The current public health emergency in the DRC and Uganda highlights the need for accelerating the development, assessment and approval of an effective vaccine and other medical countermeasures for BDBV. Further, major outbreaks of filoviruses are an ongoing and increasing global health concern (14). Therefore, the evidence maps here are not only aimed to provide a resource for each individual pathogen, but the four filovirus maps also allow rapid analysis of what has been achieved in similar studies across-pathogens. For example, the original licensure of the Zabdeno / Mvabea vaccine against EBOV by the European Medicines Agency (EMA) (6, 7) was supported by data on antibody binding and protection in NHP, and antibody binding after vaccination in human clinical trials. To assess the potential for a similar pathway in Bundibugyo, a researcher may choose to review all NHP vaccination and challenge studies with BDBV, as well as the assays used to measure antibody responses in these studies. These NHP infection models and antibody binding assays can also be compared with those used in Ebola by reference to the Ebola evidence map. In the case of Bundibugyo, the maps can be used to identify gaps in the current evidence landscape where new primary research may be required to provide evidence for immunity or vaccine effectiveness, avoiding duplication, and focusing future research efforts.

Systematic reviews of the available evidence to address particular questions are of enormous value, and in cases with limited investment can leverage the existing research to identify new insights (16). A small number of systematic reviews have been performed on filovirus immunity, for example summarising the results of clinical and NHP studies of immunity (17, 18). However, this approach requires each systematic reviewer to conduct exhaustive searches and study selection for each specific research question. Evidence maps can greatly accelerate this process by aggregating and classifying the available literature across a broad topic area. They can operate as the first step in a systematic review for multiple filovirus research questions and remove the need for individual investigators to search the literature separately to identify relevant studies. Researchers can focus on the studies in the map cells relevant to their question and focus their efforts on appraising and synthesising the studies within these cells. In addition to aggregating the available evidence, these maps will ideally be used by funders and vaccine developers to identify gaps where key studies should be prioritized. For example, evidence on human convalescent immune responses and vaccine immunogenicity for BDBV are currently lacking, and the evidence on any association between antibody binding and vaccine protection in NHP is sparse.

Regularly updating evidence syntheses (i.e. making them ‘living’) is important for research areas like filoviruses, where research evidence is emerging, current evidence is uncertain or where new research might change policy or practice. The value of updating evidence synthesis and guidelines is increasingly accepted in clinical research (19). To ensure the maps continue to accurately reflect the research landscape, the researchers recommend continuing to update the evidence maps with new studies every six months, and perhaps more regularly during PHEIC such as the present one. To date, one update to include new literature of the evidence maps has been completed (outlined in the methods) for EBOV, MARV and SUDV.

An important limitation of evidence maps is that a higher number of publications in a given cell within a map does not necessarily indicate a high quantity or quality of evidence. It can be that (for example) a large number of very small studies, or studies lacking necessary controls, may be found in a given cell of the map and which do not provide the required evidence or quality of evidence to address a given question. How immunological evidence can be evaluated for quality is also area that requires ongoing investment (20).

Evidence mapping is a comparatively new research methodology, and has most often been applied to clinical literature (21). To our knowledge, these are the first evidence maps categorising studies reporting measures of immune response to infection or vaccination. The application of evidence mapping approaches to preclinical research presents several challenges. Firstly, clinical studies often include structured abstracts and standardised reporting tools that allow easy identification of the relevant study details. By contrast, immunological studies in animals often include narrative abstracts and limited reporting of experimental procedures. Standardised reporting tools such as the ARRIVE guidelines 2.0 (22) attempt to address this issue by identifying the minimum information necessary to report in publications describing in vivo experiments, however these approaches are not universally applied. More structured, detailed, and standardised reporting would substantially improve the efficiency of mapping of preclinical studies.

Another major challenge is that preclinical publications often present multiple research studies in different animal models and with more than one outcome and/or measurement tool. This can make it difficult to precisely map an individual publication. As an example, a manuscript may include studies of ELISPOT numbers after vaccination in NHPs, as well as a separate study where NHPs were infected and only the infection outcome (survival) was reported. Due to the limitations of existing evidence mapping software, for this publication both the columns (‘vaccination’ and ‘infection challenge’) would appear to intersect both the rows (e.g. ELISPOT, Outcome survival) – even though ELISPOT was not reported in the infection challenge study and survival was not reported in the vaccination study. Thus, it is important to recognise that the goal is to include all the publications relevant to a specific cell on the map, so inclusivity was prioritised. This ensures that all research relevant to answering a question will be available in a cell on the map. However, occasionally, some of the publications listed may include non-relevant studies (for the reasons above). We acknowledge that the trade-off for this is that a cell may require further manual filtering by the user to find the research that addresses their question.

Artificial Intelligence (AI) was not used to create the evidence maps beyond what is already available in the Covidence® Systematic review Software, which includes some automation features to increase the speed of the screening phase of evidence review (23). AI and automation may ultimately play a role in accelerating evidence syntheses. However, it is essential to ensure the AI system or tool used is methodologically sound and will not undermine the reliability of the synthesis (24). Importantly, for this work, we could find no examples of how AI has been validated or piloted to ensure that it is appropriate for use in this relatively innovative area of immunological evidence synthesis (24). We recommend a direct comparison of the human curated versus AI developed map to validate the use of AI tools and to assess their readiness to be employed for future pathogen evidence mapping or updates. Moreover, it is worth noting that with the infrastructure in place, even without AI tools, the Bundibugyo evidence map could be generated rapidly because of the existing pre-registered methods, limited literature, and experienced reviewers. That is, the PHEIC was declared on 17 May 2026, and the generation of the map of available evidence on immunity and vaccination was completed by 20 May.

Filoviruses pose an ongoing threat both to local communities and globally as pathogens of pandemic potential. These evidence maps provide the filovirus community with a curated overview of preclinical immunological research for BDBV, MARV, SUDV and EBOV, to help inform future research priorities and define key questions for more focused evidence synthesis. The evidence maps will also allow researchers to confidently assess the totality of available evidence on their research topic, as well as facilitate future evidence synthesis and meta-analysis. Ultimately, these maps provide a resource to support the research, development and deployment of effective vaccines and treatments for filoviruses.