|Home | About | Journals | Submit | Contact Us | Français|
Background: Drug-related infectious diseases are among the major health consequences of drug use, and any existing drug-related infection may predispose injecting drug users (IDUs) to other infections. Methods: We assessed among IDUs in Budapest, Hungary the prevalence of and vulnerability to selected drug-related infections and co-infections. The sample consisted of 186 participants recruited between October 2005 and December 2006. Results: We found 0% HIV, 37% HCV, 24% HAV, and 14% past HBV infection. Infections with Herpes 1 or 2, tuberculosis, Chlamydia, syphilis, and gonorrhoea were 79%, 12%, 7%, 4%, and 0%, respectively. Co-infection with HAV/HCV was 12%, HBV/HCV 9%, HAV/HBV 7%, and HAV/HBV/HCV 4%. Those over age 30, the ethnic Roma, and the homeless were more likely to have any hepatitis and a higher number of drug-related infections. Amphetamine injectors were more likely to have a higher number of drug-related infections and those who travelled within Hungary were more likely to have any STI. However, those who worked at least part time and those who were in treatment were less likely to have drug-related infections. Conclusions: These results highlight the need of interventions in Hungary to reach and focus on marginalized (Roma or homeless) IDUs and address not only injecting and sex risk, but also hygienic living and injecting conditions. Furthermore, structural interventions to increase social integration (working or being in treatment) may improve welfare and decrease drug use and infection risk tied to drug use/injection among disadvantaged, marginalized, mostly minority populations.
Drug-related infectious diseases are among the major health consequences of drug use,1 and any existing drug-related infection may predispose injecting drug users (IDUs) to other infections. For example, STIs, such as Chlamydia, syphilis, and herpes (with type 2 (HSV-2) related to most of prevalent genital herpes cases but type 1 (HSV-1) accounting for about half of new cases in developed countries2) may contribute to increased susceptibility to sexually transmitted HIV infection.3,4 Other STIs, such as gonorrhoea, may be markers for sexual risk for HIV infections. Furthermore, HIV infection is the strongest risk factor for progression from tuberculosis infection to clinical tuberculosis.5 In addition, infections with more than one type of Hepatitis infection (hepatitis co-infections), regardless of the mode of transmission (oral-faecal, sexual, blood-to-blood), result in a disease severity and mortality that are much worse than hepatitis A (HAV), hepatitis B (HBV), or hepatitis C (HCV) alone.6–9
Certain subgroups of drug injectors may be especially vulnerable for drug-related infections. Such IDU subgroups may be identified by socio-demographic characteristics, and include socially, and/or culturally marginalized populations, such as racial/ethnic or national minorities, the homeless, and women.10,11 For example, compared to other IDUs, minority IDUs, especially minority women IDUs, have higher prevalence of HIV, HBV, and HCV infections and co-infections,10,12 and homeless IDUs have higher prevalence of infections that are related not only to drug injecting, but also to housing conditions and hygiene.13 Research in Hungary indicates that ethnic minority Roma IDUs may have exceptionally high levels of hepatitis infections and co-infections.14
The prevalence and incidence of HIV in Hungary and other Central European countries have been low.15–17 Hungary had 1485 diagnosed cases of HIV infection at the end of 2007, and the estimated prevalence of HIV among IDUs was <1%.17–19 The prevalence of HCV among Hungarian IDUs, though, have increased from about 15% to about 30% in the decade following the 1990s.1,20,21 However, other than HIV and HCV, little is known about the prevalence of drug-related infectious diseases, including sexually transmitted infections (STIs), tuberculosis, other hepatitis infections and hepatitis co-infections among IDUs in Hungary, and about patterns of vulnerability to drug-related infections.
The aim of this analysis was to assess, among a sample IDUs in Budapest, Hungary, the prevalence of selected drug-related infections and co-infections that are parenterally or sexually transmitted (HIV, HBV, HCV, syphilis, herpes, Chlamydia, and gonorrhoea infections), or related to hygiene and living conditions (HAV and tuberculosis). Another aim was to examine associations of these infections with selected socio-demographic and related characteristics. Specifying these associations may assist not only in enhancing our understanding of the distribution of these infections among sub-groups of IDUs and the pattern of vulnerability to these infections, but also in developing appropriate interventions targeting specific high-risk, vulnerable drug injector groups.
