoses of HIV infection during the current calendar year. Rates age-adjusted using the U.S. 2000 standard population. Rates calculated based on consideration that analyses of data with <12 deaths are considered unstable and should be interpreted with caution. ¶ Proportion of deaths with a known underlying cause of death is <85%. compared with White persons (9), and reduced disparities during 2010-2016 in viral suppression among Black persons and Hispanic/Latino persons, compared with White persons (10). These findings highlight how successes in identifying HIV infections, initiating treatment, and achieving viral suppression among PWDH improve health outcomes.
Despite success in reducing rates of HIV-related deaths among PWDH, differences still exist by gender, race/ethnicity, age, transmission category, and region. Variation in timely diagnosis and treatment initiation, along with ongoing treatment, likely contributes to differences in HIV-related deaths. During 2015, delays in HIV diagnosis were longer among non-White racial/ethnic groups and males with HIV infection attributed to heterosexual contact (11). Timely initiation of treatment, as measured by the proportion of persons with suppressed viral loads ≤6 months after diagnosis, and receipt of ongoing, recommended treatment, as measured by the proportion of PWDH with a suppressed viral load, varied during 2017 by gender, age, race/ethnicity, transmission category, and region (8,12); populations with higher rates of HIV-related deaths were less likely to have evidence of timely initiation of treatment and ongoing treatment as demonstrated through lower proportions of viral suppression in the population.
Prevalence of HIV infection and the number of HIV-related deaths were greatest by race/ethnicity among Black persons and by U.S. region in the South (4). Rates of HIV-related deaths were also high among these two populations. Higher levels of poverty, unemployment, and persons uninsured, challenges associated with accessing care, and HIV-related stigma likely affect timely diagnosis and access to treatment and contribute to higher rates of HIV-related deaths (13,14). Expanded efforts to address these and other structural barriers are critical to improving health outcomes, including reducing differences in HIV-related death rates, especially among Black persons and persons in the South.
Although rates of HIV-related deaths were lower among younger PWDH, the proportion of HIV-related deaths among younger PWDH (ages 13-44 years) was higher than that among older PWDH; this is concerning because HIV-related deaths are preventable. Higher proportions of undiagnosed HIV infections and lower levels of viral suppression are more common among younger persons (8,15). Additional efforts are needed to ensure younger persons are aware of their infection and able to access and adhere to recommended, ongoing HIV treatment to improve health outcomes.
CDC supports numerous activities for identifying HIV infections: initiating treatment as quickly as possible and ensuring ongoing treatment; addressing social barriers to HIV prevention and treatment efforts; and expanding opportunities for persons to test for HIV infection and receive the results on their own (i.e., self-testing), which allows persons who might not otherwise take a test to learn their HIV status (16). CDC's Integrated HIV Surveillance and Prevention Programs for Health Departments, initiated in 2018, includes critical activities to enable state and local health departments to improve identification of HIV infections and increase viral suppression among PWDH (17). CDC's national campaign, Let's Stop HIV Together, supports efforts to end HIV stigma and promote HIV testing, prevention, and treatment (18). Ending the HIV Epidemic: A Plan for America is an initiative for reducing HIV infections in the United States by ≥90% by 2030; it focuses on strategies regarding diagnosis, treatment, prevention, and response to HIV infection in communities most affected by HIV (19). In addition to decreasing the risk for ongoing HIV transmission, prompt diagnosis and improving timely and continuing access to HIV treatment should also improve health outcomes for PWDH and prevent HIV-related deaths.
The findings in this report are subject to at least two limitations. First, cause-of-death information on death certificates is typically completed by funeral directors, attending physicians, medical examiners, or coroners (3). HIV-related deaths might be underreported because of lack of knowledge about the correct documentation needed or reluctance to include HIV on the death certificate because of possible stigma (5). An assessment of Florida's HIV surveillance data for 2000-2011 indicated that HIV-related deaths were underestimated in the surveillance system by approximately 9% (20). Second, the proportion of deaths with a known cause was <100%. Overall, the proportion of deaths with a known cause was high for the United States (94.6% in 2010 and 96.7% in 2017); however, the proportion of deaths with a known cause was lower for certain demographic groups (e.g., Asian persons) and for certain jurisdictions (e.g., Hawaii during 2017). Deaths among persons with HIV have decreased, and by 2018 had surpassed the 2020 national target, primarily because of a reduction in HIV-related deaths. Deaths caused by HIV infection have likely decreased because of improvements in diagnosing infections and in treatment and medical care. However, differences in HIV-related death rates still exist for multiple populations. Diagnosing HIV infection early, treating it promptly, and maintaining access to high-quality care and treatment over a lifetime can improve life expectancy and reduce differences in rates of deaths across all populations.