Stakeholder Priorities for ART Initiation and Early Retention Interventions in Malawi: A Qualitative Study Comparing International and National Perspectives

Background New or returning ART clients are often ineligible for differentiated service delivery (DSD) models, though they are at increased risk of treatment interruption and may benefit greatly from flexible care models. Stakeholder support may limit progress on development and scale-up of interventions for this population. We qualitatively explored stakeholder perceptions of and decision-making criteria regarding DSD models for new or returning ART clients in Malawi. Methods We conducted in-depth interviews with internationally based stakeholders (from foundations, multilateral organizations, and NGOs) and Malawi-based stakeholders (from the Malawi Ministry of Health and PEPFAR implementing partners). The interviews included two think-aloud scenarios in which participants rated and described their perceptions of 1) the relative importance of five criteria (cost, effectiveness, acceptability, feasibility, and equity) in determining which interventions to implement for new or returning ART clients and 2) their relative interest in seven potential interventions (monetary incentives, nonmonetary incentives, community-based care, ongoing peer/mentor support and counseling, eHealth, facility-based interventions, and multimonth dispensing) for the same population. The interviews were completed in English via video conference and were audio-recorded. Transcriptions were coded using ATLAS.ti version 9. We examined the data using thematic content analysis and explored differences between international and national stakeholders. Results We interviewed twenty-two stakeholders between October 2021 and March 2022. Thirteen were based internationally, and nine were based in Malawi. Both groups prioritized client acceptability but diverged on other criteria: international stakeholders prioritized effectiveness, and Malawi-based stakeholders prioritized cost, feasibility, and sustainability. Both stakeholder groups were most interested in facility-based DSD models, such as multimonth dispensing and extended facility hours. Nearly all the stakeholders described person-centered care as a critical focus for any DSD model implemented. Conclusions National and international stakeholders support DSD models for new or returning ART clients. Client acceptability and long-term sustainability should be prioritized to address the concerns of nationally based stakeholders. Future studies should explore the reasons for differences in national and international stakeholders’ priorities and how to ensure that local perspectives are incorporated into funding and programmatic decisions.

international stakeholders' priorities and how to ensure that local perspectives are incorporated into funding and programmatic decisions.

Background
Promoting sustained engagement in antiretroviral therapy (ART) services is a major focus of HIV programs throughout sub-Saharan Africa [1].New or returning ART clients (de ned as those on ART < 3 months) are at increased risk of treatment interruption [2][3][4][5][6].The rst few months after (re)engagement represent a critical period.With targeted interventions, access-related barriers can be reduced, and clients can develop the problem-solving skills and external support needed to maintain ongoing engagement in care [7,8].
Differentiated service delivery models (DSDs) are a leading strategy for easing access to lifelong services for stable clients (de ned as virally suppressed and/or 6 + months post-ART initiation) [1].Examples include multimonth dispensing, community-based ART delivery, and tailored ongoing counseling and peer support.DSD models have a range of bene ts, including increased access to care [9], greater acceptability among clients [9,10], and improved or noninferior retention and viral suppression [11][12][13][14].However, new and returning ART clients have largely been excluded from DSD models [15], even though they may bene t most from convenient, private, and lower-cost services [16].Several DSD models, including incentives [17,18], peer support [19,20], community-based ART initiation [13,14], and e-health interventions [21][22][23], have been tested among this population in study settings but are only now being considered for scale and national policy [1].
Stakeholder input can improve intervention design, translation, adoption, and scale-up of successful strategies.However, stakeholder perceptions are not systematically explored, and multilevel stakeholders are often not consulted [24][25][26][27].A deeper understanding of stakeholders' priorities and constraints can also help researchers offer decision-relevant information, such as data on expected costs and outcomes for different implementation and scale-up scenarios [28,29].
To our knowledge, no published literature on stakeholder perceptions of DSD models for new or returning ART clients in sub-Saharan Africa exists.Despite its highly resource-constrained health system, Malawi has been the vanguard of innovative public health strategies in the region (such as Option B + and HIV self-testing), making it an ideal study setting.[30][31][32][33].
We qualitatively explored stakeholder perceptions of DSD models for new or returning ART clients and decision-making criteria for scaling DSD models and compared the views of internationally based and Malawi-based stakeholders.

