Research that rescues: investigating policy impact of humanitarian logistics and disaster supply chain management research through machine learning Open Access

Environmental degradation due to accelerated global warming, destruction of forests, growth of industrialization, increased greenhouse gas emissions, rapid urbanization and a major shift in recent weather patterns has induced a high risk of occurrence and increase in natural and man-made disasters. Analyzing of historical events and current situation, the forecasts indicate five-fold increment in the number of disasters in the upcoming 50 years of time span (Chukwuka et al., 2023). Resilient, economic, health-focused and community-centered Humanitarian Logistics and Disaster Supply Chain Management (HLDSCM) design research capable of addressing disaster preparedness, response and recovery challenges is essential (Mena, 2025) to mitigate the adverse effects of the upcoming disasters. The integration of UN Sustainable Development Goals, SDG-3 (Good Health and Well-being) and SGD-11 (Sustainable Cities and Communities) are fully linked with this research domain and supports its practical applicability by ensuring that these supply chain designs are qualified for undertaking the challenges posed (Harpring et al., 2021).

Efficient and responsive availability of critical vaccines, protective equipment and other pharmaceuticals are crucial during health emergencies including pandemics and disease outbreaks, which can be achieved through robust medical supply chain managements. It is intrinsically associated with SDG-3 objectives on universal health coverage (SDG-3.8) and epidemic preparedness (SDG-3.d.). Furthermore, continuity of care amidst disaster including maternal and child health services among others necessitates adaptive and resilient healthcare infrastructure, where SDG-11 becomes significantly relevant owing to its emphasis on disaster risk reduction and urban resilience. It indicates that HLDSCM must address disruptions and damages to the water, sanitation, power systems and transportation related infrastructures. The degree of resilience of these infrastructures, especially in the densely populated urban areas prone to recurrent disasters determines the speed, accuracy and efficacy of emergency response operations. Another key objective of SDG-11 is to ensure equitable access to the essential services, a major consideration in HLDSCM, as it prioritizes the vulnerable populations including displaced persons, refugees and people with disabilities (Rodr′ıguez-Pereira et al., 2024). Prepositioning and design of relief supply operations under disaster preparedness phase in absolute alignment with SDG-11.5 and SDG-11.b. are critical to mitigate the postdisaster impacts and enhance recovery efficacy. Incorporation of these community resilience and health-centric frameworks in sustainable HLDSCM design research and development require a coordinated and multisectoral approach (Ahmad et al., 2025).

The existing research in this field has provided several technology-based, data-driven forecasting and decentralized storage-based strategies to improve disaster relief operation’s responsiveness and align the overall supply chain managements with overarching goals of SDGs to reduce the loss of life and economic crises under disaster situations (Van Wassenhove, 2022). The literature has emphasized the development of resilient humanitarian and disaster relief supply chains as a strategic mechanism to achieve long-term sustainability in global health and economic eco-systems. Disaster management studies initially emerged in 1980s with an emphasis on minimizing life losses under severe large-scale environmental and industrial disasters. The Indian Ocean tsunami in 2004 ignited rigorous research in this domain and the researchers and practitioners incorporated robust tools for effective predisaster preparedness and postdisaster relief operations (Ahmad et al., 2025). Further disasters including earthquake in Pakistan in 2005, Myanmar’s Cyclone Nargis in 2008, Wenchuan earthquake in China in 2008, floods in Pakistan in 2008 and Japan’s tsunami in 2011 fueled to the research and development in the HLDSCM toward robust methods and policies for responsive rescue operations (Shah et al., 2022). Recently, COVID-19 pandemic has transformed this research stream by emphasizing resilience, local stakeholder engagement and the integration of technologies in relief operations. This major shift in the research reflects broader understanding of the inherent complexities and vulnerabilities in HLDSCM, leading to practical strategies for the government and nongovernment agencies responsible for carrying out relief operations (Ivanov, 2024). Figure 1 presents general disaster supply chain process structure.

Government agencies and organizations play a focal function in humanitarian and disaster relief operations by serving as the policy developers, planners, coordinators, facilitators and implementers of the regulations. Their involvement is critical to ensuring effective communication among stakeholders, emergency resource allocation and postdisaster logistical efficiency for timely and effective response. These agencies, for instance, Federal Emergency Management Agency (FEMA) in US and National Disaster Management Authority (NDMA) India, are influential in managing disaster relief efforts (Hong, 2023) by maintaining clear communication channels and addressing network coverage limitations to enhance operational efficacy as per the applied policies and SDGs.

These agencies further administer resource allocation and logistics by prepositioning essential emergency supplies and coordinating transportation to confirm a timely response during emergencies (Habib et al., 2022). The importance of developing and implementing effective logistic policies cannot be overstated, as delays in supply chain operations can lead to significant loss of life and resources (Ahmad et al., 2025). In addition to this, government agencies collaborate closely with other nongovernmental organizations (NGOs) and several international entities to expedite and streamline supplies distribution, as established during the Klang Valley floods in Malaysia. This collaboration among agencies with different aims and objectives is essential for adhering to operational policies and maximizing the effectiveness of relief efforts and ultimately strengthening the disaster response mechanisms (Ab Malik and Omar, 2024).

The said sustainable development goals (SDGs) and government policies pertaining to humanitarian logistics and disaster relief management have prompted the relief agencies and nongovernmental organizations to reorganize their practices to effectively respond to disaster situations (Dostal, 2015). This increased impact of regulatory authorities has in turn fostered a two-way relationship between research efforts and policy development which underscores the importance of continued collaboration among the key stakeholders. In response to it, the researchers and academicians have actively proposed a range of innovative methodologies, strategies and policies with the objectives of incorporating sustainability and resilience within the preparedness and relief operations. Certain research areas focused at identifying the barriers and enablers of sustainable HLDSCM (Patil et al., 2021; Tasnim et al., 2022; Anjomshoae et al., 2025), frameworks for integration of agility, flexibility and contingency planning (Zarei et al., 2019; Tukamuhabwa et al., 2024), humanitarian logistics performance measurement (Shafiq and Soratana, 2020; Agarwal et al., 2022), collaboration and partnerships management (Singh, 2024), humanitarian logistics network configuration for sustainable supply chain design (Malmir and Zobel, 2021; Jamali et al., 2022), reverse flows and circularity in humanitarian supply chain design (Jilani et al., 2018), integration of advanced digital technologies and artificial intelligence (Karuppiah et al., 2025; Corbett et al., 2022) and development of resilient and self-reliant relief projects to contribute to SDGs (Kawane et al., 2024) have gained attention of the scholars.

