Drug and Alcohol Dependence 145 (2014) 180–184
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Differences in HIV risk behaviors among people who inject drugs by gender and sexual orientation, San Francisco, 2012 Harry Jin, Emalie Huriaux, Eileen Loughran, Tracey Packer, H. Fisher Raymond ∗ San Francisco Department of Public Health, United States
a r t i c l e
i n f o
Article history: Received 13 August 2014 Received in revised form 15 October 2014 Accepted 16 October 2014 Available online 23 October 2014 Keywords: People who inject drugs Gender Sexual orientation HIV
a b s t r a c t Background: Sharing of drug injection equipment is a well-established risk factor for the transmission of viral infections, such as human immunodeficiency virus (HIV). However, there are multiple mechanisms through which people who inject drugs (PWID) can acquire and transmit HIV. Differences in drug using and sexual behaviors among heterosexual males, males who have sex with males (MSM), and females who inject drugs may explain health disparities. Methods: Data were collected in San Francisco by the National HIV Behavioral Surveillance (NHBS) System of PWID in 2012, and were analyzed to compare the sexual behaviors, drug use behaviors, and prevalence of viral infections among heterosexual males, MSM, and females. Results: Using a weighted analysis for the RDS sampling design, we estimate that 3.7% of heterosexual males who inject drugs, 24.0% of MSM, and 13.0% of females who inject drugs are living with HIV. Females and heterosexual males primarily injected heroin, while MSM primarily injected methamphetamine. MSM were most likely to have received goods or money for sex and have unprotected intercourse. Conclusion: These data demonstrate differences in risk behaviors and prevalence of viral infections among heterosexual males, MSM, and females. The results also suggest that public health programs prioritizing the different populations of PWID are necessary. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Sharing of drug injection equipment is a well established mode of transmission of viral infections such as human immunodeficiency virus (HIV; Mathers et al., 2008). The San Francisco Department of Public Health reported that by the end of 2012, there were 15,705 people living with HIV in San Francisco, 3298 of whom were people who inject drugs (PWID; San Francisco Department of Public Health, 2013). Additionally, 12% of all newly reported HIV cases in 2012 were among PWID. National HIV Behavioral Surveillance data collected in San Francisco over the past decade estimates that the percent of PWID who tested for HIV in the past 6 months dropped from 62.4% in 2005 to 37.8% in 2012 (Coffin et al., 2014), and 42.7% of HIV-infected in 2012 were unaware of their status. The low uptake of HIV testing and high prevalence of undiagnosed HIV among PWID in San Francisco may be due to the increased emphasis on HIV prevention programs for males who have sex with males (MSM) in San Francisco, due to fact that 90% of new infections are among MSM in the city (Coffin
∗ Corresponding author. Tel.: +1 415 554 9093. E-mail address: hfi[email protected] (H.F. Raymond). http://dx.doi.org/10.1016/j.drugalcdep.2014.10.015 0376-8716/© 2014 Elsevier Ireland Ltd. All rights reserved.
et al., 2014). This highlights the need to address population specific barriers to HIV testing and treatment, particularly for PWID. Although there have been studies reporting differences between male and female PWID (Coffin et al., 2014; Young et al., 2014), no studies to our knowledge have examined differences in risk behaviors and viral infections among heterosexual males, MSM, and females who inject drugs living in San Francisco. This may present a significant gap in understanding of the mechanisms through which different subpopulations of PWID acquire viral infections and whether the health care needs of the subpopulations have been met. Therefore, the purpose of the present study was to use most recent data to confirm patterns in differences across the groups that have been suggested in other analyses. 2. Methods These data were collected by the San Francisco Department of Public Health as part of the National HIV Behavioral Surveillance (NHBS) system, which conducts interviews on triennial cycles with PWID, MSM, and heterosexuals at increased risk for HIV (Frajzyngier et al., 2007). The third PWID cycle was conducted between July 10, 2012 and November 1, 2012, and used respondent-driven sampling (RDS), a peer-referral sampling method, to recruit PWID (Magnani et al., 2005; Lansky et al., 2012). Recruitment chains began with seven seeds, who broadly represented the racial, gender, and social-economic status of the population of PWID in San Francisco. Participants who completed the survey and were eligible to recruit
H. Jin et al. / Drug and Alcohol Dependence 145 (2014) 180–184 Table 1 Crude and adjusted demographics of people who inject drugs (PWID) in San Francisco, CA, 2012.
Age ≤35 36–45 46–50 51+ Race/ethnicity White Black Latino Other Country of birth U.S. Mexico Puerto Rico Other Educational attainment Never attended school Grades 1–8 Grades 9–11 Grade 12 or GED Some college College Post-graduate Employment Employed full time Employed part time Homemaker Full-time student Retired Disabled Unemployed Other Income $0–9999 $10,000–29,999 $30,000–49,999 $50,000–74,999 $75,000+ Unknown/other Sexual orientation Heterosexual Gay/lesbian Bisexual Other HIV status Positive Negative
Total sample 563
Total adjusted %
% Difference
81 (14.4) 110 (19.5) 101 (17.9) 271 (48.1)
15.4 (9.6, 20.1) 18.5 (13.5, 26.0) 17.7 (13.0, 23.3) 48.3 (40.8, 56.6)
