Abstract
Background
Adolescents and young people living with human immunodeficiency virus (HIV), hereafter AYPLHIV, experience depression, anxiety, behavioural difficulties, and psychosocial distress that may undermine antiretroviral therapy (ART) adherence and long-term HIV outcomes. In Zambia, mental health services remain fragmented, centralised, and weakly integrated with HIV care. This modified Delphi study sought expert consensus on contextually relevant system, service, and digital strategies that may strengthen mental health access for AYPLHIV.
Methods
We conducted a modified two-round Delphi study with 10 Zambia-based experts in psychiatry, psychology, nursing, HIV programme management, and digital health. Twelve experts were approached; 2 did not respond and 10 completed both rounds. Round 1 collected open-ended responses on barriers, facilitators, and digital strategies. Responses were coded by the lead author, reviewed by all co-authors, documented in an audit trail, and synthesised into 29 statements. An anonymised Round 1 summary was shared before Round 2. In Round 2, experts rated each statement using a five-point Likert scale. The prespecified consensus criterion was at least 70% agreement (ratings of 4 or 5 divided by all 10 respondents) and interquartile range (IQR) of 1 or less.
Results
Of 29 statements, 24 met the prespecified consensus criterion, 4 showed high agreement with moderate dispersion, and 1 did not reach consensus. Participating experts rated stigma, poverty and transport costs, low mental health literacy, limited decentralisation, weak HIV-mental health integration, and workforce shortages as important constraints. Round 1 responses additionally raised harmful gender norms and referral-navigation challenges as implementation themes. Consensus-prioritised strategies included decentralising services, strengthening youth-friendly and confidential counselling, awareness campaigns, peer advocates, and complementary digital tools with data protection, low-data or offline functionality, multilingual and audio-visual content, and cultural adaptation.
Conclusions
This study identified expert-derived consensus among participating Zambia-based experts on barriers, facilitators, and potential complementary digital strategies for improving mental health access for AYPLHIV. Because AYPLHIV, caregivers, frontline ART providers, peer supporters, and community actors were not Delphi panellists, findings should be interpreted as expert consensus rather than service-user consensus. These findings may inform future co-design with young people and frontline providers, implementation research, and policy discussions.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12982-026-03060-0.
Keywords: Adolescents and young people living with HIV, Mental health services, Modified Delphi study, Digital health, HIV care integration, Zambia
Introduction
Adolescents and young people living with human immunodeficiency virus (HIV), hereafter referred to as AYPLHIV, experience a high burden of common mental disorders, including depression, anxiety, and behavioural difficulties [1–3]. These conditions are associated with poorer antiretroviral therapy (ART) adherence, unsuppressed viral load, and increased risk of loss to follow-up, undermining long-term HIV outcomes [4–6]. In Zambia, adolescents living with HIV report substantial psychological distress, and symptoms of depression and anxiety are linked to suboptimal ART adherence [5].
Despite this burden, mental health services in Zambia remain highly centralised and under-resourced, with few specialist providers, limited integration into HIV programmes, and marked urban-rural inequities in access to care [7]. Recent policy developments, including the Mental Health Act of 2019 and the Zambia Digital Health Strategy 2022–2026, signal national commitment to strengthening rights-based, integrated mental health care and leveraging digital tools for service delivery [8, 9]. However, implementation at district and community levels has been slow, and youth-focused mental health services remain scarce.
Digital and mobile health interventions are increasingly recognised as promising strategies to expand access to mental health support for young people in low- and middle-income countries, including through self-help tools, remote counselling, and task-sharing models [10–13]. However, recent evidence also cautions that digital mental health interventions for young people in low- and middle-income countries remain limited in number and quality, and should not be assumed to be effective, equitable, acceptable, or sustainable without further testing [14]. There is limited locally generated evidence on which digital and service-delivery strategies should be prioritised for AYPLHIV in Zambia and how they should be integrated with existing HIV service platforms.
The need for a Delphi process was informed by three considerations. First, available literature and national policy documents showed that mental health services for AYPLHIV in Zambia are under-resourced and weakly integrated into HIV care, but there is limited local evidence on which service-delivery and digital strategies should be prioritised. Second, the Delphi process was identified as a mechanism to engage experts in mental health, HIV programming, and digital health to support development of a contextually appropriate mobile health (mHealth) approach. Third, because the study sought consensus on programme and implementation priorities rather than only descriptive accounts, a Delphi approach was more appropriate than a conventional qualitative study alone. AYPLHIV, caregivers, frontline ART providers, peer supporters, and community actors were reserved for later validation and co-design because this first Delphi stage was intended to define expert-derived priorities that could subsequently be tested against lived experience and implementation realities.
This modified Delphi study, therefore, aimed to identify areas of expert consensus on the key barriers and facilitators affecting AYPLHIV access to mental health services in Zambia, and to prioritise potential digital strategies to strengthen youth mental health service delivery within HIV programmes.