The translation of academic research into policy is under developed in several research fields including HLDSCM. Some qualitative methods augment mixed-methods credibility (Hendren et al., 2023), while financial technological frameworks reveal the role of innovation in sustainable outcomes reflected in policy documents (Kihombo et al., 2021). Conceptual analyses highlight the historical interplay and barriers between researchers and policymakers, offering actionable recommendations (Soare, 2013) and sociological perspectives on boundary objects illustrate co-production mechanisms in health policymaking (Bekker et al., 2010). Together, these studies underscore strategies for integrating research findings into policy processes, informing our machine learning-driven analysis of HLDSCM research impacts. While extensive literature has examined the SDGs and policy influence on the HLDSCM research (Paciarotti et al., 2021; Sentia et al., 2023; Carnero Quispe et al., 2024; Tasnim et al., 2022; Jamali et al., 2022; Anjomshoae et al., 2025), a significant gap remains in understanding the reverse relationship, specifically how research informs and shapes humanitarian logistics and disaster management policy development. The impact of published research is conventionally evaluated through bibliometric analysis by including the indicators such as citation counts and journal impact factors (Jia et al., 2025; Tahir, 2024; Tahir et al., 2024). The humanitarian logistics and disaster management is no exception, with several studies using such bibliometric analysis to assess the research impact through citation-based metrics and other related indicators. Furthermore, the existing literature has encompassed a broad spectrum of subtopics within HLDSCM research, and have provided insights into the overall intellectual landscape of the field. For instance, big data and its applications (Akter and Wamba, 2019), collaboration, citizen partnership and risk perception (Oh and Lee, 2020), research trends in pre, during and postera of COVID-19 (Rahman et al., 2022), coevolution of social media and disaster management themes (Fauzi, 2023), resource management in disaster relief operations (Geng et al., 2024) and long-term sustainability in humanitarian supply chains (Anjomshoae et al., 2025) have been extensively explored through bibliometric analysis. Despite the usefulness of bibliometric indicators in assessing the research influence, these indicators fall short in capturing the broader societal impact of the research – particularly its role in shaping HLDSCM policy frameworks. The actual measure of societal impact lies in how research is referenced and integrated into policymaking and regulatory decisions around the globe. To the author’s knowledge, no existing studies have explored this critical dimension of research. This gap is not unique to the HLDSCM only, it is uncommon across numerous fields where the interrelationship between academic research and real-world policymaking largely remains unexplored. This oversight presents a significant opportunity to explore the impact of published research on the development of policy frameworks in this field. By deploying this innovative methodological framework, this research bridges this gap by offering a systematic approach to analyzing how research informs and influences policy decisions in the dynamic and high-stakes field of HLDSCM.

Research influence on policy is not linear, and “impact” can occur through multiple pathways. We therefore treat policy-document citations as one observable trace of research visibility in policy processes, while recognizing that citations do not capture all influence (for instance, advisory committees, informal networks and confidential procurement). To interpret why some HLDSCM research is taken up while other work is not, we draw on two complementary bodies of scholarship:

1. policy studies on how agendas and decisions evolve; and

2. science–policy interface research on how knowledge is translated and stabilized for use in governing.

First, policy studies emphasize that research can shape policy in different ways. Evidence may be used instrumentally (directly informing a choice of policy option), conceptually (shaping how actors understand problems and solutions) or symbolically/politically (legitimizing a preselected position) (Jabali et al., 2024). This implies that policy visibility is not solely a function of methodological rigor; it also depends on whether research helps policy actors frame problems, justify interventions or operationalize solutions. Second, policy process theories highlight that evidence uptake is conditioned by institutional and political dynamics.

Agenda-setting perspectives (as “policy windows”) suggest that research is most likely to be used when problem attention is high (such as disasters, pandemics) and decision venues are seeking actionable options (Warnement Wrobel and McBeth, 2025). Coalition-oriented perspectives emphasize that policy actors filter evidence through beliefs, mandates and organizational interests (Nohrstedt and Heinmiller, 2024), which helps explain why some topics gain traction in particular venues (government agencies vs IGOs vs think tanks). Third, science–policy interface research explains how knowledge crosses boundaries between academic communities and policy/practice communities. Boundary work and knowledge-brokering research suggests that uptake increases when there are credible intermediaries (individuals, institutions or journals) that translate technical results into usable forms (Mackillop et al., 2023). A widely used synthesis in this literature argues that policy-relevant knowledge must be perceived as salient (addresses the decision problem), credible (technically trustworthy) and legitimate (produced/communicated in ways that policy actors view as fair and appropriate) (Andrews et al., 2024). In HLDSCM, this often means that research is more usable when it provides implementable parameters, decision-support tools, protocols and performance metrics aligned with humanitarian standards and operational constraints.

Guided by this framework, our empirical analysis does more than rank citations. We use policy citation patterns (coverage/density/intensity), actor distributions (journals/authors/countries) and topic modeling results to interpret the mechanisms that plausibly drive policy visibility in HLDSCM:

• problem salience and policy windows (such as pandemics and emergency healthcare);

• boundary spanning and brokerage (authors/institutions engaged with humanitarian systems); and

• usability/translation (research forms that reduce the cost of adoption for policy analysts and operational planners).

This theory-informed approach enables us to move from “what is cited” to “why it is cited” and “what HLDSCM should do differently to produce policy-relevant knowledge.”

We address a significant gap in existing research by examining the “reverse dynamic” that the process through which scientific research in HLDSCM influences policy formation throughout the globe. Our research not only contributes to the science-policy dialogue but also enhances current understanding of humanitarian and disaster relief-related literature. To accomplish this, we deployed a novel methodological framework presented by Tahir et al. (2025) to measure and analyze the influence of humanitarian logistics and disaster management research on policy initiatives. This approach began with a comprehensive search for relevant publications in the Scopus database, targeting articles through keywords found in abstracts, titles and other author-designated terms. After applying a stringent inclusion and exclusion criteria, we identified a pool of 2,132 articles. These were then systematically linked to the policy documents obtained from the Overton database. This cross-referencing allowed us to quantify the extent of HLDSCM research that informs policy decisions and pinpoints the most influential entities – ranging from leading journals and prolific authors to key contributing countries and standout articles. Then we employed a machine learning technique entitled “Latent Dirichlet Allocation (LDA)” for topic modeling to dissect the policy-cited documents. This analysis revealed the primary research themes that are driving policy changes and explored existing gaps between academic research and policy implementation.

