The US Centers for Disease Control and Prevention (CDC) recently reported a decline in new HIV diagnoses in the United States.1 However, the nation's progress in preventing HIV has been uneven and has not clearly included men who have sex with men (MSM), the population most affected by HIV. Success in curbing the HIV epidemic among MSM in other high-income nations has proven similarly elusive.2
The failure of medical and public health efforts to control the US HIV epidemic led to the development of the National HIV/AIDS Strategy (NHAS) in 20103 and, more recently, to the National HIV Care Continuum Initiative.4 These efforts have 3 primary goals: (1) to reduce the number of new HIV infections, (2) to improve access to care and medical outcomes among persons living with HIV/AIDS (PLWHA), and (3) to reduce HIV-related health disparities. Here we report indices related to each of these outcomes in King County, WA. Our findings demonstrate success in achieving the goals of the NHAS, including a decline in new HIV diagnoses among MSM, including black MSM.
Data on HIV diagnoses, AIDS diagnoses, and deaths were taken from King County data collected as part of the National HIV Surveillance System. Medical providers in Washington State are required to report cases of newly diagnosed HIV infection and AIDS. The law has required laboratories to report positive HIV diagnostic test results since 1985 and to report all CD4 lymphocyte and HIV RNA test results since 2006.
Number of HIV Diagnoses and Persons Living With HIV
Calculations of new HIV diagnoses included persons with a King County residence at diagnosis and individuals who reported a prior HIV diagnosis but for whom a date and place of prior diagnosis could not be confirmed; this latter group is included in all CDC-supported surveillance calculations. The number of persons with diagnosed HIV infection living in the area (prevalent cases) included persons who moved into King County after an HIV/AIDS diagnosis, and excluded individuals who died or moved out of the county after diagnosis. We included individuals who relocated out of King County as prevalent cases through the year of their most recent laboratory test result ordered by a King County medical provider. Public Health–Seattle & King County (Public Health) routinely investigates persons for whom no laboratory data are reported and uses these investigations to identify persons who have moved out of the area.5
Populations at Risk
We used US Census data, including information on age, race/ethnicity, and nativity, to define the size and demographic composition of King County's population, extrapolating data for 2012 to 2013 based on changes in 2010 to 2011.6 We stratified HIV and AIDS rates for blacks by nativity, defining individuals with an unknown place of birth as US-born.7 Because Public Health estimates that only 277 Asian and Pacific Islanders were living with HIV in King County in 2014, we do not report data separately for that group; data on HIV in Asian and Pacific Islands and other racial/ethnic groups are presented in our annual surveillance report.8
We estimated the number of MSM in the population as 5.4% of males 15 years or older, the midpoint between the estimated percentage of US men who report ever having sex with another man (6.9%) and the percentage who report sex with another man in the prior 5 years (3.9%).9 We assumed that the percentage of men who were MSM did not change over time or vary by race/ethnicity or age. Numbers of people who inject drugs (PWID) were based on a published estimate for the period 2004–2007,10 extrapolated with a straight line projection for 2008 to 2013. The estimated proportion of residents who were PWID increased from 1.7% to 2.0% over the 10-year study period. HIV and AIDS diagnosis rates were restricted to individuals 15 years or older.
Analyses assessed trends in 3 HIV risk categories defined according to standard US surveillance definitions: MSM (including MSM/PWID), non-MSM PWID, and heterosexuals. Risk groups were mutually exclusive, with heterosexual PWID classified as PWID. We recategorized 405 cases with unknown risk (13% of all cases) to PWID, MSM, and heterosexual risk by separately reassigning male and female cases into risk groups based on the sex-specific proportion of cases in each risk category among cases with complete risk data. We removed prevalent cases from the denominator to calculate rates of new HIV and AIDS.
HIV Testing Data
We used data from publicly funded HIV tests to monitor trends in test positivity among MSM. (Test positivity data are only available for publicly funded tests; negative HIV test results are not reportable in Washington State.) Approximately 37% of newly diagnosed HIV cases in King County between 2012 and 2013 were diagnosed through health department–funded testing.8 Test positivity trends included individuals testing negative more than once in a year or across years, but excluded duplicate positive test results.