Methods
Study design
We conducted a modified two-round Delphi study to obtain expert consensus on barriers, facilitators, and digital strategies to improve mental health service access for AYPLHIV in Zambia. A modified Delphi approach was selected because local implementation evidence on integrated mental health and digital strategies for this population is limited, and the study required structured prioritisation across experts from clinical, HIV programme, policy, and digital health backgrounds. Compared with a stakeholder workshop or nominal group technique, the Delphi approach allowed experts to contribute independently and anonymously, reducing the risk of dominance by senior or more vocal participants. The ≥ 70% agreement criterion and the two-round stopping rule were specified as part of the modified Delphi design before analysis; items that did not meet the prespecified consensus criterion were retained as areas requiring further validation rather than being re-rated.
The Delphi rounds were conducted between September and November 2025. Round 1 generated qualitative input through open-ended questions, and Round 2 quantified agreement with the resulting statements using structured Likert-scale ratings. The study was reported using the Conducting and Reporting Delphi Studies (CREDES) guidance [15]. Because Round 1 involved open-ended qualitative responses, reporting of that component was additionally informed by the Consolidated Criteria for Reporting Qualitative Research (COREQ) [16]. A completed CREDES checklist is provided as Supplementary File 1.
Participants and sampling
Participants were purposively sampled to include individuals with substantial experience in mental health, HIV service delivery, adolescent or youth health, and/or digital health in Zambia. Eligibility criteria were: at least 10 years of experience in clinical, programme, policy, or digital health roles relevant to mental health or HIV; current or recent involvement in adolescent or youth health programming; and willingness to participate in two Delphi rounds.
Potential experts were identified through professional networks in mental health, HIV service delivery, adolescent or youth health, digital health, and public health programming. Initial expert identification began through Ministry of Health and professional networks, after which snowball referrals were used to identify additional experts with relevant experience. The team sought balance across psychiatry, psychology, mental health nursing, HIV programme management, youth health, policy, and digital health implementation. Eligibility was verified through known professional roles, institutional positions, documented programme experience, and self-confirmation at invitation and consent. Twelve experts were approached by email. Two experts did not respond, and none formally declined. Ten experts agreed to participate, provided consent, and completed Round 1. All 10 experts completed Round 2, giving 83.3% enrolment among those approached and 100% retention between Delphi rounds. All participating experts were Zambia-based and had direct professional experience in Zambia.
To protect anonymity, panel characteristics are reported only in aggregate and by broad professional categories. The panel included expertise in psychiatry, psychology, mental health nursing, mental health coordination, HIV programme management, and digital health implementation. Participants had between 10 and 25 years of professional experience and represented national, provincial, district, academic, programme, and digital health perspectives (Table 1).
Table 1.
Expert recruitment and panel profile
| Characteristic | Summary |
|---|---|
| Experts approached | 12 |
| Did not respond | 2 |
| Formally declined | 0 |
| Enrolled and completed Round 1 | 10 |
| Completed Round 2 | 10 |
| Retention between rounds | 100% |
| Item response rate in Round 2 | 100% for all 29 statements |
| Geographic/professional scope | All experts were Zambia-based; national, provincial, district, programme, academic, and digital health perspectives were represented. |
| Professional backgrounds represented | Psychiatry, psychology, mental health nursing, mental health coordination, HIV programme management, and digital health implementation. |
| Years of experience | 10–25 years |
Data collection tools and piloting
The Round 1 open-ended questionnaire was developed specifically for this study based on the approved Doctor of Philosophy (PhD) protocol, literature on adolescent mental health and HIV, and the study objectives. The questions covered health-system barriers, socioeconomic and cultural barriers, mental health literacy, policy and service delivery, technology adoption, mobile application features, confidentiality and privacy, multimedia content, language diversity, cultural relevance, traditional coping practices, and private-sector engagement.
The Round 1 questionnaire was piloted with three experts who were not part of the final Delphi panel and who had expertise similar to that of the Delphi panelists. These pilot experts did not participate in Round 1 or Round 2 ratings. Feedback from the pilot was used to check clarity, wording, sequencing, and relevance of questions before administration to the study panel. The Round 1 questionnaire is provided as Supplementary File 2.
Round 1 qualitative item generation
In Round 1, experts completed a structured open-ended questionnaire administered electronically using KoboToolbox [ee.kobotoolbox.org]. The questionnaire was deployed on Android tablets/smartphones using the KoboCollect app v2024.3.0 and submitted to the KoboToolbox server when internet was available. They were asked to describe the main barriers preventing AYPLHIV from accessing mental health services; key facilitators or enabling factors that could improve access and quality; and opportunities, risks, and design priorities for using mHealth and other digital tools to support AYPLHIV mental health.
