Understanding the Correlates of Attrition Associated with Antiretroviral Use and Viral Suppression Among Women Living with HIV in Canada

Abstract

Attrition along the cascade of HIV care compromises attainment of the UNAIDS 90-90-90 goals and achievement of desirable treatment outcomes for people living with HIV. Given known gender disparities in HIV care and outcomes, understanding the correlates of attrition at stages of the care cascade for women living with HIV (WLWH) is essential. Among the 1425 WLWH enrolled in the Canadian HIV Women's Sexual and Reproductive Health Cohort Study (CHIWOS), we measured the proportion who reported not being currently on combination antiretroviral therapy (cART) and the proportion who reported a detectable viral load (VL; ≥40 copies/mL) despite cART use. Correlates of these cascade indicators were examined using univariate and multivariable logistic regression. Overall, 14.8% of women were not currently on cART. Of women who were on cART, 9.0% were not virally suppressed. In multivariable analyses, age between 26 and 34, unstable housing, food insecurity, current injection drug use, higher HIV-related stigma, and racial discrimination were associated with increased odds of not being on cART. Factors associated with increased odds of reporting a detectable VL among women on cART included age ≤34 years, less than a secondary education, unstable housing, and incarceration in the previous year. Programmatic efforts to support cART use and viral suppression for WLWH in Canada should focus on social determinants of health, including housing and food insecurity, social exclusion, and education.

Introduction

For the benefits of combination antiretroviral therapy (cART) to be realized, people living with HIV (PLWH) must be fully engaged along the cascade of care—a continuum of care from the time of diagnosis to the provision of cART to ensure viral suppression.^1,2 The use of cART, with sustained viral load (VL) suppression, is associated with an exceedingly low risk of transmission;³ the concept of “Treatment as Prevention (TasP)” has been widely supported by several studies and mathematical modeling^4

–8 and is becoming more commonplace with the rise in pre- and postexposure prophylaxis (PrEP and PEP).⁹ Incomplete engagement at any stage of the cascade will compromise the full benefits of cART at both the individual and societal level, and will result in increased risk of HIV-related morbidity, mortality, and transmission.¹⁰ Thus, disparities in cART prescription, uptake and adherence, and subsequent VL suppression are of paramount importance and have been prioritized by UNAIDS in their 90-90-90 objectives.¹¹

Women comprise 50% of all PLWH worldwide, and a growing body of evidence reveals marked gender disparities along the cascade of care.^{12

–20} Recent data from the United States show that while a larger percentage of women living with HIV (WLWH) are retained in care compared with men, a smaller percentage are on cART and are virally suppressed.^13,14 Similar findings have emerged from Canada that demonstrate women are less likely than men to be engaged along all aspects of the cascade of care.^15

–19 While it appears as though overall engagement in the cascade is improving with time for both men and women, attrition at each stage of the care cascade continues.^21,22

Few studies have focused on barriers to cART use among women specifically. There are ethnoracial disparities in cART use; in the United States, women of Hispanic or black ethnicities are significantly less likely to be using cART than white/Caucasian women.^{12,13,23
–25} WLWH also experience more psychosocial barriers to cART use, including stigma, depression, substance, and alcohol use.^23,26,27 A large body of research has confirmed that the absence of health insurance compromises access to HIV care in the United States, including access to cART.^28
–30 Prior studies have shown similar barriers to VL suppression for WLWH.^12,13 Barriers have been reported along social inequities,¹² among those of younger age,^12,13 with lower levels of education,²¹ and among those lacking financial resources.¹² Psychosocial barriers precluding VL suppression include depression, substance use, and incarceration.¹²

In Canada, women's lack of access to cART has primarily been examined among women who experience marginalization, including women who inject drugs and women involved in sex work.^15,31 There are limited studies looking at the broader sociodemographic profile and barriers of cART access and VL suppression among WLWH in a setting with universal access to healthcare. The purpose of this study was to determine the proportion of Canadian WLWH who are currently not on cART, and who report a detectable VL despite cART use among a large sample of women participating in a national study. Secondly, we aimed to identify factors associated with not currently being on cART and factors associated with having a detectable VL among Canadian women who were taking cART.

Methods

Study population and design

This study utilized baseline data collected between August 2013 and May 2015 from 1425 WLWH participating in the Canadian HIV Women's Sexual and Reproductive Health Cohort Study (CHIWOS), a community-based, participatory, and prospective cohort study of WLWH from Ontario, Quebec, and British Columbia.³² The overall inclusion criteria were as follows: (1) self-identifying as a woman (trans-inclusive), (2) living with HIV, (3) being 16 years of age or older, and (4) living in Ontario, British Columbia or Quebec. For this analysis, we further restricted the inclusion criteria to women who answered the question on cART use (current or prior) (n = 1382).

Data collection

WLWH, who were trained as peer research associates (PRAs), recruited self-identified WLWH aged 16 years or older from one of the three study provinces. Purposive, nonrandom, and venue-based sampling methods, including word-of-mouth and online (e.g., Listservs for WLWH) methods, were used to recruit participants. Participants were recruited from AIDS service organizations (ASOs), HIV clinics, community-based organizations (CBOs), and ethnoracial-specific ASOs and CBOs. PRAs then administered a 90-min structured, online questionnaire either in-person (e.g., home, HIV clinic) or by phone/Skype for rural or remotely located participants. A $50 honorarium was given for participation.

Ethics approval was obtained from the respective Research Ethics Boards (REBs) of Women's College Hospital, Simon Fraser University and the University of British Columbia/Providence Health, and McGill University Health Centre. Study sites with independent REBs obtained their own approval before commencing enrolment. All participants gave informed consent before commencing the survey.

Outcome variables

Participants who reported ever accessing HIV medical care were asked about their history of cART use and categorized as never having used cART, used cART previously but not currently, and currently using cART. For the purpose of this analysis, those reporting previously but not currently using cART (n = 67) and never using cART (n = 137) were categorized as not currently using cART (n = 204) and compared with those who reported currently being on cART (n = 1178). Participants who reported ever accessing HIV medical care were asked about their most recent VL, categorized as undetectable (<40 copies/mL), detectable (≥40 copies/mL), never received, do not know, or prefer not to answer (DK/PNTA). Those who reported undetectable or detectable VLs were included in the analysis and compared (n = 1130). All DK/PNTA or missing responses were excluded.