We compared changes in HIV and AIDS diagnoses, death rates, and proportions within CD4 and viral load categories using the χ2 for trend (CDC and World Health Organization, Geneva, Switzerland. EpiInfo Version 6.04).
Mortality Rates and Viral Suppression
We calculated mortality rates as deaths among prevalent cases. Mortality rates were age adjusted by standardizing rates to the proportion of King County residents in each 10-year age group in 2010; this adjustment allows one to assess changes in mortality holding the age composition of the population constant. We adjusted for reporting delays based on historical lags between the date of death and the report of a death to HIV/AIDS surveillance, calculating expected lags as the average time between a death and the report of a death to surveillance. Deaths are presented as 3-year rolling averages.
In accordance with CDC standards, we defined viral suppression for each year as a most recent reported viral load less than 200 copies/mL.11 Analyses of viral suppression and CD4 lymphocyte counts include only years during which these tests were reportable (2006–2013).
Protection of Human Subjects
This work was conducted under surveillance authority and did not include data collection beyond that done for routine public health HIV/AIDS surveillance.
Table 1 presents characteristics of the 9539 PLWHA in King County at any point between 2004 and 2013. There were 3779 new HIV diagnoses among county residents during the study period (range of annual diagnoses, 275–436). The total number of PLWHA in the county increased from 5510 in 2004 to 7255 in 2013.
Trends in HIV Diagnoses
The rate of new HIV diagnosis among King County residents declined by 28% between 2004 and 2013 (Fig. 1). This decline was greater among men than among women (32% decline (P < 0.0001) vs. no decline (P = 0.85)). Rates of new HIV diagnoses declined significantly in all risk groups: 26% among MSM (P = 0.0002), 79% among PWID (P = 0.002), and 17% among heterosexuals (P = 0.03). Trends in HIV diagnosis varied substantially by age, dropping by 36% in persons aged 30 to 39 years and 34% in persons aged 40 to 49 years (both P < 0.0001). Younger and older persons did not experience similar declines, and individuals younger than 20 years had a 6-fold increase in HIV diagnosis rates (P = 0.02), although annual diagnosis rates in this group were low (peak rate 3/100,000).
Among MSM, the HIV diagnosis rates in 30- to 39-year-old and 40- to 49-year-old men declined by 51% and 44%, respectively (P < 0.0001 and P = 0.002, respectively). Men who have sex with men younger than 30 years experienced a 16% decrease in HIV diagnosis rates (P = 0.0003).
The number of HIV tests performed in MSM using public health funds increased by 30% between 2007 and 2013, whereas test positivity among these tests declined from 2.3% to 1.6% (P = 0.005; Fig. 2).
Trends in Viral Suppression, CD4 Counts, AIDS Diagnoses, and Mortality
The percentage of persons with a viral load test result reported to HIV surveillance increased from 58% in 2006 to 85% in 2009, and remained stable thereafter (84%–86%; Fig. 3). Among persons with any test results between 2006 and 2013, the percentage with a suppressed viral load increased from 45% to 86% (P < 0.0001). Between 2009 and 2013, the period during which surveillance included data on a stable percentage of PLWHA, viral suppression among persons with reported laboratory results increased from 74% to 86% (P < 0.0001). Among persons for whom laboratory results were available, the percentage of people with a CD4 count at least 350 cells/μL at the time of last report during the calendar year increased from 54% to 79% (Fig. 3; P < 0.0001).
Defining PLWHA without laboratory results reported to surveillance as unsuppressed, between 2009 and 2013, the percentage of King County residents with diagnosed HIV infection who were virally suppressed increased from 63% to 74% (P < 0.0001). The increase in viral suppression observed over the entire period of observation was evident among all groups defined by sex, age, and HIV risk category (data not shown). However, in 2013, levels of viral suppression were lower among persons 39 years or younger than among older persons (66% vs. 77%, P < 0.0001) and among PWID versus other risk groups (66% vs. 75% P < 0.001).
