UBC Faculty Research and Publications

Sex-for-Crack exchanges: associations with risky sexual and drug use niches in an urban Canadian city Duff, Putu; Tyndall, Mark; Buxton, Jane; Zhang, Ruth; Kerr, Thomas; Shannon, Kate Nov 15, 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12954_2013_Article_298.pdf [ 205.57kB ]
JSON: 52383-1.0223647.json
JSON-LD: 52383-1.0223647-ld.json
RDF/XML (Pretty): 52383-1.0223647-rdf.xml
RDF/JSON: 52383-1.0223647-rdf.json
Turtle: 52383-1.0223647-turtle.txt
N-Triples: 52383-1.0223647-rdf-ntriples.txt
Original Record: 52383-1.0223647-source.json
Full Text

Full Text

RESEARCH Open AccessSex-for-Crack exchanges: associations with riskysexual and drug use niches in an urbanCanadian cityPutu Duff1,2, Mark Tyndall3, Jane Buxton2,4, Ruth Zhang1, Thomas Kerr1 and Kate Shannon1,2,5*AbstractBackground: While crack cocaine has been associated with elevated sexual risks and transmission of HIV/STIs,particularly in the context of street-based sex work, few empirical studies have examined correlates of direct sex-for-crack exchanges. This study longitudinally examined the correlates of sex-for-crack exchanges and associated effectson sexual risk outcomes among street-based female sex workers (SW) who use drugs in Vancouver, Canada.Methods: Data were drawn from a prospective cohort of street-based SWs (2006–2008), restricted to those whosmoke crack cocaine. Multivariable generalized estimating equations (GEE) were employed to examine the correlates ofexchanging sex for crack. A confounding model using GEE quasi-Poisson regression modeled the independent effectof exchanging sex for crack on number of clients/week.Results: Of 206 SWs, 101 (49%) reported sex-for-crack exchanges over 18 months of follow-up. In multivariable GEEanalyses, sharing a crack pipe with a client (aOR = 1.98; 95%CI: 1.27-3.08) and smoking crack in a group of strangers(e.g., in an alley or crackhouse) (aOR = 1.70; 95% CI: 1.13-2.58) were independently correlated with sex-for-crackexchanges. In our confounding model, exchanging sex for crack (aIRR = 1.34; 95% CI: 1.07-1.69) remained significantlyassociated with servicing a greater number (>10) of clients/week.Conclusions: These findings reveal elevated sexual- and drug- risk patterns among those who exchange sex for crack.The physical and social environment featured prominently in our results as a driver of sex-for-crack exchanges,highlighting the need for gender-sensitive multilevel approaches to harm reduction, STI and HIV prevention thataddress SWs’ environment, individual level factors, and the interplay between them.Keywords: Crack cocaine, Sex work, Sexual risk, HIV, STIIntroductionThe advent of widespread use of crack cocaine in NorthAmerica in the 1980s and 1990s has been directly linkedto elevated rates for sexually transmitted infections(STIs), including HIV transmission [1-3], through in-creased sexual risk pathways (e.g. higher number of sex-ual partners and unprotected sex) [4,5]. Crack cocaineuse has been documented as a predictor for both HIVand HCV, even after adjusting for known confounderssuch as injection drug use, suggesting a non-parenteralrisk pathway [1,6]. For example, a Canadian study foundthat crack smoking was associated with a 4.01 increasedodds of HIV acquisition, while an American study re-ported a 3.87 increased odds of acquiring HCV amongparticipants who used crack [1,6]. In addition to the sex-ual and drug risk pathways, the use of non-injectioncrack cocaine has been linked to an array of adversephysical and mental health outcomes, including elevatedindividual and community-level violence and physicalhealth harms, such as oral sores and pulmonary compli-cations [5,7,8].While the population prevalence of crack use variesacross settings, a growing number of studies in high-income settings have suggested that street-involvedwomen’s crack use may exceed men’s [9-11]. For example,* Correspondence: kshannon@cfenet.ubc.ca1British Columbia Centre for Excellence in HIV/AIDS, St. Paul’s Hospital,608-1081 Burrard Street, Vancouver, BC V6Z 1Y6, CANADA2School of Population and Public Health, University of British Columbia, 2206East Mall, Vancouver, BC V6T 1Z3, CANADAFull list of author information is available at the end of the article© 2013 Duff et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Duff et al. Harm Reduction Journal 2013, 10:29http://www.harmreductionjournal.com/content/10/1/29among a sample of treatment-seeking individuals who usedrugs, Lejuez and colleagues reported that 84.5% ofwomen reported crack use compared to 63.6% of men. InFlorida, a study among drug-using inmates found that74% of females reported crack as their primary drug, com-pared to 49% of men [12]. Similar trends have also beenreported in the Vancouver context; the reported preva-lence of daily crack use among street-involved women wasfound to be 9.7% compared to 5.6% among men [1].Though it is unclear whether the higher prevalence amongstreet-involved women is related to sex work, studies else-where have noted high levels of crack cocaine among sexworkers [5,13]. Qualitative researchers in our settings havehighlighted a need to better understand the contextualfactors