Antiretroviral treatment outcomes among foreign-born and aboriginal peoples living with HIV/AIDS in northern Alberta.

Lefebvre, Megan E. ; Hughes, Christine A. ; Yasui, Yutaka 等

The HIV/AIDS epidemic disproportionately impacts vulnerable populations. In 2012, persons born in HIV-endemic countries comprised approximately 2.2% of the Canadian population but accounted for 13% of newly diagnosed HIV cases. (1) In Alberta, from 2000-2005, next to Caucasians, Aboriginal peoples made up the second-largest group of HIV cases with known ethnicity. After 2006, however, there have been more individuals from HIV-endemic countries diagnosed with HIV compared to Aboriginal peoples. (2) Antiretroviral therapy (ART) has the potential to dramatically increase healthy life expectancy, but requires lifelong treatment and a high level of adherence. Despite the well-documented benefits of ART, vulnerable HIV-infected patients in Canada appear to experience less successful ART outcomes. Recently Martin et al. (3) found that after starting ART, Canadian-born Aboriginal (CBA) patients experienced higher rates of treatment failure compared to Canadian-born non-Aboriginal (CBNA) patients. Researchers have also noted that CBA HIV-infected patients experienced higher rates of HIV-related and all-cause mortality compared to CBNA patients. (3,4)

Little attention has been given to ART outcomes among foreign-born (FB) patients in North America. A better understanding of ART outcomes in the FB population is important as this is a substantial and growing population. (1,2) The stress of resettlement after migration and lack of familiarity with the Canadian medical system can impact access to health care, resulting in poorer ART outcomes. (5) Cultural barriers, including health beliefs and expectations of a medical interaction, may also impact access to health care and ART outcomes. Further, stigma may be a major obstacle to accessing care in close-knit ethnic communities. Findings from the United States indicated that delayed presentation for HIV-related medical care was associated with belonging to an ethnic minority and HIV transmission through heterosexual contact. (6,7) Foreign-born patients, consequently, may also experience poorer HIV treatment outcomes, including virologic failure and progression of HIV disease.

To understand how FB and CBA patients served by our clinic were doing on once-daily ART (the standard of HIV care), and to determine whether there were differences in ART outcomes, we compared ART outcomes between HIV-infected patients born in Canada and those born elsewhere using a retrospective cohort. In this paper, therefore, we report the probability of achieving virological success for FB, CBA and CBNA patients prescribed once-daily ART from 2006-2012. Among those patients who achieved virological suppression, we compared the rates of subsequent virological failure.

METHODS

Our study was a two-part retrospective cohort study in which we assessed: Part 1, the odds of achieving initial virological suppression by FB status compared with CBA and with CBNA patients; and Part 2, the rate of ART failure by FB status among patients who achieved initial virological suppression in Part 1. This study was approved by the University of Alberta Health Ethics Research Board.

Study setting and data source

We conducted our research with the clinicians at the Northern Alberta HIV Program (NAP). The NAP, a HIV clinic based in Edmonton, cares for all HIV-infected individuals in the northern half of Alberta (from Red Deer to the Northwest Territories border), comprising approximately 2,000 patients. The population of HIV-infected individuals in northern Alberta is heterogeneous with approximately one third of patients self-identifying as Metis or First Nations and 15-20% of patients having been born outside of Canada.

The NAP team includes infectious disease physicians, nurses, pharmacists, social workers, psychologists and dietitians. Initial genotypic viral resistance testing is performed on each new patient. Patients are usually seen in follow-up approximately four weeks after starting ART, then every three to four months. Antiretroviral therapy in northern Alberta is dispensed from two outpatient pharmacies, and is provided free of charge for all patients with provincial health care coverage; this program also provides CD4 cell counts, HIV-1 viral load (VL) measurements, and routine genotypic sensitivity testing, and any resistance identified is taken into account in the selection of the treatment regimen. Additionally, Canada has a national program which covers ART for refugees who qualify.

Our study's primary data source was the NAP clinic database. This database contains patient-related information, including patient demographics, risk behaviours, ART prescribed, CD4 cell counts, and VL measurements and is updated following each patient's clinic visit.

