Secondary Logo

Journal Logo

Supplement Article

A Decline in HIV Testing Among Persons Who Inject Drugs in the Seattle Area, 2004–2015

Burt, Richard D. PhD*; Tinsley, Joe BA*; Glick, Sara Nelson PhD, MPH*,†

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: July 1, 2017 - Volume 75 - Issue - p S346-S351
doi: 10.1097/QAI.0000000000001409



HIV testing is one of the foundations of the public health response to HIV. Testing opens multiple opportunities to reduce the burden of HIV disease. There is convincing data that HIV-positive persons reduce their sexual risk after diagnosis and have lower sexual risk than HIV-positive persons unaware of their status,1,2 although these data do not derive specifically from populations of persons who inject drugs (PWID). Once persons become aware of being HIV-positive they can initiate antiretroviral therapy. Initiation of HIV antiretroviral treatment is associated with lowered HIV-associated morbidity and mortality.3 Antiretroviral treatment reduces the level of virus in the blood,4 and consequently the likelihood of (at least) sexual HIV transmission.5,6

In 2014 PWID, including PWID who also practice male-to-male sex (MSM), accounted for an estimated 3852 new cases of HIV in the United States, comprising 8.7% of total cases.7 Although HIV prevalence among most PWID in the Seattle area remains relatively low (2% among women and men reporting no MSM in 2015), PWID reporting MSM and reporting amphetamines as the drug they most frequently inject have a markedly higher prevalence (38%).8 The Centers for Disease Control and Prevention (CDC) has recommended that PWID be tested for HIV at least annually.9 Given the potential for HIV transmission through sharing of injection equipment, HIV testing remains an essential component of local HIV prevention efforts for PWID. Ascertaining the frequency of HIV testing among PWID can help to monitor the status of this preventive measure and identify further opportunities for reducing HIV transmission, morbidity, and mortality.

Since 2002, CDC's National HIV Behavioral Surveillance (NHBS) system has been conducting ongoing surveys of MSM, PWID, and persons thought to be at increased risk for heterosexually transmitted HIV in successive yearly surveys in approximately 20 US cities (the number varied by year).10 In addition, Public Health—Seattle & King County has been conducting surveys of Needle Exchange users since 2004, at essentially biyearly intervals. In this report, we use data from 4 NHBS and 6 Needle Exchange surveys of PWID conducted 2004–2015 to investigate time trends in HIV-testing frequency.



Anonymous NHBS surveys were conducted in 2005, 2009, 2012, and 2015 using respondent-driven sampling (RDS), a coupon-based peer recruitment refinement of snowball sampling in which initially recruited seeds receive coupons to recruit their peers, who in turn are issued coupons to recruit succeeding waves of participants.11 Eligibility for the surveys was restricted to persons who were 18 years of age or older, residents of King or Snohomish Counties, could communicate in English, were alert enough to complete the survey, and had injected drugs in the previous 12 months.

The details of recruitment in Seattle-area NHBS surveys in 2005, 2009, and 2012 have been published.12,13 In the 2015 NHBS survey, recruitment began on June 6, 2015 and continued through November 17, 2015. Participants were paid $25 to complete the survey, $25 for an HIV test and $10 for each coupon brought in by a new participant. Because previous NHBS surveys had disproportionately recruited older participants, a system of differential coupon distribution by age was instituted. Participants 18–29 years of age were issued 5 coupons; those older were issued 2 coupons. All participants who came into the study offices consented to the interview, although we have no information on how many persons given coupons did not to participate in the survey. Study procedures for the NHBS surveys were approved by the Washington State Institutional Review Board.