Between October 2005 and December 2006, IDUs were recruited in Budapest, Hungary from non-treatment settings using a combination of street outreach and chain referral methods.22 Those who self-reported injecting drugs in the past 30 days were eligible to participate. Self-report of drug use was verified by urine metabolite tests for those who reported injecting drugs in the past 2–3 days. Injecting route of drug administration was verified by inspection of injecting marks. Participants were paid HUF 2000 (about EUR 8) for participation and HUF 500 (about EUR 2) for bringing in other IDUs who were eligible to participate in the study. After providing their informed consent, eligible participants were administered a structured survey interview that took about 2 h to complete. The Institutional Review Boards at the National Development and Research Institutes, Inc. in New York, USA and the Hungarian Academy of Sciences in Budapest, Hungary approved all human subject procedures.
Participants also received counselling about disease transmission due to unsafe infecting and unprotected sex, and, if consented, provided blood and urine samples for biological testing. Blood samples were tested for HIV, HAV, HBV, HCV, tuberculosis, herpes, and syphilis infections, and urine samples were tested for Chlamydia and gonorrhoea infections. We used Abbott ELISA test confirmed by western blot for HIV antibody screening and Abbott Hepatitis B virus CORZYME immunoassay for HBV core antibody (anti-HBc), HBV surface antibody (anti-HBs), and HBV surface antigen (HbsAg) testing. Past infection with HBV was identified with positive tests for either HbsAg or anti-HBc, acute infection was identified with negative tests for anti-HBs and positive tests for HbsAg, and being immunized against HBV was identified with a positive test for anti-HBs and negative tests for both HbsAg and anti-HBc.23 Abbott HCV EIA 3.0 procedure for encoded antigens (recombinant c100-3, HC-31, and HC-34) confirmed by RIBA were used for HCV antibody screening. The presence of past HAV infection (IgG antibodies) was detected by the use of ELISA Biotest. Currently active tuberculosis infection was detected by TB-Check-1 rapid test (VEDA.LAB Cedex, France). Herpes infection was detected by Herpes Simplex 1 + 2 IgG ELISA NOVATEC (note: this test does not distinguish between HSV-1 and HSV-2). Syphilis antibodies were tested with a non-specific trepanomal antibody test (Venereal Disease Research Laboratory or VDRL) and confirmed with a specific trepanomal test (Trepanoma pallidum haemagglutination essay or TPHA). Chlamydia trachomatis and Neisseria gonorrhoeae were detected using ligase chain reaction (LCR).
Independent variables in this analysis are selected variables that were chosen to reflect patterns of vulnerability: sociodemographics, geographic mobility, and access to prevention or harm reduction services. Selected sociodemographic variables are: age (30 and under vs. above 30), gender, ethnicity (Roma vs. non-Roma), having at least a high school degree, homelessness, income (HUF 100 000 [approximately EUR 400] per month or more vs. less than HUF 100 000 per month—based on median split), and working at least part time. Selected geographic mobility variables are: travelling outside Budapest and within Hungary in the past 12 months and travelling abroad in the past 12 months. Access to prevention or harm reduction services variables are: currently being in drug treatment, and currently using a needle exchange.
Two binary composite dependent variables were created: any STIs (syphilis, Chlamydia, or gonorrhoea infections) and any hepatitis infection (HAV, HBV, or HCV infection). In addition, as an indicator to assess vulnerability to drug-related infections overall, a four-category composite dependent variable was also created by adding up the total number of infections participants have24 (of HIV, HAV, HBcAb, HCV, tuberculosis, syphilis, Chlamydia, and gonorrhoea; indeterminate test results were counted as not infected), and categorized, based on the distribution of this variable, as 0 (not infected with any of the tested infections), 1, 2, and 3 and more infections. Note: we excluded herpes from the composite variables because the test did not differentiate between oral and genital herpes.