Study patients

We assembled a cohort of patients from the NAP database using the following inclusion criteria: 1) started ART between 1 January 2006 and 1 January 2012 (baseline); 2) previously ART-naive; 3) receiving once-daily ART; and 4) [greater than or equal to] 18 years of age when starting ART. We defined ART as a combination of at least three antiretrovirals, other than ritonavir, recorded as prescribed on the same date. We excluded ritonavir assuming that during the study period (2006-2012), physicians prescribed ritonavir at low dosages intended to "boost" other protease inhibitors (PI), rather than at clinically therapeutic levels. The ART start date was the first date that an ART prescription is recorded in the NAP database, and we assumed, unless otherwise stated, that patients remained on ART. We specified January 2006 as the start of follow-up as the standard of HIV care had become once-daily ART in the vast majority of cases by 2006. (8) As a result, our analysis focused on individuals prescribed once-daily ART.

[FIGURE 1 OMITTED]

We excluded patients if they: 1) were missing country of birth data; 2) were missing baseline VL data; 3) had a baseline VL <400 copies/mL; 4) had <6 months follow-up time; or 5) started ART [less than or equal to] 26 weeks before delivering a baby. We excluded the latter group of patients in order to limit the study group to patients who started ART for the purpose of treatment rather than to prevent vertical transmission of HIV. We excluded patients with baseline VLs that were missing or <400 copies/mL under the assumption that these patients were not ART-naive when initiating ART (i.e., these patients may have been transferred from elsewhere on treatment).

Statistical analysis

The exposure variable of interest was FB status; we defined FB patients as those individuals who reported being born outside of Canada. We followed up patients from 1 January 2006 until 1 July 2012, allowing for a follow-up time of six months to six years.

Initially for Part 1 and Part 2, we tabulated and compared patient characteristics between FB, CBA and CBNA patients with other exposures using [chi square] tests for categorical variables and Kruskal-Wallis rank tests for continuous variables.

For Part 1, we utilized logistic regression to assess the odds of achieving initial virological suppression with respect to FB status, adjusting for potential confounding variables. We defined viral suppression as one VL test <400 copies/mL [less than or equal to] 6 months after starting ART. We defined patients without VL tests [less than or equal to] 6 months after starting ART as not achieving initial virological suppression.

For Part 2, we used Cox proportional hazard models to determine the rate of virological failure among those patients who achieved initial virological success in Part 1. We defined virological failure as the first of two consecutive VL tests >400 copies/mL. In unadjusted analysis for Part 2, we utilized cumulative incidence curves, as described by Vittinghoff et al., (9) to compare rates of virological failure by FB status. We compared hazard ratios of virological failure using Cox proportional hazards models, (9) adjusting for potential confounding variables. We assessed the proportional hazards assumption for FB status using a time-varying covariate (FB status by the logarithm of observation time), which we entered into an unadjusted model of virological failure.

We measured observation time for Part 2 in person-years, (10) starting on the date of initial virological suppression and, depending on the patient's outcome, ending on the earliest of the following events: 1) the date of the first (of 2 consecutive) viral loads [greater than or equal to] 400 for patients who experienced virological failure; 2) 1 January 2012 for patients who were censored; or 3) the date of death for those who died.

We considered many potential confounding variables: ART regimen (PI- or non-PI-based), baseline age, baseline CD4 cell count, baseline VL, calendar year, IDU as an HIV exposure category, and sex. We defined patients as IDU if NAP staff recorded their HIV exposure as IDU or any exposure combined with IDU; we considered patients with other exposures, including unknown/missing exposures, as "other exposures". We defined baseline CD4 cell count and VL as the closest tests taken to the ART start date. In our unadjusted analysis, we identified all potential confounding variables as associations with initial virological suppression and subsequent virological failure with two-tailed p-values of <0.20. In the final statistical models, p-values were two-tailed and associations <0.05 statistically significant. We conducted analyses with Stata (version 11.0; StataCorp LP, College Station, TX).

RESULTS

During the study period, 550 persons were prescribed ART and 322 patients met the eligibility criteria for Part 1 (initial virological suppression). We excluded 228 patients, the majority of whom were missing a baseline VL measurement (53%) or had a baseline VL <400 copies/mL (22%) (Figure 1). Table 1 shows the patient characteristics of the 322 patients in Part 1.