The rate of AIDS diagnosis declined by 42% between 2004 and 2013, from 12 to 7 diagnoses per 100,000 residents. This decline was particularly large between 2009 and 2010, and reflected a drop both in the number of AIDS cases associated with an opportunistic infection (49 and 32, respectively) and the number of cases associated only with a CD4 lymphocyte count less than 200 cells/μL, or less than 14% (166 and 121, respectively). New AIDS diagnoses among MSM, injection drug users, and heterosexuals declined by 37% (P < 0.0001), 82% (P < 0.0001), and 35% (P = 0.003; Fig. 4).
Crude annual mortality rates declined from 16 to 12 deaths per 1000 PLWHA (28% decline, P = 0.004), and age- and reporting lag–adjusted rates declined from 27 to 15 per 1000 PLWHA (42% decline, P < 0.0001; Fig. 5). Mortality declined by 32% (P < 0.0001), 20% (P = 0.81), and 21% (P = 0.78) among MSM, PWID, and heterosexuals, respectively.
Racial and Ethnic Disparities
New HIV diagnosis rates decreased significantly among blacks (39% decrease, P = 0.0001), whites (27% decrease, P = 0.01), and Hispanics (25% decrease, P = 0.01; Fig. 1D). The decline in new diagnoses observed among blacks occurred in US-born (57% decrease, P = 0.001) and foreign-born blacks (23% decrease, P = 0.02). White, Hispanic, and black MSM experienced declines of 26% (P = 0.02), 22% (P = 0.13), and 44% (P = 0.07), respectively. The estimated new HIV diagnosis rate per 1000 persons declined in white MSM from 6.2 to 4.6 (P = 0.02); in Hispanic MSM, from 10.7 to 8.3 (P = 0.13); and in black MSM, from 17.9 to 10.0 (P = 0.07). The relative risk (RR) of an HIV diagnosis for black compared with white MSM declined from 2.9 (95% confidence interval [CI], 1.9–4.3) in 2004 to 2.2 (95% CI, 1.4–3.4) in 2013. The RR for Hispanic compared with white MSM was 1.7 in 2004 (95% CI, 1.1–2.6) and did not decline over the decade (RR in 2013, 1.8 [95% CI, 1.2–2.8]).
The proportion of persons with viral load test results reported to surveillance who were suppressed increased between 2006 and 2013 for Hispanics (54% to 86%), blacks (45% to 81%), and whites (44% to 88% [increases of 59%, 79%, and 98%, respectively]; all P < 0.0001). Assuming that persons without reported viral load results were unsuppressed, viral suppression during this period increased by 116% for Hispanics (32% to 70%), 128% for blacks (30% to 68%), and 217% for whites (25% to 79%; all P < 0.0001). The proportion of PLWHA whose most recent CD4 count was at least 350 increased by 43% for blacks (51% to 73%), 42% for whites (57% to 81%), and 47% for Hispanics (51% to 75%). AIDS diagnosis rates decreased for Hispanics, whites, blacks overall, and foreign-born blacks (P = 0.0003, P < 0.0001, P < 0.0001, and P = 0.0002, respectively; Fig. 3). Among MSM, AIDS diagnoses decreased in each race/ethnicity group (46% Hispanic MSM, 24% black MSM, 48% white MSM), although this decline was smaller for black MSM than for white or Hispanic MSM (data not shown). Deaths decreased by 33% among whites (Fig. 5), 73% among blacks (US-born and foreign-born blacks combined), and 33% among Hispanics.