Correlates of cART use and viral suppression

Sociodemographic factors of interest included age, sexual orientation, immigration status, ethnicity, education, personal gross yearly income, main source of income (social assistance, jobs considered illegal [sex work, selling drugs, panhandling], savings/loan/family, and honoraria/other vs. paid job), and number of financial dependents. We assessed several other factors known to correlate with different points along the care cascade, including categorical variables, such as housing status (own, rent vs. unstable [self-contained room, housing facility, HIV care group home, transition/halfway/safe/outdoor housing or other]), food security (insecure vs. secure), incarceration (current vs. previous and never), history of injection drug use (IDU; currently vs. not currently but previously and never), method of HIV acquisition (nonconsensual sex, sharing needles/contaminated needles, blood transfusion/other medical procedures, perinatal exposure, or other vs. consensual sex) and duration of time living with HIV (6–14 years or >14 years vs. <6 years).

We also explored the association between each outcome and various continuous scaled measures such as depression, social support, HIV-related stigma, racial discrimination, and gender discrimination. We determined score ranges and Cronbach's alpha for the analytic sample (n = 1382) who responded regarding cART use. Depression was measured using the 10-item Center for Epidemiologic Studies Depression (CES-D) (score range: 0–30; Cronbach's alpha = 0.88).³³ Social support was measured using the 4-item version of the Medical Outcomes Survey Social Support Survey (MOS-SSS; score range: 4–20 with higher scores indicating higher social support; Cronbach's alpha = 0.84).³⁴ HIV-related stigma was measured using the total score of the HIV Stigma Scale^35
–37 (score range: 0–100 with higher scores indicating higher IV stigma; Cronbach's alpha = 0.84). Racial and gender discrimination were measured using two separate modified (8-item) versions of the 9-item Everyday Discrimination Scale (score range: 8–48 for both scales with higher scores indicating higher racial and gender discrimination; Cronbach's alpha = 0.96 for racism, 0.936 for sexism among cisgender women, and 0.949 for gender discrimination among transwomen and other gender-identified participants).^38,39

Statistical analyses

All outcomes and hypothesized factors associated with each outcome were described using medians and interquartile ranges (IQRs) for continuous variables and frequencies and proportions for categorical variables. In bivariable analyses, categorical variables were compared using the Pearson's chi-squared or Fisher's exact test, and continuous variables were compared using the Wilcoxon rank-sum test. Logistic regression was used to study the univariate and multivariate associations between the explanatory variables and each outcome (never or not currently using cART vs. currently using cART; not being virally suppressed vs. viral suppression). Factors significant at p < 0.05, as well as correlates chosen a priori, were considered for the multivariable model. Model selections were conducted using a backward stepwise technique based on the Type III p values and Akaike Information Criterion (AIC). The variable with the highest Type III p value was dropped at each step of the selection process until the model reached the lowest AIC (lower AIC indicates better model fit). All p values were two-sided and considered statistically significant at alpha = 0.05. Analyses were conducted using SAS version 9.4 (SAS, Cary, NC).⁴⁰

Results

Participant sociodemographic and clinical characteristics

Overall, 85.2% (n = 1178/1382) of women were on cART, while 14.8% (n = 204/1382) of women were not currently receiving cART. Of those not receiving cART, 67.2% (n = 137/204) had never received cART, and 42.2% (n = 86/204) reported a CD4 T+ cell count of >500 cells/mm³. Table 1 summarizes the sociodemographic and clinical characteristics of women with complete data regarding cART use. The median age of women on cART was 44 (IQR 37–51) years versus 36 (IQR 32–44) years for those not on cART (p < 0.001). The majority (96.3%) of women on cART self-identified as cisgender versus 93.1% for those not on cART (p = 0.04). Similarly, 88.4% of women on cART self-identified as heterosexual versus 80.9% for those not on cART (p = 0.002). Women on cART were predominantly of white (40.7%) or African, Caribbean or black (ACB) ethnicities (31.9%) versus women not on cART were primarily of white (46.6%) or indigenous (31.4%) ethnicities (p < 0.001). Among women on cART, 91.0% women reported an undetectable VL versus 39.9% for those not on cART (p < 0.001).

Table 1.

Participant Sociodemographic and Clinical Characteristics

	On cART		Not on cART^a
	N = 1178^b		N = 204^b
	N or median	% or IQR	N or median	% or IQR	p
Sociodemographic characteristics
Age	44	37–51	36	32–44	<0.001
Gender identity, cisgender	1134	96.3	190	93.1	0.040^c
Sexual orientation, heterosexual	1041	88.4	165	80.9	0.002
Ethnicity					<0.001
White/Caucasian	479	40.7	95	46.6
African/Caribbean/black	376	31.9	33	16.2
Indigenous	232	19.7	64	31.4
Other ethnicity	91	7.7	12	5.9
Education, secondary or less (n = 1175; n = 202)	690	58.7	119	58.9	0.960^d
Personal gross yearly income <$20,000 CAD (n = 1141; n = 198)	827	72.5	141	71.2	0.713^e
Housing status, rent vs. own (n = 988; n = 146)	849	85.9	130	89.0	0.307^f
Clinical characteristics
Duration of HIV diagnosis (n = 1144; n = 195)
<6 Years	227	19.8	87	44.6	<0.001
6–14 Years	480	42.0	70	35.9
>14 Years	437	38.2	38	19.5
Years on cART (n = 1056)	8.5	4.3–14.6	—	—	—
Years not on cART (n = 137)	—	—	4.5	3.1–9.5	—
Viral load, undetectable (n = 1130; n = 168)	1028		67	39.9	<0.001
CD4 count (n = 745; n = 80)	600	415–800	628	400–800	0.956

Includes those never and previously but not currently on ART.

N = 1178 and 204 unless otherwise stated.

Women vs. all others.

Secondary or less vs. postsecondary.

<$20,000 vs. ≥$20,000.

Own vs. rent.