Of the 215 participants that were interviewed, 186 (86.5%) provided biological samples. Those with at least a high school education and those who self-reported injecting amphetamine only in the past 30 days (vs. injecting heroin, injecting both heroin and amphetamine and injecting drugs other than heroin and amphetamine) were significantly less likely to provide biological samples. Gender, age, years since first injecting, daily heroin injecting, daily amphetamine injecting, Roma ethnicity, homelessness, income, work status, geographic mobility, being in treatment, and using a needle exchange were not associated with getting tested. This analysis is based on a sample of 186 participants who provided biological samples.
In univariate analysis, chi-square tests were used to assess statistically significant associations (P < 0.05). Multivariate logistic regression for the binary composite variables and multivariate multinomial regression for the four-category composite variable were used to test associations between the dependent and the independent variables and to calculate adjusted odds ratios (aOR) with corresponding 95% confidence intervals. Variables significantly associated with getting tested [at least high school education and type of injected drug (injecting amphetamine only, injecting heroin only, injecting both heroin and amphetamine and injecting only drugs other than heroin and amphetamine)] were included in the multivariate models as explanatory or control variables. Hosmer Lemeshow goodness of fit was used to evaluate the fitness of the logistic regression models and the likelihood ratio test was used to evaluate the fitness of the multinomial regression model (all models had good fit). SAS v9.1 (SAS Institute Cary, NC) was used for data management and data analysis.
The characteristics of the sample can be found in Table 1 and the prevalence of infections can be found in Table 2. In univariate analysis for individual infections and co-infections, female gender was significantly associated with syphilis infections; older age with HCV infection; Roma ethnicity with HAV, HCV, HSV infections (note: all Roma had HSV) and with HAV/HBV, HAV/HCV, and HAV/HBV/HCV co-infections (data not shown in table). Having completed at least high school was inversely associated with HBV and HCV infections and with HBV/HCV co-infections. Homelessness was associated with HAV, HBV and HSV infections, and HBV/HCV co-infections. Travelling outside Budapest was associated with Chlamydia infection, and needle exchange use with tuberculosis infection. Injecting amphetamine only was associated with HAV infection and HAV/HCV co-infection. Income level, work status, travelling abroad, and being in treatment were not associated with any of the tested infections.
In univariate analysis of the combined infection variables, older age, Roma ethnicity, not having at least high-school education, and homelessness were associated with having any hepatitis and a higher number of infections (Table 3). Not working at least part time was associated with having any hepatitis infection, and travelling outside Budapest was associated with having any STI. In multivariate logistic regression, when controlled for the other independent variables, travelling outside of Budapest within Hungary was associated with having any STI (Table 4). Older age, Roma ethnicity, and being homeless were associated with having any hepatitis infection. Multivariate multinomial analysis assessing correlates of the number of drug-related infections showed that, when controlled for the other independent variables, those who were older than 30 were significantly more likely than those who were 30 or younger to have two infections. Furthermore, Roma ethnicity was significantly associated with having any (one or more) infections, and those who reported being homeless were significantly more likely than the non-homeless to have at least two infections. In addition, those who worked at least part time were less likely to have one infection and three or more infections, and those who were in treatment were less likely to have any infections. Finally, those who injected amphetamines (but no heroin) were more likely to have three or more infections than those who injected heroin only, both heroin and amphetamines, or other drugs. There was a marginally significant (P = 0.09) indication of elevated risk for three or more infections for females.