Part 1--Initial virological suppression

In Part 1, 261 (81%) achieved initial virological suppression within six months of initiating ART (Table 1). Of the 61 patients who did not achieve initial virological suppression, 23 were FB and 24 were CBA patients. Four (6.6%) of these 61 patients died [less than or equal to]6 months after starting ART, all of whom were CBA patients. In the unadjusted analysis, there was no significant difference in terms of likelihood of achieving initial virologic suppression between FB and CBNA patients (Table 1, p=0.31). Canadian-born Aboriginal patients, however, were significantly less likely than CBNA patients to experience initial virological suppression (Table 1, p=0.01). After controlling for the effects of age, ART regimen, IDU as an exposure to HIV, and calendar year, compared to CBNA patients, there was no significant difference for the odds of achieving initial virological suppression for FB patients;, however, the odds were significantly lower for CBA patients (Table 2).

Part 2--HIV treatment failure

The 261 patients who achieved virological suppression were eligible for Part 2.

We followed these patients retrospectively for a total of 635.1 person-years; 21 (8.1%) experienced virological failure, 234 (90%) were censored, and 6 patients died, indicating that no patients left care for reasons other than death. Of the 6 patients who died, 4 (1.5%) died before experiencing any other event and 2 (9.5%) patients who experienced virological failure died. The event rate for experiencing virological failure was 3.3% per year. There was no significant difference between FB and CBNA patients in the rate of virological failure (8% vs. 4%, p=0.68), however CBA patients were significantly more likely than CBNA patients to experience virological failure (15% vs. 4%, p=0.01). Compared to CBNA patients, CBA patients experienced a significantly higher cumulative incidence of virological failure (Class p=0.01) (Figure 2).

In Cox proportional hazards models, there was no statistically significant difference in unadjusted virological failure rates for FB patients compared to CBNA patients (Table 3). Conversely, CBA patients experienced significantly higher unadjusted virological failure rates compared to CBNA patients (Table 3). Adjusting for age, sex, baseline CD4 cell count, and ART regimen, neither FB nor CBA patients had a statistically significant different rate of virological failure compared to CBNA patients after achieving initial virological suppression Table 3.

DISCUSSION

We observed that among HIV-infected patients starting ART, FB and CBNA patients had similar ART outcomes. Conversely, CBA patients were less likely to achieve initial virological suppression compared to CBNA patients. Finally, among patients who achieved initial virological suppression, rates of virological failure did not differ for FB or CBA patients.

Our findings are in accordance with a Canadian study evaluating the impact of a recent policy change to Canada's Immigrant Act. (11) This study compared initial virological suppression among Canadian-born, Sub-Saharan African (SSA)born, and FB other than SSA patients. The authors defined virological suppression as achieving a VL <40 copies/mL for two consecutive measurements. Although SSAs initially presented into care with a lower CD4 cell count than Canadian-born patients, these authors observed that the responses to ART were similar in SSA and Canadian patients. (11) Two studies in Europe compared virological failure rates by FB status and observed results comparable to our study. (12,13) After adjusting for confounding variables, both studies reported no differences in the time to virological failure among FB and non-FB patients. Similar to our study's analysis, these researchers considered age, sex, ART regimen, baseline CD4 cell count, baseline VL, and risk behaviour for HIV infection as confounding variables. However, residual confounding may be of concern; additional factors that may have an effect on ART outcomes among FB patients include differences in social, cultural and economic conditions. (14)

In fact, Gardezi et al. (15) conducted a qualitative study to understand HIV-related treatment concerns. Among HIV-positive Caribbean and East African individuals living in Toronto, Canada, these authors suggested that: support services need to be culturally specific; there is a need for community development and increased community awareness; there should be expanded effort geared towards settlement issues; and HIV providers need to have increased sensitivity and knowledge. Recently, the Canadian Observational Cohort (CANOC) collaboration suggested that female and non-White HIV patients reported higher HIV-related stigma. (16) CANOC concluded that future research should understand "contextual factors, such as culture, country, and rural/urban differences" (p. e48168) when addressing HIV-related issues.