CAD, Canadian dollars; cART, combination antiretroviral therapy; IQR, interquartile range.

Factors associated with not currently being on cART (n = 204) compared with those currently on cART (n = 1178)

Table 2 shows the unadjusted and adjusted odds ratio (aOR) of not currently being on cART compared with those currently on cART. In unadjusted analyses, being younger than or equal to 25 years of age or 26–34 years of age versus >35, of sexual minority status versus heterosexual, selling drugs, panhandling, or sex work as the main source of income versus a paid job, having less than two financial dependents versus two to three, unstable housing versus owning, food insecurity, current IDU, having greater social support and higher HIV-related stigma, and racial as well as gender discrimination, were all associated with increased odds of not being on cART. Persons of ACB ethnicity versus white/Caucasian, having less than a secondary education versus a postsecondary education, acquiring HIV through nonconsensual sex, sharing needles, or through blood transfusion/other medical procedures versus consensual sex, being diagnosed with HIV between 6 and 14 years ago or >14 years ago versus <6 years ago, and depression were associated with reduced odds of not being on cART. In multivariable logistic regression, age (being 26–34 years of age [aOR: 1.84, 95% confidence interval, CI: 1.15–2.95] versus >35), unstable housing (aOR: 2.51, 95% CI: 1.08–5.87) versus owning, food insecurity (aOR: 1.69, 95% CI: 1.01–2.83), current IDU (aOR: 2.56, 95% CI: 1.28–5.15), and racial discrimination (aOR: 1.03, 95% CI: 1.01–1.06) were associated with increased odds of not being on cART. ACB versus white/Caucasian ethnicity (aOR: 0.32, 95% CI: 0.17–0.59), having less than a secondary education versus postsecondary (aOR: 0.35, 95% CI: 0.17–0.75), having a personal income of less than $20,000 (aOR: 0.30; 95% CI: 0.15–0.59) or $20,000 to $40,000 per year (aOR: 0.31, 95% CI: 0.15–0.67) versus greater than $40,000 per year, acquiring HIV through nonconsensual sex (aOR: 0.40, 95% CI: 0.21–0.75) or sharing needles (aOR: 0.52, 95% CI: 0.29–0.93) versus consensual sex, living with HIV for 6–14 years (aOR: 0.40, 95% CI: 0.26–0.64) or >14 years (aOR: 0.28, 95% CI: 0.16–0.49) versus <6 years, and depression (aOR: 0.97, 95% CI: 0.94–0.99) were associated with decreased odds of not being on cART.

Table 2.

Factors Associated with Not Currently Being on Combination Antiretroviral Therapy (n = 204) Compared to Those Currently on Combination Antiretroviral Therapy (n = 1178), Among Women Living with HIV in Canada

Factors	Unadjusted OR (95% CI)	p	Adjusted OR ^a (95% CI)	p
Age		<0.001		0.025
≤25	2.01 (1.11–3.66)		2.02 (0.83–4.94)
26–34	3.24 (2.30–4.57)		1.84 (1.15–2.95)
≥35	Referent		Referent
Sexual orientation (sexual minority vs. heterosexual)	1.86 (1.26–2.76)	0.002	1.48 (0.88–2.49)	0.144
Ethnicity		<0.001		0.004
White/Caucasian	Referent		Referent
Indigenous	1.39 (0.98–1.98)		0.65 (0.37–1.12)
African/Caribbean/black	0.44 (0.29–0.67)		0.32 (0.17–0.59)
Other ethnicity	0.67 (0.35–1.26)		0.62 (0.28–1.35)
Education level		0.002		0.007
Less than secondary	0.48 (0.28–0.83)		0.35 (0.17–0.75)
Secondary	1.24 (0.90–1.70)		1.09 (0.69–1.72)
Postsecondary	Referent		Referent
Main source of income		0.004	Not selected
Paid job	Referent
Social assistance/pension/WCBI/EI	0.89 (0.62–1.27)
Sex work/selling drugs/panhandling	4.21 (1.80–9.82)
Savings/loan/family	0.63 (0.26–1.55)
Honoraria/other	NA
Personal gross yearly income (CAD)		0.259		0.001
≥$40,000 per year	Referent		Referent
$20,000–$40,000 per year	0.72 (0.46–1.14)		0.30 (0.15–0.59)
<$20,000/year	0.63 (0.36–1.12)		0.31 (0.15–0.67)
Financial dependents		0.004	Not selected
<2	2.00 (1.20–3.34)
2–3	Referent
>3	1.02 (0.48–2.16)
Housing status		<0.001		0.019
Own (house/apartment/condo)	Referent		Referent
Rent (house/apartment/condo)	1.33 (0.77–2.30)		1.33 (0.63–2.80)
Unstable^b	2.65 (1.46–4.81)		2.51 (1.08–5.87)
Food security (insecure vs. secure)	1.64 (1.18–2.29)	0.003	1.69 (1.01–2.83)	0.045
IDU (current vs. not current but previously and never)	1.83 (1.15–2.91)	0.011	2.56 (1.28–5.15)	0.008
Depression	0.97 (0.95–0.99)	0.011	0.97 (0.94–0.99)	0.015
Social support	1.06 (1.02–1.10)	0.002	Not selected
HIV stigma	1.01 (1.00–1.02)	0.004	1.01 (1.0–1.02)	0.050
Racial discrimination	1.02 (1.01–1.04)	0.001	1.03 (1.01–1.06)	0.004
Gender discrimination	1.02 (1.01–1.03)	0.009	Not selected
HIV acquisition method		<0.001		0.008
Consensual sex	Referent		Referent
Nonconsensual sex	0.39 (0.23–0.64)		0.40 (0.21–0.75)
Sharing needles/contaminated needles	0.51 (0.33–0.77)		0.52 (0.29–0.93)
Blood transfusion/other medical procedures	0.12 (0.03–0.50)		0.17 (0.02–1.26)
Perinatal exposure	0.28 (0.09–0.91)		0.20 (0.04–1.10)
Other	0.99 (0.11–8.95)		0.66 (0.06–6.94)
Duration of HIV diagnosis (years)		<0.001		<0.001
<6 Years	Referent		Referent
6–14 Years	0.38 (0.27–0.54)		0.40 (0.26–0.64)
>14 Years	0.23 (0.15–0.34)		0.28 (0.16–0.49)

Adjusted for age, ethnicity, education level, personal gross yearly income, housing status, HIV acquisition method, and duration living with HIV.

