Psychological predictors of adherence to antiretroviral therapy among people living with HIV in high-income countries: a systematic review protocol

Abstract

Introduction

People living with HIV (PLHIV) frequently face psychological challenges, including stigma, stress and social isolation, which can negatively affect adherence to antiretroviral therapy (ART). Even in high-income countries where treatment is accessible, poor adherence can lead to drug resistance, reduced immune function and early morbidity. This systematic review aims to synthesise evidence on the relationship between psychological and mental health factors and ART adherence among PLHIV in high-income settings.

Methods and analysis

We will include studies published in any language between January 2015 and the date of the last searches. Reports of studies published in languages other than English, and which appear to be eligible for inclusion after the first level of screening, will be translated using Google Translate.

Studies will be included if they continue to meet the inclusion criteria and the quality of the translation is sufficient to extract the relevant data. PLHIV aged ≥15 years receiving ART in high-income countries. The studies to be included must assess psychological or mental health variables and ART adherence. Peer-reviewed journal articles will be the primary source of evidence. Grey literature identified from reference lists of key articles or using Google Advanced search techniques will be included. Searches for published studies will be done in OVID Medline, PsycINFO and Embase. Cochrane CENTRAL will be used to identify clinical trials in ClinicalTrials.gov and the International Clinical Trials Registry Platform.

Two independent reviewers will assess study quality and risk of bias using the Newcastle-Ottawa Scale, National Institutes of Health (NIH) Quality Assessment Tool and Jadad Scale. Discrepancies will be resolved by a third reviewer. Synthesis of quantitative data will be primarily descriptive. Predictors that have been examined in three or more studies will be reported in detail while those assessed in fewer studies will be presented concisely.

Ethics and dissemination

This study will be a review of the literature and will not involve primary collection of patients’ data. We will include amendments to the protocol in the final review. The final study will be published in a peer-reviewed journal and presented at conferences. The results of this systematic review will inform clinical practice, guide future research and support policy development that minimise mental health barriers to ART adherence.

PROSPERO registration number

CRD420251102248.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Variations in how adherence and psychological variables are measured across studies may lead to heterogeneity of data, thus making it difficult to compare outcomes across studies and draw conclusions, ultimately reducing the internal validity of the study.

• Focusing our review on high-income countries will limit the applicability of findings to low-income and middle-income settings, where the health system, psychological stressors and adherence barriers differ significantly.

• We will involve two independent reviewers for all stages: screening, data extraction and quality assessment to reduce bias and errors, hence increasing the reliability of our study selection and data handling.

• We will follow the Preferred Reporting Items for Systematic Review and Meta-analysis Protocols 2020 checklist for reporting to ensure all methodological actions are clearly documented and reproducible.

Introduction

People living with HIV (PLHIV) tend to experience high rates of psychological problems which undermine their overall health and can lead to increased disease burden and virus transmission. Indeed, estimates of the presence of significant mental health symptoms in PLHIV range from approximately 28%–62%.¹ Based on a systematic review of 240 studies published in 2025, PLHIV tend to experience symptoms associated with depression (31%), anxiety (29%), suicidal ideation (20%), post-traumatic stress disorder (PTSD; 20%) and psychological distress (44%), as well as significant disease-related stigma (47%).² Indeed, many PLHIV have experienced traumatic life events, including sexual abuse, violence or loss, which can lead to PTSD and long-term psychological distress.²

Chronic stress from managing a stigmatised, lifelong illness may also result in adjustment disorders, characterised by emotional or behavioural symptoms in response to an identifiable stressor. Somatisation, where psychological distress is expressed through physical symptoms, is also more common in PLHIV, thereby complicating diagnosis and care.³ Cognitive impairments, including difficulties with memory, attention and executive functioning, are linked to the virus’s effects on the central nervous system, side effects of long-term antiretroviral therapy (ART) and traumatic experiences.

