Open Collections

UBC Faculty Research and Publications

Regional differences in rates of HIV-1 viral load monitoring in Canada: Insights and implications for… Raboud, Janet M; Loutfy, Mona R; Su, DeSheng; Bayoumi, Ahmed M; Klein, Marina B; Cooper, Curtis; Machouf, Nima; Rourke, Sean; Walmsley, Sharon; Rachlis, Anita; Harrigan, P R; Smieja, Marek; Tsoukas, Christos; Montaner, Julio S; Hogg, Robert S Feb 25, 2010

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

Item Metadata


52383-12879_2009_Article_1020.pdf [ 214.9kB ]
JSON: 52383-1.0132550.json
JSON-LD: 52383-1.0132550-ld.json
RDF/XML (Pretty): 52383-1.0132550-rdf.xml
RDF/JSON: 52383-1.0132550-rdf.json
Turtle: 52383-1.0132550-turtle.txt
N-Triples: 52383-1.0132550-rdf-ntriples.txt
Original Record: 52383-1.0132550-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessRegional differences in rates of HIV-1 viral loadmonitoring in Canada: Insights and implicationsfor antiretroviral care in high income countriesJanet M Raboud1,2*, Mona R Loutfy3,4,5,6, DeSheng Su2, Ahmed M Bayoumi3,5,7,8, Marina B Klein9, Curtis Cooper10,Nima Machouf11, Sean Rourke3,7,12,13, Sharon Walmsley2,3, Anita Rachlis3,14, P Richard Harrigan15,16, Marek Smieja17,Christos Tsoukas9, Julio SG Montaner15,16, Robert S Hogg16,18, the CANOC CollaborationAbstractBackground: Viral load (VL) monitoring is an essential component of the care of HIV positive individuals. Rates ofVL monitoring have been shown to vary by HIV risk factor and clinical characteristics. The objective of this studywas to determine whether there are differences among regions in Canada in the rates of VL testing of HIV-positiveindividuals on combination antiretroviral therapy (cART), where the testing is available without financial barriersunder the coverage of provincial health insurance programs.Methods: The Canadian Observational Cohort (CANOC) is a collaboration of nine Canadian cohorts of HIV-positiveindividuals who initiated cART after January 1, 2000. The study included participants with at least one year offollow-up. Generalized Estimating Equation (GEE) regression models were used to determine the effect ofgeographic region on (1) the occurrence of an interval of 9 months or more between two consecutive recordedVL tests and (2) the number of days between VL tests, after adjusting for demographic and clinical covariates.Overall and regional annual rates of VL testing were also reported.Results: 3,648 individuals were included in the analysis with a median follow-up of 42.9 months and a median of15 VL tests. In multivariable GEE logistic regression models, gaps in VL testing >9 months were more likely inQuebec (Odds Ratio (OR) = 1.72, p < 0.0001) and Ontario (OR = 1.78, p < 0.0001) than in British Columbia andamong injection drug users (OR = 1.68, p < 0.0001) and were less likely among older individuals (OR = 0.77 per10 years, p < 0.0001), among men having sex with men (OR = 0.62, p < 0.0001), within the first year of cART(OR = 0.15, p < 0.0001), among individuals on cART at the time of the blood draw (OR = 0.34, p < 0.0001) andamong individuals with VL < 50 copies/ml at the previous visit (OR = 0.56, p < .0001).Conclusions: Significant variation in rates of VL testing and the probability of a significant gap in testing wererelated to geographic region, HIV risk factor, age, year of cART initiation, type of cART regimen, being in the firstyear of cART, AIDS-defining illness and whether or not the previous VL was below the limit of detection.BackgroundViral load (VL) testing is an essential component of thecare of HIV-positive individuals, both with regard totiming of initiation of antiretroviral therapy (ART) andto monitoring of virologic response to combinationART (cART) [1]. The goal of cART is sustained virolo-gic suppression, defined as a VL below the level ofdetection of the test performed [1]. Guidelinesrecommend that HIV-positive individuals receive VLtesting at intervals of three to four months as standardof care [1]. CD4 count monitoring is important fordeciding when to start cART and for determining prog-nosis, but alone is insufficient as a marker of treatmentefficacy as it does not identify individuals experiencingvirologic rebound or failure [2]. Early determination ofvirologic rebound and failure is one of the most crucialcomponents of HIV management as it contributes tothe reduction of ART drug resistance [3]. Lastly,* Correspondence: raboud@lunenfeld.ca1Dalla Lana School of Public Health, University of Toronto, Toronto, CanadaRaboud et al. BMC Infectious Diseases 2010, 10:40© 2010 Raboud et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (, which permits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited.VL monitoring has also been shown to promote treat-ment adherence, which is additionally important formaintaining virologic suppression and reducing the evo-lution of drug resistance [4].Access to VL testing has been studied previously. Inan Ontario cohort, injection drug use, younger age andresidence in Toronto were associated with lower VLtesting rates [5]. In another study, drug users were alsofound to be at