UBC Faculty Research and Publications

The impact of drug use patterns on mortality among polysubstance users in a Canadian setting : a prospective… Hayden, Anna; Hayashi, Kanna; Dong, Huiru; Milloy, M-J; Kerr, Thomas; Montaner, Julio S; Wood, Evan Nov 6, 2014

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

Item Metadata


52383-12889_2014_Article_7262.pdf [ 272.42kB ]
JSON: 52383-1.0220623.json
JSON-LD: 52383-1.0220623-ld.json
RDF/XML (Pretty): 52383-1.0220623-rdf.xml
RDF/JSON: 52383-1.0220623-rdf.json
Turtle: 52383-1.0220623-turtle.txt
N-Triples: 52383-1.0220623-rdf-ntriples.txt
Original Record: 52383-1.0220623-source.json
Full Text

Full Text

RESEARCH ARTICLEThe impact of drug use paehnIllicit drug use has been well established as a risk factor the general population [1]. An increased mortality rateHayden et al. BMC Public Health 2014, 14:1153http://www.biomedcentral.com/1471-2458/14/1153which have adjusted for potential confounders, many2Department of Medicine, University of British Columbia, 317-2194 HealthSciences Mall, Vancouver, BC V6T 1Z3, Canadafor adverse health outcomes including premature mor-tality [1-6]. Previous studies have indicated that fataloverdose accounts for a majority of deaths in illicit drug-using populations, particularly among persons who in-ject drugs (PWIDs) [3-5,7,8]. Much of the existing litera-ture has highlighted opioid use associated with fataloverdose [1,7,9,10]. One meta-analysis found that illicithas also been demonstrated among users of other sub-stances, including cocaine and amphetamines [4,6,11,12].Earlier studies of mortality among PWIDs have oftenbeen limited by the fact that they were restricted tocross-sectional examinations of toxicology reportsamong deceased individuals where important potentialconfounders, such as human immunodeficiency virus(HIV) infection or other socio-demographic comorbidi-ties, like homelessness, were difficult to control for[2,7,13,14]. While longstanding prospective cohort stud-ies of PWIDs exist in a number of regions, some of* Correspondence: uhri-ew@cfenet.ubc.ca1British Columbia Centre for Excellence in HIV/AIDS, St. Paul’s Hospital,608 - 1081 Burrard Street, Vancouver, BC V6Z 1Y6, CanadaBackgroundAbstractBackground: Illicit drug use is a well-established risk factor for morbidity and mortality. However, few studies haveexamined the impact of different drug use patterns on mortality among polysubstance using populations. Thisstudy aimed to identify drug-specific patterns of mortality among a cohort of polysubstance using persons whoinject drugs (PWIDs).Methods: PWIDs in Vancouver, Canada were prospectively followed between May 1996 and December 2011.Participants were linked to the provincial vital statistics database to ascertain mortality rates and causes of death.We used multivariate Cox proportional hazards regression to investigate the relationships between drug usepatterns (daily alcohol use, heroin injection and non-injection use, cocaine injection, amphetamine injection andnon-injection use, crack smoking and speedball injecting) and time to all-cause mortality.Results: 2330 individuals were followed for a median of 61 months (inter-quartile range: 33 – 112). In total, 466(19.1%) individuals died for an incidence density of 3.1 (95% confidence interval [CI]: 2.8 – 3.4) deaths per 100person-years. In multivariate analyses, after adjusting for HIV infection and other potential confounders, only dailycocaine injection remained independently associated with all-cause mortality (adjusted hazard ratio [AHR] = 1.36,95% CI: 1.06 – 1.76).Conclusions: Although heroin injecting is traditionally viewed as carrying the highest risk of mortality, in thissetting, only daily cocaine injecting was associated with all-cause mortality. These findings highlight the urgentneed to identify novel treatments and harm reduction strategies for cocaine injectors.Keywords: Mortality, Injection drug use, Cocaine, Vancouver, Cohort studyopiate users have a mortality rate thirteen times that ofamong polysubstance ussetting: a prospective cohAnna Hayden1, Kanna Hayashi1, Huiru Dong1, Michael-Joand Evan Wood1,2*© 2014 Hayden et al.; licensee BioMed CentraCommons Attribution License (http://creativecreproduction in any medium, provided the orDedication waiver (http://creativecommons.orunless otherwise stated.Open Accesstterns on mortalityrs in a Canadianort studyMilloy1,2, Thomas Kerr1,2, Julio SG Montaner1,2l Ltd. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/4.0), which permits unrestricted use, distribution, andiginal work is properly credited. The Creative Commons Public Domaing/publicdomain/zero/1.0/) applies to the data made available in this article,Hayden et al. BMC Public Health 2014, 14:1153 Page 2 of 7http://www.biomedcentral.com/1471-2458/14/1153have been undertaken in areas where a limited numberof specific drugs are primarily used (e.g. primarily heroinor cocaine users) [7,10,11,15-17].The Downtown Eastside of Vancouver (DTES), Canadais well known for its concentration of poly-substance use,high prevalence of hepatitis C virus (HCV) and HIV infec-tion and high rates of homelessness [18,19]. The DTES isamong the most impoverished neighborhoods in Canadawith an estimated 5000 PWIDs residing in the area[18,19]. Various patterns of drug use in this populationhave been described in previous studies including highrates of cocaine injecting [20], crack cocaine smoking[21], amphetamine injecting [22] and heroin injecting[23]. Accordingly, it provides an excellent environmentto examine the contribution of specific drug use patternson mortality. Therefore, the present study aimed to iden-tify drug specific patterns of mortality among a cohort ofpolysubstance using PWIDs in Vancouver.MethodsThe Vancouver Injection Drug Users Study (VIDUS)and AIDS Care Cohort to Evaluate Access to SurvivalServices (ACCESS) are open prospective cohorts of drugusers in Vancouver. The recruitment and follow-up pro-cedures for the two studies are identical to allow foranalyses of merged data, with the only differences beingthat HIV-positive individuals are followed in ACCESSwhereas HIV-negative individuals are followed in VIDUSand that ACCESS includes non-injecting drug users. Inboth studies the primary modes of enrollment were self-referral, word of mouth, and street outreach. Detailedsampling and recruitment procedures for these twocohorts have been described elsewhere [20,24,25].To be eligible, participants were 18 years of age orolder, had used illicit drugs other than cannabinoids inthe previous month and resided in the greater Vancouverregion. All participants provided written informed con-sent. Participants were given a stipend ($20 CDN) at eachstudy visit for their time and transportation. The studywas approved by the University of British Columbia/Providence Healthcare Research Ethics Board.At baseline and semianually thereafter, participantscompleted an interviewer-administered questionnairethat elicited a range of data, including demographiccharacteristics, injection and non-injection drug use, andsexual risk behaviors. In addition, venous blood sampleswere drawn at each visit and tested for HIV and HCVantibodies. All participants had private interviews andwere offered both pre- and post-test counseling withtrained nurses. Referral for free healthcare was providedto those who tested HIV positive and these individualswere subsequently followed in ACCESS.We ascertained all-cause mortality rates and under-lying causes of death among participants through aconfidential record linkage with the British ColumbiaVital Statistics Agency and through ongoing follow-upwith contacts provided by participants. The Vital Statisticsdatabase recorded causes of death according to the Inter-national Classification of Diseases (ICD), 10th edition.The present study included PWIDs who were re-cruited and completed at least one follow up visit be-tween May 1996 and December 2011. To avoid potentialbias relating to long durations between the last studyvisit where behavioural information was assessed andthe date of death (i.e. loss to regular follow up), individ-uals who were identified as deceased more than12 months after the the last follow up visit were cen-sored on the date of the last follow up.The primary endpoint in this analysis was all-causemortality. The primary explanatory variables of interestincluded a number of substance use behaviors in theprevious six months, including at least daily alcohol use,at least daily cocaine injection, at least daily heroin injec-tion and non-injection use, at least daily amphetamineinjection and non-injection use, at least daily crackcocaine smoking and at least daily speedball injecting(a mixture of cocaine and heroin). Additionally, weexamined risk associated with non-daily heroin andamphetamine use defined as any report of use in the lastsix months. Potential confounders that were consideredincluded: age (per 10 years older); gender (male vs. female),time since first injection (per 10 years longer); ethnicity(Caucasian vs. others); HIV serostatus (positive vs. nega-tive); and unstable housing in the previous six months. Aspreviously, unstable housing