UBC Faculty Research and Publications

Factors associated with optimal pharmacy refill adherence for antiretroviral medications and plasma HIV… Hayashi, Kanna; Wood, Evan; Kerr, Thomas; Dong, Huiru; Nguyen, Paul; Puskas, Cathy M.; Guillemi, Silvia; Montaner, Julio; Milloy, M-J Aug 27, 2016

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

Item Metadata

Download

Media
52383-Hayashi_K_et_al_Factors_associated_with_optimal.pdf [ 464.57kB ]
Metadata
JSON: 52383-1.0340730.json
JSON-LD: 52383-1.0340730-ld.json
RDF/XML (Pretty): 52383-1.0340730-rdf.xml
RDF/JSON: 52383-1.0340730-rdf.json
Turtle: 52383-1.0340730-turtle.txt
N-Triples: 52383-1.0340730-rdf-ntriples.txt
Original Record: 52383-1.0340730-source.json
Full Text
52383-1.0340730-fulltext.txt
Citation
52383-1.0340730.ris

Full Text

RESEARCH ARTICLE Open AccessFactors associated with optimal pharmacyrefill adherence for antiretroviralmedications and plasma HIV RNA non-detectability among HIV-positive crackcocaine users: a prospective cohort studyKanna Hayashi1,2, Evan Wood1,2, Thomas Kerr1,2, Huiru Dong1, Paul Nguyen1, Cathy M. Puskas1,3, Silvia Guillemi1,Julio S. G. Montaner1,2 and Michael-John Milloy1,2*AbstractBackground: Crack cocaine use is known to contribute to poor adherence to antiretroviral medications; however,little is known about facilitators of or barriers to effective HIV treatment use among HIV-infected crack cocaineusers. We sought to identify correlates of optimal pharmacy refill adherence for antiretroviral medications andplasma HIV RNA viral load (pVL) suppression among this population.Methods: Data from a prospective cohort of HIV-positive people who use illicit drugs in Vancouver, Canada, werelinked to comprehensive HIV clinical monitoring and pharmacy dispensation records. We used multivariablegeneralized linear mixed-effects modelling to longitudinally identify factors associated with ≥95 % adherence topharmacy refills for antiretroviral medications and pVL <50 copies/mL among crack cocaine users exposed tohighly-active antiretroviral therapy (HAART).Results: Among 438 HAART-exposed crack cocaine users between 2005 and 2013, 240 (54.8 %) had ≥95 %pharmacy refill adherence in the previous 6 months at baseline. In multivariable analyses, homelessness (adjustedodds ratio [AOR]: 0.58), ≥daily crack cocaine smoking (AOR: 0.64), and ≥ daily heroin use (AOR: 0.43) wereindependently associated with optimal pharmacy refill adherence (all p < 0.05). The results for pVL non-detectabilitywere consistent with those of medication adherence, except that longer history of HAART (AOR: 1.06), receiving asingle tablet-per-day regimen (AOR: 3.02) and participation in opioid substitution therapies was independentlyassociated with pVL non-detectability (AOR: 1.55) (all p < 0.05).Conclusions: Homelessness, and daily crack cocaine and/or heroin use were independently and negativelyassociated with optimal HAART-related outcomes. With the exception of opioid substitution therapies, no addictiontreatment modalities assessed appeared to facilitate medication adherence or viral suppression. Evidence-basedtreatment options for crack cocaine use that also confer benefits to HAART need to be developed.Keywords: Crack cocaine, ART adherence, Plasma HIV-1 RNA viral load suppression, Addiction treatment* Correspondence: uhri-mjsm@cfenet.ubc.ca1British Columbia Centre for Excellence in HIV/AIDS, St. Paul’s Hospital, 608 -1081 Burrard Street, Vancouver, BC V6Z 1Y6, Canada2Division of AIDS, Department of Medicine, University of British Columbia,608 - 1081 Burrard Street, Vancouver, BC V6Z 1Y6, CanadaFull list of author information is available at the end of the article© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Hayashi et al. BMC Infectious Diseases  (2016) 16:455 DOI 10.1186/s12879-016-1749-yBackgroundHighly-active antiretroviral therapy (HAART) has sub-stantially improved the life expectancy of people livingwith HIV [1] and has also been shown to prevent HIVtransmission by reducing individuals’ viral load to un-detectable levels [2, 3]. Despite the remarkable impact ofHAART on the HIV/AIDS pandemic, its benefits on dis-ease progression and the incidence of new infectionshave yet to be fully realized among all HIV-seropositivegroups, such as people who use illicit drugs. Crack co-caine use has been shown to be an independent risk fac-tor of HIV seroconversion in some settings, presumablythrough high-risk sexual activities associated with crackuse [4–7]. While evidence-based strategies to reduceHIV transmission risk through injection drug use exist,such as needle and syringe programs [8], crack cocaineis often smoked, and interventions to reduce sexual riskbehaviour among crack cocaine smokers have proven tobe of limited effectiveness [5, 9, 10].In 2014, the Treatment-as-Prevention (TasP) approach,which aims to expand access and adherence to HAARTamong people living with HIV and thereby prevent HIVtransmission, was incorporated into the global HIV/AIDS response strategies by the Joint United NationsProgramme on HIV/AIDS [11]. This is likely apotentially effective HIV prevention approach to addressHIV transmission associated with persistent high-riskbehaviour among HIV-infected crack cocaine users [12].However, there remains concern regarding the viabilityof the intervention due to previous studies describingsub-optimal adherence to HAART in crack cocaineusing populations [13–15]. For example, a 2002 study ofHIV-infected people who use drugs reported that thestrongest predictor of poor adherence to antiretroviralmedications was cocaine/crack cocaine use [13]. A morerecent study found that stimulant users, particularlythose who used cocaine/crack cocaine, had seven timeshigher odds of HAART adherence failure (i.e., <90 % ofdoses taken) compared to patients who did not usedrugs [15].Despite many reports of poor adherence to HAARTamong HIV-infected crack cocaine users [13–15], todate, there is a paucity of research identifying factors as-sociated with HAART adherence and plasma HIV RNAviral load (pVL) suppression in this population, whichmay be amenable to intervention. This is a concern asthe success of TasP-based approaches relies on ensuringaccess and adherence to HAART. In this study, wesought to identify correlates of optimal adherence toantiretroviral