Round 1 responses were imported into LiGRE version 6.5.1 for coding and data management. Coding was led by the lead author and reviewed by all co-authors, with decisions documented in an audit trail. An inductive-deductive hybrid approach was used: inductive codes captured concepts emerging directly from expert responses, while deductive codes reflected the study objectives and domains of access, service delivery, and digital strategies. The co-author review focused on whether codes were plausible, non-duplicative, and traceable to the Round 1 responses. Disagreements or suggested refinements were resolved through team discussion, and an audit trail was maintained linking Round 1 themes to candidate Round 2 statements. A statement-generation matrix showing this transformation is provided as Supplementary File 5.
The research team transformed the Round 1 themes into 29 concise, single-concept statements suitable for Round 2 rating. Statements were reviewed by the authors to ensure that they were clear, non-leading, not double-barrelled, and traceable to the Round 1 themes. An anonymised aggregate summary of Round 1 themes and the resulting candidate statements was shared with participants before Round 2. Round 1 themes that were not converted into stand-alone Round 2 statements, such as referral-navigation challenges and harmful gender norms, were retained as qualitative implementation themes rather than treated as Round 2 consensus findings.
Round 2 consensus rating and analysis
In Round 2, the 29 statements were circulated to all 10 experts using a self-administered online questionnaire through KoboToolbox [ee.kobotoolbox.org]. For each statement, experts rated their level of agreement on a five-point Likert scale, where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree. All 10 experts rated all 29 statements.
For each item, we calculated the median, interquartile range (IQR), and the proportion of experts who rated the item as 4 or 5. Drawing on Delphi guidance and prior consensus studies [15, 17, 18], the prespecified consensus criterion was defined a priori as at least 70% agreement and an IQR of 1 or less. Agreement was calculated as the number of experts rating the item 4 or 5 divided by all 10 respondents. Items with at least 70% agreement but IQR greater than 1 were classified as high agreement with moderate dispersion. Items with less than 70% agreement and/or substantial dispersion were classified as no consensus. The classification “met prespecified criterion” refers only to this study-defined threshold and does not imply unanimity.
Round 2 ratings were completed independently and were anonymous to other panel members. Experts were free to agree, disagree, or select neutral responses. Round 2 responses were imported into R version 4.5.1 for analysis. No individual-level responses were shared with other experts. Because the modified Delphi design was planned as a two-round process, no third rating round was undertaken. The absence of a third round means that stability of ratings after further feedback was not assessed; this is treated as a limitation.
Anonymity and bias reduction
To reduce dominance effects and professional hierarchy, experts completed both rounds individually and did not meet as a group. Responses were coded using participant identifiers. Round 1 feedback was shared only in aggregate form and without attribution to individual panel members. Round 2 ratings were submitted independently through a self-administered questionnaire.
The study team used neutral prompts and allowed experts to provide open-ended responses and disagreement in Round 1 and to rate each Round 2 statement across the full Likert scale. No financial conflicts of interest were reported between researchers and Delphi participants. The authors acknowledge that some team members work in health, HIV, or digital health-related institutions; therefore, the use of anonymous individual responses, aggregate feedback, and team review of statement wording was intended to reduce potential influence on expert responses.
Ethical considerations
The study protocol was approved by the University of Zambia Biomedical Research Ethics Committee (UNZABREC; Ref. 5864 − 2024) and the National Health Research Authority (NHRA; Ref. NHRA-1679/04/11/2024). The research was conducted in accordance with the ethical guidelines and regulations of UNZABREC and NHRA for research involving human participants, including voluntary participation, informed consent, confidentiality, and secure handling of study data. All participants provided written informed consent before participation. Data were anonymised for analysis and reporting.
Results
Expert recruitment and retention
Twelve experts were approached. Two did not respond, and 10 enrolled and completed Round (1) All 10 experts completed Round 2, giving 100% retention between rounds. All 29 statements were rated by all 10 experts, giving a 100% item response rate in Round (2) The flow of participants and Delphi activities is shown in Fig. 1.
Fig. 1.
Modified Delphi process flow
Round 1 findings and development of candidate statements
Round 1 open-ended responses generated themes related to barriers to mental health service access, facilitators and system-strengthening opportunities, digital and mobile health strategies, and cultural and multisector considerations. Barriers included stigma, poverty, transport costs, low mental health literacy, limited trained providers, and centralisation of services. Facilitators included decentralisation, youth-friendly and confidential services, peer support, awareness campaigns, and non-governmental organisation (NGO) or faith-based support. Digital themes included confidentiality, data protection, low-data or offline functionality, multilingual content, audio-visual engagement, inclusive features, and referral pathways. Cultural and multisector themes included local storytelling, traditional coping practices, collaboration with traditional and faith leaders, and telecommunications partnerships. These themes were synthesised into 29 candidate statements for Round 2 rating (Table 2).
Table 2.