that drive crack use and associated outcomes [11].STIs and other health risks posed by crack use are exacer-bated by the environmental and social contexts such asinterpersonal-violence, poverty, homelessness, incarcer-ation and stigmatization that street-based SWs often con-tend with [13]. For example, street-based SWs who useillicit drugs, including crack, have reported limited accessto health and harm reduction services due to avoidance ofpolicing and client violence [14].Despite growing evidence linking street-based sexwork and crack use [15,16], as well as the numerousharms associated with using the drug, the epidemio-logical understanding of the individual and contextualcorrelates of “sex-for-crack exchanges” and associatedsexual risk outcomes among street-based SWs is limited.While existing STI and HIV research have focused pri-marily on drug-related harms [17], and to a lesser extentsexual-related risks, few have explicitly examined risksand outcomes among street-based SWs who exchangesex for crack. Furthermore, most existing epidemio-logical studies related to crack use among women havefocused on individual-level factors, such health behav-iors, drug use, ethnicity, gender and age. While thesecharacteristics are important, they do not fully capturefactors more distal to the individual such as the largercontext of crack use, including sex-for-crack exchanges,gender-relations, interpersonal violence, as well as otherenvironmental-level factors that drive and shape crack use[11]. This study therefore sought to longitudinally exam-ine the individual-, interpersonal- and environmental-levelcorrelates and outcomes of exchanging sex-for-crackamong a population of street-based SWs in Vancouver,Canada.MethodsThis study was a secondary analysis drawn on data froma community-based prospective cohort, partnered withlocal sex work and community service agencies, and hasbeen described in detail previously [18]. Briefly, between2006–2008, 252 women (inclusive of transgenderedindividuals) engaged in street-based sex work were re-cruited through outreach and participated in an in-formed consent process. The response rate of SWscontacted for interview was 94%. Time-location sam-pling and mapping by peer research team (current/former sex workers) were used to identify sex workspaces for targeted outreach and recruitment. Eligibilityfor the study included being female (inclusive of trans-gender male-to-female), 14 years of age or older, havingexchanged sex for money or resources in the last 30 days,and have used illicit drugs (other than cannabis). Base-line and 6-monthly follow-up surveys were completedby participants, and consisted of interview-administeredquestionnaires by a peer researcher (current/former street-based sex worker), nurse-administered pre-test counselingquestionnaire, and HIV screening. A $25 honoraria wasprovided to respondents at each 6-monthly visit as com-pensation for their time and expertise. Ethical approvalwas provided by UBC/Providence Health ethics reviewboard.Main outcome measureThe primary outcome for this study was havingresponded ‘yes’ to having ‘exchanged sex directly foryour next rock (crack cocaine)’ in the previous 6 months.Based on existing literature and a priori knowledge ofsexual risks of crack smokers, we developed a separateconfounding model to examine the independent effectof “exchanging sex directly for crack” (main explanatoryvariable) on number of clients/week (continuous).Explanatory variablesBased on the literature, individual, interpersonal/socialenvironment and physical environmental factors wereselected as explanatory variables in our analyses. Individ-ual factors included socio-demographic factors (e.g., age,education) and drug use patterns (any use of cocaine,heroin, crack cocaine, crystal methamphetamine in thelast 6 months). In Vancouver, people of Aboriginal an-cestry are overrepresented in street-based sex work [19],and are disproportionately affected by socioeconomic in-equities such as poverty, homelessness and substanceuse [20,21]. Given the overrepresentation of individualsof Aboriginal ancestry (inclusive of First Nations, Metis,Inuit ancestry) in street-based SW and drug use popula-tions in Canada, we adjusted for Aboriginal ancestry vs.non-Aboriginal ancestry (Caucasian). Due to the limitednumber of participants from other ethnic backgroundsamong our sample, we did not adjust/stratify by anyother ethnic groups. Given high rates of daily crack co-caine use among SWs, we stratified crack cocaine smok-ing at the mean by intensive (≥10 rocks/day) vs. lessintensive use (<10 rocks/day) [18]. Interpersonal and so-cial environmental variables considered in our analysesDuff et al. Harm Reduction Journal 2013, 10:29 Page 2 of 8http://www.harmreductionjournal.com/content/10/1/29included: using drugs with a regular client, borrowingused rigs, smoking crack cocaine in a group of strangers(e.g., in crack houses or alleys), sharing drugs with cli-ents, servicing a higher number of clients (stratified atthe mean, >10), inconsistent condom use for vaginal sexby clients (defined as ‘always’ versus ‘usually’, ‘sometimes’,‘occasionally’ or ‘never’, client perpetrated