medications and pVL non-detectabilityamong HAART-exposed crack cocaine users in Vancouver,Canada, where crack cocaine use is common and hasbeen identified as an independent predictor of HIVseroconversion [7].MethodsStudy designData for this analysis were derived from the AIDS CareCohort to evaluate Exposure to Survival Services(ACCESS), a prospective cohort study of HIV-positivepeople who use illicit drugs in Vancouver, Canada.Recruitment and data collection procedures have beendescribed in detail elsewhere [16]. In brief, beginning in1996, participants were recruited through self-referral andstreet-based outreach from Vancouver’s DowntownEastside neighbourhood, a post-industrial area with a largeopen drug market and high levels of illicit drug use,poverty, and HIV infection. Individuals were eligible toparticipate in ACCESS if they were aged 18 years or older,were HIV-seropositive, had used illicit drugs other thancannabis in the month prior to enrolment, and providedwritten informed consent. Participants were compensated$20 CAD at each study visit.At baseline and semi-annually thereafter, participantscompleted an interviewer-administered questionnairesoliciting demographic data, information on drug usepatterns, as well as other characteristics and exposures.At each visit, individuals also underwent an examinationby a study nurse and provided blood samples for sero-logic analyses. Through a confidential linkage with theprovincial Drug Treatment Program (DTP), a completeclinical profile of all CD4 T-cell counts, pVL observa-tions, exposure to specific antiretroviral agents, andtreatment outcomes for each participant were obtained.In British Columbia, all provision of HIV treatment iscentralized through a province-wide dispensation pro-gram, where treatment and related care are providedfree of charge. Data collection by the DTP has beendescribed previously [17, 18]. The ACCESS study hasbeen approved by the University of British Columbia/Providence Health Care Research Ethics Board.Participants and measuresParticipants were eligible for the present analysis if theywere recruited by ACCESS between December 1, 2005and November 30, 2013, reported having smoked crackcocaine in the six months prior to the baseline assess-ment, and had initiated HAART prior to enrolment inACCESS. As well, at least one observation of CD4 cellcount and pVL had to be completed within ±180 days ofthe day they entered the study. For a sub-analysis focus-ing on dual crack cocaine and opioid users, the samplewas further restricted to those who had a history of anyopioid use or participation in methadone maintenancetherapy at baseline.The primary outcomes of interest were optimal adher-ence to pharmacy refills for antiretroviral medications andpVL non-detectability in the previous six months. Wemeasured pharmacy refill adherence in each 6-monthHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 2 of 12period as the number of days for which medications weredispensed over the number of days since the initiation ofHAART, which was capped at 180 days. We defined opti-mal pharmacy refill adherence as equal to or greater than95 %. We have previously demonstrated the clinical valid-ity of this pharmacy refill data and shown that it reliablypredicts virologic suppression [19, 20] and survival [21].Plasma HIV RNA viral load non-detectability was definedas having achieved a HIV RNA viral load of <50 copies/mL of plasma. In the event that more than one pVL obser-vation was made within a 6-month follow-up period, weused the median of all the observations, which was thencategorized into either <50 or ≥50 copies/mL of plasma.We considered a range of explanatory variables thatmight be associated with optimal pharmacy refill adher-ence and/or pVL non-detectability. Demographic char-acteristics included: age (per 10-year older), gender(male vs. female), ethnicity (Caucasian vs. other), andeducation level (<secondary education vs. ≥secondaryeducation). Substance-using behaviours included: crackcocaine smoking (≥daily vs. <daily), heroin use (≥dailyvs. <daily), crystal methamphetamine use (≥daily vs.<daily), and heavy alcohol use (≥5 drinks per day vs. <5drinks per day). Other behavioural variables included:homelessness, engagement in sex work or drug dealing,incarceration, participation in inpatient addiction treat-ment (i.e., recovery houses, detoxification facilities, andtreatment centres), participation in outpatient detoxifica-tion services, seeing drug counsellors or participation inaddiction-related peer support meetings (e.g., NarcoticsAnonymous, Cocaine Anonymous, etc.), and participa-tion in opioid substitution therapies (i.e., methadonemaintenance therapy or Suboxone treatment) (all: yes vs.no). The use of opioid substitution therapies was consid-ered in the sub-analysis focusing on dual crack cocaineand opioid users only. Clinical variables included: CD4cell count (per 100-cells/mm3 increase); antiretroviraltablet regimen (single tablet vs. 2-3 tablets vs. >3 tabletsper day); years since HAART initiation (per year in-crease); physician experience on HIV care at the time ofantiretroviral therapy initiation (having ever treated <6patients vs. ≥6 patients). All variables except for thedemographic characteristics referred to the previous sixmonths unless otherwise indicated.Statistical analysesFirst, we examined baseline characteristics of the sample,stratified by medication adherence (≥95 % vs. <95 %) inthe six months prior to the baseline assessment.Categorical variables were analysed using Pearson’s χ2test (or Fisher’s exact test in the presence of small sellcounts) and continuous variables were analysed usingthe Wilcoxon rank-sum test. Next, we used generalizedlinear mixed-effects modelling (GLMM) with the logitlink to longitudinally identify factors associated with opti-mal pharmacy refill adherence and pVL non-detectability,respectively. This form of regression modelling accountsfor the correlation between repeated observations gath-ered over time from the same individual, and estimatesthe effect of the explanatory variables on the likeli-hood of optimal pharmacy refill adherence and pVLnon-detectability in each individual.Since our study aimed to identify the set of variablesthat best explains a higher odds of optimal pharmacy