UBC Faculty Research and Publications

Impact of unstable housing on all-cause mortality among persons who inject drugs Zivanovic, Rebecca; Milloy, MJ; Hayashi, Kanna; Dong, Huiru; Sutherland, Christy; Kerr, Thomas; Wood, Evan Feb 7, 2015

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

Item Metadata


52383-12889_2015_Article_1479.pdf [ 419.15kB ]
JSON: 52383-1.0132549.json
JSON-LD: 52383-1.0132549-ld.json
RDF/XML (Pretty): 52383-1.0132549-rdf.xml
RDF/JSON: 52383-1.0132549-rdf.json
Turtle: 52383-1.0132549-turtle.txt
N-Triples: 52383-1.0132549-rdf-ntriples.txt
Original Record: 52383-1.0132549-source.json
Full Text

Full Text

RESEARCH ARTICLEImpact of unstable housincDfnt tnationally in many settings over the last two decades [1].Understanding the factors which increase the likelihoodreported between four and six times that of age- andgender- matched peers [5]. Premature mortality in PWIDZivanovic et al. BMC Public Health  (2015) 15:106 DOI 10.1186/s12889-015-1479-xprevention and treatment of substance use disorders could608-1081 Burrard St, Vancouver, BC V6Z 1Y6, Canada2Department of Medicine, University of British Columbia, Vancouver, Canadaof premature mortality among persons who inject drugs(PWID) could inform the development of effective pro-grams to improve both individual and community health.The criminalization of people who use illicit drugsmakes accurate study of the prevalence of illicit drug useis the result of a variety of causes including both acute andchronic diseases, accidents, violence and accidental over-dose [3,6,7]. A recent systematic review and meta-analysisof mortality in PWID showed the two most commoncauses of death to be AIDS and overdose [6].Homelessness and unstable housing often co-occur withillicit drug use [8,9]. Drug-use disorders independentlyincrease the risk of first-time homelessness, suggesting* Correspondence: uhri-ew@cfenet.ubc.ca1British Columbia Centre for Excellence in HIV/AIDS, St Paul’s Hospital,housing and all-cause mortality among PWIDs living in Vancouver, Canada.Methods: PWIDs participating in two prospective cohort studies in Vancouver, Canada were followed betweenMay 1996 and December 2012. Cohort data were linked to the provincial vital statistics database to ascertain mortalityrates and causes of death. We used multivariate Cox proportional hazards regression to determine factors associatedwith all-cause mortality and to investigate the independent relationship between unstable housing and time toall-cause mortality.Results: During the study period, 2453 individuals were followed for a median of 69 months (Inter-quartile range[IQR]: 34 – 113). In total, there were 515 (21.0%) deaths for an incidence density of 3.1 (95% Confidence Interval[CI]: 2.8 – 3.4) deaths per 100 person years. In multivariate analyses, after adjusting for potential confoundersincluding HIV infection and drug use patterns, unstable housing remained independently associated with all-causemortality (adjusted hazard ratio [AHR] = 1.30, 95% CI: 1.08 – 1.56).Conclusions: These findings demonstrate that unstable housing is an important risk factor for mortality independentof known risk factors including HIV infection and patterns of drug use. This study highlights the urgent need to providesupportive housing interventions to address elevated levels of preventable mortality among this population.Keywords: Unstable housing, Injection drug use, MortalityBackgroundThe link between high-intensity illicit drug use and elevatedrates of preventable morbidity and mortality has been welldocumented. Furthermore, rates of illicit drug use anddrug–related deaths have generally increased inter-and health-related outcomes, including mortality, challen-ging [2]. Nevertheless, mortality rates in studies of opioidusers have been reported between six and 30 times that ofage- and gender- matched non-opioid using individuals[3,4]. Similarly, mortality rates in cocaine users have beenamong persons who injeRebecca Zivanovic1,2, MJ Milloy1,2, Kanna Hayashi1,2, Huiruand Evan Wood1,2*AbstractBackground: Illicit drug injecting is a well-established risknumber of prospective studies have examined the indepepersons who inject drugs (PWIDs). In this study we sough© 2015 Zivanovic et al.; licensee BioMed CentCommons Attribution License (http://creativecreproduction in any medium, provided the orDedication waiver (http://creativecommons.orunless otherwise stated.Open Accessg on all-cause mortalityt drugsong1, Christy Sutherland1, Thomas Kerr1,2actor for morbidity and mortality. However, a limiteddent effect of unstable housing on mortality amongo identify if a relationship exists between unstableral. 