Participatory system dynamics modeling: advancing equity through contextualized implementation science in global health

Abstract

Implementation science is central to advancing health equity in low- and middle-income countries (LMICs), yet many interventions continue to fall short of sustained impact. These failures are not merely technical or logistical; rather, they stem from a deeper epistemic gap: the inability of dominant implementation approaches to account meaningfully for contextual complexity. By treating “context” as a static backdrop or checklist of barriers and facilitators, existing approaches overlook how dynamic interactions, institutional logics, and local power relations shape implementation outcomes over time. We argue that Participatory System Dynamics Modeling (PSD) offers a transformative shift. Unlike conventional methods, PSD integrates diverse sources of evidence, including administrative data, prior research, expert insight, and lived experiences, into a logically consistent, interactive simulation model. PSD invites stakeholders to co-create these models, surfacing feedback loops, time delays, trade-offs, and unintended consequences that traditional frameworks often miss. We outline how PSD enables researchers and practitioners to move beyond linear planning and surface-level stakeholder consultation by co-creating dynamic models that reflect local complexity, power dynamics, and feedback loops. Drawing from examples across LMICs, we illustrate how PSD strengthens implementation by aligning with local realities and enabling systemic reflection. We recommend embedding PSD into health planning, research, and training infrastructures as a capacity-building pathway to advance its methodological uptake in implementation science and strengthen the global capacity to deliver health interventions that are context-responsive, system-informed, and grounded in LMIC realities.

Background

Despite substantial investments in global health, many interventions fail to address complex healthcare problems and to achieve their intended outcomes, often resulting in counterproductive outcomes or harm [1–3]. These failures are not merely due to resource constraints or technical limitations. Rather, they are rooted in colonial legacies that continue to shape how global health priorities are set, problems are framed, and solutions are imposed in low-and middle-income countries (LMICs). Typically, interventions are designed, guided by high-income country (HIC)–led approaches, generalized evidence, or donor priorities, often from a distance, by external actors [4].

This result is the application of top-down approaches that marginalize local voices, exclude Indigenous knowledge, and undermine ownership and long-term sustainability. In addition, current implementation approaches lack a comprehensive understanding of the complex, multi-dimensional, and multi-level context in which interventions are delivered [2, 5]. Context is a dynamic and interactive component in implementation endeavors that both actively shapes and is being shaped by the intervention, implementation, and individuals in the context [6].

In such complex contexts, interactions between interventions, implementation strategies, contextual factors, and health system actors are nonlinear, unpredictable, and influenced by multiple interdependent factors, making outcomes difficult to anticipate. Successful implementation requires intervention and implementation strategies that are adaptive and context sensitive. Prior work has increasingly recognized the value of integrating systems thinking with implementation science to better account for dynamic, multi-level contextual factors that influence implementation success [7–11].

While these approaches have shown promise in various settings, there is a particular need to adopt and adapt such methods in LMICs. In these settings, critical contextual factors, such as socio-political dynamics, cultural logic, institutional structures, and informal health practices, are frequently overlooked in intervention design and implementation strategy selection. This neglect contributes to a persistent disconnect between the design of an intervention and the lived realities of the communities it intends to serve. For example, in sub-Saharan Africa, insecticide-treated mosquito nets distributed to combat malaria were often repurposed for fishing or agricultural use. While viewed as misuse by implementers, this behavior reflected unmet economic needs and unintended trade-offs [12]. Similarly, in Uganda, the U.S.–funded Abstinence, Being Faithful, Using Condoms (ABC) strategy for HIV prevention, emphasizing abstinence, fidelity, and condom use, conflicted with prevailing gender dynamics and youth behaviors, limiting uptake and marginalizing more contextually appropriate, community-driven solutions [13].