[FIGURE 2 OMITTED]

Individual adherence behaviour occurs within an economic, social and cultural context. Accordingly, research needs to consider factors affecting HIV treatment response in a broader social, economic and political context. For example, limited access to HIV information and HIV-related stigma among Aboriginal patients delay access to health care and thus compromise effective medical care. (17) Our study results indicated that CBA patients had poorer HIV treatment outcomes compared to CBNA patients. As a result, treatment outcomes among CBA patients in northern Alberta may be related to underlying living conditions, collectively known as the social determinants of health, including; education, employment and working conditions; unemployment and job security; food insecurity; housing; income and income distribution; and social exclusion. (18)

HIV care, however, involves several stages, including identifying infected individuals, linkage to initial HIV care, long-term retention in care, and ART adherence--the "cascade of care" (COC). (19) Appropriately, the COC has become of concern for areas in public health concerned with maximizing the benefits of ART. For patients and populations to benefit from ART (reduction in HIV incidence and transmission), ART programs have to ensure the complete quality of a cascade of services, from testing and referral to care, to ensuring continued adherence to treatment. (20) In many settings, gaps remain in the COC, where few HIV-infected patients actually achieve undetectable VLs, the ultimate goal of HIV treatment and care. (19) These gaps include late HIV diagnosis, suboptimal linkage to care, low retention in care, low ART coverage, and poor ART adherence. (19) Recently, Nosyk et al. (21) from the STOP HIV/AIDS Study Group were the first group to track the longitudinal changes in the COC in British Columbia, Canada, from 1996 to 2011. This group assessed the proportions of individuals at each stage along the COC. Although their findings suggested that overall engagement in care improved from 1996 to 2011, substantial numbers of patients were lost at each stage, most notably the retention in care stage.

Understanding limiting factors in each step of the cascade is important, while recognizing that what causes patients to be lost from the cascade likely varies within regions and patients. This is particularly important for CBA patients, as CBA patients continue to be disproportionately affected by HIV/AIDS and continue to account for an increasing proportion of new HIV-positive test reports and AIDS diagnoses. (1) Moreover, our results were similar to previous research which suggests that CBA patients continue to experience worse ART outcomes compared to CBNA patients. (3,4,22) This is of extreme concern. Perhaps we need to approach caring for CBA HIV patients with a participatory approach, meaning we need to improve care by forming meaningful relationships and partnerships with the patients themselves, the clinicians, and the HIV community service workers, to ensure that patients can successfully navigate their way within the health care system.

This study has numerous limitations, including missing data (i.e., the 26 patients excluded for missing country of origin data), combining immigrants and refugees into a single category to describe FB patients, small sample size, the inherent difficulties of using a clinical database for research purposes, and the retrospective nature of our design. We acknowledge that while we classified immigrant and refugee patients into one category, this group is heterogeneous and encompasses a broad range of characteristics. The use of any classification method, however, is a simplification of reality. We classified patients by country of origin, which has the limitation of ignoring cultural differences within a country and the length of time a patient resides in their birth country or Canada. For example, if a patient self-reported their country of origin as being outside of Canada but having migrated to Canada as an infant, we classified this patient as FB even though this patient may culturally identify as Canadian. Our findings, therefore, may have overlooked important differences between ethnic groups. However, our clinic database defined immigrant and refugees as those patients born outside of Canada, and therefore we did not have access to specific country of origin information.

Our studied utilized a retrospective design and relied on a clinical data source for exposure information. Further, our strict exclusion criteria reduced our sample size. This study, however, aimed to investigate ART outcomes among patients prescribed once-daily therapy, the standard of modern HIV care. Despite these limitations, our cohort provided a representative sample of all patients who recently initiated ART in northern Alberta using the NAP database. Previous research has validated this database, (22-24) including Martin et al.'s (22) recent investigation of HIV treatment outcomes comparing CBA and CBNA patients.

Although FB patients may have potential obstacles to adherence and ART success, our analysis found that ART outcomes among FB patients were similar to those of HIV patients born in Canada. Our findings, however, indicated that CBA patients continue to have less successful ART outcomes compared to CBNA patients. As such, it is critical that researchers, clinicians and community organizations engage with Canadian Aboriginal communities and work together to find innovative and culturally-appropriate strategies to address all stages along the COC, including ART outcomes.