This study makes two contributions to HLDSCM. Methodologically, we provide a replicable approach to quantify policy-document visibility by linking Scopus-indexed HLDSCM publications (via digital object identifiers [DOIs]) to citations in the Overton policy database and combining policy-citation indicators (coverage, density, intensity) with machine learning topic modeling. Substantively, we identify which HLDSCM themes and which knowledge-brokering actors (journals, authors, institutions and countries) appear most frequently in global policy venues, and we interpret these patterns through a policy process and science–policy interface framework. The aim is to move beyond academic impact (citations within scholarship) toward a clearer understanding of how HLDSCM knowledge becomes usable in policy and practice.

Key contribution and innovative features of this study are driven by four pivotal research questions that aim to redefine the research-to-policy narrative for HLDSCM by offering significant practical insights and future application road maps:

1. How does humanitarian logistics and disaster management research supports global policy formulation? We develop a novel analytical framework that assesses the involvement of scholarly work in policy design and offer tangible indicators and topics for both researchers and policymakers to address this research question.

2. Which research themes are highly instrumental in shaping the policy outcomes under HLDSCM? We implement advanced machine learning methodology (LDA) to uncover the most influential themes and topics of research in the HLDSCM research that drive policy innovation and align with SDGs.

3. Who are the key players in bridging the research and policy in HLDSCM? We identify the influential institutions, seminal studies and leading authors that are acting as knowledge brokers (intermediary actors including institutions, authors and journals whose work crosses the boundary between academia and policy) as their academic research insights have been translated into impactful policies (Tahir et al., 2025).

4. Which methodological advances have a capability to integrate research into effective policy roadmaps? We propose a replicable diagnostic and strategic framework to better embed the real-world policy implications in the research design to enhance evidence-based decision-making in humanitarian logistics.

Figure 2 illustrates structure of this research.

Our methodological framework follows a two-phased approach to evaluate how academic research work on HLDSCM, shapes policy. Phase one involves capturing and refining relevant data, including policy references, while phase two focuses on quantitative assessment and topic modeling of the policy-cited literature.

This phase encompasses four core activities:

1. identifying pertinent literature;

2. applying the criteria for selection;

3. forming the final data set; and

4. tracing citations within policy documents.

• Step 1. Literature analysis: Scopus was chosen for its broad coverage of scholarly publications in fields such as engineering and technology (Tahir, 2024). We opted to search the title, abstract and keywords using the below query on February 14, 2025, which yielded 3,649 records:

((disaster* AND (relief* OR response*)) OR (humanitarian* AND (relief* OR response*)) OR (emergency* AND (relief* OR response*))) AND (supply AND chain* OR logistic* OR operation* OR” operations management” OR” supply chain management”)

• Step 2. Selection criteria: We narrowed the (1) timeline from 2000 to 2024, (2) document type to articles, book chapters, books, notes, letters and short surveys and (3) language to English, removing 1,380 publications.

• Step 3. Forming the final data set: During subsequent analysis, we rely on DOIs. Consequently, items missing DOIs were excluded, leaving 2,132 entries for in-depth analysis. We exported each document’s title, abstract, authorship, journal title, publication year and DOI into MS Excel for further scrutiny.

• Step 4. Tracing policy citations: The Overton database, recognized for its broad scope of governmental and institutional publications (Tahir et al., 2025), was employed to pinpoint policy citations. On February 14, 2025, our 2,132 DOIs were checked in Overton, revealing 389 documents (18%) cited at least once in policy materials. This subset, alongside the corresponding policy document details, was also exported to MS Excel for integrated analysis with referencing information.

Phase two comprises two segments: first, determining the proportion of HLDSCM research that found its way into policy, then examining the substance of these policy-linked studies through machine learning-based topic modeling:

1. Step 1. Measuring the prevalence of policy-cited research: We commenced with an overview of the complete data set of 2,132 publications by categorizing them based on publication year, document type, geography and citation count. We then shifted our focus to the 389 articles identified as policy-cited using three indicators, adapted from Haustein et al. (2015):

   • Coverage – the fraction of HLDSCM articles with at least one policy citation.

   • Density – the mean count of policy citations across all HLDSCM publications, whether cited or not.

   • Intensity – the average number of policy citations exclusively among the policy-cited works.

In addition, we pinpointed highly influential papers, core journals, key authors and associated countries. This layer of investigation revealed how specific research outputs support policy formulations across the globe:

2 Step 2. Topic modeling with latent dirichlet allocation (LDA) To delve deeper into the intellectual structure of policy-cited HLDSCM research, we used the LDA approach (Blei, 2012). This unsupervised machine learning method was implemented in RStudio via the packages, including tm, SnowballC, dplyr, servr, topicmodels, stringr and ldavis.

• Corpus creation and preprocessing: We assembled a data set of abstracts for the 389 policy-cited HLDSCM publications. Each abstract was treated as an individual text item. Data cleaning involved tokenizing text, standardizing it to lowercase and stripping out punctuation, numbers, URLs and standard English stop words. We also excluded certain frequently used yet semantically neutral words (e.g. “study,” “paper”) and performed stemming to consolidate word forms. We analyzed the 389 abstracts of policy cited HLDSCM papers (total word tokens = 86,214; unique terms = 4,972). Using R4.3.2 we relied on the packages tm, SnowballC, tokenizers and quanteda to: lower case, remove punctuation, numbers, URLs and 174 customized stop words; apply Porter stemming; and build uni and bi grams that occurred ≥ 3 times. The final vocabulary comprised 3,408 tokens, preserving 96% of the corpus information.

• Document-term matrix (DTM): We constructed a DTM capturing word frequency per abstract. Words appearing fewer than three times were removed to reduce computational load without affecting overall thematic detection.

• Implementing LDA: We evaluated varying topic counts using perplexity metrics, selecting a value of k that maximized interpretability. Through Gibbs sampling, every abstract was assigned to each topic with a probability score, forming a 389 × k matrix. The topic with the highest score was the abstract’s primary classification. On the basis of perplexity score obtained through LDA, we have chosen five most relevant research topics for further discussion and analysis (see Figure 3). The inter-topic distance mapping as shown in Figure 10 further elaborates the nonexistence of overlap among the selected five topics. While perplexity generally decreases with an increasing number of topics (k), as observed in the resulting.