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,Zivanovic et al. BMC Public Health  (2015) 15:106 Page 2 of 7have a positive effect on homelessness [10]. In addition,homelessness has been linked with subsequent increasesin drug use, injection-related risk behaviour and relapse inthose who have stopped injecting drugs suggesting thateffective interventions to promote housing have the po-tential to decrease the negative effects of drug use [11,12].The role of homelessness as a risk factor for mortalityhas been well studied [13,14]. Large cohort studies in theUnited States show age-adjusted mortality among home-less persons to be three to nine times that of the generalpopulation [13,14]. Even after removing those with sub-stance abuse from the group the difference remained threetimes greater than the general population [13]. A studyconducted in Toronto showed that homeless women be-tween the ages of 18 to 44 were 10 times more likely to diethan women in the general population [15]. More recently,homelessness was shown to be an independent risk factorfor mortality after adjustment for socioeconomic statusand morbidity [16].While studies have described the impacts of drug use pat-terns and comorbidities like HIV infection on mortalityamong persons who use drugs [17], few studies have beenable to examine the effects of unstable housing on mortalityafter adjustment for potential confounders like prospectivemeasures of drug use patterns and HIV infection especiallyin settings with universal health care systems where HIVtreatment and care is provided free of charge. Therefore, thepresent study was conducted to examine the independentrelationship, if any, between unstable housing and all-causemortality among PWID living in Vancouver, Canada.MethodsThe Vancouver Injection Drug Users Study (VIDUS) andAIDS Care Cohort to Evaluate Access to Survival Services(ACCESS) are open prospective cohorts of illicit drug usersin Vancouver. The recruitment and follow up proceduresfor the two studies are largely identical to allow for analysesof merged data, with the only key differences being thatHIV positive individuals are followed in ACCESS whereasHIV-negative individuals are followed in VIDUS. In bothstudies the primary modes of enrollment were self-referral,word of mouth, and street outreach. Detailed sampling andrecruitment procedures for these two cohorts have beendescribed elsewhere [18,19].To be eligible, participants were 18 years of age or older,had ever injected illicit drugs and resided in the greaterVancouver region. Participants with no completed followup surveys were excluded. All participants provided writ-ten informed consent. Participants were given a $20 sti-pend at each study visit for their time and transportation.The study was approved by the University of BritishColumbia/Providence Healthcare Research Ethics Board.At baseline and semianually thereafter, participantscompleted an interviewer-administered questionnairethat elicited a range of data, including demographic charac-teristics, housing status, injection and non-injection druguse, and sexual risk behaviors. In addition, venous bloodsamples were drawn at each visit and tested for HIV andhepatitis C virus (HCV) antibodies among those previouslytesting negative for these diseases. Venous blood samplestaken from HIV positive individuals were assessed for HIVdisease progression [18]. All participants had private inter-views and were offered both pre- and post-test