1.0 1.0 0.2 0.2
205 (36.4) 227 (40.3) 60 (10.7) 71 (12.6)
38.1 (29.4, 46.0) 42.9 (34.6, 52.1) 10.3 (6.6, 14.8) 8.7 (5.2, 13.2)
1.7 2.6 0.4 3.9
542 (96.3) 0 (0.0) 2 (0.4) 19 (3.4)
96.1 (93.7, 98.7) 0.0 (0.0, 0.0) 0.2 (0.0, 0.3) 3.8 (1.2, 6.2)
0.5 0.0 0.2 0.4
1 (0.2) 23 (4.1) 117 (20.8) 217 (38.5) 168 (29.8) 25 (4.4) 12 (2.1)
0.1 (0.0, 0.2) 4.7 (2.5, 10.0) 17.8 (13.1, 23.0) 38.2 (32.9, 45.0) 30.9 (23.4, 36.2) 5.8 (2.3, 9.7) 2.6 (0.7, 4.8)
0.1 0.6 3.0 0.3 1.1 1.4 0.5
22 (3.9) 44 (7.8) 3 (0.5) 2 (0.4) 29 (5.2) 238 (42.3) 211 (37.5) 14 (2.5)
2.2 (1.1, 3.4) 8.6 (4.9, 13.2) 1.0 (0.0, 2.5) 1.0 (0.0, 2.7) 4.6 (2.3, 8.9) 40.3 (33.9, 47.6) 38.8 (31.0, 45.4) 3.5 (1.1, 6.1)
1.7 0.8 0.5 0.6 0.6 2.0 1.3 1.0
2 (0.4) 225 (40.0) 282 (50.1) 29 (5.2) 16 (2.8) 9 (1.6)
0.0 (0.0, 0.1) 42.4 (36.3, 49.6) 49.4 (41.9, 55.6) 4.8 (2.6, 7.2) 2.7 (1.3, 4.9) 0.6 (0.1, 1.4)
0.4 2.4 0.6 0.4 0.1 1.0
394 (70.0) 39 (6.9) 127 (22.6) 3 (0.5)
65.7 (58.2, 72.4) 8.5 (4.6, 13.8) 25.7 (19.9, 31.8) 0.1 (0.0, 0.2)
4.3 1.6 3.1 0.4
65 (11.7) 498 (88.5)
12.7 (7.3, 16.9) 87.3 (83.1, 92.7)
0.1 0.2
others were given 3–5 recruitment coupons to distribute to members of their social networks. Surveys, designed in QDS 2.6 (Nova Research) were administered by study interviewers using hand held tablet computers. Following the survey, a rapid HIV test was conducted using Insti HIV-1 Antibody Test (Biolytical Labs, Vancouver, Canada). Preliminary positives were confirmed using oral fluid testing with the Orasure HIV-1 Western Blot (Bethlehem, PA). Hepatitis testing was not offered in this study. All data besides HIV test results were self-reported. All activities were approved by the University of California, San Francisco Committee on Human Research. Each participant who finished the survey and the HIV test was given $50 and an additional $10 for each eligible participant they recruited (Tables 1–4). Data from the surveys were uploaded into a data warehouse and then exported into SAS 9.3 (SAS Institute, Cary, NC), where unweighted frequencies and percents of the sample population were calculated. Further analyses were weighted using Respondent-Driven Sampling Analysis Tool 7.1 (RDSAT), which adjusts for recruitment bias in RDS sampling and produced adjusted population estimates (Abramovitz et al., 2009) to calculate RDS-weight population estimates of indicators and their 95% confidence intervals (95% CI). Weights were exported and used to conduct statistical tests in SAS software. Chi-square tests were used to compare differences in demographic, behaviors, and health status characteristics among heterosexual males, MSM, and females. MSM were defined as men who have had sex with at least one male in the past 12 months. Multivariate logistic regressions were created for heterosexual men, MSM, and women to calculate the odds of being HIV+.
181
3. Results A total of 563 PWID participants were recruited into the study, 49.4% (N = 278) were heterosexual males, 21.1% (N = 119) were MSM, and 29.5% (N = 166) were females, with ages ranging from 20 to 70 years. Weighted analyses were done in order to make inferences about the PWID population living in San Francisco. We estimate that 42.9% (95% CI 34.6%, 52.1%) of PWID in San Francisco are black, 38.1% (95% CI 29.4%, 46.0%) are white, and 10.3% (95% CI 6.6%, 14.8%) are Latino; almost all (96.3%, 95% CI 93.7%, 98.7%) were born in the United States; 40.3% (95% CI 33.9%, 47.6%) are on Social Security Disability Insurance, and 38.8% (95% CI 31.0%, 45.4%) are unemployed; 42.4% (95% CI 36.3%, 49.6%) make between $10,000 and $29,999 annually from all sources, and 49.4% (95% CI 41.9%, 55.6%) make $30,000–$49,999 annually. We estimate that 12.7% (95% CI 7.3%, 16.9%) of PWID are living with HIV, 39.4% (95% CI 22.1%, 61.6%) of whom do not know that they are infected with the virus. We estimate that 4.2% (95% CI 2.6%, 6.2%) of PWID have been diagnosed with Hepatitis A; 10.3% (95% CI 6.7%, 15.0%) have been diagnosed with Hepatitis B; 53.9% (95% CI 46.9%, 62.0%) have been diagnosed with Hepatitis C; and 5.8% (95% CI 3.0%, 9.7%) have been diagnosed with HIV/HCV co-infection. The population estimate of PWID who have had sex for goods or money with at least one of their last five sexual partners is 17.8% (95% CI 13.6%, 22.6%); 60.3% (95% CI 52.6%, 67.1%) had any unprotected sex with their last five sexual partners; and 39.7% (95% CI 32.9%, 47.4%) do not always use new, sterile needles when injecting drugs. We estimate that 57.7% (95% CI 50.2%, 67.4%) primarily use heroin and 31.1% (95% CI 22.3%, 38.7%) primarily use methamphetamine. Compared to females and heterosexual males who inject drugs, MSM who inject drugs have a significantly higher HIV prevalence (24.0% vs. 13.0% and 3.7%, p-value < 0.0001), are more likely to participate in commercial sex work (25.4% vs. 22.0% and 9.3%, pvalue < 0.0001), and are more likely to have intercourse without a condom (73.7% vs. 61.0% and 56.1%, p-value 0.0019). The majority of MSM (55.3%, 95% CI 38.9%, 67.9%) use methamphetamines as their primary drug, while females and heterosexual males primarily use heroin (70.0%, 95% CI 56.5%, 81.5% and 59.5%, 95% CI 52.7%, 74.9%). The multivariate logistic regressions suggested that having HIV among heterosexual men, MSM, and women were not associated with age, gender, having unprotected intercourse with his/her last five sexual partners in the past 12 months, or always using a clean needle to inject drugs in the past 12 months. Heterosexual men who have received goods or money for sex in the past 12 months were 3.8 (95% CI 1.3, 11.5) as likely to be HIV+ compared to heterosexual men who have not received goods or money for sex.