was defined as living in oneof the DTES’ single room occupancy hotels, shelters orother transitional housing, or living on the street [26,27].As an initial analysis, we used the Chi-square test andWilcoxon rank sum test to compare the baseline charac-teristics of the included and excluded individuals, andthose who did and did not report daily cocaine injection inthe previous six months among the included sample basedon an interest in the role of cocaine injecting on mortality.All-cause mortality rate and 95% confidence interval [CI]were calculated using the Poisson distribution.Next, we used bivariate and multivariate Cox propor-tional hazards regression to examine the relationship be-tween substance use patterns and time to death. Allbehavioral variables including substance use patterns,unstable housing and HIV serostatus were treated astime-varying variables. The multivariate model was fitusing an a priori defined protocol whereby all illicit druguse variables were entered into the multivariate model.Variables found to be significantly associated with time toall-cause mortality in bivariate analyses at p <0.10 were en-tered into the multivariate model as potential confounders.We also conducted a sub-analysis where we restrictedthe dependent variable to accidental poisonings (ICD-10:Hayden et al. BMC Public Health 2014, 14:1153 Page 3 of 7http://www.biomedcentral.com/1471-2458/14/1153X40–44) and other accidental causes of death to specif-ically examine if drug use patterns were associated withaccidental death. All statistical analyses were performedusing SAS software version 9.3 (SAS, Cary, NC). Allp-values were two-sided.ResultsA total of 2597 individuals were recruited between May1996 and December 2011, among whom 267 were ex-cluded as a result of no follow-up information. In com-paring the study sample to those that were excluded, theexcluded sample was younger, less likely to be HIV posi-tive and was less likely to inject cocaine and speed ball(all p <0.05).The remaining 2330 (89.7%) were followed for a me-dian of 61 months (inter-quartile range [IQR]: 33 – 112).Table 1 shows the baseline characteristics of the studysample. As shown, at baseline, 1550 (66.5%) were men,640 (27.5%) were HIV positive, and 1424 (61.1%) wereCaucasian. The median age was 37.3 years (IQR: 29.4 –43.7), and the median time since first injection was14.4 years (IQR: 6.2 – 24.2). At baseline, 550 (23.6%) ofthe study sample consumed alcohol daily in the previoussix months, 901 (38.7%) injected heroin daily, 97 (4.2%)smoked heroin daily, 1588 (68.2%) injected cocaine daily,51 (2.2%) injected amphetamines daily, 7 (0.3%) smokedamphetamines daily, 560 (24.0%) smoked crack cocainedaily, and. 288 (12.4%) injected speedball daily.Compared to non-daily cocaine injectors, those whoreported daily cocaine injection at baseline were morelikely to be younger (odds ratio [OR] = 0.76, 95% CI:0.69 – 0.83), to reside in unstable housing (OR = 1.62,95% CI: 1.32 – 1.99), and to have less time since first in-jection (OR = 0.91, 95% CI: 0.84 – 0.99). They were lesslikely to be male (OR = 0.70, 95% CI: 0.58 – 0.84) andreport Caucasian ethnicity (OR = 0.77, 95% CI: 0.64 –0.92). In terms of drug use patterns, daily cocaine injec-tion was significantly and positively associated with dailyheroin injection (OR = 1.68, 95% CI: 1.40 – 2.00) andspeedball injection (OR = 9.01, 95% CI 6.77 – 12.00),and negatively associated with daily amphetamine injec-tion (OR = 0.34, 95% CI: 0.15 – 0.76), and daily cracksmoking (OR = 0.65, 95% CI: 0.52 – 0.80).In total, 466 (19.1%) individuals died for an incidencedensity of 3.1 (95% CI: 2.8 – 3.4) deaths per 100 person-years. The primary underlying causes of death includedaccidental poisonings (22.6%) and HIV disease (18.5%).The remaining 100 classified categories including assault(2.3%) and lung malignancy (2.4%) are varied and eachindividually contributed less than 5%, with most contrib-uting less than 1% to all-cause mortality.Table 2 shows results of the bivariate and multivariateCox regression analyses of all-cause mortality. In the bi-variate analysis, daily cocaine injection was significantlyand positively associated with time to all-cause mortalitywith a hazard ratio (HR) of 1.41 (95% CI: 1.12 – 1.78)whereas daily heroin injection was significantly andnegatively associated with the ouctome with a HR of0.75 (95% CI: 0.60 – 0.95). Daily alcohol use (HR = 0.98,95% CI: 0.77 – 1.25), daily heroin non-injection use(HR = 1.12, 95% CI: 0.51 – 2.44), amphetamine injection(HR = 0.39, 95% CI: 0.10 – 1.54) and non-injection use(HR = 1.13, 