Acknowledgements: The authors wish to recognize the participants involved in this study. We thank the clinicians at the NAP for their help and support with this research, and Yutaka Yasui for his statistical support.

Conflict of Interest: None to declare.

REFERENCES

(1.) Public Health Agency of Canada. HIV and AIDS in Canada: Surveillance report until December 31, 2012. Ottawa, ON: The Agency, 2013.

(2.) Alberta Government. HIV and AIDS in Alberta: 2011 Annual Report. Edmonton, AB: Government of Alberta, 2012.

(3.) Martin L, Houston S, Yasui Y, Wild TC, Saunders LD. Rates of initial virological suppression and subsequent virological failure after initiating highly active antiretroviral therapy: The impact of Aboriginal ethnicity and injection drug use. Current HIV Research 2010;8:649-58.

(4.) Lima VD, Kretz P, Palepu A, Bonner S, Kerr T, Moore D, et al. Aboriginal status is a prognostic factor for mortality among antiretroviral naive HIV-positive individuals first initiating HAART. AIDS Research and Therapy 2006;3:14.

(5.) Canadian Task Force on Mental Health Issues Affecting Immigrants and Refugees. After the door has been opened: Mental health issues affecting immigrants and refugees in Canada. Ottawa: Ministry of Supply and Services Canada, 1998.

(6.) Zingmond DS, Wenger NS, Crystal S, Joyce GF, Liu H, Sambamoorthi U, et al. Circumstances at HIV diagnosis and progression of disease in older HIV-infected Americans. Am J Public Health 2001;91:1117-120.

(7.) Samet JH, Freedberg KA, Savetsky JB, Sullivan LM, Stein MD. Understanding delay to medical care for HIV infection: The long-term non-presenter. AIDS 2001;15:77-85.

(8.) United States Food and Drug Administration. News Release. Available at: http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/2006/uc m108689.htm (Accessed June 15, 2014).

(9.) Vittinghoff E. Regression Methods in Biostatistics: Linear, Logistic, Survival, and Repeated Measures Models. New York, NY: Springer, 2005.

(10.) Rothman KJ, Greenland S, Lash TJ. Modern Epidemiology (3rd edition). Philadelphia, PA: Lippincott Williams & Wilkins, 2008.

(11.) Krentz H, Gill MJ. The five-year impact of an evolving global epidemic, changing migration patterns, and policy changes in a regional Canadian HIV population. Health Policy 2009;90:296-302.

(12.) Perez-Molina JA, Rillo MM, Suarez-Lozano I, Casado Osorio JL, Teira Cobo R, Gonzalez PR, et al. Do HIV-infected immigrants initiating HAART have poorer treatment-related outcomes than autochthonous patients in Spain? Results of the GESIDA 5808 study. Current HIV Research 2010;8:521-30.

(13.) van den Berg JB, Hak E, Vervoot SCJM, Hoepelman IM, Boucher CAB, Schuurman R, et al. Increased risk of early virological failure in non-European HIV-1-infected patients in a Dutch cohort on highly active antiretroviral therapy. HIV Medicine 2005;6:299-306.

(14.) Rhodes T, Simic M. Transition and the HIV risk environment. BMJ2005;331:220-23.

(15.) Gardezi F, Calzavara L, Husbands W, Tharao W, Lawson E, Myers T, et al. Experiences of and responses to HIV among African and Caribbean communities in Toronto, Canada. AIDS Care 2008;20:718-25.

(16.) Cescon A, Patterson S, Chan K, Palmer AK, Margolese S, Burchell AN, et al. Gender differences in clinical outcomes among HIV-positive individuals on antiretroviral therapy in Canada: A multisite cohort study. PLoS One 2013;8:e83649.

(17.) Pottie K, Greenaway C, Feightner J, Welch V, Swinkels H, Rashid M, et al. Evidence-based clinical guidelines for immigrants and refugees. CMAJ2011;183:E824-E925.

(18.) Mikkonen J, Raphael D. Social Determinants of Health: The Canadian Facts. Toronto: York University School of Health Policy and Management, 2010. Available at: http://www.thecanadianfacts.org (Accessed June 15, 2014).