4. Discussion This study of PWID in San Francisco provides a comprehensive comparison of viral infection prevalence, sexual risk behaviors, and drug use behaviors among heterosexual males, MSM, and females who inject drugs. We estimate that 39.4% of PWID who are living with HIV/AIDS are unaware of their status. Low rates of HIV testing among PWID have been attributed to barriers such as a lack of knowledge of how HIV is transmitted (Moyer et al., 2008) and the belief that they do not have access to treatment (Spielberg et al., 2003). To address these common issues, in 2012, San Francisco Department of Public Health’s HIV Prevention Planning Council recommended that programs serving people who use drugs integrate HIV prevention services, HIV and sexual health education, and linkage to services to help people engage with medical and mental health care (HIV Prevention Planning Council, 2012). Additionally, syringe access
182
H. Jin et al. / Drug and Alcohol Dependence 145 (2014) 180–184
Table 2 RDS-weighted demographics by gender and sexual orientation among people who inject drugs (PWID) in San Francisco, CA, 2012. Heterosexual men % (95% CI) Age ≤35 36–45 46–50 51+ Race/ethnicity White Black Latino Other Country of birth U.S. Mexico Puerto Rico Other Educational attainment Never attended school Grades 1–8 Grades 9–11 Grade 12 or GED Some college College Post-graduate Employment Employed full time Employed part time Homemaker Full-time student Retired Disabled Unemployed Other Income $0–9999 $10,000–29,999 $30,000–49,999 $50,000–74,999 $75,000+ Unknown/other Sexual orientation Heterosexual Gay/lesbian Bisexual Other HIV status Positive Negative
MSM % (95% CI)
Women % (95% CI)
9.4 (3.9, 14.4) 14.7 (5.7, 28.7) 21.2 (14.1, 29.0) 54.7 (43.3, 66.7)
23.0 (8.6, 31.3) 26.9 (19.7, 40.1) 13.1 (4.6, 22.8) 36.9 (25.3, 54.6)
14.2 (6.6, 25.3) 17.8 (10.3, 26.4) 17.7 (8.1, 24.8) 50.3 (37.9, 64.3)
40.3 (27.3, 53.1) 43.2 (32.1, 55.4) 7.7 (3.6, 13.5) 8.8 (4.5, 12.5)
35.0 (22.0, 45.5) 43.4 (29.9, 59.5) 8.6 (2.3, 18.1) 13.0 (3.0, 24.7)
36.9 (25.0, 50.5) 42.5 (29.6, 57.3) 14.5 (6.9, 23.0) 6.0 (2.1, 8.7)
95.3 (90.9, 98.6) 0.0 (0.0, 0.0) 0.3 (0.0, 0.6) 4.5 (1.2, 8.9)
99.4 (99.3, 100.0) 0.0 (0.0, 0.0) 0.3 (0.0, 0.7) 0.3 (0.0, 0.0)
94.4 (89.8, 99.9) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 5.3 (0.1, 10.2)
0.2 (0.0, 0.6) 5.8 (1.3, 16.2) 19.8 (12.9, 27.0) 35.6 (26.4, 44.7) 30.1 (18.7, 40.5) 5.7 (0.8, 12.2) 2.8 (0.6, 6.8)
0.1 (0.0, 0.0) 4.7 (1.4, 9.4) 12.7 (6.0, 23.2) 38.2 (3.0, 52.9) 33.5 (21.1, 41.7) 8.4 (1.2, 16.0) 2.5 (0.0, 5.9)
0.0 (0.0, 0.0) 3.8 (0.8, 8.0) 18.4 (8.5, 30.2) 41.5 (28.6, 51.1) 30.4 (20.7, 43.6) 3.1 (0.5, 7.4) 2.8 (0.0, 7.9)
2.5 (0.6, 4.4) 10.4 (4.3, 18.8) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 7.8 (2.9, 16.4) 33.9 (25.2, 42.7) 42.8 (31.9, 53.4) 2.6 (0.4, 5.2)
1.2 (0.0, 3.1) 5.1 (1.1, 11.5) 0.0 (0.0, 0.0) 2.3 (0.0, 7.0) 6.1 (2.2, 12.3) 36.6 (25.8, 52.2) 44.5 (27.4, 57.2) 4.2 (0.0, 11.2)
2.8 (0.9, 5.6) 8.2 (1.9, 16.4) 3.2 (0.0, 8.5) 1.7 (0.0, 5.5) 0.1 (0.0, 0.3) 51.1 (40.2, 65.1) 28.7 (17.8, 39.4) 4.2 (0.3, 8.4)
0.1 (0.0, 0.4) 43.6 (33.1, 53.8) 46.8 (36.5, 56.9) 5.3 (2.5, 10.0) 3.3 (0.9, 7.0) 0.8 (0.0, 2.1)
0.1 (0.0, 0.0) 35.4 (23.6, 51.4) 57.6 (41.6, 69.9) 4.7 (1.5, 9.2) 2.1 (0.0, 5.7) 0.1 (0.0, 0.0)
0.1 (0.0, 0.0) 43.7 (30.8, 54.5) 48.5 (36.6, 61.9) 4.6 (0.5, 9.7) 2.6 (0.2, 6.4) 0.6 (0.0, 2.3)
100.0 (99.9, 100.0) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 0.0 (0.0, 0.1)
11.3 (3.0, 13.5) 26.1 (15.4, 44.5) 62.6 (48.8, 76.3) 0.0 (0.0, 0.0)
66.9 (54.6, 77.8) 5.1 (0.5, 11.7) 27.8 (17.9, 39.7) 0.2 (0.0, 0.7)
3.7 (1.7, 6.7) 96.3 (93.3, 98.3)
24.0 (12.9, 38.0) 76.0 (62.0, 87.1)
13.0 (5.2, 22.4) 87.0 (77.6, 94.8)
and disposal programs and pharmacies that sell non-prescription syringes are required to provide information to customers about where they can get HIV testing if they themselves do not offer HIV testing on site (HIV Prevention Planning Council, 2012). Task forces are currently being formed to ensure that these recommendations become implemented as standards of care. In this study, we found that MSM who inject drugs were most likely to engage in behaviors that put them at risk for HIV and have the highest prevalence of HIV, which makes them a priority population for public health efforts to address HIV. Previous studies conducted in San Francisco have suggested that certain PWID populations, like MSM who inject drugs, may be harder to reach than others and are further disconnected from care. For example, the Party and Play Study, which recruited high risk MSM exclusively during late night hours, found that a third of the sample was living with HIV/AIDS, and a quarter of those who were living with HIV were unaware of their status (Pendo, 2003). The results of the Party and Play Study became the impetus for the Late Night Breakfast Buffet (LNBB) Study, which tested the efficacy of providing free harm reduction services to drug-using MSM from 1:00 am to 5:00 am out of a van which stopped in neighborhoods drugusing MSM were likely to frequent (Rose et al., 2006). These services
