Impact of multi-level factors and partner characteristics on antiretroviral therapy adherence and access to HIV care during the COVID-19 pandemic

Abstract

Background

Adherence to antiretroviral therapy (ART) remains a challenge for many people with HIV and was exacerbated during the COVID-19 pandemic. This paper examines factors associated with ART adherence among people with HIV who inject drugs (PWHWID) in Almaty, Kazakhstan during the COVID-19 pandemic.

Methods

A cross-sectional survey was conducted from October 2020 to August 2022 with 66 PWHWID and their treatment support partners (n = 66) to assess associations between ART adherence and sociodemographic, COVID-19 related, social support, and other factors. Multilevel generalized linear mixed models were used to examine factors associated with optimal (≥90%) and standard (≥80%) adherence levels.

Results

We found low medication adherence rates: only 55.8% took ≥80% of their doses, and just 14.7% took ≥90%. People were more likely to take their medication consistently if they had a partner with HIV, experienced less household conflict during COVID-19, or met with addiction professionals. Poor mental health and negative emotional impacts from COVID-19 were associated with low adherence.

Conclusion

These findings suggest that HIV treatment programs should consider the importance of patients’ relationship factors, mental health, and home environment impacts during public health crises.

Keywords

HIV antiretroviral therapy adherence people who inject drugs COVID-19 Kazakhstan

Introduction

One of the key factors to achieve viral suppression within the UNAIDS 95-95-95 strategy for people with HIV (PWH) is antiretroviral therapy (ART) adherence. Viral suppression is key to improving population health (e.g., reduced HIV transmission, decreased drug resistance)^1,2 and individual health (e.g., reduced morbidity and mortality).^3,4 The COVID-19 pandemic may have made adherence more difficult, particularly in low- and middle-income countries (LMICs), through factors such as changes in healthcare access and social support.⁵ The impacts were more severe for marginalized populations, such as people who inject drugs (PWID), resulting in disrupted addiction treatment and harm reduction services.⁶

Kazakhstan has one of the fastest growing HIV epidemics globally,⁷ driven primarily by drug injection, with over half of PWH acquiring HIV infection through this route.⁸ Kazakhstan experienced several waves of lockdowns during the COVID-19 pandemic, which restricted access to healthcare services and overburdened healthcare systems. These changes resulted in altered social dynamics among PWID such as changes in drug use patterns and behaviors including isolation during drug use and shifts in drugs used leading to overdoses⁹; alterations in the way PWID access and share drug injection equipment, potentially leading to increased risk of blood-borne infections like HIV and HCV⁶; and disruptions in the illicit drug supply chain which may affect the availability, purity, and price of drugs.¹⁰ Increasing and maintaining ART adherence among PWID in Kazakhstan, especially during public health crises, is critical to reducing the spread of HIV in the country. Yet there is limited research on factors associated with ART adherence among people with HIV who inject drugs (PWHWID) in Kazakhstan.

COVID-19-related factors that may impact ART adherence include increases in partner or family conflict, potential social isolation, and lack of social support.¹¹ The immediate environment, especially the household environment and close social circles, play a large role in creating a safe space and facilitating health promotion behaviors. During the COVID-19 lockdown periods, many individuals were confined to their households, and there was an increase in substance use and poor mental health,¹² all of which could adversely affect ART adherence.¹³ During the COVID-19 pandemic, social relationships may have had an increased impact on medication adherence for PWH during this stressful period.

Understanding factors associated with ART adherence among PWID during public health crises is key to maintaining their health and reducing further spread of the virus when such crises inevitably arise in the future. Thus, the purpose of this paper was to examine multi-level factors (e.g., partner HIV status, employment, mental health, level of negative financial and emotional impact of the COVID-19 pandemic) associated with ART adherence during the COVID-19 pandemic among PWHWID in Kazakhstan.

Methods

Study design, Recruitment and procedures

Data are from the baseline survey of a study focused on adapting a dyad-based intervention for PWHWID to improve ART adherence.¹⁴ This survey was conducted from October 2020 to August 2022 with 66 PWHWID and their treatment support partners (n = 66). An epidemiologist at the AIDS center reviewed patient records to determine which patients potentially met eligibility requirements (i.e., were diagnosed with HIV, on ART for ≥3 months, not virally suppressed, had a history of injection drug use). AIDS Center nurses informed potentially eligible patients about the study and screened interested individuals. After consent, nurses provided research staff from the Global Health Research Center of Central Asia (GHRCCA) with contact information. Research staff explained the study, answered questions, and conducted confirmatory screening and consent procedures. Eligible participants ('index cases') were asked to invite a treatment support partner (e.g., spouse/romantic partner, family member, friend) to participate. Screening and consent procedures were conducted separately with support partners.