(19.) Gardner EM, McLees MP, Steiner JF, Del Rio C, Burman WJ. The spectrum of engagement in HIV care and its relevance to test-and-treat strategies for prevention of HIV infection. Clin Infect Dis 2011;52:793-800.

(20.) Burns DN, Dieffenbach CW, Vermund SH. Rethinking prevention of HIV type-1 infection. Clin Infect Dis 2010;51:725-31.

(21.) Nosyk B, Montaner JSG, Colley G, Lima VD, Chan K, Heath K, et al., for the STOP HIV/AIDS Study Group. The cascade of HIV care in British Columbia, Canada, 1996-2011: A population-based retrospective cohort study. Lancet Infect Dis 2013; published online Sept 27. http://dx.doi.org/10.1016/S1473-3099(13)70254-8.

(22.) Martin L, Houston S, Yasui Y, Wild TC, Saunders LD. Poorer physical health-related quality of life among Aboriginals and injection drug users treated with highly active antiretroviral therapy. Can J Public Health 2012;104(1):e33-e38.

(23.) Bowker SL, Soskolne CL, Houston SC, Newman SC, Jhangri GS. Human immunodeficiency virus (HIV) and hepatitis C virus (HCV) in a Northern Alberta population. Can J Public Health 2004;95:188-92.

(24.) Henson LD, Hughes CA, Schapansky LM. The relationship between pharmacy refill record adherence and viral load in HIV patients [Oral Presentation]. 9th Annual Canadian Conference on HIV/AIDS Research. Montreal, QC, April 27-30, 2000.

Received: September 17, 2013

Accepted: June 22, 2014

Megan E. Lefebvre, MSc, [1] Christine A. Hughes, BScPharm, PharmD, [2] Yutaka Yasui, PhD, [1] L. Duncan Saunders, MBBCh, PhD, [1] Stan Houston, MD, DTM&H, FRCPC [3]

[1.] School of Public Health, University of Alberta, Edmonton, AB

[2.] Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, AB

[3.] Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB

Correspondence: Megan Lefebvre, School of Public Health, University of Alberta, 3-300 Edmonton Clinic Health Academy, 11405-87 Ave, Edmonton, AB T6G 1C9, Tel: 780-492-9954, E-mail: mejohnst@ualberta.ca

Table 1. Patient characteristics by country of origin for
Part 1: Initial virological suppression (n=322; 81%
achieved initial viral suppression)

Characteristic               Foreign-born     Canadian-born
                               (n=128)         Aboriginal
                                                 (n=87)

Proportion achieving       105 (82)            63 (72)
  initial VS, n (%)
Age (years) at baseline,    38 (34.5-44.5)     42 (34-47)
  median (IQR)
Age at baseline, n (%)
  17-30                     18 (14)            14 (16)
  31-40                     58 (45)            25 (29)
  41-50                     36 (28)            33 (38)
  51-77                     16 (13)            15 (17)
Sex, n (%)
  Female                    68 (53)            41 (47)
  Male                      60 (47)            46 (53)
HIV exposure category,
  n (%)
  Injection drug use         3 (2)             38 (44)
  Other exposure           125 (98)            49 (56)
CD4 cells/[micro]L at      240 (150-340)      225 (90-360)
  baseline, median (IQR)   (n=122)            (n=86)
CD4 cells/[micro]L at
  baseline, n (%)
  5-50                      13 (10)            13 (15)
  51-200                    35 (27)            27 (31)
  201-350                   46 (36)            24 (28)
  351-1130                  28 (22)            22 (25)
  Missing                    6 (5)              1 (1)
HIV RNA copies/mL at       27,000             32,000
 baseline, median (IQR)    (4836.5-135,000)   (8600-98,000)
HIV RNA copies/mL at
  baseline, n (%)
  0-9999                    40 (31)            26 (30)
  10,000-44,999             33 (26)            23 (26)
  45,000-139,999            23 (18)            20 (23)
  140,000-750,000           32 (25)            18 (21)
Initial HaART regimen,
  n (%)
  PI-based                  68 (53)            46 (53)
  Not PI-based              60 (47)            40 (47)
Year starting HAART,
  n (%)
  2006-2008                 68 (54)            30 (34)
  2009-2012                 59 (46)            57 (66)