95% CI: 0.29 – 4.50), daily crack cocainesmoking (HR = 0.91, 95% CI: 0.74 – 1.12) and dailyspeedball injection (HR = 1.02, 95% CI: 0.69 – 1.52) werenot significantly associated with time to all-cause mortal-ity. A bivariate analysis of non-daily heroin use (HR =0.99, 95% CI: 0.72 -1.37) and amphetamine (HR = 0.47,95% CI: 0.23 – 1.02) use, which was not included in thetables showed no significant association with mortality.In the multivariate analysis, after adjustment for po-tential confounders including HIV serostatus, age andunstable housing, daily cocaine injection remained inde-pendently and positively associated with time to all-cause death (adjusted hazard ratio [AHR] = 1.36, 95% CI:1.06 – 1.76) whereas the remaining substance use vari-ables were not significantly associated with the outcome.The sub-analysis whereby the dependent variable wasrestricted to accidental mortality showed that none ofthe substance use variables was significantly associatedwith accidental mortality. The only variable where therewas a trend towards an association with accidental mor-tality was daily crack cocaine injecting, though this wasnot statistically significant (AHR = 1.36, 95% CI: 0.91–2.10). Full data are available from the correspondingauthor.DiscussionIn the present study, we found that daily cocaine injec-tion was independently associated with all-cause mortal-ity among PWIDs in Vancouver, after adjusting forpotential confounders including age, unstable housingand HIV serostatus. These variables are known risk fac-tors for mortality and were included in the analysis inorder to adjust for their potential confounding effects[8]. The most common causes of death included acci-dental poisonings and HIV disease. We did not observean independent mortality risk associated with daily alco-hol use, daily heroin injection or non-injection use, dailyamphetamine injection or non-injection use, daily crackcocaine smoking and daily speedball injecting.Our findings that daily cocaine injection was the onlydrug use behaviour independently associated with mor-tality were somewhat unique in the context of previousliterature indicating a significant mortality risk associ-ated with heroin injection most commonly due to fataloverdose [5]. Heroin use has been shown to carry thehighest mortality risk amongst illicit drugs in someatiHayden et al. BMC Public Health 2014, 14:1153 Page 4 of 7http://www.biomedcentral.com/1471-2458/14/1153Table 1 Baseline demographics of the study population strstudies [14,28,29]. Particularly, non-daily heroin use isthought to carry mortality risk due to diminished toler-ance to respiratory depression effects with sporadic use(n = 2330)Characteristic Total (%)DaYes(n = 2330) (n = 1HIV serostatusPositive 640 (27.5) 212 (2Negative 1688 (72.5) 515 (7GenderMale 1550 (66.5) 444 (6Female 780 (33.5) 284 (3EthnicityCaucasian 1424 (61.1) 413 (5Other 906 (38.9) 315 (4Unstable housing*Yes 1637 (70.3) 556 (7No 677 (29.1) 163 (2Median age (IQR)Per 10 year older 37.3 (14.3) 35.3 (Median time since first injection (IQR)Per 10 year longer 14.4 (18.0) 14.1 (Daily alcohol use*Yes 550 (23.6) 179 (2No 1775 (76.2) 547 (7Daily heroin injection*Yes 901 (38.7) 345 (4No 1424 (61.1) 383 (5Daily heroin (non-injection)*Yes 97 (4.2) 30 (4No 2230 (95.7) 697 (9Daily amphetamine injection*Yes 51 (2.2) 7 (1No 2274 (97.6) 721 (9Daily amphetamine (non-injection)*Yes 7 (0.3) 1 (02322 (99.7) 727 (9Daily crack cocaine smoking*Yes 560 (24.0) 137 (1No 1767 (75.8) 589 (8Daily speedball injection*Yes 288 (12.4) 216 (2No 2038 (87.5) 512 (7*Refers to activities in the six months prior to interview. IQR = Interquartile range. †Ffied by daily cocaine injection in the past 6 monthsleading to increased of accidental overdose [30]. How-ever, our analysis showed no significant increase in mor-tality risk in both daily and non-daily heroin users.ily cocaine injection(%) No (%)Odds ratio (95% CI) p value588) (n = 728)9.1) 426 (26.8) 1.12 (0.92 – 1.36) 0.2470.7) 1161 (73.1)1.0) 1098 (69.1) 0.70 (0.58 – 0.84) <0.0019.0) 490 (30.9)6.7) 1001 (63.0) 0.77 (0.64 – 0.92) 0.0043.3) 587 (37.0)6.4) 1072 (67.5) 1.62 (1.32 – 1.99) <0.0012.4) 509 (32.1)13.4) 38.4 (14.1) 0.76 (0.69 – 0.83) <0.00116.3) 14.6 (18.8) 0.91 (0.84 – 0.99) 0.0334.6) 368 (23.2) 1.10 (0.88 – 1.33) 0.4505.1) 1217 (76.6)7.4) 555 (35.0) 1.68 (1.40 – 2.00) <0.0012.6) 1033 (65.1).1) 66 (4.2) 0.99 (0.64 – 1.54) 0.9695.7) 1520 (95.7).0) 44 (2.8) 0.34 (0.15 – 0.76) 0.0069.0) 1542 (97.1).1) 6 (0.4) 0.36 (0.04 – 3.02) 0.445†9.9) 1581 (99.6)8.8) 420 (26.5) 0.65 (0.52 – 0.80) <0.0010.9) 1167 (73.5)9.7) 71 (4.5) 9.01 (6.77 – 12.00) <0.0010.3) 1516 (95.5)isher’s Exact Test.rdrdr HHayden et al. BMC Public Health 2014, 14:1153 Page 5 of 7http://www.biomedcentral.com/1471-2458/14/1153While the underlying reasons for the discrepancy be-tween these previous studies and our study findings arenot entirely clear, a contributing factor may be related tofrequency of drug injecting, which was not accountedfor in the previous studies [4,8]. Due to the short half lifeof cocaine, injectors will often use greater than 20 timesin a day, predisposing them to increased risk of infec-tion/bacteremia and other associated negative health ef-fects, while local heroin injectors typically inject only 2to 4 times per day [20]. Rates of injecting may increaserisk of bacterial infection, needle sharing and other po-tential risks of mortality (e.g. air embolus, cellulitis, etc.)[12,20,28].We also found a differential risk of mortality betweencocaine injectors and crack smokers. Some cohort stud-ies have demonstrated an independent mortality risk as-sociated with cocaine use [4,12] in addition to significantmorbidity associated with the use of this substance, in-Table 2 Univariate and multivariate Cox proportional hazapeople who inject drugs in Vancouver, Canada (n = 2330)Unadjusted hazaVariable HR (95% CI)Daily alcohol use* 0.98 0.77 – 1.25Daily heroin injection* 0.75 0.60 – 0.95Daily heroin (non-injection)* 1.12 0.51 – 2.44Daily cocaine injection* 1.41 1.12 – 1.78Daily amphetamine injection* 0.39 0.10 – 1.54Daily amphetamine (non-injection)* 1.13 0.29 – 4.50Daily crack cocaine smoking* 0.91 0.74 – 1.12Daily speedball injectionn 1.02 0.69 – 1.52*Refers to activities in the six months prior to interview. †Model was adjusted focluding cardiovascular, psychiatric, neurologic disordersand unintentional injuries [16]. Much of the mortalitydata on cocaine use however does not distinguish be-tween injecting and smoking crack cocaine [16,17,28].Furthermore, the majority of these studies present datain standardized mortality ratios, which allow for a com-parison with the general population but not betweensub-groups of cocaine-using populations, as performedin the present study. Though both cocaine injectionand crack smoking have been shown to be associatedwith HIV infection in this setting [20,31], our multivari-ate analyses suggest that daily cocaine injectors appearto be at an elevated risk of mortality due to the risksassociated with injection practices other than HIVinfection.Harm reduction strategies, including a supervisedinjecting facility, in Vancouver’s DTES have been shownto be successful at attracting cocaine injectors, reducingrates of fatal overdose [32,33] and reducing syringesharing and HIV risk behavior [34,35] among the localPWID population. It is noteworthy that both cocaineand heroin injectors utilize this unique program, [35]which is widely accessible for extended hours. It alsoconnects PWIDs to addiction services, contributingto quicker entry into detoxification programs [36]and leading to increased likelihood of stopping druginjecting [37].The findings of this study highlight a need to furtheridentify addiction treatment and public health strategiestailored for cocaine injectors. Currently, there is nostandard pharmacotherapy proven effective for cocaineaddiction, though multiple therapeutic agents, includinganticonvulsants and stimulants have been investigated aspotential treatments [38,39]. A recent comprehensive re-view of human clinical trials for potential novel therapiesfor treatment of cocaine dependence outlined burgeon-ing research in this field. It identified multiple promisingregression analyses of the time to all-cause death amongratio (HR) Adjusted† hazard ratio (AHR)p-value AHR (95% CI) p-value0.854 1.00 0.78 – 1.28 0.9790.018 0.92 0.71 – 1.19 0.5020.779 1.44 0.68 – 3.07 0.3400.003 1.36 1.06 – 1.76 0.0170.177 0.41 0.10 – 1.72 0.2240.859 0.94 0.13 – 6.88 0.9490.358 0.83 0.67 – 1.04 0.0990.917 0.98 0.62 – 1.56 0.944IV serostatus, age, and unstable housing.pharmocotherapies including dopamine agonists, seroto-nergic agents and GABA-ergic medications and a cocainevaccine. One promising randomized controlled trial from2013 demonstrated that topiramate was more efficaciousthan placebo at reducing weekly cocaine use [40]. Add-itionally, contingency management strategies have beenpresented as possible interventions. One such studyshowed that employment-based abstinence reinforce-ment could lead to increased cocaine abstinence [41].However, these strategies have been found to be challen-ging to implement [38]. Future research should continueto