re-fill adherence and pVL non-detectability, respectively,we used a priori-defined statistical protocol based onthe examination of the Akaike information criterion(AIC) and Type III p-values to construct multivariableGLMM logistic regression models. This procedure bal-ances model selection on finding the best explanatorymodel with best model fit, as described previously [22].In brief, we first included in the full multivariablemodels all explanatory variables that were significantlyassociated with optimal pharmacy refill adherence andpVL non-detectability, respectively, at the p < 0.10 levelin the univariable analyses. After examining the AICvalue of the models, we removed the variable with thelargest p-value and built a reduced model. We continuedthis iterative process and selected the multivariablemodels with the lowest AIC value. Through this process,some variables, even though being significantly associ-ated with an outcome at the level of p < 0.05 in bivari-able analyses, were dropped from the final multivariablemodel. For the sub-analysis focusing on dual crack co-caine and opioid users, we added participation in opioidsubstitution therapies to the final multivariable modelsas a covariate. All p-values were two-sided. All statisticalanalyses were performed using the SAS software version9.3 (SAS, Cary, NC).ResultsIn total, 438 HAART-exposed crack cocaine users wereincluded in this study, of whom 293 (66.9 %) were maleand 239 (54.6 %) self-reported Caucasian ancestry. Ofthe 438 participants, 406 (92.7 %) had returned for atleast one follow-up visit, contributing a median follow-up duration of 48.0 months (interquartile range [IQR]:35.0 – 75.9). At baseline, 240 (54.8 %) exhibited optimalpharmacy refill adherence in the previous six months,and a total of 390 (89.0 %) individuals attained optimaladherence at some point during the study period.Table 1 presents sample characteristics at baseline,stratified by medication adherence levels. A substantialproportion of the sample was homeless (29.5 %), smokedcrack cocaine at least on a daily basis (40.0 %), and en-gaged in sex work or drug dealing (37.4 %) in the previ-ous six months. Very few individuals accessed any typesof addiction treatment except for opioid substitutionHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 3 of 12Table 1 Baseline characteristics of HAART-exposed crack cocaine users in Vancouver, Canada, stratified by optimal pharmacy refilladherence (n = 438)Characteristic Total(n = 438)n (%)≥95 % pharmacy refill adherence for antiretroviralsa p-valueYes240 (54.8 %)No198 (45.2 %)Age (median, IQR) 44 (38–49) 45 (40–51) 42 (37–46) <0.001GenderMale 293 (66.9 %) 171 (71.3 %) 122 (61.6 %) 0.033Female 145 (33.1 %) 69 (28.7 %) 76 (38.4 %) referenceAncestryCaucasian 239 (54.6 %) 147 (61.3 %) 92 (46.5 %) 0.002Non-Caucasian 199 (45.4 %) 93 (38.7 %) 106 (53.5 %) referenceEducation< Secondary education 228 (52.1 %) 116 (48.3 %) 112 (56.6 %) 0.069≥ Secondary education 196 (44.7 %) 117 (48.8 %) 79 (39.9 %) referenceHomelessaYes 129 (29.5 %) 57 (23.8 %) 72 (36.4 %) 0.003No 306 (69.9 %) 182 (75.8 %) 124 (62.6 %) referenceCrack cocaine smokinga≥ Daily 175 (40.0 %) 78 (32.5 %) 97 (49.0 %) 0.001< Daily 263 (60.0 %) 162 (67.5 %) 101 (51.0 %) referenceHeroin usea≥ Daily 48 (11.0 %) 19 (7.9 %) 29 (14.7 %) 0.024< Daily 389 (88.8 %) 221 (92.1 %) 168 (84.9 %) referenceCrystal methamphetamine usea,c≥ Daily 9 (2.1 %) 5 (2.1 %) 4 (2.0 %) 1.000< Daily 428 (97.7 %) 235 (97.9 %) 193 (97.5 %) referenceHeavy alcohol use (≥5 drinks per day)a,cYes 13 (3.0 %) 5 (2.1 %) 8 (4.0 %) 0.266No 425 (97.0 %) 235 (97.9 %) 190 (96.0 %) referenceEngagement in sex work or drug dealingaYes 164 (37.4 %) 78 (32.5 %) 86 (43.4 %) 0.018No 270 (61.6 %) 160 (66.7 %) 110 (55.6 %) referenceIncarcerationaYes 58 (13.2 %) 32 (13.3 %) 26 (13.1 %) 0.951No 380 (86.8 %) 208 (86.7 %) 172 (86.9 %) referenceHIV physician experiencebPreviously treated <6 patients 48 (11.0 %) 25 (10.4 %) 23 (11.6 %) 0.397Previously treated ≥6 patients 340 (77.6 %) 199 (82.9 %) 141 (71.2 %) referenceCD4 cell counta (median, IQR) 300 (180–450) 340 (223.5–495) 235 (120–360) <0.001Years since HAART initiationa (median, IQR) 5.7 (2.2–8.5) 4.9 (1.6–8.0) 6.9 (3.2–8.8) <0.001RegimenbSingle tablet per day 34 (7.8 %) 24 (10.0 %) 10 (5.1 %) 0.2782-3 tablets per day 121 (27.6 %) 76 (31.7 %) 45 (22.7 %) 0.723> 3 tablets per day 230 (52.5 %) 140 (58.3 %) 90 (45.5 %) referenceHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 4 of 12therapies (1.1–9.8 %), while the majority (63.5 %) of 293dual crack cocaine and opioid users participated in opi-oid substitution therapies in the previous six months.Compared to those who had suboptimal pharmacy re-fill adherence in the previous six months, individualswho had optimal pharmacy refill adherence were older,more likely to be male, self-report Caucasian ancestry,possess a high school diploma, have higher CD4 cellcount and a shorter length of exposure to HAART, andhave seen drug counsellors or participated in addiction-related peer support groups in the previous six months,while they were less likely to be homeless, smoke crackcocaine or use heroin at least daily, engage in sex workor drug dealing in the previous six months (all p < 0.05).Additionally, among a sub-sample of dual crack cocaineand opioid users, those who accessed opioid substitutiontherapies were more likely to have exhibited optimalpharmacy refill adherence (p = 0.005). Table 2 presentsbaseline characteristics stratified by pVL non-detectability.The observed significant associations were generally simi-lar to those found with optimal pharmacy refill adherence,except that those who were on a 2-3 tablet-per-dayregimen were more likely to have pVL non-detectabilitycompared to those who were on a >3 tablet-per-day regi-men (p = 0.008).Table 3 shows the results of univariable and multivari-able GLMM logistic regression analyses of factors asso-ciated with optimal pharmacy refill adherence. Asshown, in univariable analyses, older age (odds ratio[OR]: 2.29; 95 % confidence interval [CI]: 1.84 – 2.86),male gender (OR: 1.50; 95 % CI: 1.02 – 2.20), Caucasianethnicity (OR: 1.61; 95 % CI: 1.12 – 2.32), higher CD4count (OR: 1.39; 95 % CI: 1.30 – 1.48) and longer historyof HAART (OR: 1.08; 95 % CI: 1.04 – 1.12) were signifi-cantly