violence (baddates), and physical or sexual violence by intimate part-ners. Economic dependence on one’s partner (defined as“having a partner who scores drugs for you”, and having aregular partner that supports you financially) was alsoconsidered as interpersonal risk factors for our analyses.Physical environmental factors considered were: smokingcrack in public spaces; homelessness (defined as sleepingon the street); police affecting where you get drugequipment; servicing clients in outdoor/indoor spaces;and working in main/commercial areas or side-streets/alleyways/industrial areas.Statistical analysisThe sample was restricted to 206 SWs who smokedcrack in the last 6 months and completed baseline andat least one follow-up visit. As it was an open cohortstudy with staggered enrolment, all participants had atleast 6 months of observation, with a few having 12 and18 months of observation (for a maximum of 4 timepoints). Descriptive statistics, including frequencies, pro-portions, medians, interquartile ranges [IQR] were providedfor baseline individual, interpersonal and environmentalfactors and were stratified by whether or not the participanthad exchanged sex for crack within the past 6 months.Baseline and follow-up data capturing socio-demographiccharacteristics (e.g., ethnicity, education) were treated asfixed covariates, and all other variables (e.g., age, housingand drug use status) were treated as time-updated co-variates of occurrences within the past six months ofthe interview. Bivariate and multivariable logistic re-gression using Generalized Estimating Equations (GEEs)were conducted and included information from each par-ticipant’s baseline and follow-up questionnaires. We usedgeneralized estimating equations (GEE) with a logit linkfor our binary outcome to take into account correlationsarising from repeated measures on the same individualsover the follow-up period (this also accounted for varyingobservations lengths between participants). Standard er-rors adjusted by repeated observations per person wereobtained using an exchangeable correlation structure.Missing data and intermittent data were handled usingthe GEE estimating mechanism, which draws on datafrom non-missing pairs for the estimators of its workingcorrelation matrix. Variables significantly associated withexchanging sex for crack in the bivariate analyses at thep < 0.10 level were subsequently fitted in a multivariableGEE model. Two-sided p-values, bivariate and adjustedodds ratios (OR and aOR) with 95% confidence intervals(95% CI) were reported.To determine independent associations with exchan-ging sex for crack, bivariate screening of a priori and hy-pothesized confounders was conducted, with variablesassociated with exchanging sex for crack at p < 0.10 con-sidered for inclusion in the multivariable explanatorymodel. Akaike Information Criteria (AIC) selection wasused to arrive at the final multivariable model. The finalmodel was assessed for multicollinearity. To assess if ex-changing sex for crack was independently associatedwith number of clients per week, a confounding modelwas constructed using an approach described by Rothmanand Greenland [22]. Confounders were chosen based ona priori knowledge of associations with sex-for-crack-exchanges and number of commercial partners. Thesepotential confounders underwent bivariate screening,and those that retained significance at p < 0.10 wereconsidered potential confounders and were includedin the multivariable confounding model. As in previ-ous studies [23,24], all potential confounders were in-cluded in a full model, and subjected to a manual stepwiseapproach, where variables that altered the association ofinterest by less than 10% were systematically removedfrom the model. As in a previous analysis [24], age wasforced into the multivariable confounding model and notsubjected to the manual stepwise approach due to thewell-established confounding effects of this variable. SASstatistical software package version 9.2 was used for alldata analyses (SAS Institute, Cary, NC, USA).ResultsOf a total of 252 participants enrolled in our open pro-spective cohort between 2006–2008, 206 (82%) had re-ported smoking crack within the follow-up period. As inTable 1, 101 (49%) reported exchanging sex for crack,and the median age of participants who exchanged sexfor crack was 35 years [IQR: 25.0-40.0] compared to37 years [IQR: 25.0-40.0] among those who did not re-port having exchanged sex for crack-cocaine, with justunder half (48.5%) who were of Aboriginal ancestry (in-clusive of First Nations, Metis, Inuit ancestry and non-status First Nations). The median age of initiation intocrack use was slightly younger (20 years; IQR: 16–27)among those who exchanged sex for crack compared tothose who did not (21 years; IQR:16–30).The results of bivariate and multivariable GEE model arepresented in Table 2. In our multivariable GEE explanatorymodel, sharing a crack pipe with a regular or one-time client[aOR= 1.98; 95% CI: 1.27-3.09] and smoking crack with agroup of strangers [aOR = 1.70; 95% CI: 1.13-2.58] remainedsignificantly correlated with exchanging sex-for-crack.The results of our confounding model