risk for irregular VL monitoring [6]. In astudy of individuals who initiated ART between 1994and 2000, individuals with low CD4 counts and highVLs had the highest rates of laboratory testing [7].In this study, we examine whether there are regionaldifferences in patterns of VL testing among individualswho initiated cART therapy since January 1, 2000 inCanada, where VL testing is available without charge toall HIV-positive residents as part of the provincial uni-versal health insurance plans. Furthermore, we identifieddemographic and clinical factors associated with subop-timal frequency of VL testing.MethodsThe Canadian Observational Cohort (CANOC) colla-boration is a Canadian cohort study of antiretroviralnaïve HIV-positive patients initiating cART since Janu-ary 1st 2000. The study was established in March 2008with funding from the Canadian Institutes of HealthResearch (grant# 711098) and the CIHR Canadian HIVTrials Network (CTN242) and includes cohorts andinvestigators from across the country (listed at the endof the manuscript). The collaboration is open to allCanadian HIV treatment cohorts with more than 100eligible patients.Participating cohortsData used in this analysis were from nine cohorts ofHIV-positive individuals in British Columbia (BC),Ontario, and Quebec, including the BC Centre forExcellence in HIV/AIDS Drug Treatment Program,Montreal Chest Institute Immunodeficiency Cohort,The Electronic Antiretroviral Therapy, Clinique Médi-cale l’Actuel, The Canadian HIV/HCV Co-infectionCohort, Ontario Cohort Study, Maple Leaf MedicalClinic, Toronto General Hospital and Ottawa HospitalHIV/HCV Cohort [8].Patient selection and data extraction were performedat the data centres of the participating cohort sites. Inprovinces with multiple cohorts, VL data were enteredfrom each cohort site and not from a provincial datasource. Non-nominal data from each cohort on a prede-fined set of demographic, laboratory, and clinical vari-ables were then pooled and analyzed at the Project DataCentre in Vancouver. All participating cohorts havereceived approval from their institutional ethics boardsand governance committees to contribute non-nominalpatient-specific data to CANOC. Ownership of indivi-dual cohort data remains with the contributing cohortand cohort data can only be used for studies approvedby the CANOC Steering Committee.Eligibility CriteriaEligibility criteria for inclusion in CANOC were docu-mented HIV infection, residence in Canada, aged 18years and over, initiation of three or more antiretroviraldrugs for the first time (i.e. ART-naïve cART start) afterJanuary 1, 2000, and a viral load measurement and CD4cell count within 6 months of the start of therapy. Tobe included in this analysis, individuals had to have atleast one year of follow-up. Viral load measurementswere available both before and after starting cART.Statistical MethodsThree measures were used to assess the frequency of VLtesting. First, the primary outcome of interest wasdefined as a gap between VL tests in excess of 9 months,corresponding to at least three missed or delayed testsassuming the optimal frequency between tests is threemonths. This was felt to be a clinically important gap inVL testing. Second, the annual rate of VL testing wascalculated by dividing the total number of tests for anindividual by the duration of follow-up for that indivi-dual in years. Third, the time interval, defined as thenumber of days between two successive VL tests for asubject, was examined.Demographic and clinical characteristics such as gen-der, race, HIV risk factors, age, geographic region, CD4count and type of cART regimen were compared amongregions with chi square tests for categorical variablesand Wilcoxon rank sum tests for continuous variables.The proportion of individuals with at least one subop-timal interval and the annual rate of VL testing experi-enced were compared among regions and bydemographic and clinical characteristics with the chisquare test and Wilcoxon rank sum test respectively.Generalized Estimating Equation (GEE) logistic regres-sion models, which account for correlation among mul-tiple observations within subjects, were used todetermine factors associated with the occurrence of asuboptimal testing interval[9]. An exchangeable correla-tion structure was assumed for this model. As a form ofsensitivity analysis, we repeated the analysis by defininga suboptimal interval as six months or more, corre-sponding to at least two missed or delayed tests if theoptimal frequency between tests is three months.The relationships between the length of the VL inter-test interval and individual characteristics were exam-ined using GEE linear regression models. Plasma VLlevels, CD4 counts, characteristics of the antiretroviralRaboud et al. BMC Infectious Diseases 2010, 10:40 2 of 9regimen and the calendar year of VL tests were treatedas time-varying covariates in all regression models. Cov-ariates with a p < 0.10 in the univariate regression mod-els were considered as candidates for inclusion in themultivariable model.Data cleaningWhen there was a gap between two consecutive VLtests