included syringe access and disposal, HIV testing and counseling, and condom distribution. The goal of the LNBB was to engage MSM who were not utilizing these services at fixed sites during day light hours. The LNBB study provided 2000 syringes, distributed 4500 condoms/lubricants, and provided 21 HIV tests, and 12 STI tests, demonstrating the willingness of MSM who use drugs to utilizing these services. This study provides evidence that some drug-using populations may be more amenable to utilize services that cater to their specific needs. An example of a successful program designed to service a niche population is Ladies’ Night, a harm reductionbased drop-in program that provides services to homeless and marginalized housed women in San Francisco (Magee and Huriaux, 2008). This program is tailored to address the structural and social barriers homeless and marginally housed women often face when seeking healthcare and social services. The results also reveal that MSM who inject drugs have unique drug using behaviors, sexual behaviors, and prevalence of viral infection. This suggests that MSM who inject drugs may operate in different social networks from females and heterosexual males who inject drugs. The choice of drug may illustrate the different social network MSM who inject drugs inhabit as compared to females and heterosexual males who inject drugs. As our
H. Jin et al. / Drug and Alcohol Dependence 145 (2014) 180–184
183
Table 3 RDS-weighted prevalence of viral infections and risk factors by gender and sexual orientation among people who inject drugs (PWID) in San Francisco, CA, 2012. Total
Heterosexual men (%)
HIV 12.7 (7.3, 16.9) 3.7 (1.8, 6.6) Yes No 88.3 (83.0, 92.8) 96.3 (93.4, 98.2) Knew that he/she was infected with HIV among HIV+ 60.6 (38.4, 77.9) 59.4 (28.9, 93.3) Yes 39.4 (22.1, 61.6) 40.6 (6.7, 71.1) No Ever been diagnosed with Hepatitis A 4.2 (2.6, 6.2) 4.8 (2.4, 8.4) Yes No 95.8 (93.8, 97.4) 95.2 (91.6, 97.6) Ever been diagnosed with Hepatitis B 10.3 (6.7, 15.0) 11.6 (5.8, 19.0) Yes 89.7 (85.0, 93.3) 88.4 (81.0, 94.2) No Ever been diagnosed with Hepatitis C 53.9 (46.9, 62.0) 57.3 (46.1, 67.7) Yes No 46.1 (38.0, 53.1) 42.7 (32.3, 53.9) HIV/HCV Co-infection 5.8 (3.0, 9.7) 2.1 (0.9, 3.8) Yes 94.2 (90.3, 97.0) 98.9 (96.2, 99.1) No Received goods or money for sex in the past 12 months 17.8 (13.6, 22.6) 9.3 (5.3, 14.0) Yes No 82.2 (77.4, 86.4) 90.7 (86.0, 94.7) Any unprotected intercourse with his/her last 5 sexual partners in the past 12 months 62.6 (54.9, 69.2) 56.1 (44.1, 67.4) Yes 37.4 (30.8, 45.1) 43.9 (32.6, 55.9) No Always uses a clean needle in the past 12 months 60.3 (52.6, 67.1) 59.1 (49.7, 71.9) Yes 39.7 (32.9, 47.4) 40.9 (28.1, 50.3) No Primary drug Heroin 57.7 (50.2, 67.4) 59.5 (52.7, 74.9) Cocaine 1.0 (0.1, 3.1) 0.1 (0.0, 0.4) Speedball 6.1 (3.2, 9.2) 5.4 (2.1, 9.4) Crack 0.2 (0.0, 0.8) 0.6 (0.0, 1.7) Methamphetamine 31.1 (22.3, 38.7) 26.0 (15.6, 33.5) Other 4.0 (0.7, 9.3) 8.5 (0.3, 17.9)
data reveal, MSM primarily inject methamphetamine. Research has shown that MSM who use methamphetamine to enhance sexual pleasure are more likely to engage in sexual behaviors that are known to increase risk of HIV transmission, including condomless anal intercourse, sex with multiple partners, and frequent sexual activity. The association between methamphetamine use and condomless anal intercourse has been illustrated among HIV negative MSM and MSM living with HIV. We also found that females who inject drugs have an elevated risk of acquiring HIV and/or HCV relative to heterosexual males who inject drugs (Wagner et al., 2013). Studies have demonstrated that females who inject drugs may have an elevated risk for HIV
MSM (%)
Women (%)
P-value
24.0 (13.4, 37.1) 76.0 (62.9, 86.6)
13.0 (5.4, 22.3) 87.0 (77.7, 94.6)
46.6 (17.5, 74.0) 53.4 (26.0, 82.5)
83.8 (55.3, 96.3) 16.2 (3.7, 44.7)
4.8 (1.8, 10.0) 95.2 (90.0, 98.2)
2.7 (0.5, 6.5) 97.3 (93.5, 99.5)
9.5 (3.8, 18.0) 90.5 (82.0, 96.2)
9.2 (3.9, 15.4) 90.8 (84.6, 96.1)
39.1 (29.0, 56.1) 60.9 (43.9, 71.0)
60.2 (49.4, 72.9) 39.8 (27.1, 50.6)
6.3 (2.3, 13.3) 93.7 (86.7, 97.7)
10.8 (3.2, 19.4) 89.2 (80.6, 96.8)
25.4 (17.6, 38.4) 74.7 (61.6, 82.4)
22.0 (13.3, 30.5) 78.0 (69.5, 86.7)
73.7 (60.4, 84.9) 26.3 (15.1, 39.6)
61.0 (49.0, 73.1) 39.0 (26.9, 51.0)
64.8 (50.2, 75.2) 35.2 (24.8, 49.8)
59.0 (46.3, 71.2) 41.0 (29.7, 54.6)
33.7 (20.5, 50.0) 2.7 (0.0, 10.3) 6.1 (1.4, 13.0) 0.0 (0.0, 0.0) 55.3 (38.9, 67.9) 2.1 (0.7, 4.4)
70.0 (56.5, 81.5) 0.6 (0.1, 1.5) 6.4 (1.3, 10.8) 0.0 (0.0, 1.6) 22.9 (12.0, 37.6) 0.0 (0.0, 0.1)
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Differences in HIV risk behaviors among people who inject drugs by gender and sexual orientation, San Francisco, 2012 Harry Jin, Emalie Huriaux, Eileen Loughran, Tracey Packer, H. Fisher Raymond ∗ San Francisco Department of Public Health, United States
a r t i c l e
i n f o
Article history: Received 13 August 2014 Received in revised form 15 October 2014 Accepted 16 October 2014 Available online 23 October 2014 Keywords: People who inject drugs Gender Sexual orientation HIV