LDA scores, selecting the optimal k should not solely rely on achieving the absolute lowest perplexity. A monotonic decrease often occurs because models with more topics can better fit the training data, potentially leading to overfitting and topics that are less distinct or interpretable (Tahir et al., 2025). Upon analyzing the perplexity scores, we observed a sharp initial decrease in perplexity up to approximately k = 5. Specifically, the perplexity drops from 866.63 at k = 2–588.28 at k = 5. Beyond k = 5, the rate of decrease in perplexity becomes less substantial and begins to plateau, for instance, dropping only from 588.28 at k = 5–563.7 at k = 7, and then to 508.39 at k = 15. This suggests that adding more topics beyond k = 5 yields diminishing returns in terms of model fit improvement and could introduce redundant or overly specific topics, making the model harder to interpret. Therefore, k = 5 was selected as the optimal number of topics, representing an effective balance between model fit (low perplexity) and the interpretability and coherence of the generated topics, aligning with the common practice of identifying an elbow point in the perplexity curve (Tahir et al., 2025) where the marginal gains in perplexity reduction become minimal as shown in Figure 3. Additional diagnostics confirm robustness:

• Mean Cv = 0.54 (good semantic coherence, > 0.50 threshold) Mean FREX exclusivity = 0.63 (topics well separated) (iii) 5 fold cross validated log likelihood: 1.87 × 105 (σ = 0.02%). (iv) Bootstrap stability: 87% of documents retained the same modal topic across ten 90% resamples.

• Interpreting and validating key topics: We reviewed the most frequent terms in each topic (see Table 3) and revisited manuscript titles, abstracts and keywords of the research papers falling in each category to finalize the category’s higher level descriptive labels. Furthermore, to validate the topics identification, we used the ldavis library, an inter-topic distance map was generated to portray topic clusters visually to further elaborate the nonexistence of the overlap among these topics (see Figure 10). We reran the model with 90% of the corpus in five bootstrap samples; 87% of documents retained their modal topic, confirming assignment stability.

Figure 4 presents graphical flowchart of aforementioned two-phased approach for the assessment of research impact om HLDSCM policy formation. By combining systematic data collection with in-depth quantitative and thematic analysis, this two-step method offers a revealing lens on how scholarly work on HLDSCM resonates within policy documents. The findings not only gauge the real-world traction of HLDSCM research but also illuminate the thematic drivers of policymaking in this rapidly evolving domain.

The field of HLDSCM has experienced steady growth since the turn of the millennium, reflecting an increasing global focus on emergency preparedness, rapid relief distribution and resilient supply chain structures. This section offers an in-depth bibliometric portrait of this domain, encompassing publication patterns, citation dynamics, influential journals, prolific authors and country-level contributions.

A Scopus-based query yielded 2,133 records on HLDSCM published between 2000 and 2024. One document lacked a DOI, leaving 2,132 items for detailed examination. Approximately 94.75% of these are research papers, with the rest comprising book chapters, letters, notes and other scholarly forms. These documents collectively accrued 58,882 citations over the 25-year period, translating to an average of 27.62 citations per publication. The corpus was distributed across 949 distinct journals and involved 1,885 authors (using the first author’s affiliation as a reference point).

From a modest count of 5 papers in the inaugural year (2000) to 304 publications in 2024, the annual number of articles has followed a consistent upward trajectory. Key milestones include surpassing 10 articles in 2006, rising to 36 in 2010, reaching 80 in 2014 and crossing the 200 mark by 2021. Total yearly citations have similarly fluctuated, though with recognizable surges in 2004 (986), 2006 (1,752), 2008 (2,547), 2010 (2,998), 2014 (4,349), 2018 (5,416) and 2021 (5,251). Notably, 2018 witnessed the greatest citation accumulation (5,416), whereas 2024 saw the highest publication volume (304). On average, the field has produced roughly 85.28 papers and garnered 2,355.28 citations annually over the examined timeline.

Figure 5 illustrates publications and citations trend of the HLDSCM research.

Although HLDSCM research extends across nearly a thousand journals, certain outlets have emerged as particularly influential, especially when considered through the lens of a multiattribute ranking process including number of publications, number of citations and citation efficiency (average citation per document). At the forefront stands Annals of Operations Research, which exhibits weighted score of 1159.73. Trailing closely is the International Journal of Production Economics (weighted score of 1130.39), followed by the Journal of Humanitarian Logistics and Supply Chain Management (weighted score of 880.32). The high impact of these publications sources suggests that they host seminal contributions to topics such as decision models for disaster relief, coordination mechanisms in humanitarian operations and resilience oriented frameworks. Figure 6 graphically presents top journals in the field of HLDSCM ranked through multiattribute weighted approach.

An impressive 1,885 authors have shaped HLDSCM literature, yet a select few have made especially prominent contributions. At the top of this group is Rameshwar Dubey, who has authored 14 papers amassing 1,683 citations – an indication of significant scholarly influence. Burcu Balcik with four publications and 1,255 citations is at the second position. Further, L.N. V an Wassenhove has secured the third place because of his single paper which has obtained 1,227 citations. This analysis indicates the existence of both prolific authors with continued contributions and others whose singular studies have a substantial influence.

Researchers from 91 countries have contributed to the HLDSCM research which illustrates the field’s global interest and impact. The USA holds the largest contribution with a total of 585 publications that has obtained 18,866 citations – about 27.44% of the overall citation pool. China ranks second by producing 299 publications that collectively earned 4,570 citations (14.02% of the total). In total, 147 studies from Iran have obtained 5,313 citations (6.89%) closely trailed by the UK with 114 research outputs totaling 4,070 citations (5.35%), and India with 113 publications accounting for 3,163 citations (5.30%). The prominence of these countries points to vigorous institutional frameworks, research funding systems and a robust drive to address global disasters. Figure 7 presents geographical analysis of the contribution toward HLDSCM research considering publications count from the year 2000–2024.

On the basis of aforementioned analysis, the HLDSCM research within the time interval of 2000–2024 can be divided into three distinct phases:

1. Incipient foundations (2000–2004): Though the initial ideas had emerged in the 1980s, early publications were scarce (for instance, just five papers in 2000). Efforts concentrated on minimizing casualty figures under large-scale disasters, gaining momentum following the 2004 Indian Ocean tsunami (Ahmad et al., 2025).

2. Dynamic evolution (2005–2014): This period saw increased research spurred by severe catastrophes such as the 2005 Pakistan earthquake, Cyclone Nargis in 2008 and the Wenchuan earthquake the same year (Shah et al., 2022). The community focused on applying sophisticated risk mitigation tools and pre/postdisaster strategies. Publications grew steadily – from 10 in 2006 to 80 in 2014 – reflecting heightened collaboration between governments, NGOs and academia.

3. Holistic convergence (2015–2024): With Japan’s 2011 tsunami and the recent COVID-19 pandemic, the field embraced resilience-based approaches, stakeholder engagement and advanced technologies. By 2021, publications exceeded 200 annually, culminating in 304 papers in 2024. This era underscores a deepened understanding of systemic risks and interdependencies, leading to robust, integrated frameworks in HLDSCM.