examining the in-dependent effect of exchanging sex for crack on numberDuff et al. Harm Reduction Journal 2013, 10:29 Page 3 of 8http://www.harmreductionjournal.com/content/10/1/29of clients/week are shown in Table 3. In our confoundingmodel (adjusting for age, servicing clients in public spaces,police affecting access to drug equipment), having ex-changed sex for crack in the past six months was associ-ated with a 34% increased risk of greater than averagenumber of clients (>10 clients/week) (Incidence Rate Ratio(IRR): 1.34 [95% Confidence Interval 1.07- 1.69]). We alsoconstructed a multivariable confounding model for therelationship between exchanging sex for crack and in-consistent condom use, which yielded statistically non-significant results.DiscussionThe results of this study demonstrate that among street-based sex workers who smoke crack, a large proportion(49%) reported non-monetary, direct sex-for-crack ex-changes. These findings highlight the importance ofintersecting social and physical contexts in driving sex-for-crack exchanges and sexual and drug risks amongstreet-based SWs. These results provide epidemiologicaldata to confirm qualitative reports among drug users inour setting that suggest marginalized physical spaces,such as alleys, are ‘niche’ settings for illicit drug use,where using in groups, and sharing drug-paraphernaliais the norm [25]. While people who use drugs have de-scribed these ‘niches’ as far from an ideal setting fordrug use [25], it is important to acknowledge that theyare a by-product of a number of structural factors in-cluding homelessness, lack of access to safe spaces tosmoke, stigma associated with drug-use and sex work,and avoidance of law enforcement [11]. Crack cocaine isa common feature of the street economy, with crack use‘niches’ (e.g., alleys or crack houses) often concentratedaround sex work strolls, social housing and vacant lots;spaces where some of the most marginalized and stigma-tized populations live and congregate [26].Table 1 Socio-demographic, interpersonal/social environment and physical-environment characteristics of sex-for-crack exchanges among a cohort of street-based female SWs in Vancouver, CanadaCharacteristic Total (%) Sex-for-crack exchanges p - valueYes (%) No (%)(n = 206) (n = 101) (n = 105)Socio-demographicAge (med, IQR) 35 (25–40) 37 (28–41) 0.091Aboriginal ancestry 100 (48.5) 57 (57.00) 43 (43.00) 0.175Caucasian 106 (51.5) 48 (45.28) 58 (54.72) –Age first used crack (median, IQR) 20 (16–27) 21 (16–30) –Individual level- drug risksCocaine injection* 73 (35.40) 46 (22.30) 27 (13.11) 0.050Heroin injection* 104 (50.49) 62 (59.62) 42 (40.38) 0.043Crystal Meth injection/non-injection* 24 (11.65) 6 (25.00) 18 (75.00) 0.186Intensive crack use (>10 rocks/day)* 57 (27.67) 25 (43.86) 32 (56.14) 0.021Interpersonal/ social environmentUsed drug with regular client* 72 (35.00) 41 (56.94) 31 (43.06) 0.110Receptive sharing of used syringe * 16 (7.77) 10 (62.50) 6 (37.50) 0.003Shared used pipe with regular client/john* 95 (46.12) 59 (62.11) 36 (37.89) <0.001Higher number of clients (>10)* 71 (37.60) 42 (22.20) 29 (15.34) <0.001Inconsistent condom use with client (for vaginal sex)* 19 (9.22) 12 (63.16) 7 (36.84) 0.031Intimate partner uses drugs 74 (35.92) 38 (51.35) 36 (48.65) 0.352Intimate partner provides drugs 55 (26.70) 30 (54.55) 25 (45.45) 0.260Economic dependence on intimate partner 11 (5.34) 6 (54.55) 5 (45.45) 0.490Physical/sexual violence by client* 45 (21.84) 29 (64.44) 16 (35.56) 0.002Physical environmentSmoke crack in groups with strangers (e.g., crack houses, alleys)* 165 (80.10) 89 (53.94) 76 (46. 06) 0.001Homeless* 88 (42.7) 50 (56.82 38 (43.18) <0.003Work in alleyways, industrial areas* 134 (65.00) 69 (51.49) 65 (48.51) <0.001Services clients in public spaces* 141 (68.45) 74 (52.48) 67 (47.52) 0.001*In the last 6 months.Duff et al. Harm Reduction Journal 2013, 10:29 Page 4 of 8http://www.harmreductionjournal.com/content/10/1/29The physical features of drug using ‘niches’ may act asa ‘site of social and cultural reproduction’ [27], wherethe dynamic interplay between individual-level factorsand their environments foster and perpetuate sex-for-crack exchanges. Ethnographic accounts describe howthe addictive and stigmatized nature of crack creates acycle of use that quickly deteriorates street-based SWs’work environments and becomes entrenched as a centralfeature of street-based sex work [27]. Street-based SWs’exposure to high-risk environments such as smoking ingroups of strangers in isolated, unsanitary public spacessuch as alleys or crack houses may facilitate the creationof social ties with other drug users, intensive daily crackuse and sharing of paraphernalia, that has been positedto link crack use and STI transmission [1,6]. Further-more, ‘niches’ such as alleys and crack houses oftenreinforce a culture where sex-for-crack exchanges repre-sent a highly gendered power dynamic. For example,sex-for-crack exchanges in these settings often occur inthe context of intense cravings/withdrawals that may fa-cilitate sex-for-crack exchanges while high, and exacer-bate female SWs’ vulnerability to gender-based violence,STI and HIV transmission, including