a b s t r a c t Background: Sharing of drug injection equipment is a well-established risk factor for the transmission of viral infections, such as human immunodeficiency virus (HIV). However, there are multiple mechanisms through which people who inject drugs (PWID) can acquire and transmit HIV. Differences in drug using and sexual behaviors among heterosexual males, males who have sex with males (MSM), and females who inject drugs may explain health disparities. Methods: Data were collected in San Francisco by the National HIV Behavioral Surveillance (NHBS) System of PWID in 2012, and were analyzed to compare the sexual behaviors, drug use behaviors, and prevalence of viral infections among heterosexual males, MSM, and females. Results: Using a weighted analysis for the RDS sampling design, we estimate that 3.7% of heterosexual males who inject drugs, 24.0% of MSM, and 13.0% of females who inject drugs are living with HIV. Females and heterosexual males primarily injected heroin, while MSM primarily injected methamphetamine. MSM were most likely to have received goods or money for sex and have unprotected intercourse. Conclusion: These data demonstrate differences in risk behaviors and prevalence of viral infections among heterosexual males, MSM, and females. The results also suggest that public health programs prioritizing the different populations of PWID are necessary. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Sharing of drug injection equipment is a well established mode of transmission of viral infections such as human immunodeficiency virus (HIV; Mathers et al., 2008). The San Francisco Department of Public Health reported that by the end of 2012, there were 15,705 people living with HIV in San Francisco, 3298 of whom were people who inject drugs (PWID; San Francisco Department of Public Health, 2013). Additionally, 12% of all newly reported HIV cases in 2012 were among PWID. National HIV Behavioral Surveillance data collected in San Francisco over the past decade estimates that the percent of PWID who tested for HIV in the past 6 months dropped from 62.4% in 2005 to 37.8% in 2012 (Coffin et al., 2014), and 42.7% of HIV-infected in 2012 were unaware of their status. The low uptake of HIV testing and high prevalence of undiagnosed HIV among PWID in San Francisco may be due to the increased emphasis on HIV prevention programs for males who have sex with males (MSM) in San Francisco, due to fact that 90% of new infections are among MSM in the city (Coffin
∗ Corresponding author. Tel.: +1 415 554 9093. E-mail address: hfi[email protected] (H.F. Raymond). http://dx.doi.org/10.1016/j.drugalcdep.2014.10.015 0376-8716/© 2014 Elsevier Ireland Ltd. All rights reserved.
et al., 2014). This highlights the need to address population specific barriers to HIV testing and treatment, particularly for PWID. Although there have been studies reporting differences between male and female PWID (Coffin et al., 2014; Young et al., 2014), no studies to our knowledge have examined differences in risk behaviors and viral infections among heterosexual males, MSM, and females who inject drugs living in San Francisco. This may present a significant gap in understanding of the mechanisms through which different subpopulations of PWID acquire viral infections and whether the health care needs of the subpopulations have been met. Therefore, the purpose of the present study was to use most recent data to confirm patterns in differences across the groups that have been suggested in other analyses. 2. Methods These data were collected by the San Francisco Department of Public Health as part of the National HIV Behavioral Surveillance (NHBS) system, which conducts interviews on triennial cycles with PWID, MSM, and heterosexuals at increased risk for HIV (Frajzyngier et al., 2007). The third PWID cycle was conducted between July 10, 2012 and November 1, 2012, and used respondent-driven sampling (RDS), a peer-referral sampling method, to recruit PWID (Magnani et al., 2005; Lansky et al., 2012). Recruitment chains began with seven seeds, who broadly represented the racial, gender, and social-economic status of the population of PWID in San Francisco. Participants who completed the survey and were eligible to recruit
H. Jin et al. / Drug and Alcohol Dependence 145 (2014) 180–184 Table 1 Crude and adjusted demographics of people who inject drugs (PWID) in San Francisco, CA, 2012.
Age ≤35 36–45 46–50 51+ Race/ethnicity White Black Latino Other Country of birth U.S. Mexico Puerto Rico Other Educational attainment Never attended school Grades 1–8 Grades 9–11 Grade 12 or GED Some college College Post-graduate Employment Employed full time Employed part time Homemaker Full-time student Retired Disabled Unemployed Other Income $0–9999 $10,000–29,999 $30,000–49,999 $50,000–74,999 $75,000+ Unknown/other Sexual orientation Heterosexual Gay/lesbian Bisexual Other HIV status Positive Negative
Total sample 563
Total adjusted %
% Difference
81 (14.4) 110 (19.5) 101 (17.9) 271 (48.1)
15.4 (9.6, 20.1) 18.5 (13.5, 26.0) 17.7 (13.0, 23.3) 48.3 (40.8, 56.6)
1.0 1.0 0.2 0.2
205 (36.4) 227 (40.3) 60 (10.7) 71 (12.6)
38.1 (29.4, 46.0) 42.9 (34.6, 52.1) 10.3 (6.6, 14.8) 8.7 (5.2, 13.2)
1.7 2.6 0.4 3.9
542 (96.3) 0 (0.0) 2 (0.4) 19 (3.4)
96.1 (93.7, 98.7) 0.0 (0.0, 0.0) 0.2 (0.0, 0.3) 3.8 (1.2, 6.2)
0.5 0.0 0.2 0.4
1 (0.2) 23 (4.1) 117 (20.8) 217 (38.5) 168 (29.8) 25 (4.4) 12 (2.1)
0.1 (0.0, 0.2) 4.7 (2.5, 10.0) 17.8 (13.1, 23.0) 38.2 (32.9, 45.0) 30.9 (23.4, 36.2) 5.8 (2.3, 9.7) 2.6 (0.7, 4.8)
