Bridging evidence gaps to prevent and control healthcare-associated infections (HAIs).

Healthcare-associated infections (HAIs) are a major source of morbidity, mortality, and AMR drivers [22]. Despite its importance, evidence on the effectiveness and cost-efficiency of specific IPC interventions, such as cohorting, isolation practices, staffing ratios, and the presence of IPC focal points, remains limited, especially in humanitarian settings. Tools and strategies showing cost-effectiveness in high-income countries often cannot be used in humanitarian settings due to a lack of data [23]. This underscores the urgent need for pragmatic research to evaluate combinations of IPC interventions under real-world conditions [22,24–29]. While the WHO’s prioritisation exercise reinforces the central role of IPC globally, further operational guidance tailored to the specific constraints of humanitarian settings would help translate global recommendations into field-level impact [19].

Likewise, simplified, standardised surveillance tools to monitor HAIs across formal and informal health systems are urgently needed. Monitoring IPC adherence through audits and feedback is essential for preventing and managing HAIs and outbreaks [29,30]. Establishing surveillance systems, including risk-based outbreak prediction algorithms, can greatly enhance preparedness. Developing and evaluating context-adapted monitoring tools and indicators is critical to optimise data collection and reporting, particularly outside formal health systems such as refugee camps. Moreover, although WHO supports microbiology-informed surveillance, it could benefit from further guidance on when and how to implement laboratory-based surveillance in humanitarian settings. Additionally, the lack of reliable patient tracking systems further complicates surveillance efforts in these environments. As such, IPC-informed surveillance must be carefully prioritised and supported through capacity building, infrastructure investments, and practical implementation strategies.

Re-evaluate isolation protocols for multidrug-resistant organisms (MDROs).

In humanitarian settings, isolation capacity is often limited or absent. Moreover, there is limited information on the community-based transmission of numerous multidrug-resistant organisms (MDROs) in many regions of the world [31]. The rationale for isolation protocols should be re-evaluated critically, particularly in settings with high community transmission of specific MDROs. Clear guidance should be established for when isolation measures can be modified or discontinued, based on local epidemiology and feasibility.

As mentioned in the WHO agenda, isolation of MDRO-infected or colonised patients, decolonisation of healthcare workers, and routine catheter changes are three areas that could greatly impact the incidence of HAIs [19]. However, operational clarity is needed on when and how to apply these measures, especially in crisis contexts. Guidelines must also be adapted to high surgical volumes and vulnerable populations, including neonates, malnourished children, people living with HIV, and those with war-related injuries.

Assess the feasibility and effectiveness of humanitarian-tailored IPC care bundles.

A range of IPC care bundles have been recommended, yet in humanitarian settings, the evidence base guiding which bundles are most feasible and effective remains thin [32]. Operational guidance on how to adapt and prioritise these bundles in resource-limited and conflict-affected settings is limited [32]. In acute crisis contexts, there is little clarity on which bundle components are most impactful or how to adapt existing bundles to local constraints. There is a pressing need for humanitarian-specific operational evidence to inform global IPC standards and context-adapted standard operating procedures (SOPs). Surveillance systems should be aligned with implementable guidelines; this way, humanitarian actors could build more resilient and effective IPC programmes.

MSF’s experience highlights the need for longitudinal research to assess the long-term impact and sustainability of IPC strategies. Therefore, IPC research must focus on practical, context-adapted interventions instead of idealised frameworks. Supporting the reduction of structural barriers, behaviour change, leadership in daily practice, and institutional accountability for IPC is a priority.

Assess the feasibility and effectiveness of AMS programmes.