Summary of Round 1 themes and transformation into Round 2 domains
| Round 1 theme | Examples of content generated | Round 2 domain |
|---|---|---|
| Health-system barriers | Fragmented and under-resourced services; weak integration with HIV care; centralisation; workforce shortages. | Barriers to access |
| Socioeconomic and cultural barriers | Poverty, transport costs, stigma, gender norms, witchcraft beliefs, and low mental health literacy. | Barriers to access |
| Service-delivery facilitators | Decentralisation, youth-friendly and confidential care, awareness campaigns, peer advocates, non-governmental organisation (NGO) and faith-based initiatives. | Facilitators to access |
| Digital and mobile health features | Confidentiality, personal identification numbers (PINs)/biometrics, encryption, low-data or offline use, multilingual and audio-visual content, and referral-navigation needs raised during Round 1. | Digital strategies |
| Cultural and multisector implementation | Zambian values and storytelling, traditional coping, traditional and faith leader collaboration, telecom partnerships, multisector collaboration. | Cultural and multisector considerations |
Round 2 consensus results
Of the 29 statements rated in Round 2, 24 met the prespecified consensus criterion of at least 70% agreement and IQR of 1 or less. Four additional statements met the agreement criterion but showed moderate dispersion (IQR > 1), and 1 statement did not reach consensus. Table 3 presents the item-level median, IQR, agreement percentage, and classification for each statement.
Table 3.
Item-level Round 2 Delphi consensus results
| No. | Domain | Statement | Median | IQR | Agreement (ratings 4–5) | Classification |
|---|---|---|---|---|---|---|
| 1 | Barriers to access | The mental health programme in Zambia is fragmented and under-resourced. | 5 | 0.5 | 100% | Met prespecified criterion |
| 2 | Barriers to access | Integration of mental health within HIV care remains inadequate. | 5 | 1.0 | 100% | Met prespecified criterion |
| 3 | Barriers to access | Poverty and transport costs hinder access for adolescents and young people living with HIV. | 4 | 1.0 | 100% | Met prespecified criterion |
| 4 | Barriers to access | Unemployment and school dropout rates contribute to poor mental health and HIV care outcomes. | 4 | 1.0 | 100% | Met prespecified criterion |
| 5 | Barriers to access | Stigma related to HIV and mental illness prevents care-seeking. | 4 | 1.0 | 100% | Met prespecified criterion |
| 6 | Barriers to access | Cultural beliefs attributing mental illness to witchcraft reduce help-seeking. | 4 | 1.0 | 100% | Met prespecified criterion |
| 7 | Barriers to access | Lack of mental health education in schools contributes to low awareness. | 5 | 1.0 | 100% | Met prespecified criterion |
| 8 | Barriers to access | Healthcare providers are inadequately trained to manage adolescent mental health needs. | 5 | 1.0 | 100% | Met prespecified criterion |
| 9 | Barriers to access | Mental health services are limited in availability and quality, especially in rural areas. | 5 | 1.0 | 100% | Met prespecified criterion |
| 10 | Barriers to access | Implementation of the Mental Health Act 2019 remains weak due to funding shortages. | 5 | 1.5 | 80% | High agreement / moderate dispersion |
| 11 | Facilitators to access | Public health facilities provide the main source of counselling for youth. | 4 | 2.0 | 60% | No consensus |
| 12 | Facilitators to access | Peer- and NGO-led initiatives are critical sources of support. | 4 | 1.5 | 80% | High agreement / moderate dispersion |
| 13 | Facilitators to access | Faith-based organisations play an important role in community support. | 4 | 1.5 | 80% | High agreement / moderate dispersion |
| 14 | Facilitators to access | Decentralisation of services would improve access. | 5 | 0.5 | 100% | Met prespecified criterion |
| 15 | Facilitators to access | Youth-friendly, confidential, and stigma-free services are essential. | 5 | 1.0 | 100% | Met prespecified criterion |
| 16 | Facilitators to access | Awareness campaigns can reduce stigma and encourage early help-seeking. | 4 | 1.0 | 90% | Met prespecified criterion |
| 17 | Facilitators to access | Training adolescents as peer advocates enhances engagement. | 5 | 1.0 | 100% | Met prespecified criterion |
| 18 | Digital and mobile health strategies | A mobile application can complement in-person counselling. | 4 | 1.0 | 90% | Met prespecified criterion |
| 19 | Digital and mobile health strategies | The application should function offline and with low data requirements. | 5 | 1.0 | 100% | Met prespecified criterion |
| 20 | Digital and mobile health strategies | Confidentiality through personal identification numbers (PINs), biometrics, and encryption is critical for trust. | 5 | 1.0 | 100% | Met prespecified criterion |
| 21 | Digital and mobile health strategies | The application must comply with the Data Protection Act and allow anonymous use. | 5 | 1.0 | 100% | Met prespecified criterion |
| 22 | Digital and mobile health strategies | Short videos and social-media integration can enhance engagement. | 5 | 1.0 | 100% | Met prespecified criterion |
| 23 | Digital and mobile health strategies | The application should support multiple languages and audio-visual content. | 5 | 0.5 | 100% | Met prespecified criterion |
| 24 | Digital and mobile health strategies | Sign-language or gesture-based features would promote inclusivity. | 5 | 0.5 | 100% | Met prespecified criterion |
| 25 | Cultural and multisector considerations | The application should reflect Zambian values and storytelling. | 4 | 1.0 | 100% | Met prespecified criterion |
| 26 | Cultural and multisector considerations | Collaboration with traditional and faith leaders can increase acceptance. | 5 | 1.0 | 100% | Met prespecified criterion |
