HIV/AIDS and Injection Drug Use in the Neighborhoods of Dar Es Salaam, Tanzania

Home , Needle sharing

Drug and Alcohol Dependence 82 Suppl. 1 (2006) S23–S27 www.elsevier.com/locate/drugalcdep Short communication HIV/AIDS and injection drug use in the neighborhoods of Dar es Salaam, Tanzania

Sheryl A. McCurdya, °, Michael W. Rossa, Gad Paul Kilonzob, M.T. Leshabarib, Mark L. Williamsa aUniversity of Texas Health Science Center at Houston, School of Public Health, WHO Center for Health Promotion and Prevention Research, Houston, Texas 77030, USA bMuhimbili University College of Health Sciences, University of Dar es Salaam, Dar es Salaam, Tanzania Received 1 May 2005; received in revised form 23 August 2005; accepted 15 October 2005

Abstract This study examines the intersection between needle-sharing practices and HIV recovered from used syringes collected from 73 heroin injection drug users (IDUs) in Dar es Salaam, Tanzania, between October 2003 and January 2004. To extract blood residue, syringes were flushed and 10 microliters of solution mixed with 120 microliters of a latex solution was placed on a Capillus HIV-1/2 slide. Thirty-five (57%) of the useable syringes tested positive for HIV antibodies. Results varied significantly: 90% of syringes tested HIV positive in a mixed-income neighborhood 2 kilometers from the city center; 0% of syringes tested HIV positive in the outlying areas. In addition, semistructured interviews were conducted with 51 IDUs. The interviews were content coded, and codes were collapsed into emergent themes regarding syringe-use practices. Injecting is a recent practice, particularly among heroin users in neighborhoods far from the city center. Sharing syringes has resulted in a high proportion of used syringes containing HIV-positive blood residue. Geographic distance is an indicator of recent adoption of IDU in neighborhoods and correlates strongly with the distribution of syringes containing HIV-positive blood residue. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Heroin; Injection; Needle sharing; HIV risk; Africa; Urban

1. Introduction We conducted this mixed method study in response to the lack of data on HIV transmission by needle- and syringe- Researchers have recently debated the role of both licit and sharing practices among IDUs in Tanzania. Instead of using illicit contaminated injection equipment in the transmission RNA testing to assess HIV antibodies, we used quality- of HIV in Africa (Schmid et al., 2004; Gisselquist et al., assured serological testing, which is the standard of care 2003). Although the significant role of injection drug use for voluntary counseling and testing sites in Tanzania. (IDU) in the transmission of HIV is widely recognized in Geographic distribution of syringe-sharing practices was the Americas, Asia, and Europe (Aceijas et al., 2004; Mc- assessed by analyzing data from semistructured interviews. Coy and Rodriguez, 2005), information is scarce for most of Findings from each component of the study were trian- Africa. To date there are no published data on HIV and IDU gulated to assess consistency of findings and draw out in Africa. This study is the first attempt to provide indicator implications. data of the intersection of IDU with HIV transmission and During 2003, HIV prevalence among antenatal clinic represents a first step in the analysis of the situation. attendees in Tanzania was 9.1%; it was 8.8% among Residual blood in drug injection paraphernalia has been blood donors. Across various Tanzanian regions, the HIV used to estimate the potential for HIV transmission through prevalence for youth between the ages of 15 and 24 who shared use (Shapshak et al., 2000; Shah et al., 1996). provided blood donations ranged from 0% to 19%. In the Shapshak and colleagues (2000) detected antibodies to Dar es Salaam region, HIV prevalence among adolescent HIV-1 polypeptides in 94% of visibly contaminated needles blood donors was between 5% and 10% (National AIDS and syringes collected from shooting galleries in Miami. Control Programme, 2004). Overall, the HIV prevalence * Corresponding author. Tel.: +1 713 500 9633; fax: +1 713 500 9750. rate for those attending national voluntary counseling and E-mail address: [email protected] (S. McCurdy). testing sites was 18.4%.