and positively associated with optimal pharmacyrefill adherence. Homelessness (OR: 0.44; 95 % CI: 0.34 –0.58), at least daily crack cocaine smoking (OR: 0.49; 95 %CI: 0.39 – 0.62), at least daily heroin use (OR: 0.32; 95 %CI: 0.22 – 0.47), and engagement in sex work or drugdealing (OR: 0.59; 95 % CI: 0.47 – 0.76) were significantlyand negatively associated with optimal pharmacy refilladherence.In the final multivariable model, older age (adjustedodds ratio [AOR]: 1.65; 95 % CI: 1.33 – 2.04) and higherCD4 count (AOR: 1.31; 95 % CI: 1.23 – 1.41) remainedindependently and positively associated with optimalpharmacy refill adherence, while homelessness (AOR:0.58; 95 % CI: 0.44 – 0.77), at least daily crack cocainesmoking (AOR: 0.64; 95 % CI: 0.50 – 0.81), and at leastdaily heroin use (AOR: 0.43; 95 % CI: 0.29 – 0.65)remained independently and negatively associated withoptimal pharmacy refill adherence. When the samplewas restricted to dual crack cocaine and opioid users,the same set of variables remained significantly associ-ated with optimal pharmacy refill adherence. Participa-tion in opioid substitution therapies was notindependently associated with optimal pharmacy refilladherence (AOR: 1.39; 95 % CI: 0.99 – 1.96).Table 4 presents the results univariable and multivari-able GLMM logistic regression analyses of factors asso-ciated with pVL non-detectability. As shown, in additionto the covariates that were included for medicationadherence, the final multivariable models includedyears since HAART initiation and antiretroviral tabletTable 1 Baseline characteristics of HAART-exposed crack cocaine users in Vancouver, Canada, stratified by optimal pharmacy refilladherence (n = 438) (Continued)Participation in inpatient addiction treatmentaYes 35 (8.0 %) 19 (7.9 %) 16 (8.1 %) 0.950No 403 (92.0 %) 221 (92.1 %) 182 (91.9 %) referenceParticipation in outpatient detoxification or other outpatient addiction treatmenta,cYes 5 (1.1 %) 4 (1.7 %) 1 (0.5 %) 0.384No 433 (98.9 %) 236 (98.3 %) 197 (99.5 %) referenceSeeing drug counsellors or participation in addiction-related peer support meetingsaYes 43 (9.8 %) 31 (12.9 %) 12 (6.1 %) 0.016No 395 (90.2 %) 209 (87.1 %) 186 (93.9 %) referenceRestricted to dual crack cocaine and opioid users (n = 293)Participation in opioid substitution therapiesaYes 186 (63.5 %) 115 (70.6 %) 71 (54.6 %) 0.005No 107 (36.5 %) 48 (29.4 %) 59 (45.4 %) referenceHAART highly active antiretroviral therapy, CI confidence interval, IQR interquartile rangea denotes activities/events in the past 6 monthsb denotes at the time of antiretroviral therapy initiationc p-values were derived from Fisher’s exact testNot all cells added up to 100 % due to some missing valuesHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 5 of 12Table 2 Baseline characteristics of HAART-exposed crack cocaine users in Vancouver, Canada, stratified by pVL non-detectability (n= 438)Characteristic Total(n = 438)n (%)pVL <50 copies/mLa p-valueYes212 (48.4 %)No226 (51.6 %)Age (median, IQR) 44 (38–49) 45 (41–51) 42 (36–47) <0.001GenderMale 293 (66.9 %) 149 (70.3 %) 144 (63.7 %) 0.145Female 145 (33.1 %) 63 (29.7 %) 82 (36.3 %) referenceAncestryCaucasian 239 (54.6 %) 122 (57.5 %) 117 (51.8 %) 0.225Non-Caucasian 199 (45.4 %) 90 (42.5 %) 109 (48.2 %) referenceEducation< Secondary education 228 (52.1 %) 107 (50.5 %) 121 (53.5 %) 0.462≥ Secondary education 196 (44.7 %) 99 (46.7 %) 97 (42.9 %) referenceHomelessaYes 129 (29.5 %) 36 (17.0 %) 93 (41.2 %) <0.001No 306 (69.9 %) 174 (82.1 %) 132 (58.4 %) referenceCrack cocaine smokinga≥ Daily 175 (40.0 %) 72 (34.0 %) 103 (45.6 %) 0.013< Daily 263 (60.0 %) 140 (66.0 %) 123 (54.4 %) referenceHeroin usea≥ Daily 48 (11.0 %) 16 (7.5 %) 32 (14.2 %) 0.026< Daily 389 (88.8 %) 196 (92.5 %) 193 (85.4 %) referenceCrystal methamphetamine usea,c≥ Daily 9 (2.1 %) 6 (2.8 %) 3 (1.3 %) 0.326< Daily 428 (97.7 %) 206 (97.2 %) 222 (98.2 %) referenceHeavy alcohol use (≥5 drinks per day)a,cYes 13 (3.0 %) 4 (1.9 %) 9 (4.0 %) 0.263No 425 (97.0 %) 208 (98.1 %) 217 (96.0 %) referenceEngagement in sex work or drug dealingaYes 164 (37.4 %) 72 (34.0 %) 92 (40.7 %) 0.145No 270 (61.6 %) 138 (65.1 %) 132 (58.4 %) referenceIncarcerationaYes 58 (13.2 %) 25 (11.8 %) 33 (14.6 %) 0.386No 380 (86.8 %) 187 (88.2 %) 193 (85.4 %) referenceHIV physician experiencebPreviously treated <6 patients 48 (11.0 %) 21 (9.9 %) 27 (11.9 %) 0.418Previously treated ≥6 patients 340 (77.6 %) 170 (80.2 %) 170 (75.2 %) referenceCD4 cell counta (median, IQR) 300 (180–450) 360 (260–500) 230 (110–360) <0.001Years since HAART initiationa (median, IQR) 5.7 (2.2–8.5) 5.9 (2.3–8.7) 5.5 (2.1–8.4) 0.243RegimenbSingle tablet per day 34 (7.8 %) 18 (8.5 %) 16 (7.1 %) 0.7132-3 tablets per day 121 (27.6 %) 78 (36.8 %) 43 (19.0 %) 0.008> 3 tablets per day 230 (52.5 %) 114 (53.8 %) 116 (51.3 %) referenceHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 6 of 12regimen. Longer history of HAART (AOR: 1.06; 95 %CI: 1.01 – 1.12) and receiving a single tablet-per-dayregimen (AOR: 3.02; 95 % CI: 1.67 – 5.46) were inde-pendently and positively associated with pVL non-detectability. The results for the other covariates wereconsistent with those found in the multivariablemodel of medication adherence except that amongdual crack cocaine and opioid users, participation inopioid substitution therapies was independently andpositively associated with pVL non-detectability(AOR: 1.55; 95 % CI: 1.02 – 2.34).DiscussionAmong our sample of HAART-exposed crack cocaineusers, we found that homelessness, and high-intensitycrack cocaine and heroin use were independently andnegatively associated with optimal pharmacy refill adher-ence and pVL non-detectability. Except for opioid sub-stitution therapies, none of the addiction treatmentmodalities assessed, including inpatient treatment, out-patient detoxification services, drug counselling, andparticipation in peer support groups such as NarcoticsAnonymous, were significantly associated with optimalpharmacy refill adherence or pVL non-detectability. Spe-cifically among a sub-sample of dual crack cocaine andopioid users, enrolment in opioid substitution therapieswas independently associated with pVL non-detectabilitybut not with optimal pharmacy refill adherence.We found an independent association between at leastdaily use of crack cocaine