0.1 0.6 3.0 0.3 1.1 1.4 0.5
22 (3.9) 44 (7.8) 3 (0.5) 2 (0.4) 29 (5.2) 238 (42.3) 211 (37.5) 14 (2.5)
2.2 (1.1, 3.4) 8.6 (4.9, 13.2) 1.0 (0.0, 2.5) 1.0 (0.0, 2.7) 4.6 (2.3, 8.9) 40.3 (33.9, 47.6) 38.8 (31.0, 45.4) 3.5 (1.1, 6.1)
1.7 0.8 0.5 0.6 0.6 2.0 1.3 1.0
2 (0.4) 225 (40.0) 282 (50.1) 29 (5.2) 16 (2.8) 9 (1.6)
0.0 (0.0, 0.1) 42.4 (36.3, 49.6) 49.4 (41.9, 55.6) 4.8 (2.6, 7.2) 2.7 (1.3, 4.9) 0.6 (0.1, 1.4)
0.4 2.4 0.6 0.4 0.1 1.0
394 (70.0) 39 (6.9) 127 (22.6) 3 (0.5)
65.7 (58.2, 72.4) 8.5 (4.6, 13.8) 25.7 (19.9, 31.8) 0.1 (0.0, 0.2)
4.3 1.6 3.1 0.4
65 (11.7) 498 (88.5)
12.7 (7.3, 16.9) 87.3 (83.1, 92.7)
0.1 0.2
others were given 3–5 recruitment coupons to distribute to members of their social networks. Surveys, designed in QDS 2.6 (Nova Research) were administered by study interviewers using hand held tablet computers. Following the survey, a rapid HIV test was conducted using Insti HIV-1 Antibody Test (Biolytical Labs, Vancouver, Canada). Preliminary positives were confirmed using oral fluid testing with the Orasure HIV-1 Western Blot (Bethlehem, PA). Hepatitis testing was not offered in this study. All data besides HIV test results were self-reported. All activities were approved by the University of California, San Francisco Committee on Human Research. Each participant who finished the survey and the HIV test was given $50 and an additional $10 for each eligible participant they recruited (Tables 1–4). Data from the surveys were uploaded into a data warehouse and then exported into SAS 9.3 (SAS Institute, Cary, NC), where unweighted frequencies and percents of the sample population were calculated. Further analyses were weighted using Respondent-Driven Sampling Analysis Tool 7.1 (RDSAT), which adjusts for recruitment bias in RDS sampling and produced adjusted population estimates (Abramovitz et al., 2009) to calculate RDS-weight population estimates of indicators and their 95% confidence intervals (95% CI). Weights were exported and used to conduct statistical tests in SAS software. Chi-square tests were used to compare differences in demographic, behaviors, and health status characteristics among heterosexual males, MSM, and females. MSM were defined as men who have had sex with at least one male in the past 12 months. Multivariate logistic regressions were created for heterosexual men, MSM, and women to calculate the odds of being HIV+.
181
3. Results A total of 563 PWID participants were recruited into the study, 49.4% (N = 278) were heterosexual males, 21.1% (N = 119) were MSM, and 29.5% (N = 166) were females, with ages ranging from 20 to 70 years. Weighted analyses were done in order to make inferences about the PWID population living in San Francisco. We estimate that 42.9% (95% CI 34.6%, 52.1%) of PWID in San Francisco are black, 38.1% (95% CI 29.4%, 46.0%) are white, and 10.3% (95% CI 6.6%, 14.8%) are Latino; almost all (96.3%, 95% CI 93.7%, 98.7%) were born in the United States; 40.3% (95% CI 33.9%, 47.6%) are on Social Security Disability Insurance, and 38.8% (95% CI 31.0%, 45.4%) are unemployed; 42.4% (95% CI 36.3%, 49.6%) make between $10,000 and $29,999 annually from all sources, and 49.4% (95% CI 41.9%, 55.6%) make $30,000–$49,999 annually. We estimate that 12.7% (95% CI 7.3%, 16.9%) of PWID are living with HIV, 39.4% (95% CI 22.1%, 61.6%) of whom do not know that they are infected with the virus. We estimate that 4.2% (95% CI 2.6%, 6.2%) of PWID have been diagnosed with Hepatitis A; 10.3% (95% CI 6.7%, 15.0%) have been diagnosed with Hepatitis B; 53.9% (95% CI 46.9%, 62.0%) have been diagnosed with Hepatitis C; and 5.8% (95% CI 3.0%, 9.7%) have been diagnosed with HIV/HCV co-infection. The population estimate of PWID who have had sex for goods or money with at least one of their last five sexual partners is 17.8% (95% CI 13.6%, 22.6%); 60.3% (95% CI 52.6%, 67.1%) had any unprotected sex with their last five sexual partners; and 39.7% (95% CI 32.9%, 47.4%) do not always use new, sterile needles when injecting drugs. We estimate that 57.7% (95% CI 50.2%, 67.4%) primarily use heroin and 31.1% (95% CI 22.3%, 38.7%) primarily use methamphetamine. Compared to females and heterosexual males who inject drugs, MSM who inject drugs have a significantly higher HIV prevalence (24.0% vs. 13.0% and 3.7%, p-value < 0.0001), are more likely to participate in commercial sex work (25.4% vs. 22.0% and 9.3%, pvalue < 0.0001), and are more likely to have intercourse without a condom (73.7% vs. 61.0% and 56.1%, p-value 0.0019). The majority of MSM (55.3%, 95% CI 38.9%, 67.9%) use methamphetamines as their primary drug, while females and heterosexual males primarily use heroin (70.0%, 95% CI 56.5%, 81.5% and 59.5%, 95% CI 52.7%, 74.9%). The multivariate logistic regressions suggested that having HIV among heterosexual men, MSM, and women were not associated with age, gender, having unprotected intercourse with his/her last five sexual partners in the past 12 months, or always using a clean needle to inject drugs in the past 12 months. Heterosexual men who have received goods or money for sex in the past 12 months were 3.8 (95% CI 1.3, 11.5) as likely to be HIV+ compared to heterosexual men who have not received goods or money for sex.