Study Design
We conducted in-depth interviews with internationally based stakeholders (from foundations, multilateral organizations, and NGOs) and Malawi-based stakeholders (from the Ministry of Health and local implementing partners) to explore perspectives and priorities regarding interventions targeting new or returning ART clients.The study was embedded in the Identifying E cient Linkage Strategies for Men (IDEaL) randomized controlled trial, which aimed to develop and test the impact of male-tailored differentiated models of care on men's ART initiation, reinitiation, and early retention in Malawi.The trial is described elsewhere [34].

Theoretical Framework
We used the Assessing Cost-Effectiveness (ACE) approach as our theoretical framework for data collection and analysis [35].The ACE is a structured method for conducting health policy priority-setting studies.It combines rigorous economic evaluation with a qualitative assessment of other implementation factors that in uence policy adoption.Stakeholders provide guidance and feedback at every stage of the study process [35].

Population and Recruitment
We aimed to recruit policy and programmatic stakeholders who were experts on DSDs and/or who were willing to provide care for new or returning ART clients.We sampled purposively to represent different organizations, including international foundations, multilateral organizations, nongovernmental organizations (NGOs), the Malawi Ministry of Health (MoH), and local implementing partners within Malawi.Although clients and community-based organizations are important stakeholders, they were not included in the sample, as this study focused on policy and guideline decision-making.
We identi ed twenty-nine potential participants through internet searches and the research team's professional networks.We also included eight additional contacts through snowball sampling.We made two direct outreach attempts to each stakeholder by email.We continued recruitment until we reached saturation of the themes were saturated within in-depth interviews.We intentionally attempted to sample similar numbers of individuals from the international and national stakeholder categories to facilitate understanding of the similarities and differences between the two groups.

Data collection and analysis
We developed an interview guide based on the ACE framework as well as the literature on interventions to improve engagement in care among new or returning ART clients in sub-Saharan Africa [36,37].The interview guide included open-ended questions about stakeholders' HIV-related priorities, perceptions of factors in uencing noninitiation or attrition immediately following initiation, challenges and solutions for nancing and implementing relevant interventions, and desired data to inform decision-making.
Stakeholders also completed two interactive tasks.First, they rated ve scale-up decision-making criteria commonly used in the Assessing Cost-Effectiveness (ACE) approach (cost, effectiveness, equity, feasibility, acceptability) as "lower," "moderate" or "high" priority.We encouraged them to rate no more than three criteria as high priority and at least one as lower priority.We asked stakeholders to verbalize their thoughts as they completed the task, following the "think-aloud" method," for qualitative research [38].After the task was complete, we offered participants the option to rate any additional decisionmaking criteria and engaged them in retrospective re ection about the activity and the reasons for their ratings.
In accordance with the same procedure as in the previous task, the stakeholders rated the relative priority of seven ART initiation and early retention interventions.Options included monetary incentives, nonmonetary incentives, community-based care, ongoing peer/mentor support and counseling, eHealth, facility changes, multimonth dispensing, and an "other" category.To construct an average rating for each intervention, we assigned each high-priority rating ten points, each moderate-priority rating 5 points, and each lower-priority rating 0 points [39].
We conducted two pilot interviews to re ne and nalize the data collection tool.Two researchers conducted each interview in English via video conference.We recorded the interviews and transcribed them verbatim.
We developed a codebook using a priori codes informed by the literature and the theoretical framework [35], as well as inductive codes based on emergent themes [40].Using Atlas.tiv9 [41], two researchers (KP and KH) piloted the codebook with six interviews.Four researchers (KD, SH, KH, and KP) reviewed the coded transcripts, discussed discrepancies, and re ned the codebook.One researcher (KH) coded the remaining data, and other researchers performed spot checks on half of the transcripts to ensure consistency.We extracted coded texts and performed thematic content analysis [42].In the present analysis, we compared and explored differences in themes between international and national stakeholders.We prioritized themes mentioned by many participants and explored divergent views.