and poor medication adher-ence and having a detectable viral load. This is consist-ent with previous reports that crack cocaine users areless likely to enjoy gains from HAART [13–15]. Inaddition, the finding that at least daily use of heroin wasalso independently associated with suboptimal pharmacyrefill adherence and outcome indicates a need to addresspoly-substance use among HIV-infected crack cocaineusers. Previous research suggested that poly-substanceuse is quite common among people who use drugs, al-though those who primarily use cocaine may continueto use it consistently over an extended period of timewhile keeping the use of other non-primary drugs atrelatively lower levels [23]. While current literature sug-gests no clinically significant pharmacokinetic interac-tions between crack cocaine and antiretroviral agents,methamphetamines may have possible lethal interactionswith antiretroviral agents that inhibit the CYP450 system[24]. In our sample of HIV-positive crack cocaine users,high-intensity use of crystal methamphetamine was notcommon or associated with optimal pharmacy refill ad-herence or pVL non-detectability; however, in a settingwhere co-use of crack cocaine and crystal methampheta-mine is common, a greater caution is required when pre-scribing HAART. Collectively, these findings emphasizethe need to ensure that a variety of addiction treatmentoptions be integrated with HIV treatment and care for in-dividuals who use crack cocaine.The finding that none of the addiction treatment mo-dalities except for opioid substitution therapies were as-sociated with optimal pharmacy refill adherence or viralsuppression serves to underscore previous calls for iden-tifying more evidence-based addiction treatment ap-proaches that also confer benefits on HIV diseasemanagement among HIV-positive crack cocaine usersTable 2 Baseline characteristics of HAART-exposed crack cocaine users in Vancouver, Canada, stratified by pVL non-detectability (n= 438)(Continued)Participation in inpatient addiction treatmentaYes 35 (8.0 %) 14 (6.6 %) 21 (9.3 %) 0.300No 403 (92.0 %) 198 (93.4 %) 205 (90.7 %) referenceParticipation in outpatient detoxification or other outpatient addiction treatmenta,cYes 5 (1.1 %) 4 (1.9 %) 1 (0.4 %) 0.203No 433 (98.9 %) 208 (98.1 %) 225 (99.6 %) referenceSeeing drug counsellors or participation in addiction-related peer support meetingsaYes 43 (9.8 %) 30 (14.2 %) 13 (5.8 %) 0.003No 395 (90.2 %) 182 (85.8 %) 213 (94.2 %) referenceRestricted to dual crack cocaine and opioid users (n = 293)Participation in opioid substitution therapiesaYes 186 (63.5 %) 98 (72.6 %) 88 (55.7 %) 0.003No 107 (36.5 %) 37 (29.4 %) 70 (45.4 %) referenceHAART highly active antiretroviral therapy, CI confidence interval, IQR interquartile range, pVL plasma HIV RNA viral loada denotes activities/events in the past 6 monthsb denotes at the time of antiretroviral therapy initiationc p-values were derived from Fisher’s exact testNot all cells added up to 100 % due to some missing valuesHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 7 of 12Table 3 Univariable and multivariable GLMM logistic regression analyses of factors associated with optimal pharmacy refilladherence for antiretroviral medications among HAART-exposed crack cocaine users in Vancouver, Canada, 2005–2013Characteristic Total sample of crack cocaine users(n = 438)Dual crack cocaine and opioid users(n = 293)Unadjusted OR(95 % CI)p-value Adjusted OR(95 % CI)p-value Unadjusted OR(95 % CI)p-value Adjusted OR(95 % CI)p-valueAgeb(per 10-year older) 2.29 (1.84 – 2.86) <0.001 1.65 (1.33 – 2.04) <0.001 2.75 (2.06 – 3.68) <0.001 1.72 (1.30 – 2.26) <0.001Gender(male vs. female) 1.50 (1.02 – 2.20) 0.041 1.10 (0.68 – 1.79) 0.696Ethnicity(Caucasian vs. other) 1.61 (1.12 – 2.32) 0.011 1.40 (0.87 – 2.24) 0.167Education(<secondary vs. ≥secondary) 0.83 (0.57 – 1.20) 0.327 0.98 (0.61 – 1.58) 0.943Homelessa,b(yes vs. no) 0.44 (0.34 – 0.58) <0.001 0.58 (0.44 – 0.77) <0.001 0.39 (0.28 – 0.54) <0.001 0.56 (0.40 – 0.79) <0.001Crack cocaine smokinga,b(≥daily vs. <daily) 0.49 (0.39 – 0.62) <0.001 0.64 (0.50 – 0.81) <0.001 0.41 (0.31 – 0.55) <0.001 0.57 (0.43 – 0.75) <0.001Heroin usea,b(≥daily vs. <daily) 0.32 (0.22 – 0.47) <0.001 0.43 (0.29 – 0.65) <0.001 0.30 (0.20 – 0.46) <0.001 0.45 (0.29 – 0.70) <0.001Crystal methamphetamine usea,b(≥daily vs. <daily) 1.07 (0.50 – 2.30) 0.858 0.79 (0.32 – 1.94) 0.609Heavy alcohol usea,b(≥5 drinks per day vs. <5 drinksper day)1.17 (0.65 – 2.10) 0.596 1.37 (0.60 – 3.10) 0.452Engagement in sex work or drug dealinga,b(yes vs. no) 0.59 (0.47 – 0.76) <0.001 0.54 (0.41 – 0.71) <0.001Incarcerationa,b(yes vs. no) 0.71 (0.49 – 1.02) 0.063 0.63 (0.41 – 0.96) 0.031HIV physician experiencec(<6 patients vs. ≥6 patients) 0.74 (0.40 – 1.35) 0.322 0.60 (0.29 – 1.24) 0.169CD4 cell counta,b(per 100-cell increase) 1.39 (1.30 – 1.48) <0.001 1.31 (1.23 – 1.41) <0.001 1.43 (1.32 – 1.55) <0.001 1.34 (1.23 – 1.45) <0.001Years since HAART initiationa,b(per year increase) 1.08 (1.04 – 1.12) <0.001 1.11 (1.06 – 1.16) <0.001Regimena,b(1 pill per day vs. >3 pills per day) 1.23 (0.80 – 1.90) 0.339 1.29 (0.76 – 2.19) 0.347(2-3 pills per day vs. >3 pills per day) 1.01 (0.79 – 1.30) 0.920 1.07 (0.78 – 1.48) 0.659Participation in inpatient addiction treatmenta,b(yes vs. no) 0.82 (0.57 – 1.18) 0.294 0.84 (0.54 – 1.32) 0.454Participation in outpatient detoxification or other outpatient addiction treatmenta,b(yes vs. no) 0.65 (0.20 – 2.09) 0.472 0.43 (0.09 – 2.02) 0.284Hayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 8 of 12[5, 14, 25]. While a previous pilot study suggested that avideo-based behavioural intervention employing theInformation-Motivation-Behavioural Skill model mightbe efficacious in reducing crack cocaine use and improv-ing medication adherence among HIV-positive crack co-caine users [26], the efficacy of the intervention needs tobe determined with a larger sample. Also, the long-termpost-intervention effect on crack cocaine use and medi-cation adherence is an important outcome to be exam-ined, as a Cochrane Collaboration systematic review hasreported that many behavioural interventions for sub-stance use suffer from a lack of the sustained effect fol-lowing the intervention [27]. Further, unlike the case ofopiate addiction, there is currently no