Includes self-contained room, housing facility and HIV care group home, transition/halfway/safe/outdoor housing or other.

CAD, Canadian dollars; cART, combination antiretroviral therapy; CI, confidence interval; EI, employment insurance, IDU, injection drug use; OR, odds ratio; WCBI, worker's compensation board insurance.

Factors associated with having a self-reported detectable VL, among those on cART

Of the 1178 women on cART, 91.0% had an undetectable VL and 9.0% had detectable VLs. Table 3 shows the unadjusted and aOR of having a self-reported detectable VL compared with an undetectable VL among women on cART. In univariate logistic regression analyses, factors associated with higher odds of having a self-reported detectable VL among women on cART included the following: being 25 years of age or younger versus 35 years of age, of sexual minority status versus heterosexual, indigenous versus white/Caucasian, having less than a secondary or a secondary education versus postsecondary, having any source of income (except savings/loan/family) versus a paid job, earning between $20,000 and $40,000 per year versus greater than $40,000 per year, renting or unstable housing versus owning, food insecurity, incarceration ever or last year versus never, current IDU, and depression. Factors associated with lower odds of having a detectable VL among women on cART included the following: ACB versus white/Caucasian ethnicity and having more than three financial dependents versus two to three. In multivariable analysis, factors associated with significantly increased odds of having a self-reported detectable VL among women on cART included age (being younger than the age of 25 [aOR: 3.11, 95% CI: 1.26–7.66] or between the ages of 26 and 34 [aOR: 1.93, 95% CI: 1.05–3.54] versus over 35), less than a secondary education (aOR: 2.50, 95% CI: 1.25–4.79) versus postsecondary, unstable housing (aOR: 5.21, 95% CI: 1.15–23.60) versus owning, and incarceration in the previous year (aOR: 2.89, 95% CI: 1.32–6.34) versus never.

Table 3.

Factors Associated with a Self-Reported Detectable Viral Load Among Women on Combination Antiretroviral Therapy (n = 1130)

Factors	Unadjusted OR (95% CI)	p	Adjusted OR^a (95% CI)	p
Age		0.022		0.011
≤25	2.50 (1.21–5.15)		3.11 (1.26–7.66)
26–34	1.55 (0.90–2.68)		1.93 (1.05–3.54)
≥35	Referent		Referent
Sexual orientation (sexual minority vs. heterosexual)	1.84 (1.07–3.18)	0.029	Not selected
Ethnicity		<0.001		0.077
White/Caucasian	Referent		Referent
Indigenous	1.72 (1.06–2.79)		0.89 (0.50–1.56)
African/Caribbean/black	0.41 (0.22–0.74)		0.54 (0.26–1.10)
Other ethnicity	1.23 (0.60–2.56)		1.80 (0.82–3.97)
Education level		<0.001		0.032
Less than secondary	5.08 (2.85–9.06)		2.50 (1.25–4.79)
Secondary	2.64 (1.54–4.52)		1.60 (0.87–2.94)
Postsecondary	Referent		Referent
Main source of income		<0.001		0.057
Paid job	Referent		Referent
Social assistance/pension/WCBI/EI	6.13 (2.46–15.29)		2.56 (0.94–6.96)
Sex work/selling drugs/panhandling	11.07 (1.89–64.93)		2.25 (0.32–16.09)
Savings/loan/family	3.11 (0.72–13.46)		2.45 (0.52–11.61)
Honoraria/other	12.02 (3.58–40.32)		7.20 (1.99–26.06)
Personal gross yearly income (CAD)		0.002	Not selected
≥$40,000 per year	Referent
$20,000–$40,000 per year	6.85 (1.66–28.21)
<$20,000/year	2.73 (0.58–12.86)
Financial dependents		0.036
<2	1.18 (0.67–2.10)
2–3	Referent
>3	0.26 (0.07–0.92)
Housing status		<0.001		0.042
Own (house/apartment/condo)	Referent		Referent
Rent (house/apartment/condo)	6.15 (1.49–25.35)		3.21 (0.74–13.88)
Unstable^b	13.64 (3.2–58.15)		5.21 (1.15–23.60)
Food security (insecure vs. secure)	2.26 (1.40–3.66)	0.001	1.52 (0.88–2.61)	0.130
Incarceration
Never	Referent	<0.001	Referent	0.028
Ever	2.57 (1.65–4.00)		1.30 (0.73–2.30)
Last year	5.55 (2.86–10.79)		2.89 (1.32–6.34)
IDU (current vs. not current but previously and never)	2.45 (1.32–4.55)	0.005	Not selected
Depression	1.04 (1.01–1.06)	0.013	Not selected

Adjusted for age, ethnicity, education level, main source of income, housing status, and history of incarceration.

Includes self-contained room, housing facility and HIV care group home, transition/halfway/safe/outdoor housing or other.

Discussion

Our study utilized the cascade of HIV care framework to identify sociodemographic, psychosocial, and clinical care factors associated with attrition at the points of cART use and VL suppression. In this study of 1382 WLWH in Canada, 85.2% of women were taking cART, among whom 91.0% reported suppressed VLs. Among the remaining 14.8% of women who were not taking cART, only 39.9% reported undetectable VLs. While these proportions are somewhat better than women in other industrialized countries such as the United States,⁹ they are still less than the 90% cART use and viral suppression targets recommended by UNAIDS.¹⁰ These findings have important implications for both individual WLWH and for healthcare policies and practices in Canada.