The “index case” (partner initially recruited from the Almaty AIDS Center) must 1) have been diagnosed with HIV by the Almaty AIDS Center; 2) have been on ART for at least 3 months; 3) was not virally suppressed within the past year according to the AIDS Center medical records (≤500 copies/ml), and 4) have injected drugs in the past year. In addition, dyads were eligible to participate if both partners were 1) ≥18 years old; 2) both identified each other as their intimate partner, family member, friend, or other support person within their social network; 3) their relationship had existed for at least 3 months; 4) both reported feeling safe participating with their partner in the study; 5) neither reported any severe physical or sexual violence perpetrated by the other partner in the past year; 6) both were able to provide informed consent and follow study procedures; and 7) both were fluent in Russian. “Partners” may or may not have HIV and may or may not be a PWID. Upon enrollment, the research assistant (RA) sent participants a Qualtrics link to complete the baseline survey.

Measures

Surveys were administered to participants in Russian and contained items on socio-demographics, measures for ART adherence, access to HIV care, COVID-19 related factors, and medication specific social support.

Dependent variable-ART adherence

ART adherence was assessed using a standardized, brief medication adherence scale¹⁵ which contains three items: 1) Days taken -- number of days the patient missed at least one dose of their medication (number of days, 0-30); and two five-point Likert items of 2) Frequency – how often the patient took ART in the past 30 days as prescribed (range: never to always); and 3) Rating – how good a job the patient did in taking their medication in the way they were supposed to in the past 30 days (range: very poor to excellent). Responses were then linearly transformed into a scale of 0–100 (higher scores indicate greater adherence). The average score across the three items was calculated to represent overall adherence. Scale reliability among this sample was high (Cronbach’s α = 0.90). Adherence cutoffs were analyzed at both the optimal (≥90%) and standard (≥80%) thresholds,^16,17 with 90% adherence being our primary threshold.¹⁸ With newer ART regimens, lower levels of ART adherence at 80% are sufficient to achieve viral suppression; thus, factors associated with an 80% threshold were examined as secondary outcomes.^19,20

Independent variables

The independent variables for the index case included sociodemographic factors, including race/ethnicity, age, gender identity, employment status, having children, and HIV status (determined through medical record). For dyadic outcomes, we used scales and questionnaires to assess potential factors associated with ART adherence across the dyads and calculated the reliability of the scales in our sample. These scales included the Patient Clinic Relationship Scale to assess the participants' perceived quality of relationship with their health-care provider (α = 0.93)²¹; Life Events Questionnaire to assess exposure to a broad range of potentially traumatic events ranging from home life, financial problems, to health concerns (α = 0.66)^22,23; the Substance Abuse and Mental Illness Symptoms Screener (SAMISS) to assess mental health and substance use problems (α = 0.81)^24–26; the Perceived Availability of Social Support Scale²⁷ to assess participants’ perceived availability of someone to provide support for different situations (α = 0.94)²⁸; and the HIV Treatment Related Knowledge Scale to measure HIV treatment knowledge about complex treatment issues such as adherence, side-effects, and drug resistance (α = 0.91). Additionally, COVID-19 questions were developed based on current Kazakh government COVID-19 related priorities and policies and included questions such as financial and emotional impacts of COVID-19. During the same time as this study, the Almaty AIDS Center was rolling out another program using home health nurses to try and improve ART adherence. We also asked participants if they had received a home visit from a nurse.

Analysis plan

The primary objective was to examine sociodemographic, social support, and COVID-19 related factors associated with achieving ≥90% adherence. We conducted descriptive analyses to examine the frequencies and distributions of variables. We conducted multilevel generalized linear mixed models using Bayesian estimation with weakly informative half-Cauchy priors²⁹ and a binary distribution that incorporated random effects for dyads to account for potential sources of within-dyad statistical correlations. Bayesian estimation has been shown to be superior to maximum likelihood estimation for studies with small sample size.³⁰ We first conducted bivariable analyses with each factor hypothesized to be related to adherence. We then checked for multicollinearity among variables. For variables that were highly collinear, only one variable was included in the model. Based on recommendations in the literature, factors significant at p ≤ .15 in the bivariable analyses were included in the multivariable model.³¹ Significance for multivariable analyses were two-sided at the 0.05 level. Because some research indicates that 80% adherence is sufficient to achieve viral suppression and prevent drug resistance,^19,20 we conducted secondary analyses to examine factors associated with ≥80% adherence using these same methods. Analyses were performed using the 'brms'³² package in R v4.3.3.³³