seek to identify a novel therapeutic intervention modelto reduce morbidity and mortality among cocaine injec-tors in this setting.A recent systematic review of mortality among PWIDsfound no significant differences in the risk of death bytype of primary drug injected [8]. The authors notedsome factors that reduced their capacity to detect suchdifferences, including a lack of adjustment for poly-Hayden et al. BMC Public Health 2014, 14:1153 Page 6 of 7http://www.biomedcentral.com/1471-2458/14/1153substance use, changes in participants’ drug use habitsthat may not have been accounted for, and the fact thatHIV status was often imprecisely measured [8]. Anothersystematic review of mortality among cocaine usersnoted limited data on the extent of elevated mortalityamong cocaine users [4]. They highlighted limitations ofprevious studies, including the fact that many cohortswere formed in drug treatment facilities with dependentcocaine users who were at increased risk of prematuredeath and that the prevalence of HIV infection betweenstudies was highly variable [4]. Our study addressedsome of the limitations of previous studies by accountingfor poly-substance use, following up with participantssemi-annually to monitor ongoing drug use habits, andtesting for HIV infection at each visit. Furthermore, un-like many previous studies, the present study populationwas not selected from a drug treatment facility or com-prised of cross-sectional examinations of deceased indi-vidual’s toxicology reports.There are several limitations in this study. Firstly, thestudy sample is not a random sample, and thereforegeneralizability of our findings may be limited. Secondly,much of the data, particularly regarding drug use pat-terns, were ascertained through self–reporting. There-fore, our data may have been affected by reportingbiases including recall bias and socially desirable report-ing. However, we note that this type of data has beencommonly utilized in observational studies involvingPWIDs and found to be valid [42,43]. Third, drug usepatterns directly before or at the time of death are notdescribed due to the method of self-reporting, which in-hibits our ability to ascertain data as proximal to thetime of death. Fourth, although migration rates out ofthe province have been shown to be low, mortality ratesmay be underestimated, as participants who died outsideof BC were not included in the provincial registry. Wecensored individuals with long durations (>12 months)from behavioural measurement to death on the date oflast behavioural measurement. This may have resulted inloss of precision regarding our risk factor measurement.Fifth, for many variables, including alcohol use, we did nothave measures that considered refined patterns of use orfrequency of use and were often dichotomized into dailyversus less than daily use. Lastly, when the endpoint wasrestricted to accidental mortality, daily cocaine injectingdid not remain significantly associated with the time toaccidental death, despite a trend towards an association.This may reflect that accidental deaths, including acci-dental poisonings, assault and accidental falls, are not ne-cessarily predicted by specific drug use patterns. Futureresearch should seek to identify predictors of accidentalmortality among PWID. Finally, certain patterns of druguse, such as intranasal cocaine use, are uncommon in thissample and were not measured in our study.ConclusionsIn summary, daily cocaine injecting was the only druguse pattern independently associated with all-cause mor-tality among a sample of PWIDs in Vancouver. Thesefindings highlight the need to identify novel effectivetreatments and harm reduction strategies for cocaineinjectors.AbbreviationsAHR: Adjusted hazard ratio; ACCESS: AIDS Care Cohort to Evaluate Access toSurvival Services; CI: Confidence interval; DTES: Downtown Eastside;HCV: Hepatitis C virus; HIV: Human immunodeficiency virus; PWIDs: Personswho inject drugs; VIDUS: Vancouver Injection Drug Users Study.