4. Discussion This study of PWID in San Francisco provides a comprehensive comparison of viral infection prevalence, sexual risk behaviors, and drug use behaviors among heterosexual males, MSM, and females who inject drugs. We estimate that 39.4% of PWID who are living with HIV/AIDS are unaware of their status. Low rates of HIV testing among PWID have been attributed to barriers such as a lack of knowledge of how HIV is transmitted (Moyer et al., 2008) and the belief that they do not have access to treatment (Spielberg et al., 2003). To address these common issues, in 2012, San Francisco Department of Public Health’s HIV Prevention Planning Council recommended that programs serving people who use drugs integrate HIV prevention services, HIV and sexual health education, and linkage to services to help people engage with medical and mental health care (HIV Prevention Planning Council, 2012). Additionally, syringe access
182
H. Jin et al. / Drug and Alcohol Dependence 145 (2014) 180–184
Table 2 RDS-weighted demographics by gender and sexual orientation among people who inject drugs (PWID) in San Francisco, CA, 2012. Heterosexual men % (95% CI) Age ≤35 36–45 46–50 51+ Race/ethnicity White Black Latino Other Country of birth U.S. Mexico Puerto Rico Other Educational attainment Never attended school Grades 1–8 Grades 9–11 Grade 12 or GED Some college College Post-graduate Employment Employed full time Employed part time Homemaker Full-time student Retired Disabled Unemployed Other Income $0–9999 $10,000–29,999 $30,000–49,999 $50,000–74,999 $75,000+ Unknown/other Sexual orientation Heterosexual Gay/lesbian Bisexual Other HIV status Positive Negative
MSM % (95% CI)
Women % (95% CI)
9.4 (3.9, 14.4) 14.7 (5.7, 28.7) 21.2 (14.1, 29.0) 54.7 (43.3, 66.7)
23.0 (8.6, 31.3) 26.9 (19.7, 40.1) 13.1 (4.6, 22.8) 36.9 (25.3, 54.6)
14.2 (6.6, 25.3) 17.8 (10.3, 26.4) 17.7 (8.1, 24.8) 50.3 (37.9, 64.3)
40.3 (27.3, 53.1) 43.2 (32.1, 55.4) 7.7 (3.6, 13.5) 8.8 (4.5, 12.5)
35.0 (22.0, 45.5) 43.4 (29.9, 59.5) 8.6 (2.3, 18.1) 13.0 (3.0, 24.7)
36.9 (25.0, 50.5) 42.5 (29.6, 57.3) 14.5 (6.9, 23.0) 6.0 (2.1, 8.7)
95.3 (90.9, 98.6) 0.0 (0.0, 0.0) 0.3 (0.0, 0.6) 4.5 (1.2, 8.9)
99.4 (99.3, 100.0) 0.0 (0.0, 0.0) 0.3 (0.0, 0.7) 0.3 (0.0, 0.0)
94.4 (89.8, 99.9) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 5.3 (0.1, 10.2)
0.2 (0.0, 0.6) 5.8 (1.3, 16.2) 19.8 (12.9, 27.0) 35.6 (26.4, 44.7) 30.1 (18.7, 40.5) 5.7 (0.8, 12.2) 2.8 (0.6, 6.8)
0.1 (0.0, 0.0) 4.7 (1.4, 9.4) 12.7 (6.0, 23.2) 38.2 (3.0, 52.9) 33.5 (21.1, 41.7) 8.4 (1.2, 16.0) 2.5 (0.0, 5.9)
0.0 (0.0, 0.0) 3.8 (0.8, 8.0) 18.4 (8.5, 30.2) 41.5 (28.6, 51.1) 30.4 (20.7, 43.6) 3.1 (0.5, 7.4) 2.8 (0.0, 7.9)
2.5 (0.6, 4.4) 10.4 (4.3, 18.8) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 7.8 (2.9, 16.4) 33.9 (25.2, 42.7) 42.8 (31.9, 53.4) 2.6 (0.4, 5.2)
1.2 (0.0, 3.1) 5.1 (1.1, 11.5) 0.0 (0.0, 0.0) 2.3 (0.0, 7.0) 6.1 (2.2, 12.3) 36.6 (25.8, 52.2) 44.5 (27.4, 57.2) 4.2 (0.0, 11.2)
2.8 (0.9, 5.6) 8.2 (1.9, 16.4) 3.2 (0.0, 8.5) 1.7 (0.0, 5.5) 0.1 (0.0, 0.3) 51.1 (40.2, 65.1) 28.7 (17.8, 39.4) 4.2 (0.3, 8.4)
0.1 (0.0, 0.4) 43.6 (33.1, 53.8) 46.8 (36.5, 56.9) 5.3 (2.5, 10.0) 3.3 (0.9, 7.0) 0.8 (0.0, 2.1)
0.1 (0.0, 0.0) 35.4 (23.6, 51.4) 57.6 (41.6, 69.9) 4.7 (1.5, 9.2) 2.1 (0.0, 5.7) 0.1 (0.0, 0.0)
0.1 (0.0, 0.0) 43.7 (30.8, 54.5) 48.5 (36.6, 61.9) 4.6 (0.5, 9.7) 2.6 (0.2, 6.4) 0.6 (0.0, 2.3)
100.0 (99.9, 100.0) 0.0 (0.0, 0.0) 0.0 (0.0, 0.0) 0.0 (0.0, 0.1)
11.3 (3.0, 13.5) 26.1 (15.4, 44.5) 62.6 (48.8, 76.3) 0.0 (0.0, 0.0)
66.9 (54.6, 77.8) 5.1 (0.5, 11.7) 27.8 (17.9, 39.7) 0.2 (0.0, 0.7)
3.7 (1.7, 6.7) 96.3 (93.3, 98.3)
24.0 (12.9, 38.0) 76.0 (62.0, 87.1)
13.0 (5.2, 22.4) 87.0 (77.6, 94.8)
and disposal programs and pharmacies that sell non-prescription syringes are required to provide information to customers about where they can get HIV testing if they themselves do not offer HIV testing on site (HIV Prevention Planning Council, 2012). Task forces are currently being formed to ensure that these recommendations become implemented as standards of care. In this study, we found that MSM who inject drugs were most likely to engage in behaviors that put them at risk for HIV and have the highest prevalence of HIV, which makes them a priority population for public health efforts to address HIV. Previous studies conducted in San Francisco have suggested that certain PWID populations, like MSM who inject drugs, may be harder to reach than others and are further disconnected from care. For example, the Party and Play Study, which recruited high risk MSM exclusively during late night hours, found that a third of the sample was living with HIV/AIDS, and a quarter of those who were living with HIV were unaware of their status (Pendo, 2003). The results of the Party and Play Study became the impetus for the Late Night Breakfast Buffet (LNBB) Study, which tested the efficacy of providing free harm reduction services to drug-using MSM from 1:00 am to 5:00 am out of a van which stopped in neighborhoods drugusing MSM were likely to frequent (Rose et al., 2006). These services