Despite the current international guidelines recommending early initiation of cART, ∼15% of Canadian WLWH were not currently on cART.⁷ This may, in part, be explained by having used baseline data from WLWH as early as 2013, at which point cART was still being prescribed based on a CD4 count cutoff of <500 cells/μL.⁴¹ In fact, 42.4% of women who were not on cART reported a CD4 cell count of >500 cells/μL, indicating that there was a sizeable proportion of women who were not on cART due to medical ineligibility at the time of interview. As such, our findings may by underestimating the true proportion of WLWH who were on cART. Another plausible explanation is that some of the women not on cART represented elite controllers or long-term nonprogressors (∼40% of these women reported suppressed VLs). Since current guidelines fail to make strong recommendations for or against cART in this population, the majority of elite controllers remain off cART.⁷ However, more recently, experts have been recommending cART for elite controllers due to inflammatory cardiovascular benefits.^42,43 Other WLWH may have chosen not to take cART for various reasons such as concomitant substance abuse, self-stigma, depression, or previous negative experiences with drug toxicities.^44
–46 It may be worthwhile exploring women's and provider's knowledge and providers' prescribing practices following the revised guidelines as there may be a role for provider education regarding updated ART initiation guidelines.

Addressing some key sociodemographic and psychosocial correlates may help ensure that all WLWH are accessing and taking cART. Women who reported current IDU had the highest odds of not being on cART. While this may reflect less frequent contact with the healthcare system, evidence suggests WLWH with a known history of substance use are less likely to be prescribed cART by their providers due to provider-related stigma or fear of poor adherence and its consequences on resistance.¹² Further, in our study, WLWH who endorsed racial discrimination had higher odds of not taking cART. These findings have broader implications for ongoing efforts in reducing harm associated with substance use and racial stigma and discrimination both at a structural and at a provider level. In addition, young WLWH (aged 26–34) had an almost twofold greater risk of not being on cART. This finding raises important questions about young women's understanding of the importance of cART, as well as provider's understanding of the needs of younger women.¹⁸ In the end, food insecurity was found to be associated with higher odds of not being on cART, a finding that is likely associated with multiple factors. Food is often required for cART use and if not required, many women indicate that they get nauseated when taking cART without food.⁴⁶ Also, food insecurity has been associated with poorer decision-making.⁴⁷ Finally, food insecurity is likely a surrogate marker for poverty. A better understanding of the way poverty and food insecurity affects cART use and access may be useful to moving forward.

Our study findings outlining the correlates associated with viremia despite cART use parallel what has already been described in the literature.^12,13 We found that young WLWH (≤25 years and between 26 and 34 years of age) were two to three times more likely to report detectable VLs on cART; decreasing age was directly proportional to increased risk of VL detectability. This presumably reflects issues with medication adherence (or due to lack of cART use in the 26- to 34-year-old age category), likely due to competing factors such as substance use, depression, and access among many others. This finding reinforces the need for youth- and young adult-focused programs, as well as the prescription of single-tablet regimens containing third agents with high barriers to resistance to optimize adherence and minimize drug resistance.⁴⁸ We also found that WLWH who had less than a high school education had 2.5-fold higher odds reporting detectable VLs on cART. Exploring the role of education to assist with VL suppression for WLWH may be necessary as the downstream effects of low education such as poverty and nonadherence are palpable correlates associated with viral detectability. Further, education captures the effect of childhood and young adulthood resource disparities, which may have long-term health consequences. The negative health effects of chronic stress brought on by poverty may be one area of inquiry. Finally, a history of incarceration in the previous year was associated with an almost threefold increased risk of VL detectability. This is likely due to treatment interruption or difficulties maintaining adherence during the time of or just after incarceration, and is crucial to rectify through structural system changes.⁴⁹

A major social determinant of health affecting cART use and VL suppression that emerged from our study was housing insecurity. Women with unstable housing had 2.5 and over fivefold higher odds of not taking cART and reporting detectable VLs on cART, respectively, adding to current evidence that unstable housing is significantly associated with decreased care along the continuum of HIV care.^50,51 Housing insecurity may itself reduce access to care, and, in addition, be a reflection of the interaction of complex social forces such as substance use, as well as poverty, lower education level, presence of mental health issues, and adverse childhood experiences among other possible contributing factors.^26,52,53 Research and programming focusing on women who are marginally housed may consider developing partnerships with and providing referrals to organizations that address housing needs in addition to other intersecting factors such as substance use and food insecurity.³⁶ Recognizing the role that structural barriers have on the HIV care continuum has implications for addressing the social determinants of health when planning programmatic efforts to address attrition along the cascade of care for WLWH.

There are a few limitations to note related to this analysis and the overall study design. Importantly, a biological outcome such as VL would best be measured utilizing blood testing. However, we used self-reported VL at most recent visit, which is subject to recall and social desirability bias. To mitigate this concern, analyses were conducted to explore the validity of our self-reported VL measure, and we found a high degree of correlation between self-reported and laboratory VLs.⁵⁴ The use of venue-based sampling—whereby many participants were recruited from ASOs, CBOs, and HIV clinics—may have introduced a selection bias, whereby we overestimated the proportion of women on cART (K. Webster, et al., unpublished observations). However, several steps were taken to ensure adequate representation of women who experience multiple forms of social and structural marginalization (e.g., women who use substances, women who do sex work, ands young women) who may be outside of care.²⁷ Therefore, on the contrary, word-of-mouth and purposive sampling may have overestimated those not on cART. The results in this analysis are not generalizable to all Canadian WLWH and must be taken in context of the nonrandom sampling. In addition, women not on cART were a combination of women who had not yet initiated cART and women who had used cART previously but not currently, leading to potential for misclassification. Ideally, a separate analysis would have been conducted for each group, but we combined these two categories due to small sample sizes. Finally, the cross-sectional study design does not allow for the establishment of causality, and reverse causality is an issue with regard to some factors (e.g., having a detectable VL could be a function of poor overall HIV-related health, which may influence poverty, rather than poverty influencing VL). Thus, some results should be interpreted with caution.

In conclusion, our study highlights several important sociodemographic, psychosocial, and clinical care factors associated with a lack of cART use and VL detection despite cART use in a large cohort of Canadian WLWH. Programmatic efforts to improve cART access and use for WLWH in Canada should focus on young women (<35 years of age), women with low levels of education or food and housing insecurity, as well as those who use drugs and have a history of incarceration. If these important social determinants of health can be considered, addressed, and modified, great strides will be made to ensure that Canadian WLWH reap the full benefits of cART.