Reflecting on these shortcomings, recent calls have emphasized the need for context-sensitive methodological approaches and guidance on how to operationalize them within the complex realities of LMICs [14, 15]. In response, we argue that insights from systems science applications in implementation science can be extended to LMICs with an explicit focus on equity. Among these approaches, PSD offers a particularly promising and practical way forward to reframe implementation science in global health. PSD is an engaged research methodology for creating analog or computer-based models of complex systems, such as socio–environmental systems [16]. By centering local knowledge and promoting inclusive, systems-based co-design, PSD shifts implementation efforts toward approaches that recognize the complexity of context and empower communities to raise their voices. In doing so, PSD moves beyond one-size-fits-all solutions toward equity-focused, culturally resonant, co-designed interventions and implementation strategies, enhancing the effectiveness, sustainability, and impact of global health initiatives [17].

PSD: a promising approach to successful implementation

PSD is a valuable tool within the broader framework of systems thinking. While systems thinking emphasizes understanding the interconnectedness and dynamic relationships within a system, participatory modeling provides a practical way to apply these principles by engaging stakeholders who collaboratively explore and understand the complexity of the system [18, 19]. Systems dynamics triangulates stakeholder expertise, empirical data, and computational simulations to visualize feedback loops, delays, and constraints often invisible to conventional implementation planning [20].

PSD involves a sequence of structured stages designed to combine participatory engagement with canonical system dynamics practices [20]. Stage 1: preparation and convening, which includes establishing neutral facilitation and safe spaces, and intentionally recruiting diverse participants with lived and professional experience to ensure balanced power, inclusivity, and multiple perspectives. Stage 2: joint problem framing, where stakeholders collaboratively define system boundaries, priorities, and assumptions of how the behavior that unfolds and contributes to the problem in their system through interviews and workshops. Stage 3: model developing, where participants co-create causal loop diagrams (CLDs), which are simple visual maps that show how variables influence each other through reinforcing or balancing feedback, and then translate these into stock-and-flow diagrams. These are quantitative structures that distinguish accumulations from rates of change and allow simulation of system behavior over time to make system dynamics behavior explicit. Stage 4: robustness testing, combining model validation (checking whether the model structure and behavior are consistent with real-world knowledge, data, and expert judgment), scenario testing (experimenting with plausible policy or environmental scenarios to explore system behavior under uncertainty), and sensitivity analysis (testing how results change when key parameters vary to identify leverage points and assess model stability) [21]. Stage 5: implementation and learning, where insights are applied in practice through small tests of change, with ongoing reflection and adaptive updates to both the model and the intervention as context evolves [22].

PSD is particularly suitable for implementation planning and practice in LMICs, where resources are limited and funds are misaligned or wasted due to failed implementation, making careful, system-informed use of resources essential. PSD feeds all phases of an implementation science project to ensure interventions and implementation strategies are both contextually fitting and equity-centered (Fig. 1): (1) stakeholder engagement: PSD moves beyond traditional stakeholder involvement, as it facilitates iterative, inclusive participation and builds a shared understanding of the problem by including lived experiences and addressing structural inequalities. PSD has been shown to strengthen stakeholders’ relationships, foster trust, communication, and buy-in [20, 22, 23]. In this way, PSD can be considered an implementation strategy in itself; (2) contextual analysis: by involving those most affected, PSD helps to identify multilevel contextual factors as well as complex interactions and feedback loops that drive system behavior and influence implementation [24–26]. By integrating lived experiences and equity considerations, PSD reveals mechanisms that drive disparities, offering a deeper understanding of context than conventional analytic methods; (3) intervention/implementation strategy development: building on stakeholder insights and contextual analysis, PSD pinpoints leverage points where interventions can generate substantial and equitable change [26–28]. Scenario analysis and sensitivity testing enable stakeholders to anticipate potential synergies, risks, and unintended consequences before implementation [26, 29]. This reduces financial and operational risks of early rollout while ensuring strategies remain responsive to context and equity concerns; (4) evaluation: PSD strengthens evaluation by providing explicit, stakeholder-informed models of the mechanisms of action underlying interventions and implementation strategies, and by simulating how context adapts in response [26, 30]. This improves understanding of why strategies succeed or fail, and whether equity goals are being achieved; (5) Sustainment: PSD promotes sustainment by embedding both the model and participatory processes within local systems. By building local capacity to apply and adapt the model, stakeholders can continue monitoring contextual changes, equity impacts, and anticipate unintended consequences over time [26, 29]. Governance structures further ensure that equity remains central as contexts evolve, fostering resilience, adaptability, and long-term equity in implementation efforts.