in excess of nine months, the accompanying CD4test dates were examined to validate the VL test dates.Viral load tests within 8 days of each other and withresults within 0.1 log10 copies/mL were considered to beduplicate measurements. In such cases, the first VL datewas kept and the average of the VL measurements wasassigned as the VL value.ResultsStudy cohort and overall VL testing3,648 subjects met cohort inclusion criteria. Demo-graphic characteristics are described by region inTable 1. The median number of VL tests and the med-ian length of follow-up were 15 (interquartile range(IQR) [9-13][14-22]) and 42.9 months (IQR [25.4-64.4]),respectively. The median rate of testing was 4.3 VLmeasurements per year (IQR [3.4-5.5]). On average, 83%of the population had more than 3 tests per year and58% had more than 4 tests per year. Of 57,308 intervalsbetween VL tests, 2.3% were > 9 months and 6.8% were> 6 months. The median annual rates of VL testingwere 4.9 (IQR [3.8-6.3]) in BC, 3.9 (IQR [3.1-4.8]) inOntario and 4.0 (IQR [3.2-4.8]) in Quebec.Table 1 Comparison of Baseline Demographic and Clinical Characteristics by RegionRegionCharacteristics Total (n = 3648) BC (n = 1674) Ontario (n = 1143) Quebec (n = 831) p valueAge 40 (34-46) 40 (34-47) 39 (34-45) 40 (34-46) <0.01Male 2925 (80%) 1328 (79%) 934 (82%) 663 (80%) 0.28Race†Caucasian 956 (52%) 552 (45%) 309 (50%) 95 (84%) <0.0001Black 181 (10%) 31 (3%) 138 (26%) 12 (11%)First Nation 121 (7%) 103 (9%) 15 (3%) 3 (3%)Mixed 465 (25%) 464 (38%) 1 (<% 0 (0%)Other 128 (7%) 66 (5%) 59 (11%) 3 (3%)Missing 1837 497 622 718Risk factor‡MSM 1015 (34%) 203 (12%) 362 (66%) 450 (60%) <0.0001IDU 632 (21%) 442 (26%) 82 (13%) 108 (14%) <0.0001Heterosexual 486 (16%) 98 (6%) 188 (35%) 200 (27%) <0.0001Endemic country 233 (19%) 0 (0%) 92 (20%) 141 (19%) 0.52Blood product recipient 43 (1%) 10 (1%) 25 (6%) 8 (1%) <0.0001Other/Missing 1554 (43%) 961 (57%) 569 (45%) 147 (9%) <0.0001AIDS defining illness 481 (13%) 259 (15%) 115 (10%) 107 (13%) <0.001cART initiation date 2003 (2001-05) 2003 (2001-05) 2004 (2002 - 05) 2004 (2002-05) <0.0001Type of cARTNNRTI-based 1586 (43%) 753 (45%) 512 (45%) 321 (39%) <0.0001Boosted PI-based 1433 (39%) 726 (43%) 397 (35%) 310 (37%)Single PI-based 384 (11%) 151 (9%) 137 (12%) 96 (12%)Other cART 245 (7%) 44 (3%) 97 (8%) 104 (13%)CD4 count (cells/mm3) 190 (100-277) 170 (80-260) 202 (107-296) 208 (127-290) <0.0001CD4 count category<200 cells/mm3 1921 (53%) 981 (59%) 555 (49%) 385 (47%) <0.0001200 - 350 cells/mm3 1175 (32%) 486 (29%) 374 (33%) 315 (38%)> 350 cells/mm3 544 (15%) 207 (12%) 212 (19%) 125 (15%)VL (log10 copies/mL) 4.9 (4.4-5.1) 5.0 (4.6-5.0) 4.8 (4.3-5.3) 4.8 (4.2-5.2) <0.01VL <50 copies/mL 139 (4%) 56 (3%) 48 (4%) 35 (4%) 0.40Hepatitis C co-infection 736 (30%) 501 (44%) 124 (22%) 111 (15%) <0.0001MSM = men who have sex with men, IDU = injection drug user, cART = combination antiretroviral therapy, NNRTI = non nucleoside reverse transcriptaseinhibitor, PI = protease inhibitor, VL = viral load.†Race percentages are calculated as percentages of non-missing values. ‡Risk factors are not mutually exclusive. Results are N (%) or median (interquartile range).Missing values are only reported for variables with more than 25% missing data.Raboud et al. BMC Infectious Diseases 2010, 10:40 3 of 9Analysis of the annual rate of VL testingIn univariate analyses, higher annual rates of VL testingwere associated with residence in BC, age > 40 years,white race, male gender, later year of cART initiation,pretreatment VL ≥ 105 copies/mL, an HIV risk categoryof “men who have sex with men” (MSM), history of anAIDS-defining illness, boosted-protease inhibitor (PI)-based cART regimen, not being co-infected with Hepati-tis C and not being an injection drug user (IDU)(Table 2).Analysis of gaps of greater than nine months and sixmonthsOf the 3,648 patients eligible for the analysis, 26% and51% of the population had experienced at least onenine-month and one six-month gap during their follow-up, respectively. Proportions of patients with at leastone nine-month gap and with at least one six-monthgap are shown by demographic and clinical characteris-tics in Table 2. Results of univariate GEE logistic regres-sion models are shown in Table 3. In the multivariableGEE logistic regression model (Table 4), gaps of bothnine and six months were significantly more likely tooccur in Ontario and Quebec and among IDUs. Gaps ofboth nine and six months were significantly less likelyto occur among older individuals, among MSM, inrecent calendar years of VL test, among individuals oncART, in the first year of cART and if the VL had beensuppressed at the previous visit.Analysis of the time interval between successive testsIn univariate GEE linear regression models, covariateswhich were associated with a decrease in the numberof days between VL tests included age, male gender,later year of initiating cART, being within the first yearof cART, receiving PI-boosted cART regimen, higherbaseline VL and having been diagnosed with an AIDS-defining illness (Table 3). Covariates which were asso-ciated with an increase in the number of