Footnotes

Acknowledgments

The CHIWOS Research Team thanks WLWH for their contributions to this study. We also thank the national team of coinvestigators, collaborators, and PRAs and acknowledge the national Steering Committee, our three provincial Community Advisory Boards, the National CHIWOS Aboriginal, and African/Caribbean/Black Advisory Boards, the BC Center for Excellence in HIV/AIDS for data support and analysis, and all our partnering organizations for supporting the study. Funded by: Canadian Institutes of Health Research (MOP-111041), the CIHR Canadian HIV Trials Network (CTN 262), the Ontario HIV Treatment Network (OHTN), and the Academic Health Science Center's (AHSC) Alternative Funding Plans (AFP) Innovation Fund.

The CHIWOS Research Team: British Columbia: Aranka Anema (University of British Columbia), Denise Becker (Positive Living Society of British Columbia), Lori Brotto (University of British Columbia), Allison Carter (British Columbia Centre for Excellence in HIV/AIDS and Simon Fraser University), Claudette Cardinal (Simon Fraser University), Guillaume Colley (British Columbia Centre for Excellence in HIV/AIDS), Erin Ding (British Columbia Centre for Excellence), Janice Duddy (Pacific AIDS Network), Nada Gataric (British Columbia Centre for Excellence in HIV/AIDS), Robert S. Hogg (British Columbia Centre for Excellence in HIV/AIDS and Simon Fraser University), Terry Howard (Positive Living Society of British Columbia), Shahab Jabbari (British Columbia Centre for Excellence), Evin Jones (Pacific AIDS Network), Mary Kestler (Oak Tree Clinic, BC Women's Hospital and Health Centre), Andrea Langlois (Pacific AIDS Network), Viviane Lima (British Columbia Centre for Excellence in HIV/AIDS), Elisa Lloyd-Smith (Providence Health Care), Melissa Medjuck (Positive Women's Network), Cari Miller (Simon Fraser University), Deborah Money (Women's Health Research Institute), Valerie Nicholson (Simon Fraser University), Gina Ogilvie (British Columbia Centre for Disease Control), Sophie Patterson (Simon Fraser University), Neora Pick (Oak Tree Clinic, BC Women's Hospital and Health Centre), Eric Roth (University of Victoria), Kate Salters (Simon Fraser University), Margarite Sanchez (ViVA, Positive Living Society of British Columbia), Jacquie Sas (CIHR Canadian HIV Trials Network), Paul Sereda (British Columbia Centre for Excellence in HIV/AIDS), Marcie Summers (Positive Women's Network), Christina Tom (Simon Fraser University, BC), Lu Wang (British Columbia Centre for Excellence), Kath Webster (Simon Fraser University), Wendy Zhang (British Columbia Centre for Excellence in HIV/AIDS). Ontario: Rahma Abdul-Noor (Women's College Research Institute), Jonathan Angel (Ottawa Hospital Research Institute), Fatimatou Barry (Women's College Research Institute), Greta Bauer (University of Western Ontario), Kerrigan Beaver (Women's College Research Institute), Anita Benoit (Women's College Research Institute), Breklyn Bertozzi (Women's College Research Institute), Sheila Borton (Women's College Research Institute), Tammy Bourque (Women's College Research Institute), Jason Brophy (Children's Hospital of Eastern Ontario), Ann Burchell (Ontario HIV Treatment Network), Allison Carlson (Women's College Research Institute), Lynne Cioppa (Women's College Research Institute), Jeffrey Cohen (Windsor Regional Hospital), Tracey Conway (Women's College Research Institute), Curtis Cooper (Ottawa Hospital Research Institute), Jasmine Cotnam (Women's College Research Institute), Janette Cousineau (Women's College Research Institute), Annette Fraleigh (Women's College Research Institute), Brenda Gagnier (Women's College Research Institute), Claudine Gasingirwa (Women's College Research Institute), Saara Greene (McMaster University), Trevor Hart (Ryerson University), Shazia Islam (Women's College Research Institute), Charu Kaushic (McMaster University), Logan Kennedy (Women's College Research Institute), Desiree Kerr (Women's College Research Institute), Maxime Kiboyogo (McGill University Health Centre), Gladys Kwaramba (Women's College Research Institute), Lynne Leonard (University of Ottawa), Johanna Lewis (Women's College Research Institute), Carmen Logie (University of Toronto), Shari Margolese (Women's College Research Institute), Marvelous Muchenje (Women's Health in Women's Hands), Mary (Muthoni) Ndung'u (Women's College Research Institute), Kelly O'Brien (University of Toronto), Charlene Ouellette (Women's College Research Institute), Jeff Powis (Toronto East General Hospital), Corinna Quan (Windsor Regional Hospital), Janet Raboud (Ontario HIV Treatment Network), Anita Rachlis (Sunnybrook Health Science Centre), Edward Ralph (St. Joseph's Health Care), Sean Rourke (Ontario HIV Treatment Network), Sergio Rueda (Centre for Addiction and Mental Health [CAMH]), Roger Sandre (Haven Clinic), Fiona Smaill (McMaster University), Stephanie Smith (Women's College Research Institute), Tsitsi Tigere (Women's College Research Institute), Wangari Tharao (Women's Health in Women's Hands), Sharon Walmsley (Toronto General Research Institute), Wendy Wobeser (Kingston University), Jessica Yee (Native Youth Sexual Health Network), Mark Yudin (St-Michael's Hospital). Quebec: Dada Mamvula Bakombo (McGill University Health Centre), Jean-Guy Baril (Université de Montréal), Nora Butler Burke (University Concordia), Pierrette Clément (McGill University Health Center), Janice Dayle, (McGill University Health Centre), Danièle Dubuc, (McGill University Health Centre), Mylène Fernet (Université du Québec à Montréal), Danielle Groleau (McGill University), Aurélie Hot (COCQ-SIDA), Marina Klein (McGill University Health Centre), Carrie Martin (Native Women's Shelter of Montreal), Lyne Massie, (Université de Québec à Montréal), Brigitte Ménard, (McGill University Health Centre), Nadia O'Brien (McGill University Health Centre and Université de Montréal), Joanne Otis (Université du Québec à Montréal), Doris Peltier (Canadian Aboriginal AIDS Network), Alie Pierre, (McGill University Health Centre), Karène Proulx-Boucher (McGill University Health Centre), Danielle Rouleau (Centre Hospitalier de l'Université de Montréal), Édénia Savoie (McGill University Health Centre), Cécile Tremblay (Centre Hospitalier de l'Université de Montréal), Benoit Trottier (Clinique l'Actuel), Sylvie Trottier (Centre Hospitalier Universitaire de Québec), Christos Tsoukas (McGill University Health Centre). Other Canadian provinces or international jurisdictions: Jacqueline Gahagan (Dalhousie University), Catherine Hankins (University of Amsterdam), Renee Masching (Canadian Aboriginal AIDS Network), and Susanna Ogunnaike-Cooke (Public Health Agency of Canada). All other CHIWOS Research Team Members wish to remain anonymous.