Public Health–Seattle & King County conducted anonymous surveys at its local Needle Exchange sites including information on HIV-testing frequency in 2004, 2006, 2009, 2011, 2013, and 2015. These surveys attempted to interview every person exchanging needles at specific exchanges over defined time intervals, usually 2 weeks. In the surveys conducted 2009–2015 data were collected on the proportion of potential participants approached who completed a survey. These were 80%–89% for the 2009, 2010, and 2013 surveys and 64% in the 2015 survey. Details of the conduct and results of these Needle Exchange surveys have been published.14 The University of Washington Human Subjects Review office determined that the Needle Exchange surveys did not require institutional review.

Variable Definitions

Analysis was restricted to persons not reporting a previous HIV-positive test result. In NHBS surveys, the time since participants' last HIV test was ascertained by the question “When did you have your most recent HIV test? Please tell me the month and year.” In Needle Exchange surveys the question was “When was the last time you got tested for HIV?”

In NHBS, data were also obtained on participants' knowledge of the HIV status of their last injection partner by the question “The last time you shared with this person, did you know their HIV status?” The definition of the last injection partner, however, varied across surveys. In 2005 and 2015, sharing partners were defined by sharing injection equipment, including needles, water, cottons, or backloading (dividing up drugs through use of a common syringe). In 2009, participants were directed to consider simply “the last time you injected with someone.” In 2012, the definition of sharing partners included persons with whom drugs were shared and those with whom injection equipment was shared. Data were elicited on participants' knowledge of the HIV status of their last sex partner by questions of the form “The last time you had sex with this partner did you know his HIV status?”

We identified a collection of sociodemographic and injection-related variables that might be plausibly be considered to identify differences in the study populations that affect the likelihood of HIV testing: age, race, sex, area of residence, education, current homelessness, drug most frequently injected, daily injection frequency, and a variable combining a report of MSM in the previous 12 months with amphetamines as the drug most frequently injected. The latter has been found to be a strong determinant of HIV-risk behavior.15 Data on education, current homelessness, and injection frequency was not available from the Needle Exchange surveys. This collection of variables was used to assess differences across years in the study populations and to control for potential confounding of HIV-testing trends in multivariate analyses.

Data Analyses

Separate analyses were performed within the NHBS and Needle Exchange series of surveys. This allowed us to evaluate 2 independent sources of data and to take into account the differences in the variables available in the 2 survey series. Because both surveys were anonymous, it was not possible to determine participant overlap. Univariate differences in sociodemographic and injection-related variables across years within the NHBS and Needle Exchange surveys and between the 2 series of surveys were evaluated by Pearson χ2 statistics. Univariate tests for trend in HIV testing across calendar years were evaluated by Mantel–Haenszel χ2 tests.

Multivariate analyses using general estimating equation analyses were used to investigate possible confounding of time trends in HIV testing by sociodemographic and drug-related differences. These analyses used a robust variance estimate and a Poisson model with a log link function; grouping was on the basis of unique individual participant identification numbers, yielding an analysis essentially lacking a grouping structure. Models were evaluated by backward regression until all remaining variables were significantly associated (P < 0.05) with HIV testing, and then tested by adding back each dropped variable in turn. The latter procedure resulted in no change in the final model. Statistical tests for simple linear trend were conducted by incorporating a continuous variable for year in the multivariate models. Analyses were conducted in SPSS.

Our analyses are not adjusted by RDS-specific methods for several reasons. We wanted to ensure comparability of our data with findings reported by CDC for HIV testing among PWID recruited by RDS in the approximately 20 US cities participating in NHBS. These data were not adjusted for RDS methods. Furthermore, we wanted to present multivariate analyses assessing confounding of time trends in HIV testing by sociodemographic and injection-related variables. There are no generally accepted multivariate methods that incorporate RDS adjustments for both point estimates and variance. Finally, there remain unresolved questions concerning how well RDS-adjusted data describe the underlying target population.16


HIV Prevalence, Sociodemographics, and Drug-Related Variables Across Surveys

HIV prevalence among PWID in the NHBS surveys was 8% in 2009, 8% in 2012, and 5% in 2015. Eighty-four percent of HIV-infected participants were aware of their status. Participants who self-reported being HIV-infected were excluded from further analyses.