Finally, the widespread social stigma associated with HIV is correlated with poor mental and physical health.⁴ For PLHIV, externalised stigma from others can lead to discrimination, social isolation and fear of moral judgement or social rejection,⁵ whereas internalised stigma is associated with disadvantageous coping, self-blame, loneliness and rumination, as well as lowered resilience, perceived control and life satisfaction.⁶

Fortunately, HIV can be successfully managed long-term with antiretroviral treatment (ART). However, proper adherence to the ART regimen is necessary for the condition to become a manageable chronic disease⁷ and to stop virus transmission. For PLHIV taking ART drugs, adherence levels of 80%–90% can be adequate for viral suppression.⁸ However, for many PLHIV, consistent adherence to ART is a challenge. A 2020 study of 169,545 PLHIV in the USA found that 45% demonstrated poor (<80%) ART adherence, 30% demonstrated suboptimal (≥80% to <95%) adherence, and only 25% demonstrated optimal (>95%) adherence.⁹

Poor ART adherence can result in increased drug resistance, HIV transmission, hospitalisations and health-related costs,¹⁰ as well as increased rates of morbidity and mortality.⁷ Dosing complexity, tolerability and toxicity,⁷ along with varying levels of health insurance coverage and overall high healthcare costs, are all key factors that typically affect adherence and outcomes.¹⁰ Moreover, some individuals experience adverse effects from ART including sleep difficulties, irritability, fatigue and psychiatric issues, all of which are also associated with poor adherence.^{11 12}

Poor mental health can influence treatment adherence,¹³ leading to both deleterious consequences at the individual level as well as increased disease spread. Poor mental health can impair cognitive functioning, motivation and the ability to maintain daily routines, all of which are essential for adherence to ART.¹⁴ Psychological distress may lead to substance use, social withdrawal and decreased engagement with healthcare services, further undermining treatment adherence. Patients report that during periods of active mental illness and substance use, they can become apathetic about their health, leading to poor adherence and missed healthcare appointments.¹⁵

As described above, HIV-related stigma can also serve as a significant barrier, as it may cause increased vulnerability to mental health difficulties and a decrease in self-efficacy for medication taking.¹⁶ To remedy this, researchers have invested in understanding psychosocial interventions to improve ART adherence. A 2018 meta-analysis found that, for PLHIV, various psychosocial interventions (eg, cognitive behavioural techniques, motivational interviewing and relaxation) can improve mental health and ART adherence.¹⁷ This demonstrates a clear need for examining the existing mental health-related and support-related adherence barriers for PLHIV.

Given that the literature has elucidated many barriers to ART adherence that require research and clinical attention, and that individuals in high-income countries may have more access to HIV testing and support, one may expect that these individuals would not struggle with ART adherence and increased virus transmission. However, despite the availability of resources in high-income countries, significant barriers remain, including migration, stigma, fragmented care systems and disparities in access to mental health support.¹⁸

In high-income countries, PLHIV may experience inadequate housing, food insecurity, injection drug use, incarceration and periods of sex work that can act as additional barriers to adherence.^{7 19} These factors, as well as their associated stigma, can exacerbate or contribute to the mental health difficulties of PLHIV.²⁰ Within high-income countries, there are still medical providers who have not received sufficient training and lack knowledge surrounding HIV treatment, which can be the basis for stigmatised viewpoints and discrimination.21,23 This experience of stigma from healthcare workers can further contribute to mental health difficulties and facilitate poor quality of care, patient quality of life and ART adherence among PLHIV.

A systematic review of 66 studies published in 2014 found that over half of PLHIV globally report experiencing stigma. In high-income countries, the prevalence rate of stigma remains high at 42%.²⁴ These challenges within high-income countries underscore the need for a clearer understanding of how mental health influences ART adherence in these regions.

Our systematic review aims to synthesise and critically examine the existing literature on this topic, providing a comprehensive overview of the relationship between mental health variables such as psychiatric symptoms (eg, PTSD, depression, anxiety) as well as intrapersonal and interpersonal psychological variables (eg, stigma, discrimination, experiences of interpersonal violence) and ART adherence in PLHIV in high-income countries. To this end, we hope to provide a relevant and timely knowledge synthesis to support efforts to reduce HIV disease spread.

Research question

What mental health/psychological variables predict adherence to treatment for HIV in high income countries?

Methods

The Preferred Reporting Items for Systematic Review and Meta-analysis Protocols (PRISMA) guidelines were used in developing the protocol.²⁵ The protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) on 26 August 2025. Identifier: CRD420251102248. Available from https://www.crd.york.ac.uk/PROSPERO/view/CRD420251102248. It is anticipated that database searches will be carried out between January and June 2026, and the study will end in December 2026. Reporting of the synthesised findings will be informed by PRISMA guidelines. Important amendments to this protocol will be published along with the results of the systematic review.