approved medica-tion to treat crack cocaine addiction. While many candi-date medications have been evaluated for the efficacy inthe treatment of crack cocaine/cocaine addiction duringthe past decade, overall results have been inconclusive[28–30]. Further research is needed in this area.In contrast, we found that dual crack cocaine and opi-oid users who were enrolled in opioid substitution ther-apies were more likely to achieve virological success, andthe positive effect of opioid substitution therapiesremained significant even after the adjustment for thehigh-intensity use of crack cocaine and heroin. This addsimportant evidence to a body of research indicating theeffectiveness of these therapies for the improvement ofHIV treatment adherence and outcomes [31]. Previousstudies reported that ongoing cocaine use underminedthe effectiveness of opioid substitution therapies for HIVtreatment-related outcomes among HIV-infected pa-tients accessing these therapies [32, 33]. However, ourfindings suggest that opioid substitution therapies stillhelp HIV-infected dual crack cocaine and opioid usersachieve optimal HAART outcomes. Nonetheless, ascrack cocaine/cocaine users have been shown to havesignificant barriers to adherence and continuance of opi-oid substitution therapies [34, 35], adjunct interventionsto opioid substitution therapies that help retain crackcocaine/cocaine-using patients in treatment need to bedeveloped.Similarly to our previous investigation that indicatedthat homelessness posed a barrier to antiretroviraltherapy among HIV-infected people who use illicit drugsgenerally [36], we found that homelessness wasindependently associated with suboptimal pharmacy re-fill adherence and viral suppression failures among HIV-infected individuals who use crack cocaine in a settingwhere HAART is provided free of charge. Collectively,these findings indicate that, in addition to identificationof evidence-based addiction treatment strategies forcrack cocaine use, structural interventions to addresshomelessness in general (e.g., housing assistance) areneeded. Expanding and promoting support services forHIV-positive homeless individuals focusing on both ac-cess and adherence to HAART would maximise the ben-efits of HAART on disease progression and onward viraltransmission in this population.Our study has several limitations. First, as the studysample was not randomly selected, our findings may notbe generalizable to other populations of crack cocaineusers. Second, the self-reported data may be affected byreporting biases, including recall bias and socially desir-able responding. However, we note that this type of datahas been commonly utilized in observational studies in-volving people who use drugs and found to be valid [37].Also, as our primary outcomes were ascertained througha confidential record linkage to the provincial HIV-related clinical database, we believe that it is unlikelythat potential reporting biases have impacted the datadifferentially by the medication adherence or pVL levels.Third, as with all observational studies, the relationshipsbetween the explanatory variables and outcome assessedmay be under the influence of unobserved confounding,although we sought to address this bias with multivari-able adjustment of the key demographic, behavioural,and clinical predictors of medication adherence or viralsuppression.ConclusionsIn sum, among our sample of HAART-exposed crackcocaine users, those who were homeless and those whoengaged in high-intensity crack cocaine or heroin usewere more likely to have suboptimal adherence to anti-retroviral medications and fail to achieve viral suppres-sion. Further, none of the addiction treatment modalitiesTable 3 Univariable and multivariable GLMM logistic regression analyses of factors associated with optimal pharmacy refilladherence for antiretroviral medications among HAART-exposed crack cocaine users in Vancouver, Canada, 2005–2013 (Continued)Seeing drug counsellors or participation in addiction-related peer support meetingsa,b(yes vs. no) 1.06 (0.76 – 1.48) 0.730 0.84 (0.55 – 1.28) 0.424Participation in opioid substitution therapiesa,b(yes vs. no) ––– ––– 1.71 (1.23 – 2.38) 0.001 1.39 (0.99 – 1.96) 0.056GLMM generalized linear mixed-effect modelling, HAART highly active antiretroviral therapy, OR odds ratio, CI confidence intervala denotes activities/events in the past 6 monthsb denotes time-varying variablesc denotes at the time of antiretroviral therapy initiationHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 9 of 12Table 4 Univariable and multivariable GLMM logistic regression analyses of factors associated with pVL non-detectability amongHAART-exposed crack cocaine users in Vancouver, Canada, 2005–2013Characteristic Total sample of crack cocaine users(n = 438)Dual crack cocaine and opioid users(n = 293)Unadjusted OR(95 % CI)p-value Adjusted OR(95 % CI)p-value Unadjusted OR(95 % CI)p-value Adjusted OR(95 % CI)p-valueAgeb(per 10-year older) 5.24 (3.72 – 7.37) <0.001 1.86 (1.37 – 2.52) <0.001 5.22 (3.45 – 7.89) <0.001 1.59 (1.11 – 2.26) 0.011Gender(male vs. female) 1.86 (1.10 – 3.15) 0.020 1.08 (0.59 – 1.97) 0.804Ethnicity(Caucasian vs. other) 1.31 (0.79 – 2.17) 0.290 1.38 (0.77 – 2.49) 0.282Education(<secondary vs. ≥secondary) 1.03 (0.62 – 1.72) 0.909 1.39 (0.77 – 2.49) 0.272Homelessa,b(yes vs. no) 0.26 (0.19 – 0.36) <0.001 0.39 (0.28 – 0.54) <0.001 0.24 (0.17 – 0.35) <0.001 0.37 (0.25 – 0.56) <0.001Crack cocaine smokinga,b(≥daily vs. <daily) 0.47 (0.36 – 0.61) <0.001 0.39 (0.29 – 0.53) <0.001 0.69 (0.49 – 0.96) 0.028Heroin usea,b(≥daily vs. <daily) 0.35 (0.23 – 0.54) <0.001 0.60 (0.36 – 0.98) 0.042 0.32 (0.20 – 0.51) <0.001Crystal methamphetamine usea,b(≥daily vs. <daily) 1.09 (0.45 – 2.62) 0.853 1.15 (0.43 – 3.10) 0.779Heavy alcohol usea,b(≥5 drinks per day vs. <5 drinksper day)0.70 (0.36 – 1.37) 0.302 0.57 (0.24 – 1.37) 0.209Engagement in sex work or drug dealinga,b(yes vs. no) 0.58 (0.44 – 0.75) <0.001 0.58 (0.43 – 0.78) <0.001Incarcerationa,b(yes vs. no) 0.57 (0.38 – 0.86) 0.008 0.49 (0.31 – 0.78) 0.002HIV physician experiencec(<6 patients vs. ≥6 patients) 1.15 (0.51 – 2.62) 0.738 0.84 (0.35 – 2.00) 0.688CD4 cell counta,b(per 100-cell increase) 1.93 (1.76 – 2.13) <0.001 1.67 (1.52 – 1.83) <0.001 1.98 (1.77 – 2.21) <0.001 1.72 (1.54 – 1.91) <0.001Years