included syringe access and disposal, HIV testing and counseling, and condom distribution. The goal of the LNBB was to engage MSM who were not utilizing these services at fixed sites during day light hours. The LNBB study provided 2000 syringes, distributed 4500 condoms/lubricants, and provided 21 HIV tests, and 12 STI tests, demonstrating the willingness of MSM who use drugs to utilizing these services. This study provides evidence that some drug-using populations may be more amenable to utilize services that cater to their specific needs. An example of a successful program designed to service a niche population is Ladies’ Night, a harm reductionbased drop-in program that provides services to homeless and marginalized housed women in San Francisco (Magee and Huriaux, 2008). This program is tailored to address the structural and social barriers homeless and marginally housed women often face when seeking healthcare and social services. The results also reveal that MSM who inject drugs have unique drug using behaviors, sexual behaviors, and prevalence of viral infection. This suggests that MSM who inject drugs may operate in different social networks from females and heterosexual males who inject drugs. The choice of drug may illustrate the different social network MSM who inject drugs inhabit as compared to females and heterosexual males who inject drugs. As our
H. Jin et al. / Drug and Alcohol Dependence 145 (2014) 180–184
183
Table 3 RDS-weighted prevalence of viral infections and risk factors by gender and sexual orientation among people who inject drugs (PWID) in San Francisco, CA, 2012. Total
Heterosexual men (%)
HIV 12.7 (7.3, 16.9) 3.7 (1.8, 6.6) Yes No 88.3 (83.0, 92.8) 96.3 (93.4, 98.2) Knew that he/she was infected with HIV among HIV+ 60.6 (38.4, 77.9) 59.4 (28.9, 93.3) Yes 39.4 (22.1, 61.6) 40.6 (6.7, 71.1) No Ever been diagnosed with Hepatitis A 4.2 (2.6, 6.2) 4.8 (2.4, 8.4) Yes No 95.8 (93.8, 97.4) 95.2 (91.6, 97.6) Ever been diagnosed with Hepatitis B 10.3 (6.7, 15.0) 11.6 (5.8, 19.0) Yes 89.7 (85.0, 93.3) 88.4 (81.0, 94.2) No Ever been diagnosed with Hepatitis C 53.9 (46.9, 62.0) 57.3 (46.1, 67.7) Yes No 46.1 (38.0, 53.1) 42.7 (32.3, 53.9) HIV/HCV Co-infection 5.8 (3.0, 9.7) 2.1 (0.9, 3.8) Yes 94.2 (90.3, 97.0) 98.9 (96.2, 99.1) No Received goods or money for sex in the past 12 months 17.8 (13.6, 22.6) 9.3 (5.3, 14.0) Yes No 82.2 (77.4, 86.4) 90.7 (86.0, 94.7) Any unprotected intercourse with his/her last 5 sexual partners in the past 12 months 62.6 (54.9, 69.2) 56.1 (44.1, 67.4) Yes 37.4 (30.8, 45.1) 43.9 (32.6, 55.9) No Always uses a clean needle in the past 12 months 60.3 (52.6, 67.1) 59.1 (49.7, 71.9) Yes 39.7 (32.9, 47.4) 40.9 (28.1, 50.3) No Primary drug Heroin 57.7 (50.2, 67.4) 59.5 (52.7, 74.9) Cocaine 1.0 (0.1, 3.1) 0.1 (0.0, 0.4) Speedball 6.1 (3.2, 9.2) 5.4 (2.1, 9.4) Crack 0.2 (0.0, 0.8) 0.6 (0.0, 1.7) Methamphetamine 31.1 (22.3, 38.7) 26.0 (15.6, 33.5) Other 4.0 (0.7, 9.3) 8.5 (0.3, 17.9)
data reveal, MSM primarily inject methamphetamine. Research has shown that MSM who use methamphetamine to enhance sexual pleasure are more likely to engage in sexual behaviors that are known to increase risk of HIV transmission, including condomless anal intercourse, sex with multiple partners, and frequent sexual activity. The association between methamphetamine use and condomless anal intercourse has been illustrated among HIV negative MSM and MSM living with HIV. We also found that females who inject drugs have an elevated risk of acquiring HIV and/or HCV relative to heterosexual males who inject drugs (Wagner et al., 2013). Studies have demonstrated that females who inject drugs may have an elevated risk for HIV
MSM (%)
Women (%)
P-value
24.0 (13.4, 37.1) 76.0 (62.9, 86.6)
13.0 (5.4, 22.3) 87.0 (77.7, 94.6)
46.6 (17.5, 74.0) 53.4 (26.0, 82.5)
83.8 (55.3, 96.3) 16.2 (3.7, 44.7)
4.8 (1.8, 10.0) 95.2 (90.0, 98.2)
2.7 (0.5, 6.5) 97.3 (93.5, 99.5)
9.5 (3.8, 18.0) 90.5 (82.0, 96.2)
9.2 (3.9, 15.4) 90.8 (84.6, 96.1)
39.1 (29.0, 56.1) 60.9 (43.9, 71.0)
60.2 (49.4, 72.9) 39.8 (27.1, 50.6)
6.3 (2.3, 13.3) 93.7 (86.7, 97.7)
10.8 (3.2, 19.4) 89.2 (80.6, 96.8)
25.4 (17.6, 38.4) 74.7 (61.6, 82.4)
22.0 (13.3, 30.5) 78.0 (69.5, 86.7)
73.7 (60.4, 84.9) 26.3 (15.1, 39.6)
61.0 (49.0, 73.1) 39.0 (26.9, 51.0)
64.8 (50.2, 75.2) 35.2 (24.8, 49.8)
59.0 (46.3, 71.2) 41.0 (29.7, 54.6)
33.7 (20.5, 50.0) 2.7 (0.0, 10.3) 6.1 (1.4, 13.0) 0.0 (0.0, 0.0) 55.3 (38.9, 67.9) 2.1 (0.7, 4.4)
70.0 (56.5, 81.5) 0.6 (0.1, 1.5) 6.4 (1.3, 10.8) 0.0 (0.0, 1.6) 22.9 (12.0, 37.6) 0.0 (0.0, 0.1)