through reducedability to negotiate for condom use by clients and clientsinsisting on sex without a condom [28]. Though crackcocaine use has been associated with gender-based vio-lence and inconsistent condom use elsewhere [29], wedid not find a statistically significant association betweeneither client violence or inconsistent condom use andexchanging sex for crack, after adjusting for potentialTable 2 Bivariate and multivariable logistic regression models using generalized estimating equations (GEEs) forcorrelates of sex-for-crack exchanges among street-involved sex workers in Vancouver, CanadaCharacteristic Unadjusted AdjustedOdds ratio p - value Odds ratio p - value(95% CI) (95% CI)Socio-demographicAge‡ 0.98 (0.96 –1.00) 0.091 0.99 (0.96–1.01) 0.230Aboriginal ancestry vs. Caucasian 0.74 (0.48 –1.14) 0.175 – –Individual level- drug risksCocaine Injection* 1.52 (1.00 – 2.31) 0.050 1.29 (0.82 –2.03) 0.275Heroin Injection* 1.57 (1.01 – 2.42) 0.043 1.12 (0.69 –1.82) 0.653Intensive crack use*† 1.71 (1.09 – 2.69) 0.021 – –Crystal meth injection/non- injection 0.61 (0.29 – 1.27) 0.186 – –Interpersonal/ social environmentShared used pipe with regular client/john* 2.31 (1.54 – 3.47) <0.001 1.93 (1.28–2.91) 0.002Intimate partner uses drugs† 1.23 (0.80 – 1.87) 0.352 – –Intimate partner provides drugs† 1.31 (0.82 – 2.08) 0.260 – –Physical/sexual violence by client*† 2.27 (1.37 – 3.78) 0.002 – –Inconsistent condom use by client (for vaginal sex)†* 2.25 (1.08 – 4.70) 0.031 – –Serviced over 10 clients/week*† 2.18 (1.39 – 3.43) <0.001 – –Physical environmentSmoke crack in groups with strangers e.g., crack houses, alleys* 2.14 (1.44 – 3.17) 0.001 1.70 (1.13–2.58) 0.012Homeless*† 1.91 (1.25 – 2.93) <0.003 – –Work in alleys/industrial areas†* 2.30 (1.53 – 3.46) 0.001 – –Services clients in public spaces*† 2.03 (1.32 – 3.13) 0.001 – –*Last 6 months.†Variable not entered into logistic model.‡Age was forced into the model based on a priori knowledge as a confounder.Table 3 Multivariable confounding model of theindependent effect of sex-for-crack exchanges on numberof clients among a cohort of street-based female sexworkers in Vancouver, CanadaCharacteristic Unadjusted incidencerate ratioAdjusted incidencerate ratioIRR (95% CI) p - value aIRR (95% CI) p - valueExchangedsex for crack*1.55 (1.22– 1.97) <0.001 1.34 (1.07- 1.69) 0.013*Within the last 6 months.N.B. Multivariable confounder model adjusted for a priori and statisticallysignificant confounders (servicing clients in public spaces, and age). Age wasforced into the model based on well-established a priori knowledge of age asa confounder.Duff et al. Harm Reduction Journal 2013, 10:29 Page 5 of 8http://www.harmreductionjournal.com/content/10/1/29confounders. This findings is somewhat surprising, givenqualitative accounts from women who use drugs in oursetting that indicate that smoking crack, particularly inunsafe areas such as alleys (due to displacement fromhomelessness, policing, lack of safe smoking places)often increases the risk of gender-based violence [11].The exclusive focus on crack-using sex workers (versus asample of drug-using women) in our sample may havecontributed to the lack of association. The lack of associ-ation with violence may reflect the general pervasivenessof violence among women who use crack, (regardless ofwhether or not they exchange sex for crack), resulting ina similar distribution of violence between those who en-gage in sex-for-crack exchanges and those who do not.While two decades of qualitative and ethnographicwork have described the physical and social contexts ofcrack houses featured prominently in street-based SW’ssex-for-crack exchanges [30], this is among the firststudies to longitudinally examine the social and physicalfeatures independently linked with sex-for-crack ex-changes. In Inciardi et al.’s study, SWs working primarilyin crack houses reported an association between exchan-ging sex for crack and higher number of clients [17].While some studies make clear distinctions betweenstreet-based vs. crack house sex workers, others do not.Inciardi (1995) makes a clear distinction between SWswho work on the street and those who work in crackhouses, describing crack house-based SWs as highlyaddicted, desperate and reliant on crack-house pimps/managers, often accepting the lowest price for a hit ofcrack or exchanging sex for a smoke [31]. The high drug-dependency among SWs who work in crack houses,paired with low pay and high traffic in these settings aredescribed to contribute to the higher number of clientsamong street-based SWs who work in crack houses [31].In contrast, another qualitative study described sex-for-crack exchanges as occurring in the same physical settingsas sex-for-cash transactions, and did not observe a cleardistinction between those who engage in sex-for-crack ex-changes and those who do not [27]. While sex-for-crackexchanges were considered degrading and purposefullyavoided by many street based SWs, Maher’s ethnographicresearch suggests that these transactions occur underpressing circumstances, where women considered theneed for crack to outweigh the shame of exchanging sexto obtain the drug [27]. In our