days betweenVL tests included living in Quebec or Ontario com-pared to living in BC, not being on cART, havingmore drugs in the cART regimen, higher baseline CD4count, being an IDU and having a suppressed VL atthe previous test. The predicted time interval betweensuccessive tests was 85 days for a 40 year old patientliving in BC who started cART between 2002 and2004, had completed the first year of a boosted-PIbased cART regimen, with the most recent VL belowthe limit of detection, being neither a MSM nor anIDU, and taking cART at the current VL test (Table4). The predicted time intervals would be 104 or 105days for a similar patient in Ontario or Quebec,respectively.DiscussionThere was considerable regional variation in annualrates of VL measurement documented in this Canada-wide study. This variation remained significant evenafter adjusting for demographic variables such as ageand HIV risk factors and clinical variables such as yearof cART initiation, type of cART regimen, being in thefirst year of cART, AIDS-defining illness and whether ornot the previous VL was below the limit of detection. InOntario and Quebec, VL was measured quarterly onaverage. In BC, rates of VL measurement were evenmore frequent than guidelines suggest, with an averagemeasurement frequency of almost five times annually.There are a number of possible explanations for theregional differences in rates of viral load measurement.Some of the difference in measurement rate was due toregional differences in demographic factors such as theproportions of IDUs, who typically have less frequentviral load testing, or pregnant women, in whom viralload is monitored more closely. Further differencescould be due to variation in rates of VL blips amongregions, after which a repeat VL measurement is typi-cally ordered. Recent data has documented differingrates of blips by VL assay [10,11] and this may explainthe higher rates of testing in BC. Regional differences inVL testing policies may also have an impact. Whilethere are no differences among provinces in the officialguidelines for the frequency of VL measurement, it ispossible that there are differences in the implementationof the guidelines. In Ontario, a VL will not be per-formed by the laboratory if one has been done withinthe last 14 days. Furthermore, rates of VL measurementmay be higher in British Columbia due to the fact thatall antiretroviral drug distribution and VL testing iscoordinated through a single center in the province.Lastly, differences in participation rates among provincesin research studies, which may require more frequentVL testing may explain some of regional variation.Effective therapy should result in at least a 90% or10-fold (1.0 log10 copies/mL) decrease in plasma VL inthe first month and suppression to below 50 copies/mLby 24 weeks, depending on the pretreatment VL level[1]. Current guidelines suggest once VL suppression tobelow 50 copies/mL is confirmed, it should be assessedat regular intervals (e.g. every 3 or 4 months). Isolatedepisodes of low-level viremia ("blips”) are not necessarilypredictive of subsequent virologic failure, but consistentelevations of VL above 50 copies/mL meet a strict defi-nition of virologic failure. Emergence of a detectable VLin a previously suppressed patient (i.e. previously consis-tently < 50 copies/mL) mandates re-evaluation of thecase, including repeated testing to confirm whether thisrepresents a “blip” or virologic failure. Confirmed VLRaboud et al. BMC Infectious Diseases 2010, 10:40 4 of 9Table 2 Annual rates of viral load testing and proportions of participants with gaps in testing by baselinedemographic and clinical characteristicsannual rate of VL testing Subjects having ≥ 1 gap of >9 months Subjects having ≥ 1 gap of >6 monthsCharacteristics N median (IQR) p value # (%) p value # (%) p valueOverall 3648 4.3 (3.4-5.5) 943 (26%) 1848 (51%)RegionBC 1674 4.9 (3.8-6.3) <.0001 347 (21%) <.0001 672 (40%) <.0001Ontario 1143 3.9 (3.1-4.8) 357 (31%) 699 (61%)Quebec 831 4.0 (3.2-4.8) 239 (29%) 477 (57%)Age (years)≤ 40 1942 4.1 (3.2-5.3) <.0001 614 (32%) <.0001 1122 (58%) <.0001> 40 1706 4.5 (3.6-5.8) 329 (19%) 726 (43%)GenderFemale 723 3.9 (3.1-5.0) <.0001 255 (35%) <.0001 475 (66%) <.0001Male 2925 4.4 (3.5-5.7) 688 (24%) 1373 (47%)CaucasianNo 855 4.4 (3.4-5.7) 0.33 255 (30%) 0.04 463 (54%) <.001Yes 956 4.4 (3.4-5.9) 244 (26%) 439 (46%)Risk FactorIDU 632 4.0 (3.0-5.1) <.0001 247 (39%) <.0001 417 (66%) <.0001MSM 976 4.3 (3.5-5.4) 186 (19%) 421 (43%)Blood product recipient 32 4.1 (3.5-5.1) 8 (25%) 18 (56%)Endemic 209 4.1 (3.4-4.8) 60 (29%) 128 (61%)Heterosexual contact 245 4.2 (3.4-5.7) 64 (26%) 135 (55%)Not reported/Other 1554 4.5 (3.4-5.8) 378 (24%) 729 (47%)AIDS defining illnessNo 3167 4.2 (3.3-5.5) <.0001 858 (27%) <.0001 1641 (52%) <.001Yes 481 4.7 (3.8-5.9) 85 (18%) 207 (43%)Type of cARTNNRTI-based 1586 4.1 (3.2-5.0) <.0001 425 (27%) <.0001 874 (55%) <.0001Boosted PI-based 1433 5.0 (3.7-6.5) 292 (20%) 573 (40%)Single PI-based 384 4.0 (3.1-4.9) 149 (39%) 256 (67%)Other cART 245 3.9 (3.2-4.7) 77 (31%) 145 (59%)Year of initiating cART2000 482 3.7 (2.8-4.7) <.0001 219 (45%) <.0001 346 (72%) <.00012001-2004 2036 4.2 (3.3-5.3) 599 (29%) 1132 (56%)>2004 1130 4.9 (3.9-6.4) 125 (11%) 370 (33%)CD4 count<200 cells/mm3 1921 4.4 (3.4-5.7) <.0001 474 (25%) <.001 941 (49%) <.0001200 - 350 cells/mm3 1175 4.3 (3.4-5.5) 286 (24%) 560 (48%)> 350 cells/mm3 544 4.1 (3.0-5.2) 181 (33%) 341 (63%)VL (copies/mL)≥ 50 3509 4.3 (3.4-5.6) <.001 905 (26%) 0.68 1770 (50%) 0.19<50 139 3.9 (2.9-4.8) 38 (27%) 78 (56%)Hepatitis C Co-infectionNo 1712 4.4 (3.6-5.6) <.0001 346 (20%) <.0001 756 (44%) <.0001Yes 736 4.1 (3.0-5.3) 281 (38%) 473 (64%)MSM = men who have sex with men, IDU = injection drug use, cART = combination antiretroviral therapy, NNRTI = non nucleoside reverse transcriptaseinhibitor, PI = protease inhibitor.