Table 1 documents differences across years in the NHBS and Needle Exchange surveys among persons who did not self-report being HIV-infected. Within the NHBS series, we observed a statistically significant difference (all P < 0.001) across years for all variables except education and the combined MSM and amphetamine injection status variable. We have previously reported more detailed comparisons, finding substantial differences between the 2005 NHBS IDU study population with previous Seattle-area study populations of PWID,12 and finding notable differences between the 2005 and 2009 NHBS study populations.13

Sociodemographic and Drug-Related Characteristics of Seattle-Area Participants in NHBS and Needle Exchange Surveys of PWID 2004–2015: By Survey Year

Similarly, we observed statistically significant differences (all P ≤ 0.007) across the years in the Needle Exchange study populations for all variables we examined. Comparing participants aggregated across all years of the NHBS surveys with aggregated Needle Exchange participants, significant differences were seen in all variables the 2 studies had in common (P < 0.001 for each variable) (Supplemental Digital Content, Table S1

HIV Testing in the Previous 12 Months

NHBS Surveys

The proportion of participants reporting an HIV test in the previous 12 months declined from 64% in 2005% to 51% in 2009, 47% in 2012%, and 47% in 2015 [p(trend) < 0.001] (Table 2). In multivariate analyses, age was found to be significantly associated with HIV testing, with younger participants more likely to report testing. Participants who reported that the drug they most frequently injected was heroin were more likely to test than those reporting other drugs. Those reporting both MSM status and amphetamines as the drug they most frequently injected were more likely to test than other participants. Race, sex, education, area of residence, homelessness, and injection frequency were not significantly associated with HIV testing. In multivariate analysis, the prevalence ratio for HIV testing declined from 1.00 in the baseline 2005 survey to 0.72 in 2009, 0.67 in 2012, and remained at 0.67 in 2015 [p(trend) < 0.001] (Table 2).

HIV Testing in the Previous 12 Months Across NHBS Surveys of Seattle-Area PWID 2005–2015: Multivariate Results*

Needle Exchange Surveys

The proportion of Needle Exchange participants reporting an HIV test in the previous 12 months declined from 72% in 2004% to 69% in 2006, then was between 57% and 61% in the 2009, 2011, 2013, and 2015 surveys [p(trend) < 0.001] (Table 3). In multivariate analyses, significant associations with HIV testing were found for age (again with younger participants most likely to test) and race (with whites less likely to test than blacks or Hispanics). The prevalence ratio for HIV testing declined from 1.00 in 2004 to 0.95 in 2005, 0.80 in 2009, then stayed between 0.80 and 0.84 in 2011, 2013, and 2015 [p(trend) < 0.001].

HIV Testing in the Previous 12 Months Across Needle Exchange Surveys of Seattle-Area PWID 2004–2015: Multivariate Results*

Knowledge of the HIV Status of Sex and Injection Partners

In the NHBS surveys, the proportion of participants reporting that they did not know the HIV status of their last injection partner was 38% in 2005, 37% in 2009, 40% in 2012%, and 45% in 2014 [p(trend) = 0.036] (Table 4). The proportion that reported not knowing the HIV status of their last sex partner increased from 27% in 2005% to 34% in 2009, 33% in 2012%, and 38% in 2015 [p(trend) = 0.011].

Knowledge of the HIV Status of Participants' Last Injection Partner and Last Sex Partner Across NHBS Surveys of Seattle-Area PWID 2005–2015


We found a decline over the time period 2004–2015 in the proportion of Seattle-area PWID reporting a test for HIV in the previous 12 months. The decline was observed in 2 independent series of surveys, one recruited by RDS, the other among Needle Exchange clients. Testing levels were somewhat higher in the Needle Exchange surveys, perhaps because Needle Exchange users have demonstrated by their presence at the exchange a readiness to undertake efforts to reduce risk of HIV transmission and because testing was offered at the Needle Exchanges. These declines were statistically significant and persisted after control for differences in sociodemographic and drug-related characteristics of the study populations.