| 27 | Cultural and multisector considerations | The application should acknowledge traditional coping strategies. | 4 | 1.0 | 100% | Met prespecified criterion |
| 28 | Cultural and multisector considerations | Partnerships with telecommunications and technology companies can improve affordability. | 5 | 1.5 | 90% | High agreement / moderate dispersion |
| 29 | Cultural and multisector considerations | Multi-sector collaboration is essential for sustainable implementation. | 5 | 1.0 | 100% | Met prespecified criterion |
The prespecified consensus criterion was defined as at least 70% agreement and IQR of 1 or less. Agreement was calculated as the number of experts rating the item 4 or 5 divided by all 10 respondents
Barriers to mental health access
Most barrier statements met the prespecified consensus criterion. Participating experts rated mental health services in Zambia as fragmented and under-resourced, rated integration within HIV care as inadequate, and rated poverty, transport costs, stigma, low mental health literacy, and rural service limitations as constraints to care access. Harmful gender norms were identified in Round 1 open-ended responses as an implementation theme but were not included as a separate rated Round 2 item; they are, therefore, not reported as a Round 2 consensus finding.
The statement on weak implementation of the Mental Health Act reached high agreement (80%) but showed moderate dispersion (IQR = 1.5). This suggests broad support for the concern but less stability in ratings, possibly reflecting differences in experts’ roles, policy exposure, or interpretation of implementation progress.
Facilitators and enabling strategies
The prespecified consensus criterion was met for statements on decentralisation, youth-friendly and confidential services, awareness campaigns, and training adolescents as peer advocates. These priorities suggest that participating experts viewed service proximity, confidentiality, stigma reduction, and peer engagement as important implementation directions to be tested and validated with service users and frontline providers.
Two facilitator items, peer- and NGO-led initiatives and faith-based organisational support, showed high agreement but moderate dispersion. These ratings suggest that experts broadly recognised their potential value, but differed in how strongly they endorsed these sources of support. One item did not reach consensus: “Public health facilities provide the main source of counselling for youth” (median 4, IQR 2.0, 60% agreement). The mixed ratings may reflect differences in experts’ professional roles, facility exposure, or interpretation of what constitutes an adequate or main counselling source rather than definitive evidence of national variation in service availability.
Digital and mobile health strategies
All seven digital and mobile health items met the prespecified consensus criterion. Participating experts rated a mobile application as able to complement in-person counselling and rated low-data or offline functionality, strong confidentiality safeguards, compliance with Zambia’s Data Protection Act, short videos, social-media integration, multiple languages, audio-visual content, and accessibility features such as sign-language or gesture-based options as important. Optional anonymous use should be understood as an implementation-design caveat to be assessed clinically and ethically, rather than as a standalone finding about intervention feasibility or effectiveness.
Experts envisioned digital tools as complementary supports that could provide psychoeducation, coping skills, self-screening, and linkage to counsellors or peer groups. Referral-navigation challenges were raised during Round 1 as a design and implementation theme, but no separate Round 2 item specifically rated referral pathways. The strong consensus on low-data or offline functionality and confidentiality indicates that participating experts considered affordability, privacy, and trust central to any digital strategy. Accessibility features such as sign-language or gesture-based functions should be co-designed with young people with disabilities and disability inclusion stakeholders before implementation.
Cultural and multisector considerations
The prespecified consensus criterion was met for statements that digital tools should reflect Zambian values and storytelling, collaborate with traditional and faith leaders to increase acceptance, acknowledge traditional coping strategies, and be supported through multisector collaboration. Partnerships with telecommunications and technology companies showed high agreement (90%) but moderate dispersion (IQR = 1.5), indicating broad interest but less stable agreement on this implementation approach.
Together, these findings show that participating experts prioritised context-sensitive digital strategies that are confidential, culturally grounded, inclusive, and connected to broader implementation partnerships. Design features not directly rated in Round 2, including referral navigation, should be interpreted as Round 1 themes and design implications rather than as items that met the prespecified consensus criterion.
Discussion
This modified Delphi study identified expert-derived consensus among participating Zambia-based experts on barriers, facilitators, and potential complementary digital strategies for improving mental health service access for AYPLHIV. The rated barriers that met the prespecified consensus criterion, including stigma, poverty and transport costs, low mental health literacy, limited decentralisation, weak HIV-mental health integration, and workforce shortages, align with broader evidence that mental health care in low-resource settings is constrained by social stigma, financial barriers, workforce shortages, and weak integration into routine HIV and primary care platforms [7, 19, 20].