Evidence on the cost-effectiveness of AMS interventions and how significantly these interventions reduce the burden of AMR and reduce patient mortality and morbidity primarily comes from high-income settings [34,35], however, evidence from LMIC is scarce [21]. Research is needed in humanitarian settings where resources are limited, and AMR interventions can be perceived as an unnecessary cost and burden. External factors, including non-bacterial outbreaks (e.g., respiratory viruses, malaria, scabies), climate change, and conflict, may indirectly influence antimicrobial prescribing practices and resistance trends [7,36–39]. Further evaluation of these drivers is necessary to inform AMS strategies adapted to dynamic and fragile contexts.

Improve AMS monitoring tools.

A significant challenge in AMS implementation is the lack of accessible, user-friendly tools to monitor antimicrobial use and consumption, clinical decision-making, and stock availability, especially in settings with fragmented information systems [19]. Developing simplified, robust, and standardised monitoring tools is essential to track the progress of AMS programmes, including antimicrobial use, consumption, access and shortages. The feasibility and acceptability of these tools should be assessed following implementation.

Special attention must be given to paediatric antimicrobial use, as traditional Defined Daily Dose (DDD) metrics are poorly suited for children and neonates [40,41]. Therefore, alternative user-friendly and adaptable measurement approaches, coupled with learning and development (L&D) initiatives for healthcare workers (including pharmacists), are needed to accurately capture paediatric and neonatal prescribing, particularly in outpatient departments [2]. Lack of systems to routinely collect patient-level data to monitor antimicrobial use remains a major limitation regarding AMS in LMIC – periodic audits of patient files are necessary to have visibility on prescribing practices. Integrating stockout data in AMS monitoring could provide valuable information on the antibiotic access gaps and prescribing shifts driven by medication shortages.

Increase evidence for guidelines, including prescription algorithms.

A critical need remains for the development of tailored prescription algorithms for settings with limited access to medicine, diagnostics and healthcare. WHO has highlighted the importance of investigating criteria and strategies to optimise empirical antimicrobial therapy for the main infectious syndromes, particularly in such settings [19]. These should incorporate patient risk factors and biomarkers [42,43]. Evidence is required to identify the nuanced contribution of malnutrition, malaria, tuberculosis, and HIV as risk factors for particular MDROs, which could be further explored in the WHO agenda [19,44]. Additionally, integrating electronic AMS systems implementable in humanitarian settings could further support these efforts [45,46].

Local resistance patterns should guide empirical guideline development, though the availability of quality-assured microbiology results is scarce in LMIC. Sentinel laboratories should be identified and supported for quality assurance; moreover, microbiological data from these sites should be linked to clinical information to enable the meaningful interpretation of local cumulative antibiograms. Alternatively, point prevalence surveys assessing bacterial causes of sepsis should be performed and the results published to inform clinical practice. In parallel, research is needed to define resistance thresholds that should trigger updates to local empirical regimens and changes in IPC strategies.

Evaluate the safety and efficacy of second-line antibiotics for neonatal and paediatric sepsis and develop oral formulations for multidrug-resistant Gram-negative infections.

As recognised by WHO, access to appropriate second-line antibiotics remains limited in many humanitarian settings [19,47]. Clinical trials on paediatric sepsis are needed to evaluate the safety and efficacy of second-line regimens, such as carbapenem-sparing combinations (e.g., ceftriaxone-amikacin or piperacillin-tazobactam). Moreover, the development of oral formulations for multidrug-resistant Gram-negative infections is urgently needed [19,48]. Oral treatment options would immensely benefit patients, especially when prolonged courses of treatment are required or when intravenous therapy is not possible due to infrastructure or staffing constraints.

Study pharmacokinetics and pharmacodynamics (PK/PD) for bone infections and paediatric populations.

There is a critical lack of data on the optimal dosing of antimicrobials in paediatric and malnourished populations, as well as for deep-tissue and bone infections [19]. MSF calls for research to focus on conducting PK/PD studies tailored to these vulnerable groups, allowing patients to access a safe and effective dosing regimen.