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsAH analyzed the data and drafted the manuscript. KH participated in thedesign of the study and helped to draft the manuscript. HD performed thestatistical analysis and helped to draft the manuscript. MM, TK and JMparticipated in the study design and edited the drafted manuscript. EWconceived of the study, participated in its design and helped to draft themanuscript. All authors read and approved of the final manuscript.AcknowledgementsThe authors thank the study participants for their contribution to theresearch, as well as current and past researchers and staff. The study wassupported by the US National Institutes of Health (VIDUS: R01DA011591,ACCESS: R01DA021525). This research was undertaken, in part, thanks tofunding from the Canada Research Chairs program through a Tier 1 CanadaResearch Chair in Inner City Medicine which supports Dr. Wood. Dr. Milloy issupported by the Canadian Institutes of Health Research and the MichaelSmith Foundation for Health Research. Dr. Montaner is supported by theBritish Columbia Ministry of Health and by the US National Institutes ofHealth (R01DA036307). He has also received limited unrestricted fundingfrom Abbvie, Bristol-Myers Squibb, Gilead Sciences, Janssen, Merck, and ViiVHealthcare.Received: 8 May 2014 Accepted: 9 October 2014Published: 6 November 2014References1. Hulse GK, English DR, Milne E, Holman CD: The quantification of mortalityresulting from the regular use of illicit opiates. Addiction 1999,94(2):221–229.2. Coffin PO, Galea S, Ahern J, Leon AC, Vlahov D, Tardiff K: Opiates, cocaineand alcohol combinations in accidental drug overdose deaths in NewYork City, 1990–98. Addiction 2003, 98(6):739–747.3. Stoove MA, Dietze PM, Aitken CK, Jolley D: Mortality among injecting drugusers in Melbourne: a 16-year follow-up of the Victorian Injecting CohortStudy (VICS). Drug Alcohol Depend 2008, 96(3):281–285.4. Degenhardt L, Singleton J, Calabria B, McLaren J, Kerr T, Mehta S, Kirk G, HallWD: Mortality among cocaine users: a systematic review of cohortstudies. Drug Alcohol Depend 2011, 113(2–3):88–95.5. Degenhardt L, Bucello C, Mathers B, Briegleb C, Ali H, Hickman M, McLarenJ: Mortality among regular or dependent users of heroin and otheropioids: a systematic review and meta-analysis of cohort studies.Addiction 2011, 106(1):32–51.6. Singleton J, Degenhardt L, Hall W, Zabransky T: Mortality amongamphetamine users: a systematic review of cohort studies. Drug AlcoholDepend 2009, 105(1–2):1–8.7. Davidson PJ, McLean RL, Kral AH, Gleghorn AA, Edlin BR, Moss AR: Fatalheroin-related overdose in San Francisco, 1997–2000: a case for targetedintervention. J Urban Health 2003, 80(2):261–273.8. Mathers BM, Degenhardt L, Bucello C, Lemon J, Wiessing L, Hickman M:Mortality among people who inject drugs: a systematic review andmeta-analysis. Bull World Health Organ 2013, 91(2):102–123.evaluation of a pilot medically supervised safer injecting facility. CMAJ2006, 175(11):1399–1404.36. Wood E, Tyndall MW, Zhang R, Stoltz JA, Lai C, Montaner JS, Kerr T:Attendance at supervised injecting facilities and use of detoxificationservices. N Engl J Med 2006, 354(23):2512–2514.37. Debeck KKT, Bird L, Zhang R, Marsh D, Tyndall M, Montaner J, Wood E:Injection drug use cessation and use of North America’s first medicallysupervised safer injecting facility. Drug Alcohol Depend 2011,Hayden et al. BMC Public Health 2014, 14:1153 Page 7 of 7http://www.biomedcentral.com/1471-2458/14/11539. Warner-Smith M, Darke S, Day C: Morbidity associated with non-fatalheroin overdose. Addiction 2002, 97(8):963–967.10. Hickman M, Carnwath Z, Madden P, Farrell M, Rooney C, Ashcroft R, Judd A,Stimson G: Drug-related mortality and fatal overdose risk: pilot cohortstudy of heroin users recruited from specialist drug treatment sites inLondon. J Urban Health 2003, 80(2):274–287.11. Arendt M, Munk-Jorgensen P, Sher L, Jensen SO: Mortality followingtreatment for cannabis use disorders: predictors and causes.J Subst Abuse Treat 2013, 44(4):400–406.12. Tyndall MW, Craib KJ, Currie S, Li K, O’Shaughnessy MV, Schechter MT:Impact of HIV infection on mortality in a cohort of injection drug users.J Acquir Immune Defic Syndr 2001, 28(4):351–357.13. Merrall EL, Bird SM, Hutchinson SJ: Mortality of those who attended drugservices in Scotland 1996–2006: record-linkage study. Int J Drug Policy2012, 23(1):24–32.14. Bartu A, Freeman NC, Gawthorne GS, Codde JP, Holman CD: Mortality in acohort of opiate and amphetamine users in Perth, Western Australia.Addiction 2004, 99(1):53–60.15. Kaye S, Darke S: Non-fatal cocaine overdose among injecting andnon-injecting cocaine users in Sydney, Australia. Addiction 2004,99(10):1315–1322.16. Barrio G, Molist G, de la Fuente L, Fernandez F, Guitart A, Bravo MJ,Brugal MT: Mortality in a cohort of young primary cocaine users:controlling the effect of the riskiest drug-use behaviors. Addict Behav2013, 38(3):1601–1604.17. Dias AC, Araujo MR, Dunn J, Sesso RC, de Castro V, Laranjeira R: Mortalityrate among crack/cocaine-dependent patients: a 12-year prospectivecohort study conducted in Brazil. J Subst Abuse Treat 2011, 41(3):273–278.18. Maas B, Fairbairn