study, smoking in anonym-ous groups (both crack houses and alleys) was independ-ently correlated with increased likelihood of exchangingsex directly for crack. The greater number of clients (>10per week) reported by street-based SWs who exchangesex for crack in our sample likely reflects women’s need tosupport their intense crack use patterns [32], and under-lines the vicious cycle of sex work and addictions drivingthese sex-for-crack exchanges.This study also revealed increased odds of reciprocalcrack-pipe sharing with clients, and smoking crack ingroups with strangers (e.g., in crack houses or alleys)among SWs’ who exchanged sex for crack, after adjust-ing for potential confounders in multivariable analysis.Our confounding model indicates that exchanging sexfor crack was associated with an increased number ofclients, after controlling for potential confounders. Theincreased odds of sex-for-crack exchanges among street-based SWs who smoke crack with groups of strangers(namely, alleys or crack houses), and share crack pipes(reciprocally) with clients underscores the importance ofsocial and structural environments in shaping sex-for-crack exchanges and STI transmission.These findings support structural and environmentalinterventions, such as the removal of criminal sanctionsthat would enable SWs to work (and smoke) in saferindoor spaces (including low-barrier housing supportsand/or workspaces). Such interventions would removeSWs’ need to smoke in public outdoor spaces withstrangers and/or clients, where sharing of smoking para-phernalia is common. Such indoor settings may altergendered-power dynamics of sexual exchanges, poten-tially improving SWs’ ability to negotiate for higher ratesper transaction, thus reducing the need to service agreater volume of clients. Increased access to safersmoking kits may improve SWs’ choices related to whereand with whom they smoke [30]. Another alternative isthe implementation of safer smoking facilities (SSFs),particularly within close proximity to SWs’ workspaces.SSFs may increase access to clean crack pipes, reducethe risk of pipe sharing with clients and provide anenvironment for safer smoking practices. As well, SSFsmay increase exposure to health care and addictiontreatment services, reduce public smoking and movestreet-based SWs who exchange sex for crack away fromalleys and crack houses. The development and evalu-ation of SSFs that specifically cater to the needs of SWsmay also be beneficial, and could further moderate vio-lence and coercion in drug and sex work scenes. Furtherresearch is needed to identify acceptable and effectivemodels for crack use harm reduction. Finally, gender-specific programs targeting women who exchange sex forcrack should be developed that address the gendered-power dynamics present in sex-for-crack exchanges.This study has a number of limitations that shouldbe noted. The findings from this study may not begeneralizable to SWs working in other venues, such asbars, massage parlours and/or escort agencies. However,this limitation is tempered by our time-location samplingmethod, which is used to recruit hard-to-reach popula-tions by sampling at times and places where they areknown to congregate. Social mapping of spaces of ser-vicing and solicitation were conducted beforehand withDuff et al. Harm Reduction Journal 2013, 10:29 Page 6 of 8http://www.harmreductionjournal.com/content/10/1/29current and former SWs. While we employed statisticalmethods commonly used to analyze data collected usingtime-location sampling, emerging evidence suggests thereare other methods that may better account for clusteringby sampling location and variability in the probabilityof sampling among members (i.e., treating the sampleas a two-staged sample) [33]. As a result, our statisticalmethods may have underestimated the true standard er-rors, as well as affected the estimates of interest. Also,though causality cannot be inferred from this study, dueto the observational nature of the research, some potentialtemporal bias may be reduced due to the use of general-ized estimating equations (GEEs) that account for re-peated measurements on the same respondents. Thisstudy used self-report data, and women’s responses maybe subject to social desirability bias (i.e. when womenreport answers they view as being socially acceptableto the interviewer or society). However, a number ofstudies have found SWs and drug users to providetruthful accounts of their sex and drug use activitieswhen questioned in a non-threatening environment[34], and we believe the community-based nature ofour study serves to reduce the likelihood of this formof response bias. Questions pertaining to events thatoccurred within the past 6 months of the interviewmay be subject to recall bias. The arbitrary 6 monthcut-off may also result in an underestimation of ourestimates, as sex-for-crack exchanges occurring morethan 6 months prior to the interview would go unre-ported. Finally, while the data collection for this studybegan in 2005, our continued research in this setting(among street- and off-street sex workers) indicates thesedrug-use ‘niches’ and sex-for-crack exchanges persist, andthese findings remain