Pharmacovigilance systems must also be strengthened to monitor, report, and prevent issues related to antimicrobial quality, safety, and misuse [49]. These systems are especially critical for antimicrobials associated with significant side effects (e.g., glycopeptides and aminoglycosides) and antimicrobials without solid evidence for their use in neonates and children. Moreover, simplified, point-of-care (POC) Therapeutic Drug Monitoring (TDM) could greatly improve patient outcomes in humanitarian settings [50,51].

Develop/pilot point-of-care/rapid diagnostics suitable for humanitarian settings (REASSURED criteria).

Currently, the diagnostic tools available in resource-limited settings provide pathogen identification and susceptibility testing in 48–72 hours versus 4–12 hours in high-income countries. More research is needed to provide diagnostic tools based on REASSURED criteria: real-time connectivity, ease of specimen collection, affordable, sensitive, specific, user-friendly, rapid, equipment-free, and delivered to be suitable for field use in humanitarian contexts [57]. Global efforts should aim to validate and introduce diagnostic tools that meet these criteria as much as possible, particularly for syndromes like fever of unknown origin, childhood pneumonia, or suspected neonatal sepsis, in areas where diagnostic uncertainty leads to inappropriate antibiotic use. POC C-reactive protein (CRP) and procalcitonin tests, ultrasound (POCUS), and computer-aided detection (CAD) algorithms may play a crucial role, and clinical algorithms should be developed, including the use and interpretation of these POC diagnostic tests, and their impact assessed post-implementation [58].

Develop/pilot simplified and standardised passive surveillance tools that integrate clinical and microbiology data.

AMR surveillance data is typically generated through passive surveillance, which involves bacterial isolates from clinical samples of patients with suspected infections. Passive or health facility-level surveillance is the basis for monitoring resistance trends, improving empiric treatments, addressing critical gaps, and enhancing the global response. While the WHO mentions strengthening surveillance, it could benefit from specific guidance on harmonising systems or the need to increase diagnostics capacity sustainably and cost-effectively [19]. Moreover, the link between clinical and microbiological data is rarely achieved in humanitarian settings.

Existing platforms like WHONET offer analysis for microbiology data with a focus on AMR surveillance but have some limitations. For example, it lacks integration with the clinical information from electronic health record systems [59]. Limitations of WHONET should be explored and jointly addressed, for example, the integration of other health data systems or further variables. Moreover, strengthening the link between WHONET and advanced statistical tools and algorithms will facilitate more robust data analysis and improve evidence-based decision-making [60,61]. Laboratory standardisation and validation, as well as data-sharing networks, should be strengthened.

Develop/evaluate contextualised and rapidly deployable D&S bundles for humanitarian emergencies and primary healthcare in humanitarian settings.

Further research is needed to strengthen the evidence on the impact and efficiency of D&S interventions, as well as the implementation factors that influence their success. This evidence is essential to inform guideline updates and to define and prioritise essential diagnostic packages required for all three levels of healthcare: primary care, primary care with basic laboratory, and hospitals [52,62]. Guidelines that define the criteria to decide whether to implement a conventional laboratory or alternatives, such as the Mini-Lab, to ensure context-appropriate diagnostic capabilities.

Although the deployment of the Mini-Lab is faster than conventional laboratories, it still requires training the staff on sample collection and processing, as well as results interpretation. Moreover, in many settings, microbiology certification is not part of university curricula [56]. Further investment in mobile and adaptable microbiology solutions, including trained staff, is needed to support faster deployment in emergencies. These microbiology diagnostics solutions should integrate a pool of microbiology-trained staff, essential POC diagnostics and prioritise the identification and resistance profiling of high-risk pathogens among war-injured patients, where infections with MDROs are particularly concerning.

Research and development of vaccines, antimicrobials and diagnostics – Accessibility.