An examination of 2,132 scholarly works on HLDSCM reveals that 389 of these studies have been referenced within policy documents, as indexed globally. This represents a coverage of 18.24%, indicating that nearly one in five of the surveyed articles has found its way into policy discussions. A closer look at how often each article is cited underscores this influence: 213 publications garnered mentions in more than one policy document, contributing to a total of 1,218 policy citations. Overall, each article in the sample averages 0.571 citations in policy sources (density), and those works that do receive at least one policy citation register an average of 3.13 policy references each (intensity). Taken together, these metrics illustrate a moderate yet meaningful incorporation of HLDSCM research into the policymaking sphere, where a subset of studies exerts a disproportionate impact on shaping humanitarian and disaster-related agendas. Figure 8 presents year-wise number of publications, Scopus citations and policy citations over the course of 2000–2024.

Further, the policy documents retrieved from Overton database citing these 389 publications encompass a broad spectrum of topics, reflecting the cross-disciplinary nature of real-world HLDSCM challenges. These documents address health, risk and pandemic response; societal and economic sustainability; emergency and disaster management; technology and Table 1 comprehensive overview of the citing policy documents infrastructure; and public health and governance. Most of these documents align with global issues such as COVID-19, climate change and humanitarian aid, highlighting their relevance and urgency. Classified under but not limited to science and technology, health, economy, environment and politics these documents demonstrate how research informs multilevel governance and strategic decision-making. They cover specific concerns like communicable diseases and supply chain resilience, showcasing a focus on both disaster management and systemic performance improvements. Widely sourced from influential bodies such as the World Health Organization, OECD and United Nations, these documents integrate evidence-based insights into policy frameworks. Institutions like think tanks, government agencies and intergovernmental organizations further diversify their perspectives and enhance credibility. The breadth of source types from legislative research to development banks testifies to the global, multistakeholder nature of policymaking.

Research in this field comes from an array of scholarly outlets, but only 251 journals have yielded material cited in policy circles. These journals encompass wider topics that cover a wide range of interdisciplinary fields, focusing on both the theoretical and practical aspects of management, technology, economics and public health. Among these, five journals stand out. The Journal of Humanitarian Logistics and Supply Chain Management (JHLSCM) is recognized as top-ranked journal with 26 policy-cited papers and a total of 43 policy citations, accounting for 3.53% of all policy citations. Its articles also amass 1,192 Scopus citations, constituting 4.23% of the broader Scopus citation pool in the policy-cited segment. The International Journal of Disaster Risk Reduction (IJDRR) follows closely, with 13 policy-cited publications and 29 policy citations, representing 2.38% of the total. It also shows a notable presence in the academic realm, as indicated by 1,203 Scopus citations (4.27% of the total). The International Journal of Production Economics (IJPE) assumes third place by recording 37 policy citations (3.04% of the total), drawn from 12 policy-cited works and stands out for its 2,183 Scopus citations – 7.75% of the entire set’s Scopus citations, the highest proportion among these top publications. Annals of Operations Research (AnOR) exhibits a slightly smaller footprint, with 20 policy citations (1.64%) from 11 policy-cited articles, but its 1,499 Scopus citations still secure a notable 5.32% share. Finally, the International Journal of Production Research (IJPR) completes the top five with 17 policy citations (1.40%), derived from 6 policy-cited works, and a total of 957 Scopus citations (3.40% of the sample). Cumulatively, these five journals account for nearly 12% of all policy citations and nearly 25% of the total Scopus citations within policy-relevant research, while providing 17.48% of policy-cited publications. Their prominence suggests that policy-makers actively seek evidence-based insights from journals that combine rigorous scholarship with practical applications in humanitarian and disaster-related contexts.

An assessment of authors whose work garners policy citations indicates that a select group exerts a noticeable influence. The top ten authors collectively contribute 13.37% of the policy-cited publications. At the forefront is Gÿongyi Kov′acs from Hanken School of Economics, Finland, whose eight policy-cited articles reflect a strong orientation toward research that resonates with decision-makers. Following closely is Rameshwar Dubey of Montpellier Business School, France, at seven policy-cited documents. Several individuals – including Nezih Altay of DePaul University, USA, and Peter Tatham of Griffith University, Australia – have each contributed six policy-cited works, illustrating broad international engagement. Angappa Gunasekaran of California State University Bakersfield, US, Luk N. Van Wassenhove of INSEAD, France, and Benita M. Beamon of the University of Washington, USA, have five articles each cited in policy materials. Meanwhile, Marianne Jahre of BI Norwegian Business School in Norway has four, whereas both Dan Hanfling of George Washington University, USA, and Karen Spens of Hanken School of Economics, Finland, stand at three. These figures suggest that authors who balance theoretical innovation with actionable frameworks are most likely to gain traction in policymaking arenas, especially in contexts demanding efficient logistical responses and improved disaster resilience.

A geographic breakdown of first-author affiliations illustrates that policy impact is unevenly distributed worldwide. Researchers based in 55 countries collectively contribute to these policy-cited works, but the USA emerges as especially influential, claiming 142 policy-cited articles – 36.50% of the entire cohort. The UK follows with 33 documents (8.48%), while Australia ranks third at 22 (5.66%). France and Canada appear next, each providing 19 (4.88%) and 13 (3.34%) policy-cited papers, respectively. Italy joins Canada with 13 papers (3.34%) as well. Meanwhile, India, Iran, Germany and New Zealand appear in descending order of influence, with 11 (2.83%), 10 (2.57%), 9 (2.31%) and 7 (1.80%) policy-cited works. The predominance of US-based research underscores that significant resources and advanced research infrastructure can drive policy-oriented output. At the same time, smaller proportions from countries like Iran or New Zealand show that a concentrated focus on disaster or humanitarian logistics can still yield notable policy attention, even if total publication volumes are modest. Figure 9 presents comprehensive overview of the top cited, authors, journals, countries and publications in the policy documents.

Within the group of 389 articles cited by policy documents, the distribution of citation counts illuminates a discernible tier of especially impactful research. Only 1.54% of these articles (six in total) record more than 20 policy citations, reflecting an exclusive cluster of highly sought-after references. Publications with more than 10 policy citations comprise 2.83% of the total (11 articles), while half of the sample (50.39%, or 196 articles) sits in the mid-range of 2–10 policy citations. Nearly 45% have been cited only once. The most extensively referenced work overall is Rose and Liao (2005) that has obtained 65 policy citations by itself, a notable exception in this landscape. This pattern suggests that while a large portion of HLDSCM research garners modest policy notice, a small subset achieves substantial traction, likely due to its direct relevance to pressing humanitarian, logistical or disaster readiness challenges. Scholars aiming to shape policies in this space can glean valuable lessons from these highly cited contributions by examining how well they integrate practical solutions and immediate policy implications in their methodological and conceptual approaches.