Participant characteristics
We conducted 22 interviews between October 2021 and March 2022-thirteen with internationally based stakeholders and nine with nationally based stakeholders (Table 1).Participants had varied levels of experience with HIV service policies, guidelines, and practice, ranging from 2-21 years of working in the HIV eld.

Decision-making criteria
Overall, intervention effectiveness and acceptability were the highest-priority decision-making criteria (Fig. 1).Approximately half of the stakeholders explained that the decision-making criteria are interrelated, making it di cult to fully distinguish them.For example, a feasible intervention would likely also be low-cost, and an equitable intervention may have greater long-term effectiveness.

Figure 1 Ratings of decision-making criteria by national and international stakeholders
Effectiveness was a high priority for nearly all international stakeholders.International stakeholders frequently described effectiveness as their rst consideration: "There's no use in implementing something that will not help us reach the expected outcome."Fewer than half of international stakeholders rated feasibility as a high priority.They explained that implementation challenges can be overcome through Both international and national stakeholders considered acceptability, especially client acceptability, to be a high priority.As one international stakeholder summarized, "If clients do not like it, then it is just going to fail, so it does not matter how cheap and easy it is to do." Healthcare worker acceptability was also considered important due to the pervasiveness of healthcare worker burnout and the di culty of implementing programs without their buy-in: I think some things might not be acceptable to healthcare workers but be super acceptable to people living with HIV or vice versa…but I would advocate that maybe if health care workers don't love [an intervention] but it is great for communities and people living with HIV, then we've got to try [it].(International Stakeholder) Some stakeholders discussed the tradeoff between equity and e ciency, explaining that reaching key populations and serving rural areas may be more expensive but still worthwhile.A couple of national stakeholders prioritized e ciency over equity: At the end of the day [it comes down to] the number of people it impacts because that will make it more cost-effective.Yes, I may want to ensure equity by having those extended clinic hours, but I'm only seeing eight men.And will eight men improve overall retention for [the] country?(National Stakeholder)

Perspectives on Interventions for New or Returning ART Clients
Overall, international and national stakeholders had similar priorities for DSD interventions (Fig. 2).They noted that multiple models are needed to address the full range of factors causing treatment interruption among new or returning ART clients.International stakeholders were more interested in ongoing peer support/counseling than were national stakeholders.The highest-rated interventions were facility changes followed by multimonth dispensing (MMD) (Table 2).
Figure 2 DSD intervention ratings by national and international stakeholders Facility E ciencies Most international and national stakeholders prioritized facility changes (such as extending service hours and reducing wait times) because they saw these changes as feasible within existing systems and infrastructure.They saw healthcare facilities as the foundation of HIV service delivery and central to client experiences of care: "We already have the facilities in place, but what we have not done is try to look at things like the client ow, opening hours.This could be easily done."(National Stakeholder) A couple of stakeholders also supported the idea of differentiating facility-based care so that more intensive services are offered to clients with the greatest needs: We already have work ows that are stripped to the bare bones.At most sites, patient consultations take only a few minutes and are not conducted by a trained health worker.There is no room to simplify any further, so the question is what we can add for patients who need it rather than what we can simplify.(National Stakeholder) Multimonth Dispensing (MMD) MMD was also prioritized by both international and national stakeholders because it is feasible, acceptable, low-cost, and frees up facility resources to support higher-risk clients.Several stakeholders emphasized that new or returning ART clients may require several visits before being given MMD to ensure that they are adequately prepared for lifelong treatment; however, many international stakeholders believe that MMD should be offered as early as possible: I assume that people look at the rst six months of ART and dread the amount of work it involves…the less they have to do, the less onerous it will seem, and we hope that will make [HIV treatment] feel like something they want to do and want to stick with.(International stakeholder)