a relevant and important issue forthis population.These findings reveal that sex-for-crack exchanges bystreet-based SWs may increase risk for STI and HIVtransmission, through the sharing of crack pipes withclients and servicing a higher volume of client (if in thecontext of inconsistent condom use). The physical andsocial environments may be important drivers of sex-for-crack exchanges, highlighting an urgent need formultilevel approaches to harm reduction, STI and HIVprevention that address street-based SWs’ environment,individual level factors, and the interplay between them.These findings reveal that sex-for-crack exchanges bystreet-based SWs may increase risk for STI and HIVtransmission, through the sharing of crack pipes withclients and servicing a higher volume of client (if in thecontext of inconsistent condom use). The physical andsocial environments may be important drivers of sex-for-crack exchanges, highlighting an urgent need formultilevel approaches to harm reduction, STI and HIVprevention that address street-based SWs’ environment,individual level factors, and the interplay between them.These findings point to a need for alternative models forcrack use harm reduction that are gender sensitive, andserve the needs of sex workers who exchange sex forcrack. In the interim, there is a need to improve accessto clean drug use equipment (e.g., through safer smokingfacilities and the distribution of safer smoking kits).Additionally, the removal of criminal sanctions that pre-vent SWs from working (and smoking) in safer indoorspaces may help reduce sex-for-crack exchanges amongthis population.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsKS had access to the data and takes full responsibility for the integrity of thedata. PD and KS developed the analyses plan, and RZ conducted thestatistical analyses. PD wrote the first draft of the manuscript and integratedsuggestions from all co-authors. All authors made significant contributions tothe conception and design of the analyses, interpretation of the data, anddrafting of the manuscript, and all authors approved the final manuscript.AcknowledgementsWe would like to extend our thanks to the women who participated in ourproject, including our many community partners, advisory board membersand in particular the peer research team: Shari, Rose, Chanel, Laurie, Debbie,and Adrian. We would like to acknowledge our research and administrativestaff, including Peter Vann, Gina Willis, Calvin Lai, Cindy Feng, and KathleenDeering. This work was supported through an operating grant from theCanadian Institutes of Health Research (CIHR) and NIH (R01DA028648).PD is supported by PHIRNET (Population 1Health Interventions Network)an initiative of the CIHR. KS is supported through a MSFHR Scholar Award,CIHR, and National Institutes of Health Research (R01DA028648).Author details1British Columbia Centre for Excellence in HIV/AIDS, St. Paul’s Hospital,608-1081 Burrard Street, Vancouver, BC V6Z 1Y6, CANADA. 2School ofPopulation and Public Health, University of British Columbia, 2206 East Mall,Vancouver, BC V6T 1Z3, CANADA. 3Department of Medicine, GeneralCampus, University of Ottawa, The Ottawa Hospital, 501 Smyth Road, Box206, Ottawa, ON K1H 8 L6, CANADA. 4British Columbia Centre for DiseaseControl, 655 West 12th Avenue, Vancouver, British Columbia V5Z 4R4,CANADA. 5Department of Medicine, University of British Columbia, St. Paul’sHospital, 608-1081 Burrard Street, Vancouver, BC V6Z 1Y6, CANADA.Received: 3 July 2013 Accepted: 3 November 2013Published: 15 November 2013References1. DeBeck K, Kerr T, Li K, Fischer B, Buxton J, Montaner J, Wood E: Smoking ofcrack cocaine as a risk factor for HIV infection among people who useinjection drugs. CMAJ 2009, 181:585–589.2. Ross MW, Hwang L-Y, Zack C, Bull L, Williams ML: Sexual risk behavioursand STIs in drug abuse treatment populations whose drug of choice iscrack cocaine. Int J STD AIDS 2002, 13:769–774.3. Fischer B, Firestone Cruz M, Rudzinski K, Rehm J: Hepatitis C virustransmission among oral crack users: viral detection on crackparaphernalia. Eur J Gastroenterol Hepatol 2008, 20:29–32.4. Edlin BR, Irwin KL, Faruque S, McCoy CB, Word C, Serrano Y, Inciardi JA,Bowser BP, Schilling RF, Holmberg SD: Intersecting epidemics–crackcocaine use and HIV infection among inner-city young adults.Multicenter Crack Cocaine and HIV Infection Study Team. N Engl J Med1994, 331:1422–1427.5. Pallecaros A, Green A, Day S: Health issues associated with increasing useof “crack” cocaine among female sex workers in London. Sex Trans Infect2000, 76:292–293.Duff et al. Harm Reduction Journal 2013, 10:29 Page 7 of 8http://www.harmreductionjournal.com/content/10/1/296. Nurutdinova D, Abdallah AB, Bradford S, O’Leary CC, Cottler LB: Risk factorsassociated with Hepatitis C among female substance users enrolled incommunity-based HIV prevention studies. BMC Res Notes 2011, 4:126.7. Faruque S, Edlin BR, McCoy CB, Word OS, Larsen SA, Schmid DS, Bargen JCV,Serrano Y: Crack cocaine smoking and oral sores in three inner-cityneighbourhoods. J Acquir Immune Defic Syndr Hum Retrovirol 1996, 13:87–92.8. Haim DY, Lippmann ML, Goldberg SK, Walkenstein MD: The