2. Methods minutes, results were recorded as nonreactive (negative) or reactive (positive). The laboratory that did the tests has been 2.1. Syringe collection procedure carrying out quality assurance and quality control for the Between October 30, 2003, and January 8, 2004, 360 past 9 years. In the more than 100,000 tests done so far, IDUs from 7 neighborhoods in Dar es Salaam, Tanzania, there has been only 1 discrepant result. Quality assurance were recruited to participate in the study using a targeted was clearly implemented and monitored by an external snowball sampling technique (Booth et al., 1993; Watters quality assurance process. and Biernacki, 1989). Neighborhoods targeted for sampling were confirmed by interviews with local key informants 2.3. Semistructured interviews and by direct observation of drug use activities, including Semistructured, face-to-face interviews (n = 82) were con- injection, by an outreach worker. Local key informants ducted with 27 female and 54 male drug users between were interviewed and asked to refer other heroin users February and July 2003. Participants were recruited from to participate in the study. In addition, each respondent separate drug-use networks in different neighborhoods who agreed to participate was asked to bring one used with the assistance of five field workers. Seventy-two syringe. Syringes were tested instead of individuals because interviews were conducted in neighborhoods adjacent to heroin IDUs were not willing to undergo HIV testing and the city center; the rest were conducted in outlying areas. counseling procedures and did not want to know their Among the 81 interviewees, 51 were IDUs (33 men and HIV status. 18 women). Forty-eight of the 50 February interviews were Syringes presented were “children’s syringes”, consisting conducted in a converted office space; the other 2 were of a 2-milliliter syringe with a 23-gauge needle. These conducted in a Land Rover in a single family’s gated syringes are sold legally in Tanzania in sealed blister packs complex. Ten of the 32 July interviews took place in a with the needle unattached. The outreach worker collected rented room in the back of a guest hotel in one of the syringes at the point of initial contact with the respondents more distant neighborhoods. The other 22 July interviews or at the study’s storefront center. Used syringes were coded were conducted with heroin IDUs in the back rooms of with serial identification numbers to document the time, sex two different family compounds near the city center. One of the presenter, and neighborhood. Syringes could not be of those 22 participants was a woman who also had linked to individual study participants. been interviewed in February; she had reported smoking No specific questions were asked regarding the degree of heroin in February but had begun injecting heroin by July. needle or syringe sharing. Unless they inject as part of a Topics covered in the interviews included initiation into group in a shooting gallery, which is a common practice, drug use, current and past drug use, drug-use practices, users do not necessarily know how many others might have social relationships among drug users, and drug-use norms used their syringe/needle set. Individuals who prefer not to in Dar es Salaam. Interviews were audiotape-recorded share needles and syringes often leave them for safekeeping and transcribed verbatim into computer files in Swahili. with the shooting gallery owner so that they can reuse them Interviews were conducted independent of and prior to later. In their absence, however, the shooting gallery owner collecting and testing syringes, and any potential overlap might allow other IDUs to use the needle/syringe set before between the samples cannot be estimated. returning it to the original owner without disclosing the The research project was approved by the Tanzanian additional users. Other IDUs hide their needle/syringe set Commission for Science and Technology (COSTEC), the somewhere on the premises of the shooting gallery but do Tanzanian National Institute for Medical Research (NIMR) not know if another IDU might have found it, used it, and Institutional Review Board (IRB), the College Research returned it to the hiding spot. The study participants’ used and Publications Committee (CRPC) of the Muhimbili syringes were not exchanged for clean needle/syringe sets University College of Health Sciences, and the University because possession of a needle or syringe can subject an of Texas Health Sciences Center at Houston IRB. All of the IDU to arrest. Needles and syringes were sent daily to the participants verbally consented to participate in the study. Muhimbili Health Information Center (MHIC) laboratory Verbal consent allowed us to protect the study participants for testing by trained laboratory technologists. from identification and action by the legal authorities. 2.2. Testing procedure

To obtain residual blood from the used needles and 3. Results syringes, the laboratory technologists ﬂushed them with the phosphate and antimicrobial preservative chase buffer Although 360 participants were recruited, only 294 were accompanying the Determine Rapid HIV Test. The tech- asked to provide used syringes because of a delay in re- nologists used a calibrated pipette with a tip to add ceiving IRB approval. Among the 294 individuals solicited 10 microliters of that solution to 120 microliters of a to contribute needles and syringes, 73 (24.8%) did so. latex reagent on a Capillus HIV-1/2 slide. After 3−7 The response to requests for needle and syringe donation S.A. McCurdy et al. / Drug and Alcohol Dependence 82 Suppl. 1 (2006) S23–S27 S25

Table 1 HIV seroprevalence of blood residues across selected neighborhoods

Suburb n syringes % response Untestable % positive % male collected

Msasani 21 65.6 0 90.4 42.9 Kariakoo 9 14.8 0 71.4 44.4 Kinondoni 10 31.3 0 50.0 10.0 Manzese/Tandale 6 8.6 0 33.3 66.7 Mbezi 7 29.2 0 0 100 Mwananyamala 17 14.3 12 0 40.0 Sinza 3 15.8 0 0 – Total 73 24.8 12 57.4 47.5