Our findings seem consistent with national NHBS data. Across the approximately 20 US cities participating in NHBS, the overall proportion of PWID testing for HIV in the previous 12 months was 65% in 2005, 49% in 2009%, and 53% in 2012.17–19 In addition, we observed contemporaneous increases in the proportion of PWID unaware of the HIV status of their last sexual partner and of their last injection partner. This documents an expected repercussion of lower HIV-testing levels in the Seattle PWID population.

Our results should be interpreted in light of several limitations. Both the NHBS and Needle Exchange study populations may constitute biased samples of the full PWID population of the Seattle area. Our findings have not been adjusted to reflect the characteristics of the RDS-recruited study populations. It is difficult to predict the nature of the bias this would produce. Our data derive from self-report and are subject to social desirability and recall bias.

Much remains to be learned about the reasons for the observed decline in HIV testing among Seattle-area PWID. HIV prevalence among non-MSM PWID in the Seattle area is low, which is likely attributable to the introduction of effective antiretroviral therapy, a robust local syringe exchange program, and a reduction in syringe sharing over time.14 It is plausible that this has reduced the motivation for annual HIV testing among PWID. In our data, the youngest age group was the most likely to report an HIV test perhaps because they share injection equipment more frequently than older PWID20 and are responding to awareness of their higher risk. In the context of declining HIV testing among PWID, syringe exchange programs, in particular, remain critical interventions for reducing HIV transmission risk.

There has been no apparent trend toward lower rates of HIV testing among MSM in the Seattle area across 3 NHBS MSM surveys, in 2008 (72% tested in previous 12 months),21 2011 (64%),22 and 2014 (70%).23 Thus there does not seem to be a global reduction in HIV testing in the Seattle area. Moreover, declines in HIV testing among PWID in Seattle cannot be attributable to lack of health insurance. Insurance coverage rates among PWID have increased dramatically in recent years as a result of the Affordable Care Act, increasing from 72% among NHBS participants in 2012% to 92% in 2015.8,24

It will be important to assess the extent to which reductions in HIV testing among PWID are occurring in other areas. Efforts are warranted to understand any changes in the motivations for HIV testing among PWID and to determine the extent to which the delivery or content of public health programs addressing HIV risk have played a role. Ongoing programs to support and encourage sustained high, adequate levels of HIV testing among PWID continue to be a priority in addressing the HIV epidemic.