While the WHO AMR research agenda highlights the need for developing new antimicrobials, vaccines, and diagnostics, MSF emphasises the equally critical need to ensure the accessibility—availability, affordability, and practicality—of existing tools in humanitarian settings [2,9,10,19,63]. Stockouts, supply chain disruptions, and lack of affordability continue to hinder appropriate use, impact clinical decision-making, and contribute to the spread of AMR. Research must prioritise understanding the link between access barriers and patient outcomes and developing solutions such as supranational platforms for pooled procurement, geographically diversified manufacturing, and strengthening forecasting and supply chain management.

The innovation ecosystem should also ensure that the needs of LMICs are integrated from discovery through implementation. Public and non-profit R&D efforts should be supported, as these are most conducive to access, affordability, stewardship, and the kind of collaboration (such as sizeable global clinical trial networks with sites in LMICs) that is needed to overcome the scientific bottlenecks in antibiotic R&D. R&D prioritisation must be public‑health driven, ensuring that medicines and diagnostics are adapted for fragile contexts. For example, favouring oral over intravenous formulations, with room‑temperature stability, simplified dosing, and providing child‑friendly options [63–66].

Further investment is needed in POC and in vitro diagnostics that meet REASSURED criteria and in the expansion of molecular tools like next-generation sequencing [19,57,63,65]. Comparative studies on feasibility, cost-effectiveness, and impact are essential to validate diagnostics in humanitarian contexts. Multi-disease platforms and integration with digital health systems could also improve surveillance and outbreak response. Ultimately, tailored, stable, and context-adapted tools are essential to support effective treatment and reduce AMR in the world’s most vulnerable populations.

Understand the intersections between AMR, climate change, and conflict.

Climate change and conflict are significant drivers of infectious diseases by disrupting healthcare, water and sanitation and food systems. More than half of the global population, 4.5 billion people, face high risks from extreme weather such as floods and droughts, with 2.3 billion living in poverty [8]. These issues, though mentioned in the WHO’s agenda, are even more important in humanitarian contexts [19]. Rising temperatures and extreme weather events have been linked to increased disease, such as vector-borne, waterborne and fungal infections [7,39,67]. Simultaneously, conflict weakens healthcare systems by damaging facilities, interrupting essential supply chains, and exhausting healthcare workforce capacity [68–70]. Food insecurity, exacerbated by climate-related crop failures and conflict-induced instability, increases the risk of malnutrition, which increases vulnerability to infections [6,39,71]. Moreover, climate change, natural disasters and conflict drive population displacement, which also increases the risk of AMR as these populations face unique barriers to accessing healthcare. A deeper understanding of prevalence, drivers and outcomes in these settings is essential to develop inclusive interventions addressing AMR and its drivers [72].

However, the nuances of the interaction of climate change and conflict with AMR are poorly understood, demanding targeted operational research to fill critical knowledge gaps and inform inclusive strategies.

Global AMR research agendas and policies predominantly focus on formal healthcare systems and may not fully account for conflict zones, internally displaced people/refugee camps, and climate change-affected settings [19]. This is especially significant considering 2 billion people, a quarter of humanity, are estimated to live in conflict-affected contexts [20]. The complex challenges of population displacement, erratic supply chains, etc., are barely mentioned in the agenda. Moreover, there is an urgent need for evidence and innovation that is adaptable to the realities of humanitarian settings.

Design strategies for the engagement of patients, communities, and healthcare workers.

There is limited evidence on the effectiveness of patient, their caregivers and community engagement initiatives in reducing AMR [73–75]. More research is needed to understand how patients, their caregivers and communities can be actively involved in IPC and AMS. The WHO’s agenda could benefit from a more structured approach to integrating patient, caregivers and community perspectives into AMR strategies, ensuring that interventions are not only scientifically sound but also socially and culturally relevant and therefore better meet contextual needs.

Healthcare workers also play a crucial role in IPC and AMS, yet existing programmes often lack comprehensive, evidence-informed, context-specific training, particularly in fragile settings. Medical doctors, nurses, social workers, infection control officers, health promoters, and pharmacists should be trained and involved in IPC and AMS programmes [76].