Characteristic              Canadian-born      Class
                            non-Aboriginal    p-value
                               (n=107)

Proportion achieving        93 (87)            0.04
  initial VS, n (%)
Age (years) at baseline,    45 (38-51)        <0.001
  median (IQR)
Age at baseline, n (%)                         0.002
  17-30                     12 (11)
  31-40                     23 (22)
  41-50                     43 (40)
  51-77                     29 (27)
Sex, n (%)                                    >0.001
  Female                    16 (15)
  Male                      91 (85)
HIV exposure category,                        >0.001
  n (%)
  Injection drug use        28 (26)
  Other exposure            79 (74)
CD4 cells/[micro]L at      211 (90-360)        0.53
  baseline, median (IQR)   (n=103)
CD4 cells/[micro]L at                          0.42
  baseline, n (%)
  5-50                      20 (19)
  51-200                    30 (28)
  201-350                   27 (25)
  351-1130                  26 (24)
  Missing                    4 (4)
HIV RNA copies/mL at       73,000              0.38
 baseline, median (IQR)    (21,000-190,000)
HIV RNA copies/mL at                           0.09
  baseline, n (%)
  0-9999                    18 (17)
  10,000-44,999             25 (23)
  45,000-139,999            33 (31)
  140,000-750,000           31 (29)
Initial HaART regimen,                         0.03
  n (%)
  PI-based                  39 (36)
  Not PI-based              68 (64)
Year starting HAART,                           0.01
  n (%)
  2006-2008                 42 (40)
  2009-2012                 64 (60)

VS = viral suppression; IQR = interquartile range; PI =
protease inhibitor; HAART = highly active antiretroviral
therapy.

Table 2. Unadjusted and adjusted logistic regression models assessing
initial virological suppression after initiating ART (n=322)

Variable                              Unadjusted analysis

                             Unadjusted      95% CI      p-value
                             Odds Ratio

Country of origin                         Class p=0.04

  Canadian-born non-            1.00           --          --
    Aboriginal (ref)
  Foreign-born                  0.69       0.55-1.41      0.51
  Canadian-born Aboriginal      0.40       0.19-0.82      0.01
HIV exposure category           0.58       0.51-1.09      0.09
  (Injection  drug use
  vs. other exposures)
Sex (Female vs. male)           1.16       0.65-2.06      0.62

CD4 cells/[micro]L at baseline            Class p=0.74

  5-50 (ref)                    1.00           --          --
  51-200                        0.07       0.45-2.55      0.88
  201-550                       1.65       0.67-4.05      0.28
  351-1130                      1.04       0.45-2.54      0.95
Missing baseline CD4 count      1.25       0.21-6.74      0.80

HIV RNA copies/mL at baseline             Class p=0.76

  0-9999                        1.51       0.61-2.85      0.48
  10,000-44,999                 1.16       0.54-2.48      0.70
  45,000-159,999                1.52       0.68-5.42      0.51
  140,000-750,000 (ref)         1.00           --          --

Baseline age, years                       Class p=0.07

  17-50 (ref)                   1.00           --          --
  51-40                         2.57       1.07-5.25      0.05
  41-50                         2.70       1.21-6.02      0.02
  51-77                         2.95       1.14-7.55      0.05
Baseline ART regimen            0.64       0.57-1.11      0.12
  (PI vs. non-PI based)
Baseline calendar year                     1.05-5.25      0.04
  (2009-2012 vs.
  2006-2008) 1.82

Variable                               Adjusted analysis

                               Adjusted      95% CI     p-value
                             Odds Ratio *

Country of origin

  Canadian-born non-             1.00          --         --
    Aboriginal (ref)
  Foreign-born                   0.75       0.55-1.75    0.51
  Canadian-born Aboriginal       0.44       0.20-0.96    0.04
HIV exposure category            0.58       0.27-1.25    0.15
  (Injection  drug use
  vs. other exposures)
Sex (Female vs. male)             --           --         --