Results

The final analytic sample of 132 participants (66 dyads) ranged in age from 21 to 76 (mean = 45) years (Table 1), with over half (55%) of the participants male. Most had a high school degree or higher (30% had a high school or general education degree and 40% had a technical or college degree). More than half (57%) were not working. Among the employed participants, an equal proportion engaged in full-time (50%) and part-time (50%) employment. While all index cases were living with HIV, 29 (44%) of the treatment support partners were also living with HIV. Of those 95 PWH, most of them (85%) reported not achieving ≥90% adherence; however, a little over half (56%) reported ≥80% ART adherence.

Table 1.

Descriptive Statistics of the sample, stratified by index and partner cases (N = 132).

Variable	Overall N = 132	Index case N = 66	Partner N = 66	p-value
Variable	Mean (SD)	Mean (SD)	Mean (SD)	p-value
Age, years	45.0 (10.36)	43.2 (6.96)	46.9 (12.69)	0.040
	N (%)	N (%)	N (%)
Sex				0.054
Male	73 (55.3%)	42 (63.6%)	31 (47.0%)
Female	59 (44.7%)	24 (36.4%)	35 (53.0%)
Maximum Educational level achieved				0.165
8-9^th grade	39 (29.5%)	24 (36.4%)	15 (22.7%)
High school	40 (30.3%)	20 (30.3%)	20 (30.3%)
Beyond high school	53 (40.2%)	22 (33.3%)	31 (47.0%)
Currently working				0.219
No	75 (56.8%)	41 (62.1%)	34 (51.5%)
Yes	57 (43.2%)	25 (37.9%)	32 (48.5%)
Employment status (n = 57)				0.060
Full-time	28 (50%)	9 (36.0%)	19 (61.3%)
Part-time	28 (50%)	16 (64.0%)	12 (38.7%)
Family status				0.262
Not married/partnered	42 (31.8%)	24 (36.4%)	18 (27.3%)
Married/Living with partner	90 (68.2%)	42 (63.6%)	48 (72.7%)
Have children				0.199
No	45 (34.1%)	26 (39.4%)	19 (28.8%)
Yes	87 (65.9%)	40 (60.6%)	47 (71.2%)
Ever received methadone or opioid agonist treatment (OAT)				0.301
Yes, currently receiving methadone or OAT	6 (4.5%)	5 (7.6%)	1 (1.5%)
Yes, previously received methadone or OAT, but not currently	4 (3.0%)	2 (3.0%)	2 (3.0%)
HIV status				<.001
Positive	95 (72%)	66 (100%)	29 (43.9%)
Negative	37 (28%)	0 (0%)	37 (56.1%)
Standard adherence ≥80% (n = 95)				0.206
No	42 (44.2%)	32 (48.5%)	10 (34.5%)
Yes	53 (55.8%)	34 (51.5%)	19 (65.5%)
Optimal adherence ≥90% (n = 95)				<.001
No	81 (85.3%)	62 (93.9%)	19 (65.5%)
Yes	14 (14.7%)	4 (6.1%)	10 (34.5%)

In the bivariate analysis for ≥90% adherence (Table 2), participants who had partners with HIV were significantly more likely to achieve ≥90% adherence compared to those with partners without HIV (OR = 10.26, 95% CI: 1.82-128.73, p = .01). Currently employed individuals had higher odds of high adherence compared to unemployed participants (OR = 4.55, 95% CI: 1.27-19.28, p = .02). A higher substance use score was associated with a decreased likelihood of ≥90% adherence (OR = 0.45, 95% CI: 0.18-1.00, p = .05). Other factors, such as sex, age, health status, and various COVID-19-related impacts, did not show statistically significant associations with adherence in the bivariate analysis.

Table 2.

Unadjusted and adjusted logistic regression for factors associated with optimal ≥90% adherence (N = 95).