† Risk factors were grouped hierarchically for comparison purposes: IDU, MSM, blood product recipient, origin/residence in an HIV-endemic area, heterosexualtransmission.Raboud et al. BMC Infectious Diseases 2010, 10:40 5 of 9rebound should prompt a careful evaluation of regimentolerability, drug-drug interactions, resistance andpatient adherence. CD4+ cell counts should generally beassessed in concert with VL.In terms of the demographic findings, our findings aresimilar to those of the study conducted in Ontario sev-eral years ago and could likely be generalizable to othersettings with universal health care programs. In settingswith user-pay programs for virologic monitoring, itseems likely that socioeconomic factors will have aconsiderably greater influence on rates of testing. Inresource limited settings, availability and travel distancesare likely to remain significant barriers to regular moni-toring of VL levels.Our results have important implications for guidelinesand for research into the monitoring of VL. For patientsfor whom VL is measured less frequently, virologicrebound will be detected later on average and patientswill remain on failing regimens longer than is necessary.In some circumstances, however, it is safe to measureTable 3 Univariate GEE regression models of interval (days), probability of >9 and >6 months between viral loadmeasurementsInterval (days) Gaps of >9 months Gaps of >6 monthsCovariates Estimate (95% CI) p value OR (95% CI) p value OR (95% CI) p valueRegionQuebec 24.1 (20.2, 28.0) <.0001 1.65 (1.37, 1.98) <.0001 1.71 (1.50, 1.96) <.0001Ontario 25.6 (22.2, 28.9) <.0001 1.77 (1.51, 2.08) <.0001 2.14 (1.90, 2.41) <.0001BC (reference) 0 1 1Age (per 10 years) -7.0 (-8.6, -5.3) <.0001 0.70 (0.65, 0.76) <.0001 0.76 (0.72, 0.81) <.0001Male -12.9 (-16.9, -8.8) <.0001 0.58 (0.50, 0.68) <.0001 0.62 (0.55, 0.69) <.0001Caucasian -2.1 (-6.6, 2.4) 0.36 0.78 (0.64, 0.95) 0.01 0.83 (0.71, 0.97) 0.02Risk factorMSM 2.2 (-1.1, 5.6) 0.19 0.74 (0.61, 0.89) <.01 0.86 (0.75, 0.97) 0.02IDU 18.5 (13.1, 23.9) <.0001 2.06 (1.74, 2.44) <.0001 1.80 (1.58, 2.04) <.0001Heterosexual 14.3 (9.4, 19.3) <.0001 1.55 (1.26, 1.89) <.0001 1.57 (1.36, 1.81) <.0001Endemic country -0.1 (-6.3, 6.1) 0.98 1.05 (0.79, 1.39) 0.74 1.08 (0.89, 1.31) 0.44Blood products 6.3 (-3.8, 16.4) 0.22 0.93 (0.52, 1.65) 0.79 1.19 (0.76, 1.89) 0.45Year of initiating cART>2004 -31.0 (-37.4, -24.6) <.0001 0.36 (0.28, 0.46) <.0001 0.49 (0.41, 0.57) <.00012001-2004 -18.4 (-24.7, -12.1) <.0001 0.62 (0.52, 0.73) <.0001 0.70 (0.61, 0.80) <.00012000 (ref.) 0 1 11st year of cART -36.1 (-37.6, -34.6) <.0001 0.15 (0.13, 0.18) <.0001 0.27 (0.25, 0.30) <.0001Type of cARTNNRTI-based 0 1 1Boosted PI-based -16.7 (-20.0, -13.4) <.0001 0.76 (0.64, 0.89) <.001 0.71 (0.63, 0.80) <.0001Single PI-based 5.4 (-0.3, 11.0) 0.06 1.37 (1.12, 1.68) <.01 1.19 (1.02, 1.40) 0.03Other cART 6.2 (-0.2, 12.6) 0.06 1.17 (0.89, 1.53) 0.26 1.05 (0.86, 1.27) 0.65Boosted PI (Y/N) -18.3(-21.4, -15.2) <.0001 0.70 (0.60, 0.81) <.0001 0.68 (0.61, 0.76) <.0001On cART at current visit -32.2 (-34.9, -29.6) <.0001 0.31 (0.28, 0.35) <.0001 0.38 (0.35, 0.42) <.0001On cART at previous visit -12.4 (-14.5, -10.3) <.0001 0.58 (0.52, 0.65) <.0001 0.65 (0.61, 0.70) <.0001Baseline CD4 count>350 cells/mm3 12.7 (7.8, 17.7) <.0001 1.43 (1.19, 1.72) <.001 1.51 (1.31, 1.74) <.0001200-350 cells/mm3 2.7 (-0.7, 6.1) 0.12 0.94 (0.80, 1.10) 0.44 0.97 (0.86, 1.09) 0.57<200 cells/mm3 (ref.) 0 1 1Baseline VL <50 copies/mL 17.3 (7.2, 27.3) <.001 1.17 (0.82, 1.67) 0.38 1.47 (1.13, 1.90) <.01Baseline VL (log10 copies/mL) -7.2 (-9.2, -5.3) <.0001 0.85 (0.79, 0.91) <.0001 0.84 (0.79, 0.88) <.0001VL <50 copies/mL at previous test 18.3 (16.5, 20.0) <.0001 0.71 (0.64, 0.78) <.0001 1.18 (1.11, 1.26) <.0001VL (log10 copies/mL) at previous visit -3.1 (-3.9, -2.3) <.0001 1.23 (1.18, 1.28) <.0001 1.02 (0.99, 1.04) 0.24AIDS defining illness -14.1 (-17.5, -10.6) <.0001 0.53 (0.42, 0.68) <.0001 0.59 (0.50, 0.70) <.0001Hepatitis C co-infection 17.9 (13.0, 22.7) <.0001 2.45 (2.06, 2.91) <.0001 1.98 (1.74, 2.26) <.0001Positive for HCV antibodies 14.0 (9.3, 18.7) <.0001 2.13 (1.81, 2.50) <.0001 1.64 (1.45, 1.85) <.0001MSM = men who have sex with men, IDU = injection drug use, cART = combination antiretroviral therapy, NNRTI = non nucleoside reverse transcriptaseinhibitor, PI = protease inhibitor, VL = viral load, HCV = hepatitis C virus.Raboud