pulmonarycomplications of crack cocaine. Chest 1995, 107:233–240.9. Lejuez CW, Bornovalova MA, Reynolds EK, Daughters SB, Curtin JJ: Riskfactors in the relationship between gender and crack/cocaine. Exp ClinPsychopharmacol 2007, 15:165–175.10. Team CP: Community Health and Safety Evaluation (CHASE) Project: FinalReport. British Columbia Centre for Excellence in HIV/AIDS and ProvidenceHealth Care: Vancouver, British Columbia; 2005.11. Bungay V, Johnson JL, Varcoe C, Boyd S: Women’s health and use of crackcocaine in context: structural and “everyday” violence. Int J Drug Pol2010, 21:321–329.12. Peters RH, Strozier AL, Murrin MR, Kearns WD: Treatment of substance-abusing jail inmates. Examination of gender differences. J Subst AbuseTreat 1997, 14:339–349.13. Surratt HL, Inciardi J: HIV risk, seropositivity and predictors of infectionamong homeless and non-homeless women sex workers in Miami,Florida, USA. AIDS care 2004, 16:594–604.14. Shannon K: Mapping violence and policing as an environmental-structural barrier to health service and syringe availability amongsubstance-using women in street-level sex work. Int J Drug Policy 2008,19:140–147.15. Jones DL, Irwin KL, Inciardi J, Bowser B, Schilling R, Word C, Evans P,Faruque S, McCoy HV, Edlin BR: The high-risk sexual practices of crack-smoking sex workers recruited from the streets of three American cities.The Multicenter Crack Cocaine and HIV Infection Study Team. Sex TransmDis 1998, 25:187–193.16. Edwards JM, Halpern CT, Wechsberg WM: Correlates of exchanging sex fordrugs or money among women who use crack cocaine. AIDS Educ Prev2006, 18:420–429.17. Inciardi JA: Crack, crack house sex, and HIV risk. Archives of Sexual Behav1995, 24:249–269.18. Shannon K, Bright V, Parsad D, Alexson D, Allinott S, Gibson K, Tyndall MW:Community-based HIV prevention research among substance-usingwomen in survival sex work: the Maka Project Partnership. Harm Reduc J2007, 4:20.19. Craib KJ, Spittal PM, Wood E, Laliberte N, Hogg RS, Li K, Heath K, TyndallMW, O’Shaughnessy MV, Schechter MT: Risk factors for elevated HIVincidence among Aboriginal injection drug users in Vancouver.CMAJ 2003, 168:19–24.20. Gracey M, Crey E: Indigenous health part 1: determinants and diseasepatterns. Lancet 2009, 374:65–75.21. Duff P, Deering K, Gibson K, Tyndall M, Shannon K: Homelessness amonga cohort of women in street-based sex work: the need for saferenvironment interventions. BMC Public Health 2011, 11:643.22. Rothman K, Greenland S: Modern Epidemiology. 2nd edition. Philadelphia:Lippincott Williams & Wilkins; 1998.23. Lima VD, Geller J, Bangsberg DR, Patterson TL, Daniel M, Kerr T, MontanerJSG, Hogg RS: The effect of adherence on the association betweendepressive symptoms and mortality among HIV-infected individuals firstinitiating HAART. AIDS (London, England) 2007, 21:1175–1183.24. Marshall BDL, Kerr T, Shoveller JA, Patterson TL, Buxton JA, Wood E:Homelessness and unstable housing associated with an increased risk ofHIV and STI transmission among street-involved youth. Health Place 2009,15:783–790.25. Small W, Rhodes T, Wood E, Kerr T: Public injection settings in Vancouver:physical environment, social context and risk. Int J Drug Pol 2007, 18:27–36.26. Bourgois P: Crack and the political economy of social suffering. Addict ResTheory 2003, 11:31–37.27. Maher L: Sexed Work: Gender, Race and Resistance in a Brooklyn Drug Market.Oxford: Clarendon Press; 1997.28. Shannon K, Strathdee S, Shoveller J, Rusch M, Kerr T, Tyndall MW: Structuraland environmental barriers to condom use negotiation with clientsamong female sex workers: Implications for HIV prevention strategiesand policy. Am J Public Health 2009, 99:659–665.29. Booth RE, Kwiatkowski CF, Chitwood DD: Sex related HIV risk behaviors:differential risks among injection drug users, crack smokers, andinjection drug users who smoke crack. Drug Alcohol Depend 2000,58:219–226.30. Wood E, Kerr T, Spittal PM, Small W, Tyndall MW, Li K, Hogg RS,O’Shaughnessy MV, Schechter MT: The potential public health andcommunity impacts of safer injecting facilities: evidence from a cohortof injection drug users. J Acquir Immune Defic Syndr 2003, 32:2–8.31. Inciardi JA, Lockwood D, Pottiger A: Women and Crack-Cocaine. New York:Macmillan; 1993.32. Hoffman J, Klein H, Eber M, Crosby HJA: Frequency and intensity of crackuse as predictors of women’s involvement in HIV-related sexual riskbehaviors. Drug Alcohol Depend 2000, 58:227–236.33. Karon J, Wejnert C: Statistical methods for the analysis of time-locationsampling data. J Urban Heal 2012, 89:565–586.34. De Irala J, Bigelow C, McCusker J, Hindin R, Zheng L: Reliability of self-reported human immunodeficiency virus risk behaviors in a residentialdrug treatment population. Am J Epidemiol 1996, 143:725–732.doi:10.1186/1477-7517-10-29Cite this article as: Duff et al.: Sex-for-Crack exchanges: associationswith risky sexual and drug use niches in an urban Canadian city.Harm Reduction Journal 2013 10:29.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitDuff et al. Harm Reduction Journal 2013, 10:29 Page 8 of 8http://www.harmreductionjournal.com/content/10/1/29


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items