counselingwith trained nurses. Referral for free healthcare was pro-vided to those who tested HIV positive and these individ-uals were subsequently followed in ACCESS.The present study included PWID who were recruitedand completed at least one follow up visit between May1996 and December 2012. To avoid potential bias relatingto long durations between the last study visit where be-havioural information was assessed and the date of death,individuals who were identified as deceased more than24 months after their last follow up visit were censored onthe date of the last follow up.The primary endpoint in this analysis was all-cause mor-tality. Here, we ascertained all-cause mortality rates amongparticipants through a confidential record linkage usingpersonal health numbers with the British Columbia VitalStatistics Agency and through ongoing follow up with con-tacts provided by participants. The primary explanatoryvariable of interest was unstable housing in the previous sixmonths. As previously, unstable housing was defined asliving in a single room occupancy hotel, shelter or othertransitional housing, or living on the street [20,21].Potential confounders that were considered includedgender (male vs. female); age (per year older); ancestry(Caucasian vs. non-Caucasian); HIV serostatus (positivevs. negative); and sex work involvement (yes vs. no). Anumber of substance use behaviors (in the previous sixmonths) were also considered, including ≥ daily heroin in-jection (yes vs. no), ≥ daily cocaine injection (yes vs. no), ≥daily crack cocaine smoking (yes vs. no), and current en-rolment in a methadone program (yes vs. no). Other co-variates that were considered included incarceration (yesvs. no) and HCV serostatus (positive vs. negative). Withthe exception of age, gender and ancestry, all variableswere in reference to the prior six months and measured ateach semiannual follow up visit and were treated as time-updated.As a first step, we used Chi-square test and Wilcoxonrank sum test to compare the baseline characteristics ofthe participants who did and did not report unstablehousing at baseline. Those who did not report unstablehousing at baseline were maintained as the referencegroup. All-cause mortality rate and 95% confidenceinterval [CI] were calculated using the Poisson distribu-tion. Survival probabilities from all-cause mortalitywere estimated using the Kaplan-Meier product limitmethod, and compared using the two-sample log-ranktest.Next, we used bivariate Cox proportional hazards re-gression to examine the associations between each ex-planatory variable and time to all-cause mortality. To fitthe multivariate model, we employed a conservative step-wise backward selection approach which considered themagnitude of change in the coefficient of unstable housing[22]. Specifically, we included all variables found to beassociated with time to all-cause mortality in bivariateanalyses at p < 0.10 in a multivariate model and used astepwise approach to fit a series of reduced models. Aftercomparing the value of the coefficient associated withunstable housing in the full model to the value of thecoefficient in each of the reduced models, we dropped thesecondary variable associated with the smallest relativechange. We continued this iterative process until the mini-mum change exceeded 5%. Remaining variables were con-sidered as potential confounders in a final multivariatemodel. All statistical analyses were performed usingTable 1 Baseline characteristics of the study participants stratified by unstable housing at baseline (n = 2453)1Unstable housingCharacteristic Total n (%) Yes 1713 (69.8) No2 722 (29.4) Odds ratio (95% CI) p-valueGenderMale 1620 (66.0) 1147 (67.0) 462 (64.0) 1.14 (0.95-1.37) 0.158Female 833 (34.0) 566 (33.0) 260 (36.0)Age, in years (median IQR) 37.7 (29.7 – 44.1) 37.8 (30.0 – 44.1) 37.6 (29.6 – 44.0) 1.00 (0.99-1.01) 0.399AncestryCaucasian 1496 (61.0) 1058 (61.8) 430 (59.6) 1.10 (0.92-1.31) 0.308Other 957 (39.0) 655 (38.2) 292 (40.4)HIV serostatusPositive 758 (30.9) 551 (32.2) 200 (27.7) 1.23 (1.02-1.50) 0.032Negative 1692 (69.0) 1161 (67.8) 520 (72.0)Sex work involvement3Yes 560 (22.8) 410 (23.9) 143 (19.8) 1.27 (1.03-1.58) 0.026No 1883 (76.8) 1296 (75.7) 576 (79.8)Daily heroin injection3Yes 909 (37.1) 660 (38.5) 242 (33.5) 1.24 (1.04-1.49) 0.020No 1538 (62.7) 1049 (61.2) 478 (66.2)Daily cocaine injection3Yes 733 (29.9) 560 (32.7) 164 (22.7) 1.65 (1.35-2.02) < 0.001No 1704 (69.5) 1143 (66.7) 552 (76.5)Daily non-injection cocaine use3Yes 590 (24.1) 496 (29.0) 90 (12.5) 2.87 (2.25-3.66) < 0.001No 1860 (75.8) 1214 (70.9) 632 (87.5)Enrolment in a methadone program3e qZivanovic et al. BMC Public Health  (2015) 15:106 Page 3 of 7Yes 559 (22.8) 373 (21.8)No 1885 (76.8) 1332 (77.8)Incarceration3Yes 326 (13.3) 255 (14.9)No 2122 (86.5) 1454 (84.9)HCV serostatusPositive 2025 (82.6) 1450 (84.7)Negative 411 (16.8) 252 (14.7)1. Not all cells add up to 2453 as participants may choose not to answer sensitiv2. The reference group is those who do not report unstable housing at baseline.3. Refers to behaviors in the last six months.IQR = inter-quartile range.184 (25.5) 0.82 (0.67-1.00) 0.051537 (74.4)67 (9.3) 1.71 (1.29-2.28) <0.001654 (90.6)562 (77.8) 1.59 (1.27-1.98) <0.001155 (21.5)uestions.SAS software version 9.3 (SAS, Cary, NC). All p-valueswere two-sided.ResultsA total of 2742 eligible individuals were recruited betweenMay 1996 and December 2012. Two hundred eighty-nineparticipants were excluded based on the eligibility criteriadescribed above leaving a sample of 2453 (89.5%) for fur-ther analysis. These individuals were followed for a me-dian of 69 months (Inter-quartile range [IQR]: 34 to113 months). Compared to the analytic sample, those noteligible were younger, less likely to be HCV seropositive,and less likely to use crack cocaine (all p <0.05), but therewas no difference by housing status (p = 0.338). Amongthe study sample, 515 (21.0%) individuals died for an inci-dence density of mortality of 3.1 (95% CI: 2.8 – 3.4) deathsper 100 person years.At baseline, 1602 participants (66.0%) were men, 758(30.9%) were HIV positive, 2025 (82.6%) were HCV anti-body positive and 1496 (61.0%) reported Caucasian an-ratio [OR] = 1.23, 95% CI: 1.02 – 1.50) and to be HCV sero-positive (OR = 1.59, 95% CI: 1.27-1.98). They were alsomore likely to have been involved in sex work (OR = 1.27,95% CI: 1.03-1.58) and have been incarcerated (OR = 1.71,95% CI: 1.29-2.28). They were less likely to be enrolled in amethadone program (OR= 0.82, 95% CI: 0.67 – 1.00). Interms of drug use patterns, unstable housing was statisti-cally significant and positively associated with daily heroininjection (OR = 1.24, 95% CI: 1.04-1.49), daily cocaine injec-tion (OR = 1.65, 95% CI: 1.35 – 2.02) and daily crack smok-ing (OR = 2.87, 95% CI: 2.25 – 3.66).Figure 1 shows the results of the Kaplan-Meier analysisof time to all-cause mortality stratified by baseline housingstatus. As shown, individuals reporting unstable housingwere significantly more likely to die during follow up thanindividuals reporting stable housing at baseline (p < 0.001).Table 2 shows results of the bivariate and multivariate Coxregression analyses of time to all-cause mortality. In the bi-variate analysis, unstable housing was statistically significantand positively associated with time to all-cause mortalitydependently associated with all-cause mortality (adjustedZivanovic et al. BMC Public Health  (2015) 15:106 Page 4 of 7cestry. The median age was 37.7 years (IQR: 29.7-44.1).Five hundred sixty (22.8%) had histories of sex work and326 (13.3%) had ever been incarcerated. At baseline, 909(37.1%) of the study sample injected heroin at least dailyin the previous six months, 733 (29.9%) injected cocaineat least daily, 590 (24.1%) smoked crack cocaine daily,and 559 (22.8%) were enrolled in a methadone program.Table 1 reports the baseline characteristics of the studyparticipants stratified by unstable housing. Compared tothose with stable housing, those who reported unstablehousing were more likely to be HIV seropositive (oddsFigure 1 Probability of survival for PWID with stable housing