		Bivariate			Multivariable
Variable	Values	Odds Ratio	95% CI	p-value	Adjusted odds Ratio	95% CI	p-value
Partner HIV status	HIV negative	Ref.			Ref.
Partner HIV status	HIV positive	10.26	1.82, 128.73	0.01	14.94	1.51, 420.77	0.02
Sex	Male	0.33	0.08, 1.17	0.09	0.34	0.05, 1.67	0.19
Sex	Female	Ref.			Ref.
Age		1.03	0.93, 1.13	0.61
Currently employed	No	Ref.			Ref.
Currently employed	Yes	4.55	1.27, 19.28	0.02	4.36	0.92, 26.44	0.07
Has kids	No	Ref.
Has kids	Yes	0.88	0.26, 3.0	0.85
Met with professional to talk about alcohol or drug use	No	Ref.
Met with professional to talk about alcohol or drug use	Yes	1.26	0.33, 4.51	0.71
Had an AIDS center nurse deliver ART medications during COVID?^a	No	Ref.
	Yes	0.28	0.05, 1.15	0.08
Mean patient-clinic relationship scale score		1.61	0.58, 4.73	0.36
Mean life events questionnaire score		0.01	0, 1.62	0.10	0.02	0.00, 5.84	0.24
Mean substance use score		0.45	0.18, 1.00	0.05	0.65	0.16, 1.96	0.46
Mean mental health score		0.25	0.03, 1.48	0.11	0.28	0.02, 3.05	0.32
Mean social support scale score		0.76	0.38, 1.59	0.46
Mean HIV treatment knowledge scale score		0.58	0.19, 1.60	0.33
Were you unable to pay for COVID-19 related medical care?	No	Ref.
Were you unable to pay for COVID-19 related medical care?	Yes	0.59	0.15, 2.27	0.42
Did you experience negative financial impacts due to COVID-19?	Not at all/ a little	Ref.
	Somewhat/ quite a bit/very mu	0.60	0.16, 2.47	0.44
Did you experience negative emotional impact due to COVID-19?	Not at all/ a little	Ref.
	Somewhat/ quite a bit/very mu	0.62	0.17, 2.17	0.44
Did you experience conflict in the home due to COVID-19?	Decreased	4.63	0.75, 33.27	0.11	22.77	1.62, 841.84	0.02
Did you experience conflict in the home due to COVID-19?	Stayed same/increased	Ref.			Ref.

^aVariable was excluded from the multivariable model due to high multicollinearity with other variables in the model.

In the multivariable model, partner HIV status and conflict in the home due to COVID-19 were significantly associated with ≥90% adherence. The odds of achieving ≥90% adherence were nearly 15 times higher among participants who had partners with HIV compared to those who had partners without HIV (AOR = 14.94, 95% CI: 1.51-420.77, p = .02). Participants reporting decreased conflict at home during COVID-19 (compared to before COVID-19) had significantly higher odds of ≥90% adherence compared to those reporting the same or increased conflict (AOR = 22.77, 95% CI: 1.62-841.84, p = .02). Currently employed individuals had marginally higher odds of achieving ≥90% adherence (AOR = 4.36, 95% CI: 0.92-26.44, p = .07). Since the test only approached statistical significance, this finding should be cautiously interpreted as a potential statistical trend rather than conclusive evidence.

In the bivariate analysis (Table 3), being currently employed was associated with higher odds of ≥80% adherence (OR = 3.14, 95% CI: 1.11-11.22, p = .03). Conversely, having an AIDS center nurse deliver ART medications during COVID-19 (OR = 0.27, 95% CI: 0.08-0.77, p = .01), poorer mental health (OR = 0.13, 95% CI: 0.02-0.51, p < .001), higher HIV treatment knowledge (OR = 0.40, 95% CI: 0.15-0.93, p = .03), and experiencing more negative emotional impact due to COVID-19 (OR = 0.25, 95% CI: 0.07-0.74, p = .01) were associated with lower odds of ≥80% adherence.

In the multivariable analysis, individuals who had met with a professional to discuss alcohol or drug use were more likely to have ≥80% adherence (AOR = 6.78, 95% CI: 1.25-117.61, p = .03). Poorer mental health (AOR = 0.09, 95% CI: 0.005-0.76, p = .03) and experiencing more negative emotional impact due to COVID-19 (AOR = 0.18, 95% CI: 0.002-0.98, p = .0485) continued to be associated with lower odds of adherence. Having kids was marginally associated with decreased adherence (AOR: 0.30, 95% CI: 0.06,-0.21, p = .08) and having a stronger patient-clinic relationship was marginally associated with increased adherence (AOR: 3.40, 95% CI: 0.93-24.88, p = .06).