Our findings suggest that the goals of the NHAS are achievable and are being realized in at least one US county. Between 2004 and 2013, the rate of new HIV diagnoses in King County, WA, declined by 28%; the rate of AIDS declined by 42%; and age-adjusted mortality among individuals with HIV/AIDS declined by 42%. These declines affected virtually all major subgroups defined by HIV risk factor and race/ethnicity, including black MSM, and led to diminished disparities in newly diagnosed HIV, a trend that contrasts with the increase in racial disparities observed among MSM in the United States as a whole.12
Our findings demonstrate both similarities and differences from national US data. As in the country as a whole, we observed a decline in the rate of new HIV diagnosis.1 However, although we found that the rate of diagnosis declined by 26% among MSM between 2004 and 2013, national data indicate that the number of new diagnoses among MSM between 2001 and 2011 was stable. The disparity between our findings and those reported nationally reflects, in part, the fact that we reported rates, whereas CDC reports national MSM diagnosis data as case counts. The US population is growing approximately 0.7% per year,13 and accordingly, the size of the MSM population is likely increasing. The population of King County increased almost 15% over our study period,14,15 necessitating the use of rates to accurately evaluate trends. Because the US population has been growing while case counts in MSM have remained essentially stable, it seems likely that the HIV diagnosis rate among US MSM is declining, although the magnitude of that decline seems to be larger in King County than in the United States as a whole.
The King County experience is not entirely anomalous. To our knowledge, the sole published scientific report of a decline in HIV diagnoses among MSM is from San Francisco. Das et al.16 reported a dramatic 46% reduction in new HIV diagnoses between 2004 and 2008. That decline occurred concurrent with the institution of named-based reporting in California, and some of the observed trend may have reflected inaccuracies associated with an evolving surveillance system.17 However, more recent surveillance reports from San Francisco support the conclusion that the rate of new HIV diagnosis among MSM in the city is declining, though at a more gradual pace than that initially reported.18 Surveillance from some other states and cities have likewise reported declines in new HIV diagnoses among MSM,19,20 although these trends are by no means universal.21,22
We also observed encouraging clinical trends. Among PLWHA with viral load results reported to surveillance, viral suppression increased by 89% between 2006 and 2013, and in 2013 an estimated 74% of all persons with diagnosed HIV infection were virologically suppressed. AIDS diagnoses and mortality among PLWHA declined over the same period. Although the simultaneous increase in viral suppression and decrease in new HIV diagnoses are consistent with the idea that improved population-level viral suppression can diminish HIV transmission—an association observed in 2 prior studies16,23—that ecological association may not be causal.24 In contrast, improved viral suppression almost certainly played a causal role in the 42% decline in age-adjusted mortality we observed. This decline is smaller than the 56% decline in age-adjusted mortality in PLWHA observed in the United States between 2003 and 2010,25 but the unadjusted mortality rate among PLWHA in King County (14 per 1000 in 2010) is lower than the national mortality rate (22 per 1000).
The improvements we observed affected diverse subgroups of PLWHA. Of particular note, new HIV diagnoses declined substantially in black MSM, from an estimated 18 to 10 new cases per 1000 MSM. The mortality rate among blacks with HIV in King county is approximately half that observed nationally26 and is comparable to that observed in whites. In interpreting these trends, it is important to note that only 6.7% of King County residents are black, compared with 13.2% of all residents of the United States. This demographic difference may affect the magnitude of racial disparities in the area as well as the difficulty in surmounting them. Although we are encouraged that King County is making progress toward the goal of diminishing disparities, significant racial disparities persist, as evidenced by a higher rate of new diagnoses in black and Hispanic MSM relative to white MSM and differences in levels of viral suppression by race/ethnicity. Additional efforts to address this issue are needed.