vs thoshazard ratio [AHR] = 1.30, 95% CI: 1.08 – 1.56).DiscussionThe aim of this study was to investigate the relationshipbetween unstable housing and risk of death among peoplewith a relative hazard (RH) of 1.37 (95% CI: 1.14-1.63).As also shown in Table 2, in multivariate analyses, afteradjusting for potential confounders including HIV infec-tion and drug use patterns, unstable housing remained in-e with unstable housing.who inject drugs. We demonstrated that among a largecohort of PWID, unstable housing was independently asso-ciated with all-cause mortality. This association persistedafter adjustment for a range of prospectively measured po-tential confounding variables including high-intensity(greater than or equal to daily) drug-use behaviours andHIV serostatus.When stratified by housing status the baseline character-istics of the two groups showed that those in the unstablehousing group were more likely to be HCV seropositive,HIV seropositive, involved in sex work or to have been in-carcerated. These individuals with unstable housing werealso less likely to be involved in a methadone program andmore likely to report at least daily heroin injection, at leastdaily cocaine injection and at least daily crack smoking.These characteristics illustrate specific examples of the dailychallenges faced by those living with unstable housing com-pared to their housed counterparts. These social, behav-ioural and medical factors associated with the group ofPWID reporting unstable housing also attests to the com-plexity of both living as and studying this population.Interestingly, even after adjustment for these competingrisks of death, our analyses demonstrated a positive associ-ation between unstable housing and mortality compared totheir housed counterparts suggesting that, despite all of thefactors at play in the challenging lives of PWID, housingstability alone can have a substantial impact on survival.Our findings of increased mortality in PWID with un-stable housing are consistent with the available literaturesupporting increased mortality in homeless populations ingeneral [13,15]. Specifically, the present research adds im-portant data on the subset of homeless individuals whoalso use drugs and demonstrates the important role ofhousing on survival. Our data supports existing literaturehighlighting the importance of framing homelessness as ahealth issue, and creating public health interventions aimedat improving the health and longevity of drug users by ad-dressing various social and environmental factors, includingTable 2 Bivariate and multivariate Cox proportional hazards analyses of the time to death among 2453 PWIDUnadjusted AdjustedRelative hazard (RH) Relative hazard (ARH)Variable RH (95% CI) p-value ARH (95% CI) p-valueUnstable housing*(Yes vs. No) 1.37 1.14-1.63 <0.001 1.30 1.08-1.56 0.005Gender(Male vs. Female) 1.09 0.91-1.30 0.353Age(per 1 year older) 1.03 1.02-1.04 <0.00188200877Zivanovic et al. BMC Public Health  (2015) 15:106 Page 5 of 7Ethnicity(Caucasian vs. Other) 1.07 0.90-1.2HIV serostatus(Positive vs. Negative) 2.67 2.24-3.1Sex work involvement*(Yes vs. No) 0.76 0.57-1.0Daily heroin injection*(Yes vs. No) 0.80 0.64-1.0Daily cocaine injection*(Yes vs. No) 1.44 1.16-1.8Daily Non-injection cocaine use*(Yes vs. No) 0.89 0.73-1.0Enrolment in a methadone program*(Yes vs. No) 0.81 0.68-0.9Incarceration*(Yes vs. No) 0.92 0.72-1.1HCV serostatus(Positive vs. Negative) 1.95 1.30-2.94*Behaviours refer to activities in the last six months.0.440<0.001 2.56 2.14-3.07 <0.0010.0710.051 0.83 0.66-1.04 0.1080.001 1.36 1.08-1.71 0.0090.2380.0200.4860.001Zivanovic et al. BMC Public Health  (2015) 15:106 Page 6 of 7homelessness, that worsen the negative consequences ofdrug use [23].Housing is included with food, clothing, medical careand necessary social services under Article 25 of theUnited Nations Declaration of Human Rights [24]. Howeverprovision lags behind and often comes with restrictions andstipulations around sobriety, health status, employment andincome. One specific example of a novel approach at