Experts endorsed digital and mobile health approaches as complementary tools for expanding access to psychosocial support, but these findings should not be interpreted as evidence that digital interventions will necessarily be effective, equitable, acceptable, or sustainable. A recent systematic review of digital mental health interventions for young people in low- and middle-income countries found that evidence remains limited and heterogeneous, with important gaps around equity, safety, usability, and long-term implementation [14]. The modified Delphi results, therefore, identify priorities for co-design and testing, not proof of intervention effectiveness.
Items that met the prespecified consensus criterion on youth-friendly, confidential services, awareness campaigns, and peer advocates resonate with studies showing that stigma, fear of disclosure, and lack of trust in providers may deter AYPLHIV from seeking care, while peer-based and adolescent-responsive approaches may improve engagement and adherence [1–6]. Integration of mental health within ART clinics emerged as an important concern through the barrier item on inadequate HIV-mental health integration and through Round 1 qualitative themes; however, specific ART-clinic integration models were not rated as distinct Round 2 implementation statements. These models should therefore be treated as priorities for co-design and implementation research rather than as established consensus recommendations.
The non-consensus and moderate-dispersion items should be interpreted cautiously. The item on public health facilities as the main source of counselling did not reach consensus, and several items showed high agreement with moderate dispersion. These patterns may reflect differences in experts’ professional roles, facility exposure, or interpretation of the statements. They do not demonstrate national or sub-national variation in service quality or availability, but they do identify topics requiring further validation and district-level assessment.
The findings may be transferable to similar low-resource settings where adolescent HIV services and mental health services are weakly integrated, but they should not be interpreted as nationally representative or directly generalisable. The panel consisted of Zambia-based experts with relevant clinical, programme, and digital health experience, which supports contextual relevance for Zambia while limiting the ability to infer consensus among all possible stakeholders.
Implementation considerations and next steps
The consensus priorities can inform practical next steps for policy, programme design, and implementation research, but they should be treated as priorities to co-design and test rather than definitive programme directives. At the service level, future work could test approaches that integrate mental health screening, basic psychosocial support, and referral navigation into ART clinics and adolescent-friendly services. At the workforce level, training and supervision models for non-specialist providers and peer advocates could be developed and evaluated. At the digital level, mobile tools could be considered as complementary mechanisms for psychoeducation, confidential access, appointment support, and navigation, but feasibility, acceptability, safety, equity, and cost should be assessed before scale-up.
Next steps should include validation of the expert-derived priorities with AYPLHIV, caregivers, frontline ART providers, peer supporters, community actors, policymakers, and programme implementers. The mHealth-related priorities should be translated into a youth-centred prototype through participatory co-design. Subsequent implementation research should assess reach, usability, acceptability, safety, equity, referral completion, cost, and integration with existing HIV and mental health service platforms.
Strengths and limitations
This study has several strengths. It used a structured modified Delphi methodology with a prespecified consensus criterion, included experts with clinical, programme, and digital health experience, and systematically transformed Round 1 qualitative input into Round 2 rating statements. The study also reports item-level medians, IQRs, agreement proportions, and classifications, improving transparency of the consensus process.
This study also has limitations. First, the panel included 10 experts, which is acceptable for Delphi studies but limits representativeness. Second, purposive and snowball recruitment may have introduced selection bias by favouring experts already engaged in mental health, HIV, or digital health programming. Third, AYPLHIV, caregivers, frontline ART providers, peer supporters, and community actors did not participate in the Delphi panel; therefore, the priorities reflect expert judgement rather than direct service-user consensus. Fourth, coding was conducted by the lead author and reviewed by co-authors rather than independently coded by multiple coders throughout, which may have introduced interpretive bias despite team review and audit trail documentation. Fifth, the study used a modified two-round Delphi and did not include a third rating round after group-level rating feedback, which may have limited further refinement of ratings. Sixth, consensus does not establish feasibility, acceptability, equity, effectiveness, or cost-effectiveness in real-world settings. The findings should, therefore, be interpreted as a preliminary expert-derived agenda requiring further validation and implementation testing.
Conclusions
This modified Delphi study identified expert-derived priorities from participating Zambia-based experts for improving mental health service access among adolescents and young people living with HIV (AYPLHIV). The findings suggest that access barriers are not limited to the availability of mental health services alone, but are shaped by interacting social, economic, cultural, health-system, and service-delivery factors. Items that met the prespecified consensus criterion emphasised stigma reduction, affordability and transport barriers, low mental health literacy, decentralisation of services, stronger HIV–mental health integration, workforce capacity, youth-friendly and confidential counselling, peer advocacy, and community-level awareness as important areas for action.