Table 3.

Unadjusted and adjusted logistic regression for factors associated with standard ≥80% adherence (N = 95).

		Bivariate			Multivariable
Variable	Values	Odds Ratio	95% CI	p-value	Adjusted odds Ratio	95% CI	p-value
Partner HIV status	HIV negative	Ref.			Ref.
Partner HIV status	HIV positive	2.23	0.74, 8.26	0.15	2.13	0.46, 12.33	0.31
Sex	Male	0.95	0.30, 2.53	0.96
Sex	Female	Ref.
Age, years		1.02	0.94-1.12	0.68
Currently working^a	No	Ref.
Currently working^a	Yes	3.14	1.11, 11.22	0.03
Has kids	No	Ref.			Ref.
Has kids	Yes	0.35	0.10, 1.03	0.06	0.30	0.06, 1.21	0.08
Met with professional to talk about alcohol or drug use	No	Ref.
Met with professional to talk about alcohol or drug use	Yes	2.77	0.83, 12.72	0.09	6.78	1.25, 117.61	0.03
Had an AIDS center nurse deliver ART medications during COVID?	No	Ref.
	Yes	0.27	0.08, 0.77	0.01	0.25	0.03, 1.25	0.10
Mean patient-clinic relationship scale score		2.67	0.99, 12.11	0.05	3.40	0.93, 24.88	0.06
Mean life events questionnaire Score^a		0.00	0.00, 0.44	0.016
Mean substance use score		0.74	0.35, 1.41	0.37
Mean mental health score		0.13	0.02, 0.51	<0.001	0.09	0.005, 0.76	0.03
Mean social support scale score		1.06	0.52, 2.01	0.82
Mean HIV treatment knowledge scale score		0.40	0.15, 0.93	0.03	0.37	0.03, 1.34	0.16
Were you unable to pay for COVID-19 related medical care?	No	Ref.
Were you unable to pay for COVID-19 related medical care?	Yes	0.32	0.07, 1.05	0.06	0.34	0.01, 1.98	0.25
Did you experience negative financial impacts due to COVID-19?^a	Not at all/ a little	Ref.
	Somewhat/ quite a bit/very much	0.24	0.03, 1.03	0.05
Did you experience negative emotional impact due to COVID-19?	Not at all/ a little	Ref.
	Somewhat/ quite a bit/very much	0.25	0.07, 0.74	0.01	0.18	0.002, 0.98	0.049
Did you experience conflict in the home due to COVID-19?	Decreased	1.03	0.16, 7.11	0.98
Did you experience conflict in the home due to COVID-19?	Stayed same/increased	Ref.

^aVariable was excluded from the multivariable model due to high multicollinearity with other variables in the model.

Discussion

Adherence to ART is crucial for effective management of HIV, yet achieving high adherence rates remains challenging. This study aimed to identify factors associated with ART adherence among PWHWID and their treatment support partners during COVID-19 disruptions. An important finding is the association between having a partner with HIV and optimal (≥90%) ART adherence, suggesting the potentially important role of partner characteristics. There is limited data examining differences in support for ART adherence between partners with HIV and partners without HIV, particularly among PWHWID. Many studies have reported partners with HIV on ART may not necessarily provide support for adherence to each other.^34–36 However, our findings suggest that among PWHWID, having a partner with HIV was associated with a greater likelihood of achieving ≥90% adherence. Partners with HIV may have aligned priorities, such as shared health-related responsibilities and the need to care for each other. Seroconcordant partnerships may have more open communication about their HIV status and treatment due to their shared treatment experience and understanding of ART adherence importance. This may increase their motivation to adhere to their medication schedule and provide direct support; better equipping them to overcome adherence challenges together. Further research is needed to clarify the relationship between partner HIV status and ART adherence across different populations and settings.

Given that partners can improve ART adherence, clinicians may wish to involve partners in treatment discussions,³⁷ though inclusion should depend on relationship quality. Research shows dyad-based interventions in clinic and community settings may be particularly effective in promoting shared responsibility for adherence.³⁸ These interventions can foster mutual accountability and provide a supportive environment for medication adherence. Dyad-based approaches may also address relationship dynamics that impact health behaviors, such as communication patterns and shared decision-making processes. By engaging both partners, these interventions can potentially create a more sustainable support system for long-term ART adherence. Future research should aim to identify best practices for further engaging partners in adherence support efforts to optimize treatment outcomes for PWH. This could involve exploring various models of partner involvement, from education and counseling to more intensive support roles. Researchers might consider investigating the effectiveness of different partner engagement strategies across diverse relationship types and cultural contexts. For those with support partners without HIV, providing additional information or skills-building activities on supporting their partner may help increase adherence. Educational initiatives could focus on enhancing understanding of HIV treatment, addressing misconceptions, and developing practical skills for supporting adherence. These efforts might also incorporate strategies for managing the emotional aspects of being in a serodiscordant relationship, which could indirectly benefit adherence by reducing stress and improving overall relationship quality.