workis “Housing First” which involves providing access to per-manent housing and the services people may desire tomaintain that housing. This program works on the philoso-phy shared by the UN that housing is a basic human rightand then the many factors that may have contributed to aperson’s homelessness can be best addressed once theperson has stable housing [25].Our study has several limitations. First, the study sam-ple was not randomly selected and our findings may notbe generalizable to all PWID. Second, although the self-reported data may have been affected by reporting bias,we do not believe that individuals could have differen-tially reported housing status based on their time to all-cause-mortality, which was ascertained through adminis-trative data. Self-reported data to control for potentialconfounding has been commonly used and found to bevalid in studies involving PWID [26]. Third, as with allobservational studies, the relationship observed betweenmortality and unstable housing may be under the influ-ence of unobserved confounding. Finally, mortality ratesmay have been underestimated as participants who diedoutside of the province were not included in the provin-cial registry and therefore not accounted for. However,this is unlikely to impact our findings as previous studieshave shown that migration rates out of province are verylow in this setting [27].ConclusionsIn summary, the present study highlights the significantrole of unstable housing in increasing premature mortal-ity in PWID. This demonstrates the need for implemen-tation of supportive housing programs to be an integralpart of public health initiatives aimed at improving thehealth of populations of PWID.AbbreviationsACCESS: AIDS Care Cohort to Evaluate Access to Survival Services;AHR: Adjusted Hazard Ratio; AIDS: Acquired Immunodeficiency Syndrome;CI: Confidence Interval; HCV: Hepatitis C Virus; HIV: Human ImmunideficiencyVirus; IQR: Inter-quartile range; PWID: Persons who inject drugs; OR: OddsRatio; UN: United Nations; VIDUS: Vancouver Injection Drug Users Study.Competing interestsThe study was supported by the US National Institutes of Health (VIDUS:R01DA011591, ACCESS: R01DA021525). This research was undertaken, in part,thanks to funding from the Canada Research Chairs program through a Tier1 Canada Research Chair in Inner City Medicine which supports Dr. EvanWood. Dr. Milloy is supported in part by the United States National Institutesof Heath (R01DA02152525). The authors declare that they have nocompeting interests.Author’s contributionsRZ was the primary author of this manuscript, drafting all sections andcoordinating with all authors. MM provided critical appraisal and hasapproved the submitted manuscript. HD provided the statistical analysis,guidance over methodology and overall appraisal. She has approved thesubmitted manuscript. CS, KH and TK provided critical appraisal andfeedback and have approved the submitted manuscript. EW madesubstantial contributions to methodology and content. He also providedcritical appraisal and feedback and has approved the submitted manuscript.All authors read and approved 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. Evan Wood. Dr. Milloyis supported in part by the National Institutes of Health (R01 DA021525).Dr. Hayashi is supported by the Canadian Institutes of Health Research.Received: 9 October 2014 Accepted: 27 January 2015References1. Darke S, Degenhardt L, Mattick R. Mortality Amongst Illicit Drug Users.Cambridge: Cambridge University Press; 2007.2. Degenhardt L, Hall W. Extent of illicit drug use and dependence, and theircontribution to the global burden of disease. Lancet. 2012;379(9810):55–70.3. Oppenheimer E, Tobutt C, Taylor C, Andrew T. Death and survival in acohort of heroin addicts from London clinics: a 22-year follow-up study.Addiction. 1994;89(10):1299–308.4. Clausen T, Waal H, Thoresen M, Gossop M. Mortality among opiate users:opioid maintenance therapy, age and causes of death. Addiction.2009;104(8):1356–62.5. Degenhardt L, Singleton J, Calabria B, Kerr T, Mehta S, Kirk G, et al. Mortalityamong concaine users: a systematic review of cohort studies. Drug AlcoholDepend. 