Fig. 1.

Integration of participatory system dynamics modeling (PSD) stages across implementation science project phases

Note. The central cycle represents the iterative process of PSD, encompassing five key stages. The square boxes surrounding the pentagon depict the phases of an implementation science project. The heptagon illustrates how PSD supports and informs the implementation science process

Advancing contextual understanding through PSD

To advance the understanding of implementation contexts, implementation efforts should begin with a rigorous contextual analysis grounded in the three pillars: (1) empirical evidence, including scientific evidence, local data and information, end-user knowledge, experiences, and preferences; (2) stakeholder engagement, i.e., engagement of local actors across sectors and levels; and (3) the use of relevant theories, models, and frameworks (TMFs) [31, 32]. Although the evidence based on LMICs’ specific contextual factors is growing, critical gaps remain in addressing complex interactions in LMICs contexts. These include competing stakeholder interests (e.g., researchers, funders, national governments), limited resources, governance instability, workforce shortages, and infrastructure fragility [33, 34]. These factors are also underrepresented in widely used TMFs, originally developed for the Global North [4, 35]. Existing TMFs are increasingly being refined to improve their applicability in LMICs, which often involves adding or adapting context-specific constructs. Examples include adaptations of the Consolidated Framework for Implementation Research (CFIR) [36] or the Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM) [37]. In addition, new TMFs have been developed specifically for use in LMICs such as the Context Assessment for Community Health (COACH) [38]. These developments represent meaningful progress; however, broader application of these TMFs in real-world practice is still needed to evaluate their usefulness and identify any further adaptations that may be required.

PSD offers a methodological approach to address these limitations. By actively involving local stakeholders in the co-creation of dynamic models that identify, prioritize, and map locally relevant contextual factors, PSD fosters a deeper, system-oriented understanding of context and how its elements interact within the health system. Unlike linear or static frameworks, systems modeling captures feedback loops, time delays, and cross-level influences, features essential for reflecting the complexity of context in LMIC settings. PSD thus complements existing TMFs by surfacing additional, context-specific factors and relationships that may be overlooked in traditional frameworks, creating opportunities for further TMF development and refinement. In turn, TMFs can guide the elicitation and structuring of variables during PSD workshops in order to develop CLDs, thereby enabling a mutually reinforcing integration between theory-driven and participatory-driven approaches [28, 30, 39]. Further, PSD allows the monitoring of contextual dynamics over time, supporting adaptive implementation strategies.

Supporting equity through PSD

Equity in global health requires more than expanding access to services; it demands that diverse perspectives shape the very processes through which problems are defined and solutions are developed. PSD addresses persistent structural and contextual inequities by ensuring that decision-making power and knowledge creation are shared across all stakeholders, particularly those most affected by health challenges. Through PSD, local and Indigenous perspectives are embedded alongside technical expertise, enabling solutions that are not only contextually relevant but also fairly distributed in their benefits and burdens. By fostering inclusive dialogue among community members, frontline healthcare workers, and policymakers, PSD promotes collaborative governance, mutual trust, and shared accountability, creating pathways for more equitable health outcomes.