The prominence of JHLSCM, IJDRR and IJPE in policy documents can be interpreted through an epistemic community lens. The concentration of policy citations in this small set of journals, authors and countries is consistent with research utilization and science–policy interface theory. From a usability perspective, journals that sit close to the science–practice boundary tend to publish work that reduces “translation costs” for policy analysts: clear problem framing, operational metrics, implementable protocols and decision-support models that can be parameterized with available data. In HLDSCM, policy actors often seek guidance under time pressure and uncertainty; research that offers actionable levers such as coordination architectures, triage protocols, prepositioning logic, continuity planning is therefore more likely to be cited in guidelines, toolkits and strategy documents. At the actor level, highly policy-visible authors and institutions can be understood as knowledge brokers or boundary spanners, that is, actors who interact with humanitarian agencies, ministries, hospitals, NGOs or IGOs and can translate operational problems into research questions and translate research outputs back into usable forms including frameworks, standards, training materials and planning tools. Such brokerage increases salience as the research matches decision problems and legitimacy as policy users recognize the context and constraints reflected in the work, which jointly increases the probability of policy citation. Geographic concentration is also expected. Policy visibility is shaped by (i) language and indexing as many policy outlets publish in English, (ii) funding and institutional incentives for translational research and (iii) access to policy venues and operational data. Importantly, this pattern should not be read as a simple quality hierarchy. Rather, it reflects unequal structures of visibility in the research-to-policy pipeline and motivates the need for HLDSCM mechanisms that elevate usable evidence from disaster-prone and lower-resource settings such as through partnerships, open toolkits and co-produced operational data sets. We return to these implications in Section 5 when proposing HLDSCM-specific recommendations.

Among the 389 articles cited in policy documents, 17 top cited articles (Rose and Liao, 2005; De Mel et al., 2012; Kendzerska et al., 2021; Gillespie et al., 2016; Van Wassenhove, 2006; Castañeda-Navarrete et al., 2021; Rossi et al., 2006; Stroud et al., 2013; Smith and Notaro, 2009; Farrell et al., 2020; Wright et al., 2002; Reinecke and Donaghey, 2015; Balcik et al., 2010; Strang et al., 2013; Galbusera and Giannopoulos, 2018; Eisenman et al., 2014; Pugh, 2014) in order of ranking have recorded more than ten policy citations each, and collectively accrued 346 citations representing 28.41% of the total policy citations. We make a thematic analysis of these publications to understand the nature of their research, innovation and overall contributions. A review of these highly cited studies reveals four overarching themes:

1. Economic and supply chain resilience;

2. Public health and community resilience;

3. Coordination in humanitarian logistics; and

4. Environmental and governance perspectives.

• Several papers (Rose and Liao, 2005; De Mel et al., 2012; Castañeda-Navarrete et al., 2021; Galbusera and Giannopoulos, 2018; Reinecke and Donaghey, 2015) investigate how disasters disrupt economies and global value chains, stressing adaptive capacities that help firms and regions recover. From computable general equilibrium and input–output modeling to examining microenterprise capital constraints and postdisaster labor rights, they highlight strategies such as access to finance, policy interventions and stronger buyer–supplier relations to build economic robustness.

• Another set (Kendzerska et al., 2021; Gillespie et al., 2016; Rossi et al., 2006; Smith and Notaro, 2009; Farrell et al., 2020; Strang et al., 2013; Eisenman et al., 2014) focuses on health systems and social factors amid crises. These works underscore the importance of uninterrupted chronic disease management, community engagement (e.g. Ebola and COVID-19 responses) and the specific vulnerabilities of certain populations (people with disabilities, malnourished communities or individuals at high risk of overdose). They emphasize multilevel interventions from telemedicine and targeted aid to social mobilization to safeguard well-being and foster local ownership in emergency preparedness and recovery.

• Three articles (Van Wassenhove, 2006; Balcik et al., 2010; Stroud et al., 2013) address logistics, resource allocation and planning standards in disaster settings. They examine coordination mechanisms among humanitarian actors, agility in supply chains, crisis triggers for implementing scarce-resource protocols and the benefits of cross-sector partnerships. Such coordination, including standardized operating procedures and precrisis preparations, emerges as critical for effective relief operations.

• Finally, two contributions (Wright et al., 2002; Pugh, 2014) expand the lens to include ecological and political dimensions. Wright et al. (2002) assess how droughts alter aquatic habitats, linking environmental stressors to water quality and macroinvertebrate community shifts. Pugh (2014)’s critical theory approach explores peacekeeping’s role in maintaining prevailing global power structures, suggesting that emancipatory reforms in governance and more accountable international institutions could transform crisis responses.

Table 2 presents the contribution of aforementioned highly policy cited research themes to the sustainable development goals (SDGs).

We applied a machine learning-based LDA topic modeling approach to identify research topics in HLDSCM that influence policy. Initially, we determined the optimal topic count by evaluating perplexity scores across values ranging from 2 to 10, where k = 5 emerged as ideal score based on a notable decline in perplexity. Using this parameter, LDA generated clusters of frequently cooccurring terms that represent distinct topics. Each policy cited document was then assigned to the topic for which it had the highest probability of association. The resulting visualizations including an inter-topic distance map with topic size indicating publication volume revealed clear and differentiated clusters. Finally, based on the frequency and structure of key terms, each topic was assigned a descriptive label to highlight the diverse topics in HLDSCM policy research as (i) cross-sector collaboration for humanitarian operations, (ii) public health and food security in crisis response, (iii) emergency health care resource coordination, (iv) pandemic-driven system adaptation and (v) integrated relief system design and optimization (see Table 3). Figure 10 illustrates (i) inter-topic distance mapping and (ii) distribution of topics using machine learning-based LDA approach, and Figure 11 depicts evolution of the policy cited HLDSCM research topics over time.

In this section, we present a comprehensive topic analysis of the HLDSCM related top policy cited research topics identified through machine learning-based LDA approach and their relevance to the SDG-3 and SDG-11, as presented in Table 3. We further discuss the future research directions on the basis of these research topics to guide the researchers to well-integrate their forthcoming scholarly research with policymaker’s interests to enhance its practical applicability.