2011;113(2–3):88–95.6. 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):102–23.7. Gossop M, Stewart D, Treacy S, Marsden J. A prospective study of mortalityamong drug misusers during a 4-year period after seeking treatment.Addiction. 2002;97(1):39–47.8. Aidala A, Cross JE, Stall R, Harre D, Sumartojo E. Housing status and HIV riskbehaviors: implications for prevention and policy. AIDS Behav. 2005;9(3):251–6.9. Tompsett CJ, Domoff SE, Toro PA. Peer substance use and homelessnesspredicting substance abuse from adolescence through early adulthood.Am J Community Psychol. 2013;51(3–4):520–9.10. Thompson Jr RG, Wall MM, Greenstein E, Grant BF, Hasin DS. Substance-usedisorders and poverty as prospective predictors of first-time homelessnessin the United States. Am J Public Health. 2013;103 Suppl 2:S282–8.11. Topp L, Iversen J, Baldry E, Maher L, Collaboration of Australian NSP.Housing instability among people who inject drugs: results from theAustralian needle and syringe program survey. J Urban Health.2013;90(4):699–716.12. Linton SL, Celentano DD, Kirk GD, Mehta SH. The longitudinal associationbetween homelessness, injection drug use, and injection-related risk behavioramong persons with a history of injection drug use in Baltimore, MD. DrugAlcohol Depend. 2013;132(3):457–65.13. Hibbs J, Benner L, Klugman L, Spencer R, Macchia I, Mellinger A, et al.Mortality in a cohort of homeless adults in Philadelphia. N Engl J Med.1994;331(5):304–9.14. Baggett T, Hwang S, O'Connell J, Porneala B, Stringfellow E, Orav J, et al.Mortality among homeless adults in Boston: Shifts in causes of death over a15-year period. JAMA Intern Med. 2013;173(3):189–95.15. Cheung A, Hwang S. Risk of death among homeless women: a cohort studyand review of the literature. Can Med Assoc J. 2004;170(8):1243–7.16. Morrison DS. Homelessness as an independent risk factor for mortality:results from a retrospective cohort study. Int J Epidemiol. 2009;38(3):877–83.17. Tyndall M, Craib K, Currie S, Li K, O'Shaughnessy M, Schechter M. Impact ofHIV infection on mortality in a cohort of injection drug users. J AcquirImmunodeficiency Syndr. 2001;28(4):351–7.18. Strathdee SA, Palepu A, Cornelisse PG, Yip B, O'Shaughnessy MV, MontanerJS, et al. Barriers to use of free antiretroviral therapy in injection drug users.JAMA. 1998;280(6):547–9.19. Tyndall MW, Currie S, Spittal P, Li K, Wood E, O'Shaughnessy MV, et al.Intensive injection cocaine use as the primary risk factor in the VancouverHIV-1 epidemic. AIDS. 2003;17(6):887–93.20. Craib KJ, Spittal PM, Wood E, Laliberte N, Hogg RS, Li K, et al. Riskfactors forelevated HIV incidence among Aboriginal injection drug users in Vancouver.Can Med Assoc J. 2003;168(1):19–24.21. Spittal P, Craib K, Wood E, Laliberté N, Li K, Tyndall M, et al. Riskfactors forelevated HIV incidence rates among female injection drug users inVancouver. Can Med Assoc J. 2002;166(7):894–9.22. Maldonado G, Greenland S. Simulation study of confounder-selectionstrategies. Am J Epidemiol. 1993;138(11):923–36.23. Galea S, Vlahov D. Social determinants and the health of drug users:socioeconomic status, homeless and incarceration. Public Health Rep.2002;117(1):S135–45.24. The Universal Declaration of Human Rights [http://www.un.org/en/documents/udhr/index.shtml]25. Palepu A, Patterson M, Moniruzzaman A, Frankish J, Somers J. Housing firstimproves residential stability in homeless adults with concurrent substancedependence and mental disorders. Am J Public Health. 2013;103(2):e30–6.26. Darke S. Self-report among injecting drug users: a review. Drug AlcoholDepend. 1998;51(3):253–63.27. Rachlis B, Wood E, Li K, Hogg R, Kerr T. Drug and HIV-related risk behaviorsafter geographic migration among a cohort of injection drug users.AIDS Behav. 2010;14(4):854–61.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 redistributionZivanovic et al. BMC Public Health  (2015) 15:106 Page 7 of 7Submit 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