fluctuated widely between neighborhoods (Table 1). In syringes with HIV-positive blood residue (90.4%, n = 21). Msasani, a high proportion provided a needle and syringe; Poor and rich youth spent time together in mageto (shooting in Kunduchi and Temeke, no one did. Response rates were galleries [plural]) in Msasani, where almost all of the users better in neighborhoods with larger and well-established appeared healthy and were observed to rinse needles and groups of drug users. In Sinza and Kijitonyama, where syringes with one flush of water. Some of the IDUs lived there had been recent police activity, users were in hiding. with their higher socioeconomic status families and others Among the 73 needles and syringes collected, 12 could not shared a room with other heroin users (e.g., three men be tested because they were old and no blood residue and two women or two women). IDUs were a mix of could be obtained. Among the remaining 61 syringes, standard seven, form-four, and form-six graduates. One 29 were collected from men and 32 from women. Residual held a diploma from a technical institute. One interviewee blood from 57.4% was HIV seropositive. Table 1 shows a bragged that 100% of the Msasani heroin users injected. breakdown of HIV-positive residue by neighborhood. Another interviewee noted that IDU was “the fashion these Semistructured interviews with IDUs confirmed that days”. syringes were shared widely. Of the 45 IDUs who discussed IDUs from Kariakoo, an old Swahili neighborhood a sharing needles, 16 denied that they or other IDUs couple of kilometers from the harbor (Kigamboni Ferry), share syringes “these days”. Eight others denied sharing also had presented syringes with a high rate of HIV-positive syringes themselves, but admitted that others share. Seven blood residue (71.4%, n = 9). Many sex workers and drug respondents acknowledged passing used syringes to others dealers congregate in Kariakoo, historically the African who injected after them, and 14 admitted that people share quarter of the colonial city and now the site of the largest syringes. Four from this last group stated that they had produce and retail market in the city. This neighborhood shared with others in groups of three or four. A fifth person is where heroin first began to be sold in volume in Dar es noted that as many as 10 people would share 1 syringe. Salaam. Shared syringes were almost always rinsed once with water Kinondoni (rate 50%, n = 10), formerly a residential area before reuse. for middle-level government employees, is now consid- Sexual contact between members of a geto (shooting ered a trendy neighborhood. Its population is of mixed gallery) was common, and reportedly most women even- socioeconomic status, and the area has become a center tually have sex with men in the injection group over the for cultural innovations that spread to other neighborhoods. course of a year (McCurdy et al., 2005). All but one of the Kinondoni is densely populated, with lots of on-premises female IDUs were engaged in sex work. alcoholic beverage outlets, nighttime entertainment places, Most individuals reported that they primarily used and sex workers. Interviewees claimed that about 75−80% shooting galleries that were in close proximity to their of the sex workers working in Kinondoni inject heroin. residence for injecting. If, however, the supply of heroin in Some sex workers meet their clients by traveling to their neighborhood was low or reportedly adulterated, they Kinondoni and other areas in the downtown area near would move to other neighborhoods to obtain drugs. Many upscale bars and discos. Outreach workers reported that female IDUs either injected together in their shared rooms many IDUs in Kinondoni look very sick. or in the shooting gallery. Mwananymala, contiguous to and west of Kinondoni, houses a mixture of lower-income government employees 3.1. Neighborhood characteristics and squatter settlements. This predominantly Muslim neighborhood is poorer than Kinondoni. Manzese and Tandale are Msasani, a former fishing village that is now a mixed unplanned, informal settlement areas established during the socioeconomic community, had the greatest proportion of 1970s and 1980s. They are populated by a mixture of people S26 S.A. McCurdy et al. / Drug and Alcohol Dependence 82 Suppl. 1 (2006) S23–S27 either unemployed or working in small-scale businesses and might have biased the data toward higher levels of HIV- the informal sector. positive blood residue in those neighborhoods. Semistruc- Mbezi and Sinza are the two neighborhoods where tured interviews also were conducted with a convenience none of the collected syringes contained HIV-positive sample of IDUs, although these were distributed evenly blood residue. These neighborhoods are located much across the neighborhoods sampled. The data might not be farther away from the port. Sinza is north of the city, generalizable throughout Dar es Salaam. adjacent to Mwananymala, and was originally settled by Despite these limitations, this study has three major mid-level government workers and employees of business findings. First, injecting heroin is a comparatively new corporations. Although predominantly a residential area, phenomenon in Dar es Salaam that has spread relatively Sinza is developing businesses along the main road through rapidly from Kariakoo. Second, there are high levels the suburb. Mbezi, a less congested commuter suburb, is of HIV antibodies in the residual blood from syringes more than 20 kilometers north of the city center. collected in several of the main neighborhoods in Dar Temeke is an older, more established neighborhood es Salaam, and a high potential for HIV transmission southwest of the city center; its residents are middle class, to other neighborhoods. If syringe sharing, which occurs working class, informal sector laborers, or unemployed. primarily in the shooting galleries, is the mechanism of Drug users in this community are primarily male. Kunduchi, HIV transmission, it might explain the variation of HIV- north of the city, is an area with beach resorts catering to positive residue (0% to 90%) by neighborhoods. Finally, expatriates. IDUs from both of these areas were concerned seroprevalence appears to reflect the geographic distribution about police activity and refused to provide syringes. of IDU as a practice. Seroprevalence does not appear to be related to the socioeconomic status or religious practices of 3.2. Geographical distribution of HIV seroprevalence a neighborhood, because syringes from neighborhoods with mixed socioeconomic, predominantly Muslim, and mixed Data from the semistructured interviews indicate that the Christian and Muslim populations had both the highest and practice of injecting heroin was first introduced between lowest prevalences of HIV-positive blood residue. 1998 and 1999 in Kariakoo. For two decades before that, These data suggest that HIV is being transmitted by most East Africans had only smoked, sniffed, or inhaled heroin IDUs’ needle- and syringe-sharing practices and heroin. IDU spread from Kariakoo to Kinondoni and creates the possibility of a new wave of HIV transmission Magomeni (Fig. 1). It then spread almost simultaneously in Dar es Salaam, if not Tanzania. The data also suggest to Manzese, Msasani, and Temeke, the next set of suburbs the possible establishment of an HIV “core” group that adjacent to Kariakoo. The interviews suggest the spread of might be resistant to prevention efforts aimed at the general IDU in concentric waves from heroin’s original entry point. population. Clearly, unsafe injection practices in East Africa need to be more fully investigated.