et al. BMC Infectious Diseases 2010, 10:40 6 of 9VL less frequently [12]. A reduced monitoring schedulewould result in cost savings due to both the cost of thelaboratory testing and the cost of the physician time. AsART is a life-long commitment, cost considerations arenot insignificant. Further, less frequent VL monitoringwould reduce the inconvenience to the patient, whichmay improve overall compliance. However, while it maybe safe and cost effective to extend the period betweenVL measurements from three to six months in patientswith virologic suppression, periods in excess of ninemonths are less easily justified and place patients atunnecessary risk. Steps should be taken to facilitatemore careful VL monitoring in groups at high risk forgaps in testing.A strength of our study is that VL monitoring is pro-vided free of charge to all HIV-positive individuals inCanada, so that we were able to examine correlates ofrates of VL testing in the absence of financial barriers totesting. Further, by combining data from nine sitesacross Canada, we achieved a large sample size as wellas a good representation by region, gender, race, andother demographic characteristics. CANOC representsabout a quarter of all HIV-positive people in the countrywho are currently on ART and approximately half ofthose who have initiated on more modern regimenssince 2000 [13]. This is the largest sample of HIV-posi-tive people on ART compiled in Canada and representsone of the most representative samples put together in ahigh-income country. One limitation of this dataset isthat this information is not currently available on anation-wide basis, so regional differences in provincesother than BC, Ontario and Quebec could not be exam-ined. A further limitation is that data were missing onrace and HIV risk factors for many patients.Future plans for investigation include further examina-tion of the reasons for the regional differences in ratesof VL testing. Also, an economic analysis to determineif the less frequent testing found in Ontario and Quebecis adequate in terms of clinical outcomes and couldresult in financial savings would be informative. Particu-larly, an analysis of VL testing frequency in patientswith full virologic suppression where less testing may beadequate would be of interest and its impact on costsavings could be important.ConclusionsIn our setting, with universal health care and similarregional guidelines for viral load testing, significant varia-tion in rates of VL measurement and the probability of asignificant gap in testing were related to geographicregion, age, HIV risk factors and clinical variables such asyear of cART initiation, type of cART regimen, being inthe first year of cART, AIDS-defining illness and whetheror not the previous VL was below the limit of detection.Table 4 Multivariate GEE regression models of testing intervals and of probability of >9 and >6 months between viralload measurementsInterval (days) between successivetestsProbability of an interval>9 monthsProbability of an interval>6 monthsCovariates Estimate (95% CI) p value Odds Ratio (95% CI) p value Odds Ratio (95% CI) p valueIntercept 125.8 (118.8, 132.9) <.0001 0.09 (0.07, 0.11) <.0001 0.20 (0.16, 0.23) <.0001RegionQuebec 20.3 (15.5, 25.0) <.0001 1.72 (1.39, 2.14) <.0001 1.61 (1.37, 1.89) <.0001Ontario 18.7 (14.3, 23.0) <.0001 1.78 (1.37, 2.31) <.0001 1.77 (1.47, 2.14) <.0001BC 0 1 1Age (per 10 years) -3.8 (-5.4, -2.2) <.0001 0.77 (0.70, 0.85) <.0001 0.77 (0.72, 0.83) <.0001Risk factorMSM -9.2 (-13.1, -5.4) <.0001 0.62 (0.49, 0.78) <.0001 0.66 (0.56, 0.77) <.0001IDU 16.0 (11.1, 20.9) <.0001 1.68 (1.38, 2.05) <.0001 1.71 (1.48, 1.99) <.0001Year of initiating cART>2004 -15.8 (-21.9, -9.7) <.0001 0.62 (0.46, 0.83) 0.001 0.74 (0.60, 0.91) 0.0052001-2004 -12.4 (-18.2, -6.6) <.0001 0.70 (0.57, 0.86) 0.0006 0.75 (0.64, 0.89) 0.00072000 (ref.) 0 1 11st year of cART -24.3 (-26.0, -22.6) <.0001 0.15 (0.12, 0.20) <.0001 0.32 (0.29, 0.36) <.0001On cART at current visit -27.8 (-31.0, -24.6) <.0001 0.34 (0.29, 0.40) <.0001 0.40 (0.36, 0.45) <.0001VL <50 copies/mL at previous test 8.6 (6.6, 10.7) <.0001 0.56 (0.48, 0.65) <.0001 0.90 (0.82, 0.98) 0.02Boosted PI based cART -9.1 (-12.2, -6.0) <.0001 0.82 (0.71, 0.94) 0.006AIDS defining illness -4.9 (-8.5, -1.4) 0.007MSM = men who have sex with men, IDU = injection drug use, cART = combination antiretroviral therapy, VL = viral load, PI = protease inhibitor.Raboud et al. BMC Infectious Diseases 2010, 10:40 7 of 9List of abbreviationsAIDS: acquired immune deficiency syndrome; ART: antiretroviral therapy; BC:British Columbia; CANOC: Canadian Observational Cohort; cART: combinationantiretroviral therapy; CI: confidence interval; GEE: generalized estimatingequation; HCV: hepatitis C virus; HIV: human immunodeficiency virus; IDU:injection drug user; IQR: interquartile range; MSM: men who have sex withmen; NNRTI: non nucleoside reverse transcriptase inhibitor; NRTI: nucleosidereverse transcriptase inhibitor; OR: odds ratio; PI: protease inhibitor; VL: viralload.AcknowledgementsWe would like to thank all the participants for allowing their