Peer Support and Counseling
International stakeholders prioritized peer support highly, emphasizing its effectiveness.In contrast, national stakeholders focused on the high cost and human resource requirements of these programs, though they did support a new MoH program in which a facility-based counselor provides support to clients living with depression, anxiety, and substance use disorders.
Stakeholders believed that peer support must be provided in addition to interventions addressing service accessibility: "We use peers a lot… However, if you just make it easier to access services, then you will not need all of this additional support" (International stakeholder)

Community-Based Care
Both groups rated community-delivered ART as a moderate priority.Perceived bene ts included reduced client time and cost for accessing services and a reduced risk of unwanted status disclosure at facilities.Despite these bene ts, both international and national stakeholders were concerned about the human resources needed to implement community services.For this reason, several believed that community services should be limited to key populations or those who are especially ill: It is very di cult to imagine how this will be done at scale in the national program.We have to remember that this takes health workers away from the facilities [and] it takes many more nurses to see the same number of clients if you send them out in the community.So I think in a health system that is overall understaffed, it is a great luxury to send nurses to meet people in the community.(National Stakeholder)

E-Health Strategies
The majority of stakeholders rated e-Health interventions as a moderate priority.Most stakeholders described e-Health as a new, "untapped" platform that could increase the reach of existing interventions, such as peer support, health education, appointment reminders, SMS check-ins between visits, and tracing clients with missed visits.However, several believed that the technology and evidence were not yet su cient for widespread implementation in Malawi.
Stakeholders discussed both equity concerns and bene ts related to e-health interventions.More than half of the stakeholders discussed uneven access to phones and the internet as critical barriers to effective and equitable e-health interventions, particularly in resource-constrained settings such as rural Malawi.National stakeholders noted that individuals often share phones with relatives, creating a risk of unwanted disclosure.However, several stakeholders noted e-Health's unique potential to connect with harder-to-reach populations, such as youth, men, and mobile and rural populations.

Incentives
Monetary and nonmonetary incentives were the lowest-rated interventions.Stakeholders perceived incentives as highly effective but dismissed the possibility of implementing them because of their cost and complexity.Stakeholders also believed that incentives may reduce clients' intrinsic motivation for lifelong treatment: If you tell them, 'we can give you transport money,' and then we don't have it; the next time, they do not show up.However, if at the beginning you did not tell them that [they would] receive cash, they would nd their own means to come to the clinic.[Because of incentives], they develop a dependency syndrome.(National Stakeholder) Stakeholders noted that targeted incentives can improve equity by supporting poorer clients in meeting daily needs and overcoming socioeconomic barriers to care but were concerned that other clients would nd targeted incentives unfair.Some have suggested that free or low-cost nonmonetary incentives could instead be used to reward or appreciate clients for their engagement in care.
Overarching Service-Delivery Priority: Person-Centered Care Person-centered care (PCC) organically emerged as a high priority for both national and international stakeholders.Stakeholders described PCC as exible and tailored services that respond to clients' holistic needs.The three components of PPC discussed most frequently were 1) segmented care, 2) integrated care, and 3) positive and empowering interactions with healthcare workers.

Segmented Care
The majority of both international and national stakeholders expressed the view that services should be adapted to meet the needs of diverse groups, such as clients returning to care vs. stable clients, men, key populations, clients who prefer private services, and clients with psychosocial needs: "...it's not a one-sizets-all approach... what works for female sex workers may not necessarily work for MSM" (international stakeholder).A few international stakeholders believed that segmented services may reduce health system costs by limiting costlier interventions to the clients who would bene t from them the most.However, one national stakeholder emphasized that tailoring services for different populations is an "extra project" that is not government funded and requires additional training and staff time.