Acknowledgements This research is based on an ongoing collaboration between researchers at the University of Texas Health Science Center at Houston and Muhimbili University College of Health Sciences and was supported by a supplement to grant number R01 DA14519 from the National Institute on Drug Abuse, National Institutes of Health, USA. The opinions expressed in the manuscript are solely those of the authors. We thank Stella Mujaya, Samueli Kihore, Mary Anderson Mbwambo, and Brown Ibrahim for their assistance in the data collection and management of this project; Patrick Courtney for his assistance with generating the map; and Dr. Jessie Mbwambo for facilitating communication Fig. 1. Map of neighborhoods in Dar es Salaam, Tanzania. between the investigators.

4. Discussion References

These data are limited by the technique used to collect Aceijas C, Stimson G.V., Hickman M., Rhodes T., 2004. United Nations Reference Group of HIV/AIDS Prevention and Care among IDU in needles and syringes and by the low response rate to our Developing and Transitional Countries. Global overview of injecting requests for syringe donation. Data appear to be biased drug use and HIV infection among injecting drug users. AIDS 18, toward areas with better established groups of IDUs, which 2295–2303. S.A. McCurdy et al. / Drug and Alcohol Dependence 82 Suppl. 1 (2006) S23–S27 S27

Booth R., Watters J., Chitwood D., 1993. HIV risk-related sex behaviors P.D., Hankins C., Zaba B., Heimer R., Boerma J.T., 2004. Transmission among injection drug users, crack smokers, and injection drug users of HIV-1 infection in sub-Saharan Africa and effect of elimination of who smoke crack. Am. J. Public Health 83, 1144–1148. unsafe injections. Lancet 363, 482–488. Gisselquist D., Rothenberg R., Potterat J., Drucker E., 2003. HIV infections Shah S., Shapshak P., McCoy C., Rivers J., Weatherby N., Xin K., Stewart in Africa not explained by sexual or vertical transmission. Int. J. STD R., Chitwood D., Page J., Mash D., 1996. Detection of HIV-1 DNA AIDS 13, 657–666. in needle/syringes paraphernalia, and washes from shooting galleries in McCoy C.B., Rodriguez F., 2005. Global overview of injecting drug use Miami: preliminary laboratory studies. J. Acquir. Immune Deﬁc. Syndr. and HIV infection. Lancet 365, 1008–1009. Hum. Retrovirol. 11, 1301–1306. McCurdy S.A., Williams M.L., Kilonzo G.P., Ross M.W., Leshabari M.T., Shapshak, P. Fujimura R.K., Page J.B., Segal D., Rivers J.E., Yang J., Shah 2005. Heroin and HIV risk in Dar es Salaam, Tanzania: youth hangouts S.M., Graham G., Metsch L., Weatherby N., Chitwood D.D., McCoy mageto, and injecting practices. AIDS Care 17, S65–76. C.B., 2000. HIV-1 RNA load in needles/syringes from shooting galleries National AIDS Control Programme, 2004. HIV/AIDS/STI Surveillance in Miami: a preliminary laboratory report. Drug Alcohol Depend. 58, Report No. 18, January to December, 2003. United Republic of Tanzania: 153–157. Ministry of Health, Dar es Salaam. Watters J., Biernacki P., 1989. Targeted sampling: options for the study of Schmid G.P., Buve A., Mugyenyi P., Garnett G.P., Hayes R.J., Williams hidden populations. Soc Problems 36, 416–430. B.G., Calleja J.G., De Cock K.M., Whitworth J.A., Kapiga S.H., Ghys