The study also highlighted the potential role of complementary digital strategies in improving service discovery, psychoeducation, linkage, and referral support for AYPLHIV. However, digital tools should not be interpreted as stand-alone substitutes for facility-based mental health care. Rather, participating experts prioritised digital approaches that are privacy-protecting, low-data or offline-capable, multilingual, audio-visual, culturally adapted, and linked to existing HIV and mental health service pathways. This points to the need for digital strategies that are embedded within the health system, responsive to young people’s realities, and supported by clear referral, safeguarding, and data-protection mechanisms.
The findings should be interpreted as a preliminary expert-derived consensus agenda rather than as service-user consensus, nationally representative evidence, or proof of feasibility, acceptability, effectiveness, equity, or sustainability. AYPLHIV, caregivers, frontline ART providers, peer supporters, and community actors were not Delphi panellists; their perspectives are therefore essential before translating these priorities into policy or implementation. Future work should validate and refine these priorities through participatory co-design with young people and frontline providers, followed by pilot testing and implementation research to assess feasibility, acceptability, safety, equity, costs, effectiveness, and sustainability in routine HIV service settings.
Overall, this study contributes locally grounded expert consensus on system, service-delivery, and digital priorities for strengthening mental health access for AYPLHIV in Zambia. These priorities can inform future co-design, programme planning, and policy dialogue, while also providing a basis for developing and evaluating youth-centred, confidential, culturally responsive, and health-system-linked mental health support strategies.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1. Completed CREDES checklist.
Supplementary Material 2. Completed CREDES checklist.
Supplementary Material 3. Round 2 Delphi rating questionnaire.
Supplementary Material 4. Aggregate item-level Round 2 results.
Supplementary Material 5. Round 1 theme-to-Round 2 statement-generation matrix.
Acknowledgements
We thank all participating mental health and HIV programme experts for sharing their time and insights. We also acknowledge the Ministry of Health mental health, HIV, and digital health team and institutional partners who provided the linkages to the experts.
Use of artificial intelligence (AI)-assisted tools
ChatGPT by OpenAI (model displayed in the interface as GPT-5.5 Pro; accessed September 2026) was used to support language editing, formatting, and organisation of draft text. The tool was not used to generate study data, code raw Round 1 responses, calculate Delphi statistics, or decide interpretation of findings. No confidential raw participant responses or identifiable participant information were entered into the tool. All AI-assisted outputs were reviewed, edited, and verified by the authors, who remain responsible for the final content, analysis, interpretations, citations, and conclusions.
Abbreviations
- AI
Artificial intelligence
- ART
Antiretroviral therapy
- AYPLHIV
Adolescents and young people living with HIV
- COREQ
Consolidated Criteria for Reporting Qualitative Research
- CREDES
Conducting and Reporting Delphi Studies
- HIV
Human immunodeficiency virus
- IQR
Interquartile range
- mHealth
Mobile health
- NGO
Non-governmental organisation
- NHRA
National Health Research Authority
- NIH
National Institutes of Health
- PhD
Doctor of Philosophy
- PIN
Personal identification number
- UNZABREC
University of Zambia Biomedical Research Ethics Committee
- WHO
World Health Organization
- ZENITH
Zambia Education Network for Implementation Science Training in Health
Author contributions
CM conceived and designed the study, coordinated data collection, conducted the analysis, and drafted the manuscript. PJC, JJF, CB, CC, ND, LM, and RNL contributed to study design refinement, interpretation of findings, and critical revision of the manuscript. All authors reviewed and approved the final manuscript.
Funding
This work was funded by the Fogarty International Center of the National Institutes of Health (NIH; Award No. 5D43TW012493) through the Zambia Education Network for Implementation Science Training in Health (ZENITH) programme. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH, the Fogarty International Center, or the United States Department of Health and Human Services.
Data availability
Aggregate item-level Round 2 results are provided in Supplementary File 4. De-identified individual-level Round 2 ratings may be available from the corresponding author upon reasonable request and subject to ethics and data-governance requirements. Round 1 qualitative free-text responses are not publicly available because the small number of experts and the specialised nature of their roles could make participants indirectly identifiable.
Declarations
Ethics approval and consent to participate
The study protocol was approved by the University of Zambia Biomedical Research Ethics Committee (UNZABREC; Ref. 5864 − 2024) and the National Health Research Authority (NHRA; Ref. NHRA-1679/04/11/2024). The research was conducted in accordance with the ethical guidelines and regulations of UNZABREC and NHRA for research involving human participants, including voluntary participation, informed consent, confidentiality, and secure handling of study data. All participants provided written informed consent before participating in the Delphi process. Participants were informed about the study purpose, voluntary participation, their right to withdraw, confidentiality protections, and data-use procedures.