Integrated Services
Integrating HIV services with other healthcare was a priority for both international and national stakeholders, although several national stakeholders believed that integration would be costly and infeasible.Stakeholders described the holistic bene ts of integrated care, including reduced costs for clients (because visits are combined), reduced risk of unwanted disclosure of one's HIV status, and improved overall health outcomes: They come to the clinic, they get their ART re lled, and tomorrow they're supposed to go to the clinic to get their diabetes medication or hypertension medication, that to me is a bad idea.[If we had a] one-stop center where you get your ART re lls and your other medications, I think that would help.(National Stakeholder)

Positive and Empowering Interactions with Healthcare Workers
Positive client-healthcare worker interactions were considered key to retaining clients in care and reducing HIV-related stigma.As one national stakeholder stated, interactions with healthcare workers should be "empowering and supportive" rather than "coercive and threatening."Stakeholders believe that counseling sessions should not be "generic" but rather tailored to clients' concerns and designed to foster trust so that clients can feel comfortable discussing their barriers to treatment: What I've seen from my own experience is that most providers do not understand that we are all human beings.We have other things that we do apart from coming to the clinic to get medication.So sometimes when a client misses an appointment and comes for a re ll, he is treated like he is being punished.[If we] change the attitudes of our providers, I do not think we are going to struggle with retention or even initiation.(National Stakeholder) Approximately one-quarter of international and national stakeholders noted that the lack of support and education clients receive when initiating ART leads to treatment interruptions: If there is no time to discuss [treatment barriers] and to encourage disclosing such issues they will of course drop out because they had problems that haven't been addressed.And I think that can only be addressed if there is … more time for [clients] to fully understand [their diagnosis] and [problem-solve] with treatment supporters and family members.(National Stakeholder) Nearly all international stakeholders and a couple of national stakeholders emphasized the importance of empowering clients by giving them choices in how they receive care, such as community or facility-based care, peer support, and the choice of different facilities.A few stakeholders noted that client needs change over time, so explicitly asking about client preferences on an ongoing basis can help identify and meet evolving needs: [If] as a client I make a choice…whether it is home-based care or whether it is multimonth dispensing, that's the one that is acceptable for me and therefore I'm likely to adhere to that intervention and have improved linkage and early retention.(National Stakeholder)

Discussion
In resource-constrained settings, stakeholders make di cult trade-offs across multiple criteria (such as effectiveness, equity, budgetary and practical constraints, and political considerations) when deciding which interventions to implement and scale up.In this study, we explored how stakeholders make these decisions in the context of interventions for new or returning ART clients in Malawi.Our study suggests that both national and international stakeholders prioritize client acceptability but diverge in other areas: program effectiveness was a higher priority for international stakeholders, while ongoing costs, feasibility, and sustainability were higher priorities for national stakeholders.Despite these differences, international and national stakeholders had similar intervention preferences; they prioritized simple, low-cost, facility-based interventions that remove barriers to care, such as multimonth dispensing and extended facility hours.Most stakeholders attributed their interest in various interventions to PCC, whereby clients are provided tailored services with positive and empowering healthcare worker interactions.
Our analysis suggests notable differences in how international and national stakeholders perceive the long-term costs of interventions.International stakeholders viewed interventions as having low ongoing costs after initial investments.In contrast, national stakeholders were deeply concerned about the longterm costs and resource requirements of interventions.National stakeholders described their experiences watching promising new programs end after donors left.International stakeholders believed that factors such as cost and feasibility could be addressed through implementation strategies, a view that was not expressed by national stakeholders.
These ndings highlight differences in experiences and perceptions of both global and local history and context.There is growing consensus that external funding should be aligned with national priorities [43,44].However, transitioning from preferred narratives to actual practice may be slow.Historically, key donor funding institutions have driven the process of intervention selection and initiatives have been managed by numerous NGOs (some locally and some internationally based) rather than by national governments [45].International donor funds predominantly support vertical programs rather than infrastructure or health system strengthening efforts.
[46] A locally driven decision-making process may start with a range of intervention options selected by the Ministry of Health and community stakeholders that are then decided upon jointly.This approach would help ensure that international and national stakeholders are working together to fund programs that are country owned, sustainable, and coordinated.
Despite differences in their decision-making criteria, international and national stakeholders had similar intervention priorities, perhaps because of their ongoing discussions.Interestingly, cost and feasibility seemed to have the greatest in uence on intervention preferences, although effectiveness and acceptability received the highest ratings in the think-aloud priority-setting tasks in this study.Like in a previous qualitative study [47], high-level stakeholders favored simple interventions with minimal costs that removed structural barriers to care (e.g., extended hours, MMD) over those that were perceived as highly effective but needed additional systems and human resources (e.g., peer support and communitybased care).
Stakeholders universally agreed on the importance of PCC, despite their concerns about resource constraints and sustainability challenges.In line with ndings from previous PCC studies [48], stakeholders were more interested in the impact of PCC on health system goals (e.g., retention in care, reducing costs) than on client goals (e.g., living a full life).They viewed PCC as a strategy to reduce healthcare costs by improving the effectiveness and e ciency of programs and recognized the critical importance of client choice and tailored, respectful, holistic care in improving retention and adherence.Additional evidence on the impact of PCC and best practices in sub-Saharan Africa is urgently needed.
PCC practices were developed in high-income countries, but little quantitative evidence on PCC in LMICs exists [49].Several aspects of PCC (such as segmented and integrated care) may require additional resources to be successfully implemented and sustained in historically vertical, disease-speci c programs with scarce resources.In such contexts, innovative, low-cost strategies for offering tailored and/or integrated services may be needed.However, recent qualitative research in Malawi suggests that returning male ART clients value positive and empowering relationships with healthcare workers more than where and how ART is delivered, suggesting that key components of PCC could be taken to scale at low cost [50].
Our study has several limitations.First, the think-aloud priority-setting tasks did not fully mimic real-world situations in which stakeholders consider many nuances of a particular context.Second, there may have been social desirability bias or differences between expressed and revealed preferences.We believe this was minimal, as interviewers expressed neutrality and encouragement toward all comments and conversations were frank and casual.Third, not all stakeholders were represented in this study; critical stakeholders such as clients and community advocacy groups were not included due to the study's focus on policy and guideline decision-making.Fourth, the quantitative results of the two ranking tasks should be interpreted with caution due to the small sample size.Fifth, some ndings may not be generalizable beyond Malawi.Despite these limitations, we believe that the priorities and preferences expressed by stakeholders in this study are re ected in real-world settings.