Our findings suggest the importance of considering personal and family dynamics in achieving ART adherence during periods of mass disruption, such as during the COVID-19 pandemic. During public health crises, when external support systems may be disrupted, the stability and supportiveness of the home environment can play a crucial role in helping PWH maintain their medication regimens.^39,40 Our results indicate that PWH living in household environments with increased conflict during COVID-19 had lower ART adherence. Identifying strategies to support families of PWH during times of public health crisis could be valuable for maintaining ART adherence. Such strategies could include counseling services (potentially via telehealth) to help families cope with crisis challenges, as well as financial assistance to ensure households can meet basic needs. Additionally, offering resources on maintaining healthy relationships and communication within the family could help create a more supportive environment for PWH. Such strategies may also help reduce the personal emotional impact of crises, which we found to be associated with adherence. By recognizing the significance of family support during public health emergencies, healthcare providers and policymakers can work towards creating a more comprehensive approach to HIV care that promotes optimal ART adherence.

Our results reveal significant associations between mental health, substance use treatment, and ART adherence among PWHWID. Notably, mental health issues, particularly depression and anxiety, can have a detrimental impact on medication adherence among PWH.⁴¹ In addition, individuals who had engaged with a professional to discuss alcohol or drug use were more likely to achieve adherence. These findings emphasize the need for regular mental health screening and substance use interventions in supporting ART adherence. They suggest that integrating mental health services into HIV treatment programs and addressing substance use issues through professional help may have a positive spillover effect on medication adherence behaviors, possibly by improving overall health management skills or reducing substance-related barriers to adherence. These findings collectively emphasize the interconnectedness of mental health, substance use, and ART adherence, particularly in the context of external stressors like the COVID-19 pandemic.

There are limitations that should be noted. First, we have a small sample size of 66 dyads, and ≥90% adherence was a relatively rare outcome among this sample. Although we used Bayesian estimation to improve the precision of our estimates, the small sample size and rare outcome resulted in wide credible intervals. Additional research with larger sample sizes is needed to provide more precise estimates. Despite our small sample size, our findings point to the potential impact of public health crises on home environment factors, which could make ART adherence more challenging. Second, data are drawn from a cross-sectional survey, which precludes causality. We also relied on a measure of self-reported adherence, which can be subject to recall and social desirability bias. Self-reported data may overestimate adherence, which may obscure associations with different factors. However, even with self-reported adherence, very few responded that they had ≥90% adherence and only half had ≥80% adherence, suggesting that participants overall were not artificially inflating their adherence levels. Additionally, participants were PWID living with HIV in Almaty, Kazakhstan. Different factors may be associated with ART adherence in other populations, but our findings lend important insights into a novel population. Also, few scales have been validated in Russian, and it is not clear if Russian-speaking participants interpreted items similarly to English-speaking participants. While most scales exhibited good internal reliability in this sample (α > 0.80), the Life Events Questionnaire had less than ideal internal reliability. However, a lower alpha corresponds to higher measurement error, which causes coefficient attenuation (bias towards the null). Therefore, it is possible that improving the internal reliability of this scale (which was significant in only the bivariate model), would provide more accurate results and potentially a stronger association. Future scale development work among Russian-speaking PWHWID is needed.

In conclusion, this study sheds light on factors associated with adherence to ART among PWHWID in Kazakhstan during the COVID-19 pandemic. Treatment support partners, in particular partners living with HIV, may play a key role in promoting ART adherence. Decreased conflict in the home and lower levels of negative emotional impact during COVID-19 also facilitated adherence. Further research is needed to better understand how partner support may be leveraged to facilitate ART adherence among partners with diverse characteristics. Future studies should employ longitudinal designs with larger samples and use mixed-methods approaches to gain a more comprehensive understanding of the dynamic nature of adherence over time, particularly during periods of disruption. It may be beneficial for healthcare providers to assess social support systems, family and partner dynamics, and mental health when developing adherence support strategies.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institutes of Health; K01DA044853.