The developed system models help to recognize feedback loops, leverage points, and unintended consequences, which are often missed in top-down planning. Simulation models can be developed to enable stakeholders to explore and reflect on various scenarios, strengthening local capacity to respond to evolving challenges. In doing so, PSD improves the contextual fit of interventions and implementation strategies to align with LMICs’ realities, where health systems are shaped by complexity, constrained resources, and informality. PSD is particularly valuable in data-scarce environments, where it draws on stakeholder insights to develop a deep understanding of implementation contexts without relying solely on large datasets. More broadly, PSD supports a paradigm shift in implementation science, from delivery-focused, externally guided practice to one grounded in relational, locally led processes. Importantly, it embeds ethical reflection into implementation science by questioning the purpose, beneficiaries, and assumptions of interventions, while shifting power by giving local actors agency over the “why,” “what,” and “how” of change.

Case studies for applications of PSD in implementation science for global health

There is a growing body of practice using PSD in LMICs, with successful cases demonstrating their ability to support implementation by capturing complexity and engaging stakeholders in intervention and implementation strategy co-design. A recent systematic review of 10 studies from both HIC and LMIC countries found that PSD in community prevention strengthened community engagement and leadership, strategic planning, collaboration, development of shared visions and goals, as well as local ownership [40].

Complementing these findings, there are individual case studies that employed a range of modeling modalities, from CLDs in Sudan to full system dynamics simulations in Nigeria. In Sudan, continuity of care after hospital discharge was examined through stakeholder interviews, the development of a CLDs, and was followed by validation by those same stakeholders. The participatory process revealed how poor communication and unclear accountability between hospitals and primary care facilities created a reinforcing cycle of missed follow-ups. Weak referral mechanisms, fragmented patient information systems, and undefined roles meant that responsibility for post-discharge care was often ambiguous. This contributed to delays, patient drop-offs, and worsening chronic disease outcomes, which increased the burden on already overstretched hospital services [41]. In Nigeria, community-based participatory research was paired with participatory system dynamics to evaluate and strengthen retention in primary-care hypertension services. Diverse stakeholders (patients, community health extension workers, nurses, pharmacists, physicians, administrators, and policymakers) were engaged through two workshops to co-develop CLDs that mapped how system factors interact to drive retention, including insurance coverage, medication costs and supply, clinic accessibility and staffing, patient awareness, and trust in providers. The CLDs were iteratively refined using Photovoice insights and evidence from a systematic review, then translated into a stock-and-flow simulation model calibrated with local program and registry data. This quantitative model allowed stakeholders to test “what-if” scenarios for single and bundled implementation strategies—such as staff training, Information and Communication Technology (ICT) enabled reminders, community support, resource strengthening, workflow optimization, and patient education—and to examine how contextual constraints shape retention trajectories over 12–36 months, highlighting integrated packages as the most effective path for sustained improvement. The simulations showed that although staff training was the strongest single lever, the highest-performing scenario was a fully integrated multi-level intervention package, which maintained retention rates above 70% at 12 months and outperformed all single-strategy alternatives [39].

Another illustrative case study from Peru demonstrates how PSD not only contributed to more context-sensitive, equitable, and sustainable health system transformation in the Proyecto Precancer (PP) project, but also enhanced implementation effectiveness [42]. Gilman and colleagues adapted and implemented a cervical cancer screen-and-treat program targeting women aged 30–49 years. The project initially faced challenges, including fragmented screening systems, low screening coverage (only 31% with Pap smear or visual inspection with acetic acid [VIA]), high loss to follow-up, and significant inequities affecting rural and underserved populations. To address these barriers, PP used PSD, including group model-building workshops with stakeholders across all health system levels. This collaborative, participatory approach facilitated a nuanced understanding of the complex health system dynamics and social determinants influencing cervical cancer screening. Key adaptations included developing a hybrid electronic-paper patient registry, task-shifting treatments to nurse-midwives, and introducing patient self-collection for HPV testing. These measures de-emphasized centralized specialist care and targeted fragmentation, duplication, and delays in the care system, such as limited availability of triage appointments that created bottlenecks in the continuum of care. Further data to examine the underlying mechanisms of these adaptations will be collected by the research team using the Normalization Process Theory. As a result, the intervention increased service outcomes such as accessibility and enhanced equity by centering local decision-making and health system ownership, aligning with decolonial principles by shifting power to local stakeholders. In terms of implementation outcomes, the intervention achieved rapid adoption in 14 of 17 health facilities, surpassed World Health Organization screening coverage targets within 6 months, and improved completion of care from 30% under Pap smear or VIA–based screening to 67% following the introduction of HPV-based screening [42].