The results of this analysis point towards the potential role of social marginalization and social integration.25 Our findings that infections and co-infections, especially Hepatitis infections and co-infections, were more common among Roma participants, and that all Roma participants were infected with HSV 1 or 2, are of great concern. The higher infection and co-infection rates among the Roma in this sample suggest that Roma IDUs may be more likely than non-Roma IDUs to engage in unsafe injecting practices and/or have isolated, highly interacting injecting risk networks in which drug-related infections can spread rapidly.26 The higher rates of HAV among the Roma in our sample may indicate unhygienic living conditions,27 and thus the risk environment among Roma IDUs may be characterized by combined risk: unhygienic living conditions may contribute to unhygienic injecting conditions.14 Similarly, the higher prevalence of two or more infections, especially HAV and HBV, among the homeless, another marginalized population, may also indicate a combination of unhygienic living and injecting conditions, more exposure to other high-risk IDUs and possibly sexual risk and/or a combination of injecting and sexual risk.28 Part time work and being in treatment may be indicators of at least some level of social integration.25 Those who work may have economic resources that allow them to control their drug use, or they may wish to control their drug use so that they can continue to work.25 They are also more likely to be in social contact with non-drug users who may provide normative influence that leads to the moderation of drug-related risk among IDUs.26
The higher prevalence of syphilis among female IDUs in this sample may be a marker for sex work.29,30 This suggests that prevention interventions should counsel female IDUs about using condoms with sex partners, especially with commercial sex partners. The higher prevalence of STIs among participants with geographic mobility suggests that those who travel may have sex with different partners in different places, so they may have a greater probability of getting exposed to STIs. In addition, the significant risk for three or more infections among injectors of amphetamine only and the marginally significant indication of elevated risk for three or more infections among females suggests that these groups may be characterized by multiplicity of risk:31,32 they may be more likely than others to engage in a combination of higher-risk (and maybe higher-frequency) drug injecting and sexual activities.
One limitation of this analysis is that we did not test to assess markers for HAV immunization. Thus we may have included vaccinated participants as HAV negative in the HAV univariate analysis, thereby biasing the associations towards the null. In addition, the sample is not a random sample of IDUs and thereby may not be representative of IDUs in Budapest. Another limitation is that sex work was probably under-reported in this sample because only one participant reported having engaged in sex work, so we were unable to include a sex work variable in the analysis. Another caveat is the difficulty to assess the role and/or location of infection (oral vs. genital) of HSV 1 and 2 infections using blood tests, given that blood tests do not provide information about the location of the infection and that both HSV1 and HSV2 can cause both oral and genital herpes.2 Adding HSV to either the STD combined variable or the number of infections combined variable would have introduced bias because positive test results for oral HSV infections would have biased associations towards the null. However, in a post hoc analysis of this sample we found that over half of participants reported receiving or performing unprotected oral sex, suggesting a sizeable possible overlap between the genital transmission of ‘oral’ herpes and the oral transmission of ‘genital’ herpes.
Our findings have several implications. Geographically mobile, Roma, and homeless populations may form bridges of infection within other traveller, Roma or homeless populations in given country or in other countries. Prevention programs aimed at the general population may not be able to reach at-risk individuals who are members of socially marginalized and often geographically isolated communities.33 Thus, given the isolation of socially marginalized populations such as the Roma and the homeless, HIV prevention, and other health promotion programs may need to specifically target these groups and the areas where these groups live or congregate. The high rates of co-infections with drug-related infectious diseases and the absence of HIV in this sample suggest that there is an urgent need to address among IDUs, especially among marginalized Roma and homeless IDUs, amphetamine users, and females, injecting risk and sex risk, as well as hygienic living and drug injecting conditions, while rates of HIV are still low.34,35 Testing and counselling services among IDUs and other socially marginalized populations should also incorporate testing and counselling for STIs and bloodborne infections, as well as HAV. In addition, health care services may routinely recommend HAV and HBV vaccination to their IDU and/or marginalized clients in order to decrease the high rates of infection and co-infections with these pathogens.36 Finally, drug treatment for addiction and structural interventions that may lead to better educational and employment opportunities and access may result in decreased marginalization and increased social integration, which may improve welfare and decrease drug use and infection risk tied to drug use/injection among disadvantaged, marginalized, mostly minority populations.
National Institute on Drug Abuse (5R01DA014515-02S1 ‘Network Mixing and HIV Risk among Young IDUs in Budapest’); Ruth L. Kirschstein award, Drug Dependence Epidemiology Training Program, National Institute on Drug Abuse (T32 DA007292).
Conflicts of interest: None declared.
We would like to thank Gergő Eperjesi and all research and support staff, Korányi Hospital, Dzsumbuj Help, and all our interviewees.