Author Disclosure Statement

No competing financial interests exist.

References

Gardner

, McLees

, Steiner

, et al. 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.

Corless

, Hoyt

, Tyer-Viola

, et al. 90-90-90-Plus: Maintaining adherence to antiretroviral therapies. AIDS Patient Care STDS, 2017; 31:227–236.

Rodger

, Cambiano

, Bruun

, et al. Sexual activity without condoms and risk of HIV transmission in serodifferent couples when the HIV-positive partner is using suppressive antiretroviral therapy. JAMA, 2016; 316:171–181.

Montaner

, Lima

, Barrios

, et al. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: A population-based study. Lancet, 2010; 376:532–539.

Fang

, Hsu

, Twu

, et al. Decreased HIV transmission after a policy of providing free access to highly active antiretroviral therapy in Taiwan. J Infect Dis, 2004; 190:879–885.

Wood

, Kerr

, Marshall

, et al. Longitudinal community plasma HIV-1 RNA concentrations and incidence of HIV-1 among injecting drug users: Prospective cohort study. BMJ, 2009; 338:b1649.

Lima

, Hogg

, Montaner

. Expanding HAART treatment to all currently eligible individuals under the 2008 IAS-USA Guidelines in British Columbia, Canada. PLoS One, 2010; 5:e10991.

Granich

, Gilks

, Dye

, et al. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: A mathematical model. Lancet, 2009; 373:48–57.

Parsons

, Rendina

, Lassiter

, et al. Uptake of HIV pre-exposure prophylaxis (PrEP) in a national cohort of gay and bisexual men in the United States. J Acquir Immune Defic Syndr, 2017; 74:285–292.

10.

Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. Available at: https://aidsinfo.nih.gov/contentfiles/lvguidelines/adultandadolescentgl.pdf (Last accessed February 16, 2017 ).

11.

UNAIDS. 90–90–90—An ambitious treatment target to help end the AIDS epidemic. 2014. Available at: www.unaids.org/sites/default/files/media_asset/90-90-90_en_0.pdf (Last accessed February 16, 2017 ).

12.

Beer

, Mattson

, Bradley

, et al. Understanding cross-sectional racial, ethnic, and gender disparities in antiretroviral use and viral suppression among HIV patients in the United States. Medicine (Baltimore), 2016; 95:e3171.

13.

Hall

, Frazier

, Rhodes

, et al. Differences in human immunodeficiency virus care and treatment among subpopulations in the United States. JAMA Intern Med, 2013; 173:1337–1344.

14.

Horbery

, Hurley

, Klein

, et al. The HIV care cascade measured over time and by age, sex, and race in a large national integrated care system. AIDS Patient Care STDS, 2015; 29:582–590.

15.

Tapp

, Milloy

, Kerr

, et al. Female gender predicts lower access and adherence to antiretroviral therapy in a setting of free healthcare. BMC Infect Dis, 2011; 11:86.

16.

Wood

, Montaner

, Tyndall

, et al. Prevalence and correlates of untreated human immunodeficiency virus type 1 infection among persons who have died in the era of modern antiretroviral therapy. J Infect Dis, 2003; 188:1164–1170.

17.

Strathdee

, Palepu

, Cornelisse

, et al. Barriers to use of free antiretroviral therapy in injection drug users. JAMA, 1998; 280:547–549.

18.

Burchell

, Gardner

, Light

, et al. Implementation and operational research: Engagement in HIV care among persons enrolled in a clinical HIV cohort in Ontario, Canada, 2001–2011. J Acquir Immune Defic Syndr, 2015; 70:e10–9.

19.

Lourenco

, Colley

, Nosyk

, Shopin

, Montaner

, Lima

. High levels of heterogeneity in the HIV cascade of care across different population subgroups in British Columbia, Canada. PLoS One, 2014; 9:e115277.

20.

Lee

, Zhou

, Messerschmidt

, et al. Impact of gender on long-term treatment outcomes of highly active antiretroviral therapy (HAART) in the TREAT Asia HIV Observational Database. AIDS Patient Care STDS, 2015; 29:229–331.

21.

Nosyk

, Montaner

, Colley

, et al. The cascade of HIV care in British Columbia, Canada, 1996–2011: A population-based retrospective cohort study. Lancet Infect Dis, 2014; 14:40–49.

22.

Koole

, Houben

, Mzembe

, et al. Improved retention of patients starting antiretroviral treatment in Karonga District, northern Malawi, 2005–2012. J Acquir Immune Defic Syndr, 2014; 67:e27–e33.

23.

Lillie-Blanton

, Stone

, Jones

, et al. Association of race, substance abuse, and health insurance coverage with use of highly active antiretroviral therapy among HIV-infected women, 2005. Am J Public Health, 2010; 100:1493–1499.

24.

Cohen

, Cook

, Grey

, et al. Medically eligible women who do not use HAART: The importance of abuse, drug use, and race. Am J Public Health, 2004; 94:1147–1151.

25.

Cook

, Cohen

, Grey

, et al. Use of highly active antiretroviral therapy in a cohort of HIV-seropositive women. Am J Public Health, 2002; 92:82–87.

26.