Consent for publication
Not applicable. This manuscript does not contain individual-level identifiable data, images, videos, or case details requiring consent for publication.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Vreeman RC, McCoy BM, Lee S. Mental health challenges among adolescents living with HIV. J Int AIDS Soc. 2017;20(Suppl 3):21497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mellins CA, Malee KM. Understanding the mental health of youth living with perinatal HIV infection: lessons learned and current challenges. J Int AIDS Soc. 2013;16(1):18593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dessauvagie AS, Joerns-Presentati A, Napp AK, Stein DJ, Jonker D, Breet E, et al. The prevalence of mental health problems in sub-Saharan adolescents living with HIV: a systematic review. Glob Ment Health. 2020;7:e18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Olashore AA, Paruk S, Akanni OO, Tomita A, Chiliza B. Psychiatric disorders in adolescents living with HIV and association with antiretroviral therapy adherence in sub-Saharan Africa: a systematic review and meta-analysis. AIDS Behav. 2021;25(6):1711–28. [DOI] [PubMed] [Google Scholar]
- 5.Okawa S, Yasuoka J, Ishikawa N, Poudel KC, Ryu S, El-Sahn M, et al. Psychological well-being and adherence to antiretroviral therapy among adolescents living with HIV in Zambia. AIDS Care. 2018;30(6):734–42. [DOI] [PubMed] [Google Scholar]
- 6.UNICEF Eastern and Southern Africa. Mental health and antiretroviral treatment adherence among adolescents living with HIV: evidence on risk pathways and protective factors. Nairobi: UNICEF; 2021. [Google Scholar]
- 7.World Health Organization. Mental Health Atlas 2020: Zambia country profile. Geneva: World Health Organization; 2021. [Google Scholar]
- 8.Republic of Zambia. Mental Health Act, 2019 (Act No. 6 of 2019). Lusaka: Government of the Republic of Zambia; 2019. [Google Scholar]
- 9.Ministry of Health (Zambia). Zambia Digital Health Strategy 2022–2026. Lusaka: Government of the Republic of Zambia; 2022. [Google Scholar]
- 10.Naslund JA, Aschbrenner KA, Araya R, Marsch LA, Unuetzer J, Patel V, et al. Digital technology for treating and preventing mental disorders in low-income and middle-income countries: a narrative review of the literature. Lancet Psychiatry. 2017;4(6):486–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Donker T, Petrie K, Proudfoot J, Clarke J, Birch MR, Christensen H. Smartphones for smarter delivery of mental health programmes: a systematic review. J Med Internet Res. 2013;15(11):e247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Grist R, Porter J, Stallard P. Mental health mobile apps for preadolescents and adolescents: a systematic review. J Med Internet Res. 2017;19(5):e176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kola L, Abiona D, Adefolarin AO, Ben-Zeev D. Mobile phone use and acceptability for the delivery of mental health information among perinatal adolescents in Nigeria: survey study. JMIR Ment Health. 2021;8(1):e20314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Alagarajah J, Ceccolini D, Butler S. Digital mental health interventions for treating mental disorders in young people based in low-and middle-income countries: a systematic review of the literature. Camb Prisms Glob Ment Health. 2024;11:e74. 10.1017/gmh.2024.71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Juenger S, Payne SA, Brine J, Radbruch L, Brearley SG. Guidance on Conducting and REporting DElphi Studies (CREDES) in palliative care: recommendations based on a methodological systematic review. Palliat Med. 2017;31(8):684–706. [DOI] [PubMed] [Google Scholar]
- 16.Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care. 2007;19(6):349–57. [DOI] [PubMed] [Google Scholar]
- 17.Hasson F, Keeney S, McKenna H. Research guidelines for the Delphi survey technique. J Adv Nurs. 2000;32(4):1008–15. [PubMed] [Google Scholar]
- 18.Diamond IR, Grant RC, Feldman BM, Pencharz PB, Ling SC, Moore AM, et al. Defining consensus: a systematic review recommends methodologic criteria for reporting of Delphi studies. J Clin Epidemiol. 2014;67(4):401–9. [DOI] [PubMed] [Google Scholar]
- 19.World Health Organization. mhGAP Intervention Guide for Mental, Neurological and Substance Use Disorders in Non-specialised Health Settings. Version 2.0. Geneva: World Health Organization; 2016. [PubMed] [Google Scholar]
- 20.Patel V, Saxena S, Lund C, Thornicroft G, Baingana F, Bolton P, et al. The Lancet Commission on global mental health and sustainable development. Lancet. 2018;392(10157):1553–98. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material 1. Completed CREDES checklist.
Supplementary Material 2. Completed CREDES checklist.
Supplementary Material 3. Round 2 Delphi rating questionnaire.
Supplementary Material 4. Aggregate item-level Round 2 results.
Supplementary Material 5. Round 1 theme-to-Round 2 statement-generation matrix.
Data Availability Statement
Aggregate item-level Round 2 results are provided in Supplementary File 4. De-identified individual-level Round 2 ratings may be available from the corresponding author upon reasonable request and subject to ethics and data-governance requirements. Round 1 qualitative free-text responses are not publicly available because the small number of experts and the specialised nature of their roles could make participants indirectly identifiable.