This research topic focuses on cross-sector collaboration for humanitarian operations, emphasizing synergy among NGOs, governments, militaries and private stakeholders to optimize resource use and enhance relief efforts. By creating these strategic partnerships, such initiatives enable the transfer of critical tangible resources; financial support, technological infrastructure and material goods, as well as intangible resources; specialized knowledge, operational expertise and innovative problem-solving methodologies. This research aligns strongly with SDG-3 by ensuring health services are efficiently coordinated during emergencies and with SDG-11 by building more inclusive and resilient communities. Policymakers value these findings because robust interagency cooperation can reduce duplication, expedite relief and strengthen governance. In this domain, Balcik et al. (2010) discusses coordination challenges and emerging practices, and Akhtar et al. (2012) focuses on the role of chain coordinators and their competencies. Furthermore, Jahre (2017) examines risk mitigation strategies, highlighting collaboration’s role. This topic inherits the network coordination stream (Balcik et al., 2010; Jahre, 2017). Policy relevance is illustrated by the UK Cabinet Office’s voluntary–state interface protocol, which heavily cites these type of studies to justify shared logistics platforms during floods. Future research in this domain of HLDSCM can guide policymakers in establishing formal coordination protocols, refining funding mechanisms and ensuring agile, effective responses to disasters.

This topic addresses public health and food security challenges in humanitarian crises, highlighting how integrated approaches can protect vulnerable populations. Disaster preparedness considering food security is an indispensable proactive measure for the local food authorities in safeguarding the public health. In this manner, loss of human lives during crisis situations can be prevented or minimized and better preparation can be done to address the foreseeable food security challenges. It aligns with SDG-3 by reducing morbidity and mortality through effective health interventions, and with SDG-11 by promoting resilient communities with sustainable access to food and essential services. Policymakers recognize its importance for preventing disease outbreaks, ensuring nutritious food supplies and strengthening healthcare systems. Considering this, Gillespie et al. (2016) underscores social mobilization as central to managing Ebola, emphasizing community engagement and Rossi et al. (2006) evaluates donor strategies in DR Congo, illustrating health and nutrition program impacts. Moll et al. (2007) links water-sanitation interventions to lower disease prevalence. Rooted in integrated health logistics work, this topic informs the WHO’s emergency nutrition and health clusters guidance that draws on these types of studies for designing joint surveillance. It aligns directly with SDG 3 targeting 3.3 and 3.d. Further research can inform policy frameworks to enhance crisis-response protocols and address urgent health and nutritional needs.

This topic prioritizes emergency healthcare resource coordination, focusing on triage, medical supply distribution and facility readiness during crises. This research topic has wide-scope importance and applicability as the emergency management generally includes complex resources and tasks, and coordination among these resources is critical to address the inherent interdependencies for efficient response operation during humanitarian disaster situations. It supports SDG-3 by safeguarding health outcomes under resource constraints and SDG-11 through more resilient emergency services for communities. Policymakers find it critical to plan allocations, prevent systemic collapse and ensure equitable access to care. In this context, Hick et al. (2012) presents structured approaches for triaging scarce medical resources in disasters, and Bravata et al. (2006) models different strategies to stockpile and dispense supplies against bioterrorism, highlighting local dispensing capacity. Morchel et al. (2015) demonstrates efficient drug distribution in mobile EDs using automated systems. Echoing triage and stockpiling scholarship, the studies under this topic have largely supported the US FEMA Crisis Standards of Care playbooks. This topic heavily operationalizes SDG 11.b by stressing urban emergency preparedness. Further research can refine frameworks for rapid resource deployment, prioritization and policy-driven health security improvements.

This topic addresses pandemic-induced transformations in supply chains, healthcare triage and broader system preparedness. Studies lying under this research topic have suggested innovative workflow designs that foreseeably include pandemic-related modifications in the physical infrastructures of healthcare facilities. Several other improvements including implementation of new clinical decision support systems, electronic notification systems and telemonitoring among others are suggested in this research. It directly contributes to SDG-3 by mitigating public health crises through resilient strategies, and to SDG-11 by fostering adaptive urban systems and global value chains. Policymakers are drawn to this domain to ensure continuity of essential services, manage surges in demand and safeguard socioeconomic stability. For instance, Rahman et al. (2021) proposes agent-based modeling for supply chain recovery amid COVID-19 disruptions, and Lai et al. (2020) introduces AI-driven triage solutions that streamline care delivery, reducing hospital crowding. In addition, Castañeda-Navarrete et al. (2021) explores pandemic impacts on global apparel value chains, suggesting policies for resilient, inclusive recovery. Heavily COVID 19 focused, this topic merges supply chain resilience and AI triage. The OECD’s building resilient supply chains reports mostly cites the papers on this topic when proposing near shoring incentives. Further research can inform adaptive policies, robust technology integration, and sustainable responses to future pandemics.

This topic focuses on advanced models and strategies for integrated relief system design and optimization, enabling efficient distribution of supplies and services. These studies have considered uncertain and chaotic conditions of postdisaster situations in terms of road closures, disaster waste, healthcare emergencies and rebuilding of the infrastructure in determining optimal resource utilization and hazardous disaster waste management sites allocation. It bolsters SDG-3 by ensuring timely delivery of healthcare and SDG-11 by enhancing the resilience of communities and infrastructure. Policymakers seek cost-effective, evidence-based planning tools that reduce response times and improve coverage. Considering this, Maharjan and Hanaoka (2018) proposes a multiobjective optimization framework for selecting temporary logistics hubs, and Ozdamar et al. (2004) introduces a dynamic planning model that updates vehicle routes and supply flows in real time. Dufour et al. (2018) illustrates network optimization strategies for a regional distribution center in East Africa. Drawing on dynamic vehicle routing and hub location models, the studies under this topic are mostly referenced in the World Bank’s logistics capacity assessment toolkits, specially to justify investment in prepositioned warehouses. Ongoing research can aid in refining robust relief operations, guiding investment decisions and shaping global humanitarian policy.