Challenges and practical implications for operationalizing PSD

Despite the potential advantages of PSD for implementation science, it is important to recognize challenges that may arise in its application. Given the diversity of regional, cultural, political, economic, and health system contexts—not only in LMICs—researchers must carefully adapt their PSD design to ensure the approach is both feasible and meaningful in their specific setting.

Stakeholder diversity, power dynamics, and trust barriers

In LMICs, stakeholder groups are often highly diverse, including community members, local leaders, healthcare workers, non-governmental organizations (NGOs), policymakers, and Indigenous groups. Each group brings distinct worldviews, values, knowledge, and degrees of power, making the inclusion of all perspectives both essential and challenging. Balancing broad representation with the need to keep models manageable can be difficult because the diversity of input can quickly create complex models that are hard to translate into actionable outputs [43, 44].

While PSD has the potential to advance equity and democratize implementation science by engaging these diverse actors, it is not without risks. Research on patient engagement demonstrates that tokenistic participation, where unequal power, limited impact, or ulterior motives dominate, may leave community members with little real influence and produce no meaningful change. Inclusive governance literature, therefore, argues that meaningful co-creation must go beyond ‘invited participation’ or tokenism by building on local and Indigenous knowledge, ensuring equal opportunities for participation, offering context-specific capacity building, and recognizing participants’ rights and autonomy [43].

In many LMICs, historical colonial legacies and persistent structural inequities further undermine trust in researchers and government actors, restricting open dialogue and amplifying power asymmetries. These imbalances can reduce willingness to participate, silence less powerful voices during modeling sessions, and ultimately distort the construction of the model itself [44]. To mitigate these risks, PSD efforts should prioritize long-term trust building through sustained community engagement and employ culturally sensitive facilitation strategies, including co-facilitation with trusted local partners or stakeholder engagement champions [45]. Critically, reflection on power imbalances must inform not only in-session facilitation but also the design of the sessions themselves, for example, decisions about participant selection, group composition, discussion structures, and allocation of facilitation roles [44]. Additional safeguards—such as rotating facilitation, co-designed agendas, fair compensation, protected spaces for dissent, and shared ownership agreements—can further reduce unequal influence and prevent extractive practices. Providing context-appropriate capacity-building opportunities and using accessible, non-technical language can further enhance meaningful engagement and reduce the risk of symbolic participation.

Capacity and resource constraints

Although PSD can support more efficient, better-aligned implementation in resource-constrained settings, its operationalization demands substantial upfront investments in time, facilitation, and technical expertise, which may be justified by reduced risks of implementation failure and inefficient scale-up [46]. Effective facilitation, stakeholder engagement, and iterative model refinement demand methodological familiarity that is not always present or requires substantial time investment to build those capacities [17]. Simulation modeling itself calls for considerable technical expertise, time, and sustained commitment, as model development, calibration, and interpretation extend well beyond initial PSD workshops.