since HAART initiationa,b(per year increase) 1.27 (1.21 – 1.34) <0.001 1.06 (1.01 – 1.12) 0.012 1.29 (1.22 – 1.37) <0.001 1.08 (1.02 – 1.14) 0.011Regimena,b(1 pill per day vs. >3 pills per day) 3.11 (1.76 – 5.52) <0.001 3.02 (1.67 – 5.46) <0.001 2.84 (1.45 – 5.57) 0.003 2.60 (1.30 – 5.21) 0.007(2-3 pills per day vs. >3 pills per day) 1.70 (1.27 – 2.30) <0.001 1.24 (0.91 – 1.69) 0.179 1.60 (1.13 – 2.29) 0.009 1.14 (0.79 – 1.65) 0.493Participation in inpatient addiction treatmenta,b(yes vs. no) 0.67 (0.45 – 1.02) 0.061 0.58 (0.35 – 0.97) 0.037Participation in outpatient detoxification or other outpatient addiction treatmenta,b(yes vs. no) 1.10 (0.33 – 3.71) 0.872 1.63 (0.34 – 7.74) 0.539Seeing drug counsellors or participation in addiction-related peer support meetingsa,b(yes vs. no) 1.32 (0.90 – 1.94) 0.158 1.16 (0.73 – 1.84) 0.532Participation in opioid substitution therapiesa,b(yes vs. no) ––– ––– 2.28 (1.58 – 3.30) <0.001 1.55 (1.02 – 2.34) 0.039GLMM generalized linear mixed-effect modelling, pVL plasma HIV RNA viral load, HAART highly active antiretroviral therapy, OR odds ratio, CI confidence intervala denotes activities/events in the past 6 monthsb denotes time-varying variablesc denotes at the time of antiretroviral therapy initiationHayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 10 of 12assessed appeared to facilitate medication adherence orviral suppression, except for opioid substitution therap-ies. These findings suggest an urgent need to developevidence-based treatment options for crack cocaine usethat also confer benefits to HAART-related outcomes.AbbreviationsACCESS, the AIDS Care Cohort to evaluate Exposure to Survival Services; AIC,Akaike information criterion; AOR, adjusted odds ratio; GLMM, generalizedlinear mixed-effects modelling; HAART, Highly-active antiretroviral therapy;IQR, interquartile range; OR, odds ratio; pVL, plasma HIV RNA viral load; TasP,Treatment as PreventionAcknowledgementsThe authors thank the study participants for their contribution to theresearch, as well as current and past researchers and staff.FundingThe study was supported by the US National Institutes of Health(R01DA021525). This research was undertaken, in part, thanks to fundingfrom the Canada Research Chairs program through a Tier 1 Canada ResearchChair in Inner City Medicine which supports Dr. Wood. Dr. Hayashi issupported by the Canadian Institutes of Health Research New InvestigatorAward (MSH-141971). Dr. Milloy is supported in part by the US NationalInstitutes of Health (R01DA021525). Dr. Montaner’s TasP research hasreceived support from the BC Ministry of Health, US National Institutes ofHealth (NIDA R01DA036307), International AIDS Society, UNAIDS, WHO,ANRS, IAPAC, UNICEF, MAC AIDS Fund and Open Society Foundations.Institutional grants have been provided by Abbvie, BI, BMS, Gilead Sciences,J&J, Merck and ViiV. He has served on Advisory Boards for Teva, GileadSciences and InnaVirVax. The funding bodies had no role in the studydesign, data collection, analysis, interpretation, or manuscript writing.Availability of data and materialsThe data used for this study is not publicly available. For further informationon the data and materials used in this study, please contact thecorresponding author.Authors’ contributionsKH and MJM designed the study. PN and HD conducted the statisticalanalyses. KH drafted the manuscript, and incorporated suggestions from allco-authors. All authors made significant contributions to the conception ofthe analyses, interpretation of the data, and drafting of the manuscript. Allauthors approved the final version of the manuscript.Competing interestsThe authors declare that they have no competing interests.Consent for publicationNot applicable.Ethics approval and consent to participateAll participants provided written informed consent for study participation.The ACCESS study received annual ethics approval from the University ofBritish Columbia and Providence Health Care Research Ethics Board(H05-50233). The PI of ACCESS (MJM) granted permission to use the data forthe present study, which was part of the larger ACCESSS study activitiesapproved by the research ethics board. The dataset used for the presentstudy was de-identified. The data were located at the British ColumbiaCentre for Excellence in HIV/AIDS.Author details1British Columbia Centre for Excellence in HIV/AIDS, St. Paul’s Hospital, 608 -1081 Burrard Street, Vancouver, BC V6Z 1Y6, Canada. 2Division of AIDS,Department of Medicine, University of British Columbia, 608 - 1081 BurrardStreet, Vancouver, BC V6Z 1Y6, Canada. 3Faculty of Health Sciences, SimonFraser University, 8888 University Drive, Burnaby, BC V5A 1S6, Canada.Received: 24 September 2015 Accepted: 3 August 2016References1. Nakagawa F, May M, Phillips A. Life expectancy living with HIV: recentestimates and future implications. Curr Opin Infect Dis. 2013;26(1):17–25.2. Montaner JS, Lima VD, Barrios R, Yip B, Wood E, Kerr T, et al. Association ofhighly active antiretroviral therapy coverage, population viral load, andyearly new HIV diagnoses in British Columbia, Canada: a population-basedstudy. Lancet. 2010;376(9740):532–9.3. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC,Kumarasamy N, et al. Prevention of HIV-1 infection with early antiretroviraltherapy. N Engl J Med. 2011;365(6):493–505.4. Des Jarlais DC, Mcknight C, Arasteh K, Feelemyer J, Perlman DC, Hagan H, etal. A perfect storm: crack cocaine, HSV-2, and HIV among non-injecting drugusers in New York City. Subst Use Misuse. 2014;49(7):783–92.5. Wechsberg WM, Golin C, El-Bassel N, Hopkins J, Zule W. Currentinterventions to reduce sexual risk behaviors and crack cocaine use amongHIV-infected individuals. Curr HIV/AIDS Rep. 2012;9(4):385–93.6. Edlin BR, Irwin KL, Faruque S, McCoy CB, Word C, Serrano Y, et al.Intersecting epidemics–crack cocaine use and HIV infection among inner-cityyoung adults. Multicenter Crack Cocaine and HIV Infection Study Team. N EnglJ Med. 1994;331(21):1422–7.7. DeBeck K, Kerr T, Li K, Fischer B, Buxton J, Montaner J, et al. Smoking ofcrack cocaine as a risk factor for HIV infection among people who useinjection drugs. CMAJ. 2009;181(9):585–9.8. World Health Organization, United Nations Office on Drugs and Crime, JointUnited Nations Programme on HIV/AIDS. WHO, UNODC, UNAIDS technicalguide for countries to set targets for universal access to HIV prevention,treatment and care for injecting drug users: 2012 revision. Geneva: WorldHealth Organization; 2012.9. Shoptaw S, Montgomery B, Williams CT, El-Bassel N, Aramrattana A, Metsch L, etal. Not just the needle: the state of HIV-prevention science among substanceusers and future directions. J Acquir Immune Defic Syndr. 