Advancing HLDSCM research to shape policymaking and support sustainability and resilience, further research could be conducted in strengthening cross-sectoral collaboration frameworks where the governments, NGOs and other stakeholders share resources, information and governance structures for disaster preparedness and relief operations. This synergy among the entities can potentially improve funding mechanisms and accelerate relief operations through streamlined communication. The research in designing early warning mechanisms for food security and public health indicates high potential for enabling the policymakers to make informed decisions. This can be done through integrating social media analysis, climate data acquisition and other socioeconomic indicators integration to develop the decision support systems for accurate forecasts related to pandemic outbreaks, food scarcity and natural disasters to ensure timely implementation of the risk mitigation strategies. The research dealing with postdisaster relief operations expects a major focus on emergency healthcare resource allocation through AI driven tools to respond faster to the humanitarian crisis. Developing research frameworks around real time hospital capacity management, automated triage and strategic stockpiling in healthcare facilities are crucial for effective postdisaster relief management. Further, embedding decision-support tools including dynamic and innovative planning features in the disaster supply chain management research can support policymakers in gaining knowledge about developing strategic, highly responsive yet cost effective solutions to protect communities under routine as well as disaster situations. Through these collaborative efforts between the academic scholars and governing bodies, HLDSCM domain can continue to serve as a robust mechanism to foster resilience, equity and sustainability throughout the globe.

Before discussing theoretical and managerial contributions, we acknowledge three structural limitations. First, Overton’s coverage is skewed toward English language and Global North policy outlets; citation based impact metrics therefore underrepresent contributions originating from low- and middle-income countries. Second, policy citations may signal visibility rather than uptake as they do not discriminate between affirmative, critical or perfunctory mentions. Third, by operationalizing impact as citation frequency, we omit silent influence pathways such as expert committees or confidential tenders. These biases must be borne in mind when interpreting the quantitative patterns reported in Sections 3 and 4.

Our results suggest that policy-visible HLDSCM scholarship tends to share three design features:

1. strong problem salience linked to urgent hazards and service continuity

2. usability in terms of implementable tools, protocols or measurable performance levers; and

3. boundary connectivity as engagement with operational and policy actors. Based on these patterns, we recommend the following research design shifts:

• Coproduce research questions and outputs with users: HLDSCM studies should be designed with government/IGO/NGO counterparts to ensure outputs align with decision cycles, data availability and operational constraints such as surge capacity, last-mile access and equity mandates.

• Package models into decision products: Optimization and simulation research should provide implementable parameter ranges, sensitivity guidance and “minimum data” versions suitable for low-information contexts, alongside code or templates when feasible.

• Report policy-relevant outcomes, not only methodological performance: In addition to technical novelty, studies should report outcomes that map to policy levers: mortality/morbidity reduction, service continuity time, coverage equity, stockout probability, time-to-delivery and robustness under disruption.

• Build open and transferable evidence assets: Where ethically and legally possible, develop reusable data sets, scenario libraries and evaluation benchmarks for coordination, triage, prepositioning and early warning, especially for underrepresented regions.

The five policy-cited themes translate into operational priorities as:

1. Cross-sector collaboration: institutionalize shared logistics platforms, role clarity and prenegotiated data-sharing agreements across government, NGOs, private logistics providers and health systems.

2. Public health and food security: integrate nutrition, WASH and disease surveillance data into logistics planning; use early warning indicators to prestage supplies and staffing.

3. Emergency healthcare resource coordination: standardize triage and allocation protocols, develop scalable stockpiling/dispensing plans and embed real-time hospital capacity reporting into response playbooks.

4. Pandemic-driven system adaptation: maintain continuity strategies for essential supply chains and healthcare workflows, including surge staffing, telemonitoring options and contingency routing for critical goods.

5. Integrated relief system design and optimization: invest in prepositioned warehouses, dynamic routing capacity and rapid hub selection tools that can be activated under infrastructure disruption.

For policymakers, the findings support concrete institutional actions as:

• Adopt evidence standards for HLDSCM decision tools (minimum data requirements, validation expectations and equity metrics) for use in emergency plans and procurement.

• Fund translational HLDSCM infrastructure (shared data sets, modeling support units and training programs) so that research products can be operationalized rapidly during crises.

• Strengthen boundary organizations and brokerage roles (research–policy fellowships, embedded analysts and joint task forces) to reduce translation gaps between HLDSCM scholarship and emergency governance.

This study offers critical evidence that scholarly work in HLDSCM does more than advance academic dialogue, as it directly shapes policy decisions across diverse organizational and geographic contexts. By triangulating a rigorous cross referencing analysis, policy citation metrics and a machine learning-based LDA topic modeling approach, we reveal a clear path from research insights to the formulation of strategic governmental and institutional actions. Our findings show that 18% of the analyzed publications have been referenced in policy documents, underscoring a growing trend among policymakers to rely on academic research to guide interventions in disaster management and relief. In particular, the key thematic clusters encompassing cross-sector collaboration, public health and food security, emergency healthcare resource coordination, pandemic-driven system adaptation and integrated relief system design demonstrate the breadth of scholarly contributions that resonate in real-world policy frameworks. Notably, the Journal of Humanitarian Logistics and Supply Chain Management emerges as an instrumental platform, disseminating research that substantially influences policy discourses. USA takes a leading position in producing policy-cited HLDSCM studies, reflecting the robust infrastructure for funding, publication and institutional collaboration. Meanwhile, Gÿongyi Kov′acs stands out as one of the most cited authors for policy-relevant humanitarian logistics research, while the Hanken School of Economics, Finland excels in bridging theory with tangible policy impacts. The alignment of these scholarly contributions with the Sustainable Development Goals, particularly SDG-3 (Good Health and Well-Being) and SDG-11 (Sustainable Cities and Communities) further highlights the practical significance of HLDSCM in advancing global resilience.

We categorically quantified HLDSCM’s visibility in global policy documents by linking the available 2,132 Scopus-indexed publications to Overton-indexed policy sources and applying policy-citation indicators alongside topic modeling of policy-cited abstracts. The results show that a meaningful subset of HLDSCM scholarship is visible in policy venues and that visibility is concentrated in specific actors (journals/authors/institutions) and themes. Using policy studies and science–policy interface theory, we interpret these patterns as the outcome of (i) problem salience and policy windows such as disasters and pandemics, (ii) boundary spanning and knowledge brokering as actors positioned between research and operational systems and (iii) usability of outputs as tools, protocols and implementable guidance that reduce translation costs for policy users. These mechanisms explain why some HLDSCM work is repeatedly cited in guidelines, toolkits and strategic documents. For HLDSCM, the implication is practical: producing policy-relevant research requires more than technical novelty. It requires coproduced problem framing, decision-ready outputs and reporting of outcomes that map to policy levers and SDG-3/SDG-11 targets. Future work should extend beyond citation traces by examining how evidence is used within policy documents (instrumental vs conceptual vs symbolic) and by triangulating policy citations with additional pathways of influence (advisory roles, standards development and operational adoption).

The authors used ChatGPT 5.1 to assist with drafting and language editing of this article. All research design, data collection, analysis, interpretation and original intellectual contributions are entirely the work of the authors, who take full responsibility for the content of this publication.