Another challenge is that PSD typically involves multiple workshops, skilled facilitation, and follow-up activities—requirements that often conflict with limited stakeholder availability, constrained budgets, and competing priorities in LMIC settings [17]. Therefore, the design of PSD should always account for practical considerations, including time, resources, and feasibility as well. In resource-limited LMIC settings, a feasible way forward may be to plan a “minimum viable” PSD cycle. This could, for example, be operationalized through two iterative workshops, structured around reflexive and equity-protective routines to ensure meaningful participation while remaining manageable. The first workshop convenes diverse stakeholders in a neutral space to jointly clarify the implementation problem, map key drivers, and surface system assumptions through CLDs or stock and flow modeling. This stage embeds structured reflection activities such as power mapping, asking whose voice is missing, and prompting reflection on who benefits or may be burdened in order to safeguard against externally driven framing and ensure lived experience shapes system boundaries and priorities. Between sessions, facilitators refine the draft model, integrate empirical data, examine initial assumptions, and prepare illustrative scenario probes for stakeholder review. The second workshop validates and amends the map or model, links insights to meaningful indicators and implementation pathways, and uses scenario comparisons to explore feasibility constraints and potential unintended consequences. Together, these cycles create a shared decision logic, enable rapid learning before resource-intensive rollout, and position stakeholders rather than external experts as stewards of ongoing sense-making and adaptation. Implementation then proceeds with small tests of change and regular community-guided reflection, allowing the model and strategies to evolve as context shifts and equity considerations surface.

Fragile contexts and logistical challenges

Fragile contexts present significant logistical and methodological challenges for applying PSD in LMICs. These settings often experience rapid and unpredictable changes—driven by conflict, displacement, or weak governance—that undermine the stability required for sustained engagement. Political instability may lead to high turnover among policymakers and leaders, as well as overburdened practitioners with limited time and resources to dedicate to participatory research. Participants may also face opportunity costs, travel difficulties, safety concerns, and other practical barriers that impede sustained involvement. Fragility simultaneously exacerbates issues of inclusive participation and power imbalance, making deliberate facilitation essential to ensure that marginalized voices are heard and that trusting relationships can be built.

To address these constraints, engagement schedules should be designed flexibly, incorporating virtual tools where feasible and providing appropriate compensation or reimbursement of expenses to support continued participation. Leveraging existing community gatherings to embed modeling sessions can further reduce participant burden. Moreover, as highlighted by Jackson et al., local stakeholder engagement champions play a crucial role in cultivating trustful, context-sensitive relationships that foster sustained participation and responsiveness to emerging needs throughout the project [45].

Especially in fragile settings, traditional data systems are often fragmented or missing altogether, necessitating alternative approaches to develop CLDs or to assess the validity and robustness of developed models. In this context, triangulation of primary and secondary data sources plays a role in interviews or stakeholder workshops, whereas secondary data can encompass program reports, health surveys, administrative records, published literature, and policy documents [30]. Integrating these diverse data streams enables the identification of convergent patterns that strengthen confidence in underlying causal relationships.

Recommendations to enhance the use of PSD in implementation science for global health

PSD in LMICs remains under-supported; there is a lack of funding, training, and institutional infrastructure to embed such approaches into health system planning. As global funding priorities shift, there is an opportunity to elevate PSD as a core strategy. This shift opens the door for PSD to support localization and reduce reliance on externally driven implementation agendas. To support this, ministries can pilot participatory modeling dockets during the policy development cycle. These dockets could include short-format stakeholder workshops aligned with planning timelines, with results feeding directly into priority-setting, intervention design, or monitoring frameworks. PSD’s capacity to simulate interventions before rollout, surface unintended consequences, and save time and resources makes it an essential safeguard in environments where implementation failures are costly and sustainability is critical.

To unlock this potential, stakeholders across policy, education, research, and funding must act decisively to embed PSD into health system planning and practice.

At the policy level, governments should invest in participatory system approaches as part of national health planning strategies. Systems thinking methods have demonstrated success, even in fragile settings [47]. Further, there is a need to invest in context-responsive infrastructure, such as regional implementation labs, to support local innovation and learning in PSD practice.