2013;63 Suppl 2:S174–8.10. Meader N, Li R, Des Jarlais DC, Pilling S. Psychosocial interventions forreducing injection and sexual risk behaviour for preventing HIV in drugusers. Cochrane Database Syst Rev. 2010;1:CD007192.11. Joint United Nations Programme on HIV/AIDS. Fast-track: ending the AIDSepidemic by 2030. Geneva: Joint United Nations Programme on HIV/AIDS; 2014.12. Montaner JS, Wood E, Kerr T, Lima V, Barrios R, Shannon K, et al. Expandedhighly active antiretroviral therapy coverage among HIV-positive drug usersto improve individual and public health outcomes. J Acquir Immune DeficSyndr. 2010;55 Suppl 1:S5–9.13. Arnsten JH, Demas PA, Grant RW, Gourevitch MN, Farzadegan H, Howard AA,et al. Impact of active drug use on antiretroviral therapy adherence and viralsuppression in HIV-infected drug users. J Gen Intern Med. 2002;17(5):377–81.14. Gonzalez A, Barinas J, O’Cleirigh C. Substance use: impact on adherence andHIV medical treatment. Curr HIV/AIDS Rep. 2011;8(4):223–34.15. Hinkin CH, Barclay TR, Castellon SA, Levine AJ, Durvasula RS, Marion SD, etal. Drug use and medication adherence among HIV-1 infected individuals.AIDS Behav. 2007;11(2):185–94.16. Strathdee SA, Palepu A, Cornelisse PGA, Yip B, O'Shaughnessy MV, MontanerJSG, et al. Barriers to use of free antiretroviral therapy in injection drugusers. JAMA. 1998;280(6):547–9.17. Hogg RS, Heath KV, Yip B, Craib KJ, O'Shaughnessy MV, Schechter MT, et al.Improved survival among HIV-infected individuals following initiation ofantiretroviral therapy. JAMA. 1998;279(6):450–4.18. Hogg RS, Yip B, Chan KJ, Wood E, Craib KJP, O'Shaughnessy MV, et al. Ratesof disease progression by baseline CD4 cell count and viral load afterinitiating triple-drug therapy. JAMA. 2001;286(20):2568–77.19. Low-Beer S, Yip B, O'Shaughnessy MV, Hogg RS, Montaner JS. Adherence totriple therapy and viral load response. J Acquir Immune Defic Syndr. 2000;23(4):360–1.20. Wood E, Montaner JSG, Yip B, Tyndall MW, Schechter MT, O'ShaughnessyMV, et al. Adherence and plasma HIV RNA responses to highly activeantiretroviral therapy among HIV-1 infected injection drug users. CMAJ.2003;169(7):656–61.21. Wood E, Hogg RS, Yip B, Harrigan PR, O'Shaughnessy MV, Montaner JSG.Effect of medication adherence on survival of HIV-infected adults who starthighly active antiretroviral therapy when the CD4+ cell count is 0.200 to 0.350 × 10(9) cells/L. Ann Intern Med. 2003;139(10):810–6.22. Lima VD, Harrigan R, Murray M, Moore DM, Wood E, Hogg RS, et al.Differential impact of adherence on long-term treatment response amongnaive HIV-infected individuals. AIDS. 2008;22(17):2371–80.Hayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 11 of 1223. Brecht M-L, Huang D, Evans E, Hser Y-I. Polydrug use and implications forlongitudinal research: ten-year trajectories for heroin, cocaine, andmethamphetamine users. Drug Alcohol Depend. 2008;96(3):193–201.24. Bruce RD, Altice FL, Friedland GH. Pharmacokinetic drug interactionsbetween drugs of abuse and antiretroviral medications: implications andmanagement for clinical practice. Expert Rev Clin Pharmacol. 2008;1(1):115–27.25. Metzger DS, Woody GE, O'Brien CP. Drug treatment as HIV prevention: aresearch update. J Acquir Immune Defic Syndr. 2010;55:S32–6.26. Ingersoll KS, Farrell-Carnahan L, Cohen-Filipic J, Heckman CJ, Ceperich SD,Hettema J, et al. A pilot randomized clinical trial of two medicationadherence and drug use interventions for HIV+ crack cocaine users.Drug Alcohol Depend. 2011;116(1-3):177–87.27. Knapp WP, Soares BGO, Farrel M, Lima MS. Psychosocial interventions forcocaine and psychostimulant amphetamines related disorders. CochraneDatabase Syst Rev. 2007;3:CD003023.28. Castells X, Casas M, Perez-Mana C, Roncero C, Vidal X, Capella D. Efficacy ofpsychostimulant drugs for cocaine dependence. Cochrane Database SystRev. 2010;2:CD007380.29. Minozzi S, Amato L, Davoli M, Farrell M, Lima Reisser AARL, Pani PP, et al.Anticonvulsants for cocaine dependence. Cochrane Database Syst Rev.2008;2:CD006754.30. Kim JH, Lawrence AJ. Drugs currently in Phase II clinical trials for cocaineaddiction. Expert Opin Investig Drugs. 2014;23(8):1105–22.31. Spire B, Lucas GM, Carrieri MP. Adherence to HIV treatment among IDUsand the role of opioid substitution treatment (OST). Int J Drug Policy.2007;18(4):262–70.32. Roux P, Carrieri MP, Villes V, Dellamonica P, Poizot-Martin I, Ravaux I, et al.The impact of methadone or buprenorphine treatment and ongoinginjection on highly active antiretroviral therapy (HAART) adherence:evidence from the MANIF2000 cohort study. Addiction. 2008;103(11):1828–36.33. Gonzalez A, Mimiaga MJ, Israel J, Andres Bedoya C, Safren SA. Substanceuse predictors of poor medication adherence: the role of substance usecoping among HIV-infected patients in opioid dependence treatment.AIDS Behav. 2013;17(1):168–73.34. Roux P, Lions C, Michel L, Cohen J, Mora M, Marcellin F, et al. Predictors ofnon-adherence to methadone maintenance treatment in opioid-dependentindividuals: implications for clinicians. Curr Pharm Des. 2014;20(25):4097–105.35. Marsden J, Eastwood B, Bradbury C, Dale-Perera A, Farrell M, Hammond P,et al. Effectiveness of community treatments for heroin and crack cocaineaddiction in England: a prospective, in-treatment cohort study. Lancet. 2009;374(9697):1262–70.36. Milloy MJ, Kerr T, Bangsberg DR, Buxton J, Parashar S, Guillemi S, et al.Homelessness as a structural barrier to effective antiretroviral therapyamong HIV-seropositive illicit drug users in a Canadian setting. AIDS PatientCare STDS. 2012;26(1):60–7.37. Darke S. Self-report among injecting drug users: a review. Drug AlcoholDepend. 1998;51(3):253–63. discussion 267–8.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Hayashi et al. BMC Infectious Diseases  (2016) 16:455 Page 12 of 12

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0340730/manifest

Comment

Related Items