CD4 cells/[micro]L at baseline

  5-50 (ref)                      --           --         --
  51-200                          --           --         --
  201-550                         --           --         --
  351-1130                        --           --         --
Missing baseline CD4 count        --           --         --

HIV RNA copies/mL at baseline

  0-9999                          --           --         --
  10,000-44,999                   --           --         --
  45,000-159,999                  --           --         --
  140,000-750,000 (ref)           --           --         --

Baseline age, years

  17-50 (ref)                    1.00          --         --
  51-40                          2.11       0.89-5.00    0.09
  41-50                          2.51       1.05-6.07    0.04
  51-77                          2.61       0.94-7.27    0.07
Baseline ART regimen             0.70       0.58-1.27    0.24
  (PI vs. non-PI based)
Baseline calendar year           2.52       1.25-4.50    0.01
  (2009-2012 vs.
  2006-2008) 1.82

Ref=referent group, PI=protease inhibitor. * Variables with
p-values [less than or equal to] 0.20 were included in the
multivariable model.

Table 3. Unadjusted and adjusted Cox Proportional Hazard models
assessing virological failure after initial virological suppression
(n=261)

Variable                                Unadjusted analysis

                              Unadjusted       95% CI       p-value
                             Hazard Ratio

Country of origin                           Class p=0.03

  Canadian-born non-             1.00            --           --
    Aboriginal (ref)
  Foreign-born                   1.29         0.39-4.30      0.68
  Canadian-born Aboriginal       4.16        1.27-13.58      0.02
Sex (Female vs. male)            4.97        1.82-13.57      0.002
HIV exposure category            1.34         0.45-4.02      0.60
  (Injection drug use vs.
  other exposures)

CD4 cells/[micro]L at baseline              Class p=0.13

  5-50 (ref)                     1.00            --           --
  51-200                         0.67         0.11-4.04      0.67
  201-350                        1.71         0.36-8.06      0.50
  351-1130                       2.89        0.61-13.67      0.18

HIV RNA copies/[micro]L at baseline         Class p=0.75

  0-9999                         1.51         0.50-4.54      0.46
  10,000-44,999                  0.87         0.24-3.10      0.83
  45,000-139,999                 0.84         0.24-3.00      0.70
  140,000-750,000 (ref)          1.00            --           --

Baseline age (years)                        Class p<0.001

  17-30 (ref)                    1.00            --           --
  31-40                          0.24         0.09-0.68      0.01
  41-50                          0.11         0.03-0.35     <0.001
  51-77                          0.04         0.01-0.32      0.003
Baseline ART regimen             4.07        1.49-11.13      0.006
  (PI vs. non-PI based)
Baseline calendar year           1.25         0.77-2.06      0.39
  (2009-2012 vs.
  2006-2008)

Variable                                Adjusted analysis

                                Adjusted        95% CI     p-value
                             Hazard Ratio *

Country of origin

  Canadian-born non-              1.00            --         --
    Aboriginal (ref)
  Foreign-born                    0.49        0.11-2.20     0.35
  Canadian-born Aboriginal        1.54        0.38-6.18     0.54
Sex (Female vs. male)             2.96        0.77-11.40    0.12
HIV exposure category              --             --         --
  (Injection drug use vs.
  other exposures)

CD4 cells/[micro]L at baseline

  5-50 (ref)                      1.00            --         --
  51-200                          0.39        0.06-2.61     0.33
  201-350                         0.69        0.11-4.23     0.69
  351-1130                        0.49        0.07-3.38     0.47

HIV RNA copies/[micro]L at baseline

  0-9999                           --             --         --
  10,000-44,999                    --             --         --
  45,000-139,999                   --             --         --
  140,000-750,000 (ref)            --             --         --

Baseline age (years)

  17-30 (ref)                     1.00            --         --
  31-40                           0.32        0.09-1.10     0.07
  41-50                           0.24        0.06-0.90     0.04
  51-77                           0.10        0.01-0.96     0.05
Baseline ART regimen              3.12        0.97-10.10    0.06
  (PI vs. non-PI based)
Baseline calendar year             --             --         --
  (2009-2012 vs.
  2006-2008)

Ref=referent group, PI=protease inhibitor.
* Variables with p-values [less than or equal to]
0.20 were included in the multivariable model.