Eligibility criteria

We will include original research articles reporting on PLHIV aged 15 years and older who are currently receiving ART. These studies must be conducted in high-income countries as classified by the World Bank and assess psychological or mental health variables such as depression, anxiety, stigma, coping strategies, motivation or self-efficacy as potential predictors of ART adherence.

Included studies must report ART adherence using either objective measures (eg, pharmacy refills, Medication Event Monitoring System caps or viral load suppression) or subjective measures (eg, self-reported adherence). Eligible study designs include quantitative observational studies (cross-sectional, cohort or case–control) and intervention studies. Peer-reviewed journal articles will be the primary source of evidence. Relevant clinical trials will be identified in Cochrane CENTRAL. Grey literature such as organisational reports, conference abstracts, dissertations or theses identified from reference lists of key articles or using Google Advanced search techniques will be included in the synthesis if they meet the inclusion criteria or will provide context to the analysis and conclusions.

Information sources

The search strategy was developed to search biomedical databases to identify studies about the population of interest based on the PICO framework. Studies published in any language from January 2015 until the date of the last searches in June 2026 will be considered. The literature search was restricted to studies published from 2015 to date to ensure the review reflects recent advances in ART and psychosocial interventions influencing adherence among PLHIV. Including studies from 2015 to date captures recent relevant advances in ART and psychosocial interventions affecting adherence among PLHIV. This timeframe ensures use of updated methods and reflects changes in HIV programme policies, digital support and sociocultural factors, enhancing the review’s relevance to current practice and future research.

Studies not published in English but that are eligible for data extraction will be translated into English using Google Translate. The participants are PLHIV aged ≥15 years, intervention (exposure) is ART, comparators are studies with or without control groups while the outcome is adherence or non-adherence to ART.

An academic health sciences librarian (JL) identified concepts, subject headings and keywords to identify relevant studies. Selected databases to identify relevant studies are Medline, Embase and PsycINFO; all using the OVID platform. Of note, JL additionally searched Scopus to improve the diversity of study output and found that there were no journal titles identified that were not indexed in Medline and/or Embase. The Cochrane Central Register of Controlled Trials (CENTRAL) will be used to identify clinical trials registered with ClinicalTrials.gov and the WHO’s International Clinical Trials Registry Platform. Additional searching using Google Advanced search techniques will be used to identify relevant grey literature such as reports by government and non-government organisations, along with any theses or dissertations.

Search strategy

We used the PICO framework to guide the development of our search strategy (table 1). The search strategy was developed to include keywords and subject headings relevant to each of the three OVID databases and will be used as a basis for searching the CENTRAL database. The preliminary search strategies created for Ovid MEDLINE, PsycINFO and Embase are detailed in our (online supplemental material). It is anticipated that searches will be carried out between January and June 2026.

Table 1. Search grid with identified PICO concepts.

Population	People living with HIV aged 15 years
Intervention (exposure)	Antiretroviral therapy (ART)
Comparators	Studies with or without control groups
Outcomes	Adherence or non-adherence to ART

Study selection

Search results from the online bibliographic databases will be saved as RIS files for uploading to Covidence and duplicates will be removed. Screening at the title/abstract stage and full-text screening will be done in Covidence. AZE and JP will blindly and independently screen the titles and abstracts of all retrieved records to assess their eligibility using ‘yes, no, or maybe’ as responses while MG and AK will conduct full-text screening. BF will resolve any disagreements between the two reviewers and the final study selection decision. We will obtain full-text articles for all abstracts that meet the inclusion criteria and seek assistance from the library services at the University of Winnipeg and/or the University of Manitoba for publications that are difficult to find. We will contact the authors directly if full-text copies are not available through these libraries. We will use full-text articles of eligible studies for data extraction.

Data extraction

MG and AZE will independently extract data using a standardised data extraction form. The following data will be extracted: Citation details (title, journal, author, year), country of study, study design, age, gender, ethnicity, psychological predictors assessed, ART adherence outcome measures and key findings relevant to adherence.

Outcomes and prioritisation

The primary objective of this review is to identify psychological or mental health predictors or determinants and other patient-related and treatment-related characteristics associated with adherence to HIV treatment. Studies will not be excluded based on how treatment adherence is defined; however, studies that do not assess psychological or mental health predictors will not be included.