The reasons for the positive trends we observed are uncertain and likely reflect a combination of factors. Improvements in clinical outcomes almost certainly reflect improvements in treatment and earlier initiation of antiretroviral therapy. However, these changes do not explain differences between King County and much of the country. We believe that several factors contribute to King County's positive trends. First, the county has a well-developed public health and care infrastructure that includes community-based organizations, medical providers, academia, and state and local health departments.27,28 Work undertaken by these groups reflects a long-term investment of state and local public funding that supplements federal resources; other investigators have identified the availability of local funds for HIV prevention as important in curbing the HIV epidemic.29 Second, King County is economically and demographically dissimilar from much of the United States. The HIV epidemic in the United States disproportionately affects African Americans and is associated with low income.30 African Americans comprise a relatively small part of King County's population, and median household income in King County is 36% higher than that of the nation.15 Finally, insofar as stigma associated with male homosexuality promotes HIV transmission and inadequate engagement with HIV care, the area may be more accepting of homosexuality than some other parts of the nation,31 creating an environment that fosters HIV prevention. Of note, HIV preexposure prophylaxis use in King County, which is now increasing very rapidly, was low throughout the period included in this analysis.32 As a result, we believe that the information we report should be regarded as pre–preexposure prophylaxis trends.
Although the overall trends we observed are encouraging, King County has experienced an increase in HIV diagnoses among persons younger than 20 years. The absolute rate of infection in that population remains very low, and we cannot say whether the observed increase reflects a true increase in infections versus an increase in testing among adolescent MSM. The annual number of new HIV diagnoses among young MSM nationally increased between 2010 and 2014,33 and this local trend merits further investigation.
Our study has limitations. First, surveillance data are affected by incomplete or inaccurate reporting. Second, we used HIV diagnoses as a surrogate for HIV incidence. The rate of new HIV diagnosis reflects both true incidence and patterns of testing. We think that it is very unlikely that that downward trend in new diagnoses we observed reflects decreased testing in the populations at risk. The number of publicly funded HIV tests performed in MSM increased by 30% during the period of observation—a trend that would lead to a transient increase in diagnoses rates—whereas the median intertest interval (time between HIV tests) and the percentage of newly HIV diagnosed MSM who reported never previously HIV testing were both stable.8,27 The decline in AIDS cases concurrent with the decline in new HIV diagnoses likewise argues against decreasing testing as a cause for the observed decline in new HIV diagnoses since one would expect that a decline in testing would lead to an increase in late HIV diagnoses and AIDS. Third, our estimates of the percentage of PLWHA who were virally suppressed could have been affected by imprecise estimates of the size and composition of the population of PLWHA, which changes due to migration. We regularly update our surveillance data through investigations of persons who seem to be out of care.5 Insofar as these efforts improved over the study period, they may have exaggerated the trends we observed. Fourth, our estimates of the size of the MSM and PWID populations are imprecise. However, assuming that the percentage of men who are MSM did not change over the study period, our assumptions would not have affected our MSM trend estimates. Finally, our estimates related to viral suppression are limited to persons with diagnosed HIV infection. We estimate that 94% of all MSM in King County are diagnosed, but do not have data on how that percentage might have changed over the period of our analysis or on other populations affected by HIV.34
In summary, we found that new diagnoses of HIV and AIDS and mortality in PLWHA are all declining in King County, whereas levels of viral suppression and CD4 counts are increasing. These changes are broad based, affecting all major racial and ethnic groups and individuals with diverse HIV risk factors. Although we recognize that many areas of the United States face substantially greater challenges related to HIV compared with King County, WA, we believe that our findings should be a source of optimism and demonstrate that the goals of the NHAS strategy are achievable.
1. Johnson AS, Hall HI, Hu 1, et al. Trends in diagnoses of HIV infection in the United States, 2002–2011. JAMA 2014; 312: 432–434.
2. Sullivan PS, Jones JS, Baral SD. The global north: HIV epidemiology in high-income countries. Curr Opin HIV AIDS 2014; 9: 199–205.
3. White House Office on National AIDS Policy. National HIV/AIDS Strategy for the United States. Washington, DC: White House Office on National AIDS Policy; 2010.
4. White House. Executive Order—HIV Care Continuum Initiative 2013.
5. Buskin SE, Kent JB, Dombrowski JC, et al. Migration distorts surveillance estimates of engagement in care: Results of public health investigations of persons who appear to be out of HIV care. Sex Transm Dis 2014; 41: 35–40.