ORCID iD

Paige Neuenschwander

References

Lima

Harrigan

Murray

, et al. Differential impact of adherence on long-term treatment response among naive HIV-infected individuals. AIDS 2008; 22(17): 2371–2380.

Bangsberg

Acosta

Gupta

, et al. Adherence–resistance relationships for protease and non-nucleoside reverse transcriptase inhibitors explained by virological fitness. AIDS 2006; 20(2): 223–231.

Flynn

Rudy

Douglas

, et al. Virologic and Immunologic outcomes after 24 Weeks in HIV type 1–infected Adolescents receiving highly active antiretroviral therapy. J Infect Dis 2004; 190(2): 271–279.

McNabb

Ross

Abriola

, et al. Adherence to highly active antiretroviral therapy predicts virologic outcome at an Inner‐City Human Immunodeficiency virus clinic. Clin Infect Dis 2001; 33(5): 700–705.

Algarin

Varas-Rodríguez

Valdivia

, et al. Symptoms, stress, and HIV-related care among older people living with HIV during the COVID-19 pandemic, Miami, Florida. AIDS Behav 2020; 24(8): 2236–2238.

Vasylyeva

Smyrnov

Strathdee

, et al. Challenges posed by COVID‐19 to people who inject drugs and lessons from other outbreaks. J Int AIDS Soc 2020; 23(7): e25583.

UNAIDS . UNAIDS data 2021 [Internet]. Geneva, Switzerland: UNAIDS; 2021, [cited 2021 Jan 20]. Available from: https://www.unaids.org/en/resources/documents/2021/2021_unaids_data

The Republican AIDS Center . Estimates of HIV cases, testing, linkage to care, initiation of ART and viral suppression among PWID as of july 1, 2014. Kazakhstan: Almaty, 2014.

Whitfield

Reed

Webster

, et al. The impact of COVID-19 restrictions on needle and syringe programme provision and coverage in England. Int J Drug Policy 2020; 83: 102851.

10.

European Monitoring Centre for Drugs and Drug Addiction . European Police Office. EU drug markets: impact of COVID 19. [Internet]. LU. Publications Office, 2020. [cited 2024 Jun 9]. Available from: https://data.europa.eu/doi/10.2810/19284

11.

Hamadani

Hasan

Baldi

, et al. Immediate impact of stay-at-home orders to control COVID-19 transmission on socioeconomic conditions, food insecurity, mental health, and intimate partner violence in Bangladeshi women and their families: an interrupted time series. Lancet Glob Health 2020; 8(11): e1380–e1389.

12.

Galvez

Altice

Meteliuk

, et al. High perceived stress in patients on opioid agonist therapies during rapid transitional response to the COVID-19 pandemic in Ukraine. Front Public Health 2023; 11: 1231581.

13.

Joska

Andersen

Rabie

, et al. COVID-19: increased risk to the mental health and safety of women living with HIV in South Africa. AIDS Behav 2020; 24(10): 2751–2753.

14.

Davis

Mergenova

Landers

, et al. Implementation of a dyad-based intervention to improve antiretroviral therapy adherence among HIV-positive people who inject drugs in Kazakhstan: a randomized trial. Res Soc Work Pract 2023; 33(3): 313–324.

15.

Wilson

Lee

Michaud

, et al. Validation of a new three-item self-report measure for medication adherence. AIDS Behav 2016; 20(11): 2700–2708.

16.

Ferro

Weikum

Vagenas

, et al. Alcohol use disorders negatively influence antiretroviral medication adherence among men who have sex with men in Peru. AIDS Care 2015; 27(1): 93–104.

17.

Altice

Evuarherhe

Shina

, et al. Adherence to HIV treatment regimens: systematic literature review and meta-analysis. Patient Prefer Adherence 2019; 13: 475–490.

18.

Phillips

Brittain

Mellins

, et al. A self-reported adherence measure to screen for Elevated HIV viral load in pregnant and postpartum women on antiretroviral therapy. AIDS Behav 2017; 21(2): 450–461.

19.

Viswanathan

Detels

Mehta

, et al. Level of adherence and HIV RNA suppression in the current Era of highly active antiretroviral therapy (HAART). AIDS Behav 2015; 19(4): 601–611.

20.

Paterson

Swindells

Mohr

, et al. Adherence to protease Inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 2000; 133(1): 21–30.

21.