Measures of effect

The measures used to assess significant mental health/psychological predictors and moderators of treatment adherence are expected to vary across studies. Studies will not be excluded based on the specific measures employed.

Risk of bias assessment

MG and AZE will assess methodological quality and risk of bias of all included studies using standardised tools tailored to each study design. This will be done at the data extraction phase. Any discrepancies in ratings will be resolved by BF. For cohort and case–control studies, the Newcastle-Ottawa Scale will be applied. This tool assesses three domains: selection (up to 4 stars), comparability (up to 2 stars) and outcome or exposure (up to 3 stars), yielding a total possible score of 9 stars. Based on these scores, studies will be classified as high quality (7–9 stars), moderate quality (5–6 stars) or low quality (4 stars or fewer).²⁶

We will use the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. This tool includes 14 items covering key aspects such as sample representativeness, exposure and outcome measurement, confounding control and statistical analysis. Each item is rated as ‘yes’, ‘no’, ‘cannot determine’, ‘not applicable’ or ‘not reported’. Based on an overall judgement of the study’s strengths and limitations, studies will be categorised as good (most criteria met with low risk of bias), fair (some criteria not met or unclear, with moderate risk of bias) or poor (many criteria not met, indicating high risk of bias).²⁷

Randomised controlled trials included in the review will be assessed using the Jadad Scale, which evaluates randomisation, blinding and reporting of withdrawals and dropouts. The scale assigns a score from 0 to 5, with studies scoring three or above considered high quality and those scoring below three considered low quality. ²⁸The results of the risk of bias assessments will not be used to exclude studies but will be considered during data synthesis and interpretation of findings. A narrative summary of the risk of bias across studies will be presented to provide context for the strength and reliability of the evidence. Each reviewer will leave an audit trail with reasons for each decision taken, which will later be compared across all reviewers.

Evidence of certainty

We will use the GRADE approach (Grading of Recommendations, Assessment, Development and Evaluations) to assess the certainty (or quality) of evidence and strength of recommendations.²⁹ We will consider five domains: risk of bias, inconsistency, indirectness, imprecision and publication bias.

We will evaluate the risk of bias using appropriate tools based on the study design. Inconsistency will be assessed qualitatively by comparing the direction and magnitude of reported associations between psychological variables and ART adherence across studies. Variations in study populations, types of psychological predictors assessed, measurement tools and adherence outcomes will be explored as potential sources of heterogeneity. Where findings are conflicting or vary substantially without clear explanations, we will consider this a source of inconsistency and may downgrade the certainty of evidence according to the GRADE framework.³⁰ Indirectness will be considered when the study population, intervention, comparator or outcomes differ from those specified in the review question. Imprecision will be assessed by evaluating the width of CIs and whether they cross decision-making thresholds.³¹ Publication bias will be explored through funnel plots, if enough studies are available, and by assessing the likelihood of selective outcome reporting.

Evidence from randomised trials will be stated as high quality, while evidence from observational studies will be regarded as low quality. The quality may be downgraded by one or two levels for serious or very serious concerns in any of the five GRADE domains. Conversely, the certainty of evidence from observational studies may be upgraded if there is a large effect size, evidence of a dose–response relationship, or if all plausible confounding would reduce the observed effect.

The overall certainty of evidence for each outcome will be rated as high, moderate, low or very low. ‘High’ will be regarded as studies with no serious limitations across GRADE domains, with no downgrading, ‘moderate’ as evidence downgraded by one level due to one serious limitation across GRADE domains, ‘low’ as evidence downgraded by two levels due to two serious limitations or one very serious limitation, and ‘very low’ as evidence downgraded by three or more levels due to multiple serious limitations across GRADE domains. A high rating indicates that we are very confident in the estimate of effect, while a very low rating means there is great uncertainty about the estimate.³²

Strategy for data synthesis

Data will be synthesised descriptively. The psychological predictors of ART adherence identified in three or more studies will be described in detail, including the direction and strength of associations. Predictors examined in fewer than three studies will be reported more succinctly. Data will be organised in tables and narrative summaries.

Patient and public involvement

Patients and members of the public will not be involved in the design, conduct, reporting or dissemination plans of this systematic review.

Ethics and dissemination

This review is based on previously published data and does not require ethical approval. Findings will be disseminated via publication in a peer-reviewed journal and presentations at relevant scientific conferences.