8. Public Health–Seattle & King County and Washington State Department of Health. 2014 HIV/AIDS Epidemiology Report 2014.
9. Purcell DW, Johnson CH, Lansky A, et al. Estimating the population size of men who have sex with men in the United States to obtain HIV and syphilis rates. Open AIDS J 2012; 6: 98–107.
10. Tempalski B, Pouget ER, Cleland CM, et al. Trends in the population prevalence of people who inject drugs in US metropolitan areas 1992–2007. PLoS One 2013; 8: e64789.
11. Bradley H, Hall HI, Wolitski RJ, et al. Vital signs: HIV diagnosis, care, and treatment among persons living with HIV—United States, 2011. MMWR Morb Mortal Wkly Rep 2014; 63: 1113–1117.
12. Center for Disease Control and Prevention. HIV Surveillance in Men Who Have Sex with Men Atlanta. Atlanta, GA: CDC; 2012.
13. World Bank. Population Growth (Annual %). 2015.
14. Office of Financial Management. Washington State Growth Management Population Projections for Counties: 2010 to 2040. Olympia, WA: Office of Financial Management; 2013.
15. U.S. Census. American Factfinder. 2013.
16. Das M, Chu PL, Santos GM, et al. Decreases in community viral load are accompanied by reductions in new HIV infections in San Francisco. PLoS One 2010; 5: e11068.
17. San Francisco Department of Health. HIV/AIDS Epidemiology Annual Report 2008. San Francisco: San Francisco Department of Public Health; 2009.
18. San Francisco Department of Health. HIV Epidemiology Annual Report 2013. San Francisco: San Francisco Department of Public Health; 2014.
19. Illinois Department of Health. Incidence of HIV Disease for Year of Diagnosis. 2014.
22. Texas State Department of Health Services. Texas HIV Surveillance Report: 2013 Annual Report. 2014.
23. Montaner JS, Lima VD, Barrios R, et al. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: A population-based study. Lancet 2010; 376: 532–539.
24. Smith MK, Powers KA, Muessig KE, et al. HIV treatment as prevention: The utility and limitations of ecological observation. PLoS Med 2012; 9: e1001260.
25. Center for Disease Control and Prevention. HIV Mortality. Atlanta, GA: CDC; 2012.
26. Siddiqi AE, Hu X, Hall HI, et al. Mortality among blacks or African Americans with HIV infection—United States, 2008–2012. MMWR Morb Mortal Wkly Rep 2015; 64: 81–86.
27. Katz DA, Dombrowski JC, Swanson F, et al. HIV intertest interval among MSM in King County, Washington. Sex Transm Infect 2013; 89: 32–37.
28. Dombrowski JC, Buskin SE, Bennett A, et al. Use of multiple data sources and individual case investigation to refine surveillance-based estimates of the HIV care continuum. J Acquir Immune Defic Syndr 2014; 67: 323–330.
29. Goga AE, Dinh TH, Jackson DJ, et al. First population-level effectiveness evaluation of a national programme to prevent HIV transmission from mother to child, South Africa. J Epidemiol Community Health 2015; 69: 240–248.
30. Centers for Disease Control and Prevention. HIV infection among heterosexuals at increased risk—United States, 2010. MMWR Morb Mortal Wkly Rep 2013; 62: 183–188.
31. Glick SN, Golden MR. Persistence of racial differences in attitudes toward homosexuality in the United States. J Acquir Immune Defic Syndr 2010; 55: 516–523.
32. Hood JE, Buskin SE, Dombrowski JC, et al. Dramatic increase in preexposure prophylaxis use among MSM in King County, Washington. AIDS 2016; 30: 515–519.
33. Center for Disease Control and Prevention. HIV Surveillance in Men Who Have Sex with Men Atlanta. Atlanta, GA: CDC; 2015.
34. Fellows IE, Morris M, Birnbaum JK, et al. A new method for estimating the number of undiagnosed HIV infected based on HIV testing history, with an application to men who have sex with men in Seattle/King County, WA. PLoS One 2015; 10: e0129551.