N, Kerr T, Li K, Montaner JS, Wood E: Neighborhood andHIV infection among IDU: place of residence independently predicts HIVinfection among a cohort of injection drug users. Health Place 2007,13(2):432–439.19. Wood EKT: What do you do when you hit rock bottom? Responding todrugs in the city of Vancouver. Int J Drug Policy 2006, 17(2):55–60.20. Tyndall MW, Currie S, Spittal P, Li K, Wood E, O’Shaughnessy MV, SchechterMT: Intensive injection cocaine use as the primary risk factor in theVancouver HIV-1 epidemic. Aids 2003, 17(6):887–893.21. Werb D, Debeck K, Kerr T, Li K, Montaner J, Wood E: Modelling crackcocaine use trends over 10 years in a Canadian setting. Drug Alcohol Rev2010, 29(3):271–277.22. Fairbairn N, Kerr T, Buxton JA, Li K, Montaner JS, Wood E: Increasing useand associated harms of crystal methamphetamine injection in aCanadian setting. Drug Alcohol Depend 2007, 88(2–3):313–316.23. Wood E, Montaner JS, Li K, Barney L, Tyndall MW, Kerr T: Rate ofmethadone use among Aboriginal opioid injection drug users.CMAJ 2007, 177(1):37–40.24. Strathdee SA, Palepu A, Cornelisse PG, Yip B, O’Shaughnessy MV, MontanerJS, Schechter MT, Hogg RS: Barriers to use of free antiretroviral therapy ininjection drug users. JAMA 1998, 280(6):547–549.25. Milloy MJ, Kerr T, Buxton J, Rhodes T, Krusi A, Guillemi S, Hogg R, MontanerJ, Wood E: Social and environmental predictors of plasma HIV RNArebound among injection drug users treated with antiretroviral therapy.J Acquir Immune Defic Syndr 2012, 59(4):393–399.26. Spittal PM, Craib KJ, Wood E, Laliberte N, Li K, Tyndall MW, O’ShaughnessyMV, Schechter MT: Risk factors for elevated HIV incidence rates amongfemale injection drug users in Vancouver. CMAJ 2002, 166(7):894–899.27. 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(1):19–24.28. Arendt M, Munk-Jorgensen P, Sher L, Jensen SO: Mortality amongindividuals with cannabis, cocaine, amphetamine, MDMA, and opioiduse disorders: a nationwide follow-up study of Danish substance usersin treatment. Drug Alcohol Depend 2011, 114(2–3):134–139.29. Fugelstad A, Annell A, Rajs J, Agren G: Mortality and causes andmanner of death among drug addicts in Stockholm during the period1981–1992. Acta Psychiatr Scand 1997, 96(3):169–175.30. White JM, Irvine RJ: Mechanisms of fatal opioid overdose. Addiction 1999,94(7):961–972.113(2–3):172–176.38. de Lima MS, de Oliveira Soares BG, Reisser AA, Farrell M: Pharmacologicaltreatment of cocaine dependence: a systematic review. Addiction 2002,97(8):931–949.39. Castells X, Casas M, Perez-Mana C, Roncero C, Vidal X, Capella D: Efficacy ofpsychostimulant drugs for cocaine dependence. Cochrane Database SystRev 2010, 2, CD007380.40. Johnson BA-DN, Wang X, Penberthy K, Javors M, Seneviratne C: Topiramatefor the treatment of cocaine addiction a randomized clinical trial. JAMAPsychiatry 2013, 70(12):1338–1346.41. Silverman K, Wong CJ, Needham M, Diemer KN, Knealing T,Crone-Todd D, Fingerhood M, Nuzzo P, Kolodner K: A randomized trial ofemployment-based reinforcement of cocaine abstinence in injectiondrug users. J Appl Behav Anal 2007, 40(3):387–410.42. Weatherby NR NL, Cesari H, Booth R, McCoy CB, Watters JK, Williams M,Chitwood DD: Validity of self-reported drug use among injection drugusers and crack cocaine users recruited through street outreach.Eval Program Plann 1994, 17(4):347–355.43. Darke S: Self-report among injecting drug users: a review.Drug Alcohol Depend 1998, 51(3):253–263. discussion 267–258.doi:10.1186/1471-2458-14-1153Cite this article as: Hayden et al.: The impact of drug use patterns onmortality among polysubstance users in a Canadian setting: aprospective cohort study. BMC Public Health 2014 14:1153.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 redistribution31. 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(9):585–589.32. Marshall BD, Wood E: Putting risk compensation to rest: reframing therelationship between risk behavior and antiretroviral therapy amonginjection drug users. AIDS 2012, 26(18):2405–2407.33. Marshall BD, Milloy MJ, Wood E, Montaner JS, Kerr T: Reduction inoverdose mortality after the opening of North America’s first medicallysupervised safer injecting facility: a retrospective population-basedstudy. Lancet 2011, 377(9775):1429–1437.34. Kerr T, Tyndall M, Li K, Montaner J, Wood E: Safer injection facility use andsyringe sharing in injection drug users. Lancet 2005, 366(9482):316–318.35. Wood E, Tyndall MW, Montaner JS, Kerr T: Summary of findings from theSubmit your manuscript at www.biomedcentral.com/submit


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