Stall

Hoff

Coates

, et al. Decisions to Get HIV tested and to Accept antiretroviral therapies among Gay/Bisexual men: Implications for secondary prevention efforts. J Acquir Immune Defic Syndr Hum Retrovirol 1996; 11(2): 151–160.

22.

Roohafza

Ramezani

Sadeghi

, et al. Development and validation of the stressful life event questionnaire. Int J Public Health 2011; 56(4): 441–448.

23.

Kubany

Haynes

Leisen

, et al. Development and preliminary validation of a brief broad-spectrum measure of trauma exposure: the traumatic life events questionnaire. Psychol Assess 2000; 12(2): 210–224.

24.

Whetten

Reif

Swartz

, et al. A brief mental health and substance Abuse screener for persons with HIV. AIDS Patient Care STDS 2005; 19(2): 89–99.

25.

Breuer

Stoloff

Myer

, et al. The Validity of the substance Abuse and mental Illness Symptom screener (SAMISS) in people living with HIV/AIDS in primary HIV care in Cape Town, South Africa. AIDS Behav 2014; 18(6): 1133–1141.

26.

Breuer

Stoloff

Myer

, et al. Reliability of the Lay adherence Counsellor administered substance Abuse and mental Illness Symptoms screener (SAMISS) and the International HIV Dementia scale (IHDS) in a primary care HIV clinic in Cape Town, South Africa. AIDS Behav 2012; 16(6): 1464–1471.

27.

Barrera

. A method for the assessment of social support networks in community survey research. Connections 1980; 3(3): 8–13.

28.

Tardy

. Social support measurement. American J of Comm Psychol 1985; 13(2): 187–202.

29.

Gelman

Jakulin

Pittau

, et al. A weakly informative default prior distribution for logistic and other regression models. Ann Appl Stat 2008; 2(4): 8. [cited 2024 Sep 21] Available from: https://projecteuclid.org/journals/annals-of-applied-statistics/volume-2/issue-4/A-weakly-informative-default-prior-distribution-for-logistic-and-other/10.1214/08-AOAS191.full

30.

McNeish

. Using data-dependent priors to mitigate small sample bias in latent growth models: a discussion and Illustration using M plus. J Educ Behav Stat 2016; 41(1): 27–56.

31.

Bursac

Gauss

Williams

, et al. Purposeful selection of variables in logistic regression. Source Code Biol Med 2008; 3(1): 17.

32.

Bürkner

P-C

. Brms: An R package for Bayesian multilevel models using Stan. J Stat Soft [Internet] 2017; 80(1): 2. [cited 2024 Nov 9] Available from: https://www.jstatsoft.org/v80/i01/

33.

R Core Team. R

: A language and environment for statistical computing. R Foundation for Statistical Computing. 2024, Available from: https://cran.r-project.org/doc/manuals/r-release/fullrefman.pdf

34.

Johnson

Dilworth

Taylor

, et al. Primary relationships, HIV treatment adherence, and Virologic control. AIDS Behav 2012; 16(6): 1511–1521.

35.

Jones

Cook

Spence

, et al.

Antiretroviral therapy in Zambia: Do partners on ART enhance adherence?

J Int Assoc Provid AIDS Care 2014; 13(6): 497–500.

36.

Meyler

Stimpson

Peek

. Health concordance within couples: a systematic review. Soc Sci Med 2007; 64(11): 2297–2310.

37.

Safren

O’Cleirigh

Skeer

, et al. Demonstration and evaluation of a peer-delivered, individually-tailored, HIV prevention intervention for HIV-infected MSM in their primary care setting. AIDS Behav 2011; 15(5): 949–958.

38.

Remien

Stirratt

Dolezal

, et al. Couple-focused support to improve HIV medication adherence: a randomized controlled trial. AIDS 2005; 19(8): 807–814.

39.

Hong

Queiroz

Hoskin

. The impact of the COVID‐19 pandemic on mental health, associated factors and coping strategies in people living with HIV: a scoping review. J Int AIDS Soc 2023; 26(3): e26060.

40.

Saltzman

Hansel

Bordnick

. Loneliness, isolation, and social support factors in post-COVID-19 mental health. Psychol Trauma 2020; 12(S1): S55–S57.

41.

Mergenova

Davis

Gilbert

, et al. Mental health and cognition in relation to adherence to antiretroviral therapy among people living with HIV in Kazakhstan: a cross-sectional study. J Int AIDS Soc 2024; 27(Suppl 3): e26320.