At the education level, while PSD can ultimately save time and resources by preventing implementation failures, it requires upfront investment in capacity building, iterative learning, and institutional readiness. This transition can be eased by supporting phased implementation, offering flexible seed funding for early engagement, and embedding PSD skills within implementation teams as tools for analyzing complexity in context. This could include structured opportunities to co-facilitate workshops, develop and document CLDs, and apply modeling insights to real implementation problems. The framework for implementing research competencies in LMICs emphasizes the relevance of contextual analysis and stakeholder engagement. Integrating PSD into training programs can strengthen these implementation science competencies and support transformative change. To ensure the broader and sustainable application of PSD, diverse training opportunities, such as courses, workshops, and mentorship programs, are needed. Training should focus on systems thinking, group facilitation, equity-centered engagement, and modeling literacy (including causal loop and stock-and-flow diagramming). These may include structured training programs, regional workshops, online courses, and mentorship schemes tailored to various stages of professional development. In LMICs, emphasis should be placed on integrating local knowledge, navigating power dynamics, and building facilitation skills to support locally led implementation. Institutions such as universities, policy think tanks, and public health institutes can anchor training efforts through formal curricula, short courses, and mentorship programs. Identifying regional champions and investing in South–South collaboration will be essential to foster communities of practice and reduce reliance on external consultants. Furthermore, there is a need to develop decolonized, regionally relevant training programs that align with national health system priorities and embed systems thinking as a core competency, as exemplified by efforts to define these competencies within masters of Science (MSc) in Global Health Systems curriculum [48].

At the research level, implementation project teams should move from interdisciplinary to transdisciplinary research teams that include local researchers, community members, and practitioners to collaboratively plan, implement, and evaluate implementation projects. The application and evaluation of PSD in LMICs needs to be documented and published to build the evidence base and share learning.

For funders, the call is to prioritize support for LMIC-led projects where participatory approaches are a core criterion, and to support the iterative process nature of PSD. Seed grants or planning-phase supplements can enable stakeholder convening and model co-creation, especially in contexts where such processes are novel or under-resourced. Donors need to revise evaluation frameworks to reflect the iterative process of systems adaptability and community engagement. Across all levels, there is an urgent need to use resources wisely, decolonize, and contextualize global health interventions. It is time to reimagine how we build capacities that are fit-for-purpose and fit-for-practice to lead health systems. Embedding PSD into implementation science fosters decolonized equity-anchored practices that are grounded in local leadership, community priorities, and systemic understanding.

Conclusions

Implementation science is evolving to address the complexity and inequity of the world it aims to transform. As donor priorities shift inward and the limitations of externally driven models become more apparent, the need for context-led, participatory approaches becomes increasingly evident and timely. Integrating PSD into implementation science provides a tangible way to challenge unjust paradigms and promote methods grounded in mutual learning, shared power, and local knowledge. This is not merely about adding tools to an existing toolbox; it is about transforming the very architecture of how we understand, plan, and evaluate change. The future of implementation in LMICs will not be secured through speed or scale alone but through humility, alignment with lived realities, and enduring partnerships. To move forward, we must advocate for and commit to this principle: no implementation without a participatory, systemic understanding of the implementation context.

Abbreviations

ABC

Abstinence, Being Faithful, Using Condoms

CLDs

Causal Loop Diagrams

CFIR

Consolidated Framework for Implementation Research

COACH

Context Assessment for Community Health

HIC

High-income country

ICT

Information and Communication Technology

LMICs

Low- and middle-income countries

MSc

Master of Science

NGOs

Non-Governmental Organizations

PSD

Participatory System Dynamics Modeling

PP

Proyecto Precancer

RE-AIM

Reach, Effectiveness, Adoption, Implementation, Maintenance

TMFs’

Theories, Models, and Frameworks

Authors’ contributions

AM conceptualized the perspective, building on ideas developed through discussion with AA, JM. JM conducted the literature review. All authors wrote sections of the manuscript and contributed to subsequent drafts. All authors read and approved the final manuscript.

Funding

Not applicable.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.