Logie

, Jenkinson

, Earnshaw

, et al. A structural equation model of HIV-related stigma, racial discrimination, housing insecurity and wellbeing among African and Caribbean Black women living with HIV in Ontario, Canada. PLoS One, 2016; 11:e0162826.

27.

Aljassem

, Raboud

, Benoit

, et al. Gender differences in severity and correlates of depression symptoms in people living with HIV in Ontario, Canada. J Int Assoc Provid AIDS Care, 2016; 15:23–35.

28.

Cunningham

, Hays

, Williams

, et al. Access to medical care and health-related quality of life for low-income persons with symptomatic human immunodeficiency virus. Med Care, 1995; 33:739–754.

29.

Keruly

, Conviser

, Moore

. Association of medical insurance and other factors with receipt of antiretroviral therapy. Am J Public Health, 2002; 92:852–857.

30.

Kahn

, Haile

, Kates

, Chang

. Health and federal budgetary effects of increasing access to antiretroviral medications for HIV by expanding Medicaid. Am J Public Health, 2001; 91:1464–1473.

31.

Shannon

, Bright

, Duddy

, Tyndall

. Access and utilization of HIV treatment and services among women sex workers in Vancouver's Downtown Eastside. J Urban Health, 2005; 82:488–497.

32.

Loutfy

, Greene

, Kennedy

, et al. Establishing the Canadian HIV Women's Sexual and Reproductive Health Cohort Study (CHIWOS): Operationalizing community-based research in a large national quantitative study. BMC Med Res Methodol, 2016; 16:101.

33.

Zhang

, O'Brien

, Forrest

, et al. Validating a shortened depression scale (10 item CES-D) among HIV-positive people in British Columbia, Canada. PLoS One, 2012; 7:e40793.

34.

Gjesfjeld

, Greeno

, Kim

. A confirmatory factor analysis of an abbreviated social support instrument: The MOS-SSS. Res Soc Work Pract, 2008; 18:231–237.

35.

Berger

, Ferrans

, Lashley

. Measuring stigma in people with HIV: Psychometric assessment of the HIV stigma scale. Res Nurs Health, 2001; 24:518–529.

36.

Wright

, Naar-King

, Lam

, Templin

, Frey

. Stigma scale revised: Reliability and validity of a brief measure of stigma for HIV+ youth. J Adolesc Health, 2007; 40:96–98.

37.

Palmer

, Duncan

, Ayalew

, et al. “The way I see it”: The effect of stigma and depression on self-perceived body image among HIV-positive individuals on treatment in British Columbia, Canada. AIDS Care, 2011; 23:1456–1466.

38.

Williams

, Yan

, Jackson

, Anderson

. Racial differences in physical and mental health: Socio-economic status, stress and discrimination. J Health Psychol, 1997; 2:335–351.

39.

Krieger

, Smith

, Naishadham

, Hartman

, Barbeau

. Experiences of discrimination: Validity and reliability of a self-report measure for population health research on racism and health. Soc Sci Med, 2005; 61:1576–1596.

40.

StataCorp. Stata statistical software: Release 14. College Station, TX: StatCorp LP, 2015.

41.

US Department of Health and Human Sciences (DHHS). Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. 2013. Available at: www.aidsinfo.nih.gov/guidelines (Last accessed February 20, 2017 ).

42.

Pereyra

, Lo

, Triant

, et al. Increased coronary atherosclerosis and immune activation in HIV-1 elite controllers. AIDS, 2012; 26:2409–2412.

43.

Hsue

, Hunt

, Schnell

, et al. Role of viral replication, antiretroviral therapy, and immunodeficiency in HIV-associated atherosclerosis. AIDS, 2009; 23:1059–1067.

44.

Uthman

, Magidson

, Safren

, Nachega

. Depression and adherence to antiretroviral therapy in low-, middle- and high-income countries: A systematic review and meta-analysis. Curr HIV/AIDS Rep, 2014; 11:291–307.

45.

Nakimuli-Mpungu

, Bass

, Alexandre

, et al. Depression, alcohol use and adherence to antiretroviral therapy in sub-Saharan Africa: A systematic review. AIDS Behav, 2012; 16:2101–2118.

46.

Bhat

, Ramburuth

, Singh

, et al. Factors associated with poor adherence to anti-retroviral therapy in patients attending a rural health centre in South Africa. Eur J Clin Microbiol Infect Dis, 2010; 29:947–953.

47.

Hindin

. Women's input into household decisions and their nutritional status in three resource-constrained settings. Public Health Nutr, 2006; 9:485–493.

48.

Hudelson

, Cluver

. Factors associated with adherence to antiretroviral therapy among adolescents living with HIV/AIDS in low- and middle-income countries: A systematic review. AIDS Care, 2015; 27:805–816.

49.

Lin

, Elwood Martin

, Milloy

M-J

, et al. Factors shaping a high prevalence of ever and recent incarceration experience among women living with HIV in Canada. International AIDS Conference (AIDS2016), Durban, South Africa, July 18–22, 2016 [Abstract THPED307].

50.

Haley

, Lucas

, Golin

, et al. Retention strategies and factors associated with missed visits among low income women at increased risk of HIV acquisition in the US (HPTN 064). AIDS Patient Care STDS, 2014; 28:206–217.

51.

, MacGovern

, Bradford

. Association of ancillary services with primary care utilization and retention for patients with HIV/AIDS. AIDS Care, 2002; 14:S45–S57.

52.

McNeil

, Guirguis-Younger

, Dilley

, et al. Harm reduction services as a point-of-entry to and source of end-of-life care and support for homeless and marginally housed persons who use alcohol and/or illicit drugs: A qualitative analysis. BMC Public Health, 2012; 12:312.

53.

Boerma

, Weir

. Integrating demographic and epidemiological approaches to research on HIV/AIDS: The proximate-determinants framework. J Infect Dis, 2005; 19:S61–S67.

54.

Carter

, de Pokomandy

, Loutfy

, et al. Validating a self-report measure of HIV viral suppression: An analysis of linked questionnaire and clinical data from the Canadian HIV Women's Sexual and Reproductive Health Cohort Study. BMC Res Notes, 2017; 10:138.