Conclusions
We found that the top priorities of international and Malawi national stakeholders regarding DSD interventions for new or returning ART clients are effectiveness, feasibility, sustainability, and client acceptability.International stakeholders should recognize and act upon the greater priority that national stakeholders place on feasibility and sustainability.Person-centered care was emphasized by all stakeholders and should be incorporated into any intervention for new or returning ART clients.The ndings can inform HIV treatment intervention development and research.Further research is needed to understand how differing priorities affect public health discussions, decision-making, and impact and how to ensure that national and local needs are prioritized.

Declarations Ethical approval and consent to participate
The was approved by the National Science Research Committee (NHSRC) in Malawi and by the University of California-Los Angeles Institutional Review Board.All participants gave oral consent before completing an interview.No identi able information was collected.

Figures
Figures

Table 1
If [an intervention] is not feasible, it will not come close to success...it needs to be straightforward and t well into the rest of the programs and activities…and healthcare workers' capacities and capabilities.
innovation, change management, and quality improvement: "Even if an intervention doesn't seem feasible initially, if you can continually create evidence that... it's effective, then systems can be built to gure out how that [can] work."Internationalstakeholderswereless concerned about the costs of interventions.They believed that longterm costs are minimal after initial investments and noted that interventions and programs can become more e cient over time: "Just because [something] is expensive today, we might gure out a way to do it more cheaply later".Some noted that interventions can even be cost-saving in the long run: "In the longterm, [care] might be less costly if you do not need to be... spending that much time with people who are coming with advanced disease."Incontrast,nationalstakeholders acknowledged the importance of effectiveness but believed feasibility, cost, and sustainability were more important.National stakeholders emphasized the t of interventions within existing systems and human resource and equipment constraints:It is easy to nd funding for innovations, but it is very di cult to nd funding to sustain services.We know that every year the funding for HIV programming goes down...If we're talking about introducing a certain innovation, are we able to sustain this?When [a donor] is no longer there, this intervention...will die a natural death.(NationalStakeholder)