information tobe a part of the CANOC collaboration.Sources of funding: This work was supported by an emerging team grantfrom the Canadian Institutes of Health Research (#711098) and by the CIHRCanadian HIV Trials Network. MRL receives salary support from the CanadianInstitutes of Health Research. JMR, CC and SLW have Career Scientist Awardsfrom the Ontario HIV Treatment Network. AMB holds a Chair in AppliedHealth Services Research from the Canadian Institutes of Health Researchand the Ontario Ministry of Health and Long-Term Care. MBK is supportedby a Chercheur-Boursier clinicien senior career award from the Fonds derecherche en santé du Québec (FRSQ). JSGM is supported by an Avant-Garde Award from the National Institute of Drug Abuse, National Institutesof Health.The CANOC Collaboration includes:Community Advisory Committee: Sean Hosein (chair), Bruno Lemay, ShariMargolese, Evelyne Ssengendo, Zoran Stjepanovic.Investigators:Gloria Aykroyd (Ontario HIV Treatment Network), Louise Balfour (OntarioHIV Treatment Network, University of Ottawa), Ahmed Bayoumi (Ontario HIVTreatment Network, University of Toronto), John Cairney (Ontario HIVTreatment Network, University of Toronto), Jeff Cohen (Windsor RegionalHospital), Liviana Calzavara (Ontario HIV Treatment Network, University ofToronto), Brian Conway (Department of Pharmacology and Therapeutics,University of British Columbia), Curtis Cooper (Ontario HIV Treatment Network,University of Ottawa), Pierre Côté (Clinique Médicale du Quartier Latin),Joseph Cox (Montreal General Hospital, McGill University Health Centre, Healthand Social Services Agency of Montreal), Kevin Gough (Ontario HIV TreatmentNetwork, University of Toronto), Silvia Guillemi (British Columbia Centre forExcellence in HIV/AIDS, University of British Columbia), David Haase (CapitalDistrict Health Authority, Halifax), Shariq Haider (McMaster University),Richard Harrigan (British Columbia Centre for Excellence in HIV/AIDS,University of British Columbia), Marianne Harris (British Columbia Centre forExcellence in HIV/AIDS), George Hatzakis (McGill University), Robert Hogg(British Columbia Centre for Excellence in HIV/AIDS, Simon Fraser University),Don Kilby (Ontario HIV Treatment Network), Marina Klein (Montreal ChestInstitute Immunodeficiency Service Cohort, McGill University), Richard Lalonde(Montreal Chest Institute Immunodeficiency Service Cohort, McGill University),Viviane Lima (British Columbia Centre for Excellence in HIV/AIDS, University ofBritish Columbia), Mona Loutfy (University of Toronto, Maple Leaf MedicalClinic), Nima Machouf (Clinique Medicale l’Actuel, Université de Montréal), EdMills (British Columbia Centre for Excellence in HIV/AIDS, University of Ottawa),Peggy Millson (Ontario HIV Treatment Network, University of Toronto), JulioMontaner (British Columbia Centre for Excellence in HIV/AIDS, University ofBritish Columbia), David Moore (British Columbia Centre for Excellence in HIV/AIDS, University of British Columbia), Neora Pick (Oak Tree Clinic, University ofBritish Columbia), Janet Raboud (University of Toronto, University HealthNetwork), Anita Rachlis (Sunnybrook Health Sciences Centre, Ontario HIVTreatment Network, University of Toronto), Stanley Read (Ontario HIVTreatment Network, University of Toronto), Danielle Rouleau (Centrehospitalier de l’Université de Montréal), Sean Rourke (Ontario HIV TreatmentNetwork, University of Toronto), Marek Smieja (Ontario HIV TreatmentNetwork, McMaster University), Irving Salit (Ontario HIV Treatment Network,University of Toronto), Roger Sandre (HAVEN group, Hôpital Régional deSudbury Regional Hospital), Darien Taylor (Ontario HIV Treatment Network,Canadian AIDS Treatment Information Exchange), Benoit Trottier (CliniqueMedicale l’Actuel, Université de Montréal), Chris Tsoukas (McGill University),Mark Tyndall (Native Health Centre and St. Paul ‘s Hospital), SharonWalmsley (University Health Network, University of Toronto), Wendy Wobeser(Ontario HIV Treatment Network, Queens University).Analysts and Staff:Svetlana Draskovic ((British Columbia Centre for Excellence in HIV/AIDS), MarkFisher (Ontario HIV Treatment Network), Sandra Gardner (University ofToronto), Nada Gataric (British Columbia Centre for Excellence in HIV/AIDS),Maggie Li (University Health Network), David Milan (British Columbia Centre forExcellence in HIV/AIDS), Sergio Rueda (Ontario HIV Treatment Network), NahimiSamia (McGill University, Anya Shen (British Columbia Centre for Excellence inHIV/AIDS), Sherine Sterling (University Health Network), Benita Yip (BritishColumbia Centre for Excellence in HIV/AIDS), Hong Yang (McGill University).Author details1Dalla Lana School of Public Health, University of Toronto, Toronto, Canada.2Division of Infectious Disease, University Health Network, Toronto, Canada.3Department of Medicine, University of Toronto, Toronto, Canada. 4Women’sCollege Research Institute, Women’s College Hospital, Toronto, Canada.5Department of Health Policy, Management and Evaluation, University ofToronto, Toronto, Canada. 