1. Marks G, Crepaz N, Senterfitt JW, et al. Meta-analysis of high-risk sexual behavior in persons aware and unaware they are infected with HIV in the United States: implications for HIV prevention programs. J Acquir Immune Defic Syndr. 2005;39:446–453.
2. Khosropour CM, Dombrowski JD, Kerani RP, et al. Changes in condomless sex and serosorting among men who have sex with men after HIV diagnosis. J Acquir Immune Defic Syndr. 2016;73:475–481.
3. Moyer VA. Screening for HIV: U.S. Preventive Services Task Force Recommendation Statement. Ann Intern Med. 2013;159:51–60.
4. Vernazza PL, Troiani L, Flepp MJ, et al. Potent antiretroviral treatment of HIV infection results in suppression of the seminal shedding of HIV. AIDS. 2000;14:1121.
5. Attia S, Egger M, Muller M, et al. Sexual transmission of HIV according to viral load and antiretroviral therapy: systematic review and meta-analysis. AIDS. 2009;23:1397–1404.
6. Cohen MS, Chen VQ, McCauley M, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
7. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report. 2014. Vol. 26. Available at: Accessed September 1, 2016.
8. Glick SN, Burt RD, Moreno C, et al. Highlights from the 2015 Seattle area National Behavioral Surveillance survey of injection drug users (NHBS-IDU). Washington State/Seattle-King County HIV/AIDS Epidemiology Report, 2016. 2017. Available at: Accessed March 14, 2017.
9. Branson BM, Handsfield HH, Lampe MA, et al. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep. 2006;44:1–17.
10. Gallagher K, Sullivan PS, Lansky A, et al. Behavioral surveillance among people at risk for HIV infection in the U.S.: the National HIV Behavioral Surveillance System. Public Health Rep. 2007;122:32–38.
11. Heckathorn DD. Respondent-driven sampling: a new approach to the study of hidden populations. Social Probl. 1997;44:174–199.
12. Burt RD, Hagan H, Sabin K, et al. Evaluating respondent-driven sampling in a major metropolitan area: comparing injection drug users in the 2005 Seattle area National HIV Behavioral Surveillance system survey with participants in the RAVEN and Kiwi studies. Ann Epidemiol. 2010;20:159–167.
13. Burt RD, Thiede H. Evaluating consistency in repeat surveys of injection drug users recruited by respondent-driven sampling in the Seattle area: results from the NHBS-IDU1 and NHBS-IDU2 surveys. Ann Epidemiol. 2012;22:354–363.
14. Burt RD, Thiede H. Evidence for a reduction in the sharing of injection equipment among Seattle-area injection drug users, 1994–2013. Am J Public Health. 2016;106:301–307.
15. Burt RD, Thiede H. Evidence for risk reduction among amphetamine-injecting men who have sex with men: results from National HIV Behavioral Surveillance surveys in the Seattle area 2008-2012. AIDS Behav. 2014;18:1998–2008.
16. McCreesh N, Frost S, Seeley J, et al. Evaluation of respondent-driven sampling. Epidemiology. 2012;23:138–147.
17. Centers for Disease Control and Prevention. Risk, prevention and testing behaviors related to HIV and hepatitis behaviors—National HIV Behavioral Surveillance system: injection drug users, May 2005-Fegruary 2006. HIV Special Surveillance Report 7. Available at: Accessed July 14, 2011.
18. Centers for Disease Control and Prevention. HIV infection, risk, prevention and testing behaviors among persons who inject drugs - National HIV Behavioral Surveillance: injection drug use, 20 U.S. Cities, 2012. HIV Surveillance Special Report 11. Available at: Accessed September 1, 2016.
19. Broz D, Wejnert C, Pham HT, et al. HIV infection and risk, prevention, and testing behaviors among injecting drug users—National HIV Behavioral Surveillance system, 20 US cities, 2009. MMWR Surveill Summ. 2014;63:1–51.
20. Burt RD, Thiede H. Comparing injection drug users who do and do not report male-to-male sex. Washington State/Seattle-King County HIV/AIDS Epidemiology Report, 1st Half 2012. 2013. Available at: Accessed September 20, 2016.
21. Finlayson TJ, Le B, Smith A, et al. HIV risk, prevention and testing behaviors among men who have sex with men—National HIV Behavioral Surveillance system, 21 U.S. cities, United States, 2008. MMWR Surveill Summ. 2011;60:1–40.
22. Centers for Disease Control and Prevention. HIV infection, risk, prevention and testing behaviors. National HIV Behavioral Surveillance system. Men who have sex with men: 20 U.S. cities, 2011. HIV Surveillance Report, Special Report. Available at: Accessed September 20, 2016.
23. Centers for Disease Control and Prevention. HIV infection, risk, prevention and testing behaviors. National HIV Behavioral Surveillance system. Men who have sex with men: 20 U.S. cities, 2014. HIV Surveillance Report, Special Report. Available at: Accessed September 20, 2016.
24. Thiede H, Burt R, Shriver C. Highlights from the 2012 Seattle area NHBS survey of injection drug users. Washington State/Seattle-King County HIV/AIDS Epidemiology Report, 1st Half 2013. 2014. Available at: Accessed March 14, 2017.

HIV; HIV testing; persons who inject drugs

Supplemental Digital Content

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.