6Maple Leaf Medical Clinic, Toronto, Canada.7Centre for Research on Inner City Health, The Keenan Research Centre inthe Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, Canada.8Division of General Internal Medicine, St. Michael’s Hospital, Toronto,Canada. 9Department of Medicine, McGill University Health Centre, Montreal,Canada. 10University of Ottawa, Ottawa, Canada. 11Clinique médicale l’Actuel,Montreal, Canada. 12Department of Psychiatry, University of Toronto,Toronto, Canada. 13Ontario HIV Treatment Network, Toronto, Canada.14Division of Infectious Disease, Sunnybrook Health Sciences Centre, Toronto,Canada. 15Faculty of Medicine, University of British Columbia, Vancouver,Canada. 16The British Columbia Centre for Excellence in HIV/AIDS, Vancouver,Canada. 17Department of Pathology and Molecular Medicine, McMasterUniversity, Hamilton, Canada. 18Faculty of Health Sciences, Simon FraserUniversity, Vancouver, Canada.Authors’ contributionsDS conducted the statistical analyses under the direction of JMR, MRL andAMB. JMR and MRL drafted the manuscript. RSH, CC, MBK, MRL, NM, JSGM,JMR, SBR, CT are principal investigators of CANOC and contributed cohortdata to CANOC. All authors reviewed the manuscript during preparation,provided critical feedback and approved the final manuscript.Competing interestsThe authors declare that they have no competing interests.Received: 21 August 2009 Accepted: 25 February 2010Published: 25 February 2010References1. Hammer SM, Eron JJ Jr, Reiss P, Schooley RT, Thompson MA, Walmsley S,Cahn P, Fisch MA, Gatell JM, Hirsch MS, Jacobsen DM, Montaner JS,Richman DD, Yeni PG, Volberding PA, International AIDS Society - USA:Antiretroviral treatment of adult HIV infection: 2008 recommendationsof the International AIDS Society-USA panel. JAMA 2008, 300:555-70.2. Moore DM, Awor A, Downing R, Kaplan J, Montaner JS, Hancock J, Were W,Mermin J: CD4+ T-cell monitoring does not accurately identify HIV-infected adults with virologic failure receiving antiretroviral therapy.J Acquir Immune Defic Syndr 2008, 49:477-84.3. Hirsch MS, Gunthard HF, Schapiro M, Brun-Vezinet F, Clotet B, Hammer SM,Johnson VA, Kuritzkes DR, Mellors JW, Pillay D, Yeni PG, Jacobsen DM,Richman DD: Antiretroviral drug resistance testing in adult HIV-1infection: 2008 recommendations of an International AIDS Society- USApanel. Clin Infect Dis 2008, 47:266-85.4. Wilson D, Keiluhu AK, Kogrum S, Reid T, Seriratana N, Ford N, Kyawkyaw M,Talangsri P, Taochalee N: HIV-1 viral load monitoring: an opportunity toreinforce treatment adherence in a resource-limited setting in Thailand.Trans R Soc Trop Med Hyg 2009, 103:601-6.5. Raboud JM, Abdurrahman ZB, Major C, Millson P, Robinson G, Rachlis A,Bayoumi AM: Nonfinancial factors associated with decreased plasma viralload testing in Ontario, Canada. J Acquir Immune Defic Syndr 2005,39:327-32.6. Laine C, Zhang D, Hauck WW, Turner BJ: HIV-1 RNA viral load monitoringin HIV-infected drug users on antiretroviral therapy; relationship withoutpatient care patterns. J Acquir Immune Defic Syndr 2002, 29:270-4.7. Druyts EF, Yip B, Lima VD, Burke TA, Lesovski D, Fernandes KA, McInnes CW,Rustad CA, Montaner JSG, Hogg RS: Health care services utilizationRaboud et al. BMC Infectious Diseases 2010, 10:40 8 of 9stratified by virological and immunological markers of HIV: evidencefrom a universal health care setting. HIV Medicine 2009, 10:88-93.8. Palmer AK, Klein MB, Raboud JM, Cooper C, Loutfy M, Machouf N,Montaner JSG, Rourke SB, Smieja M, Tsoukas C, Yip B, Milan D, Hogg RS, theCANOC Collaboration: Cohort Profile: The Canadian Observational Cohort(CANOC) Collaboration. International Journal of Epidemiology .9. Zeger SL, Liang KY: Longitudinal data analysis for discrete andcontinuous outcomes. Biometrics 1986, 42:121-30.10. Lima V, Harrigan R, Montaner JSG: Increased reporting of detectableplasma HIV-1 RNA levels at the critical threshold of 50 copies permilliliter with the Taqman assay in comparison to the Amplicor assay.J Acquir Immune Defic Syndr 2009, 51:3-6.11. Smit E, Bhattacharya S, Osman H, Taylor S: Increased Frequency of HIV-1Viral Load Blip Rate Observed After Switching from Roche CobasAmplicor to Coas Taqman Assay. J Acquir Immune Deficiency Syndr 2009,51:364-5.12. Reekie J, Mocroft A, Sambatakou H, Machala L, Chiesi A, van Lunzen J,Clumeck N, Kirk O, Gazzard B, Lundgren JD, the EuroSIDA study group:Does less frequent routine monitoring of patients on a stable, fullysuppressed cART regimen lead to an increased risk of treatment failure?.AIDS 2008, 22:2381-2390.13. Boulos D, Yan P, Schanzer D, Remis RS, Archibald CP: Estimates of HIVprevalence and incidence in Canada, 2005. Canada Communicable DiseaseReport 2006, 32:164-74.Pre-publication historyThe pre-publication history for this paper can be accessed here:[]doi:10.1186/1471-2334-10-40Cite this article as: Raboud et al.: Regional differences in rates of HIV-1viral load monitoring in Canada: Insights and implications forantiretroviral care in high income countries. BMC Infectious Diseases 201010:40.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at et al. BMC Infectious Diseases 2010, 10:40 9 of 9


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