Author	Affiliation
Prashant Mahajan, MD, MPH, MBA	University of Michigan, Department of Emergency Medicine, Ann Arbor, Michigan
Shu-Ling Chong, MBBS, MRCPCH, MCI, MPH	KK Women’s and Children’s Hospital, Department of Emergency Medicine, Singapore
,	University of Michigan, Emergency Medicine Education and Research by Global Experts (EMERGE) Network, Ann Arbor, Michigan
Vijaya Arun Kumar, MD, MPH	Wayne State University
Prerna Batra, MD	UCMS & GTP Hospita
Apoorva Belle, MA, MHA	University of Michigan
Ben Bloom, MD	Royal London Hospital
Chung-Hsien Chaou, MD, PhD	Chang Gung Memorial Hospital
Ulf Ekelund, MD, PhD	Skane University at Lund, Lund University
Sagar Galwankar, MBBS, DNB, MPH, Dip. ABEM	Florida State University
Johanna Kaartinen, MD, PhD	University of Helsinki and Helsinki University Hospital
Vimal Krishnan, MD	Kasturba Medical College
Qingbian Ma, MD	Peking University Third Hospital
Paul M. Middleton, MD	South Western Emergency Research Institute
Anna Miethke Morais, MD, MBA	Sao Paulo University
Chip Jin Ng, MD	Chang Gung Memorial Hospital
Daniel Osei-Kwame, MBChB, FGCS	Komfo Anokye Teaching Hospital
Dominik Roth, MD, PhD	University of Vienna
Rasha Sawaya, MD	American University of Beirut Medical Center
Sanjeev Singh, PhD, MPhil, DCH, MBBS	South Western Emergency Research Institute
Tej Prakash Sinha, MBBS, MS	Sao Paulo University
Mabel Vasnaik, MD	Chang Gung Memorial Hospital
Katie Walker, MBChB, Dip DHM, FACEM	Komfo Anokye Teaching Hospital
Adriana Yock, MD, MSc	University of Vienna

Introduction
Methods
Results
Discussion
Limitations
Conclusion

INTRODUCTION

Research in emergency medicine (EM) has increased in complexity and sophistication through collaborative efforts in the past two decades.¹ Specifically, research networks can provide adequate cohort size for statistical power, global representation of illnesses for generalizability, and structured research support to ensure integrity and quality of study designs.² Research networks can overcome current barriers such as varying research infrastructure support in emergency departments (ED) across the globe and allow for evidence generation.³

The Emergency Medicine Education and Research by Global Experts (EMERGE) network was formed in June 2018, with the goals of generating evidence, translating knowledge to improve emergency care for patient populations, and strengthening EM research capacity globally.⁴ In the spirit of being inclusive we have reduced barriers to entry; thus, all institutions with EDs regardless of their annual census or academic affiliations can become EMERGE members. EMERGE continues to attract member sites across all six World Health Organization regions and continents, with a current membership of 26 EDs in 17 countries (www.EMERGENetwork.org). Aggregating high quality data on a common platform regardless of geographic borders is critical for describing and comparing the basic epidemiology of emergency care globally; increasing numbers to power hypothesis-generating studies; and improving generalizability of research findings.⁵ Here we report the issues and barriers faced by the network and discuss potential solutions.

METHODS

Operating on a carousel model,⁶ EMERGE is governed by an executive committee consisting of three subcommittees aligned with the missions of research, education, and data oversight (performed by a data coordinating center). EMERGE intentionally proposed the development of a global ED visit registry to demonstrate feasibility of data-sharing. The registry is a six-month, retrospective, deidentified, minimal dataset of routinely collected variables, such as patient demographics, diagnosis, and disposition. The data was collected by determining the core elements a priori, and each site communicated with the data-coordinating center to harmonize variable fields from their respective electronic health records (EHR).

RESULTS

EMERGE experienced considerable barriers ranging from variable data taxonomies, need for language translation, varying processes for data cleaning and transfer of deidentified data, and navigating numerous data protection regulations (Table 1). Institutions and countries vary substantially regarding data oversight, ranging from the need to set up individualized research agreements to legal barriers resulting in the inability to transfer deidentified data across geographic borders for some participating sites. Such data-fencing has precluded some sites from participating in the EMERGE registry.

We have overcome many of these challenges through creating detailed data-sharing agreements with bilateral regulatory oversight, developing relationships with and training site health informaticians to ensure secure transfer of deidentified data, and formalizing a transfer process ensuring data privacy. Currently, 18/19 EMERGE sites in 13 countries have institutional review board approval, 13/18 have provided initial sample data while 8/13 have provided complete six-month data (Supplement Table 1). In most instances, after meeting site regulatory requirements for data transfer, sites reported difficulties in extracting data from their EHRs. The EMERGE data coordinating center had to take on roles beyond data curation and analyses and is now working with each individual institution’s information technology teams across EMERGE sites to ascertain data quality and authenticity. We also found substantial variation in each participating institution’s research infrastructure including training in research-related activities (eg, good clinical practices, study design and statistical analyses, grant writing). and resources including statisticians and research associates.

DISCUSSION

EMERGE encountered substantial challenges in obtaining high quality data across its participating sites in a timely manner, which is an inherent barrier toward generating evidence and improving emergency care globally. Barriers encountered ranged from restrictive regulatory data governance to lack of time and support for emergency clinicians to participate. Despite the barriers we identified in this article, EMERGE was able to quickly respond and conducted a pandemic preparedness study among 26 member and 103 non-member sites. By leveraging the EMERGE network, many more sites were recruited via referrals and direct solicitation, and all participants who were approached agreed to participate in the study.⁷ This supports the notion that EDs across the globe want to participate in endeavors to generate evidence.

Because it is an unfunded network, such intense data efforts from EMERGE will require substantial resources and are unsustainable. We believe the future viability of international research networks will depend on developing a federated data model in which the data is collected using standardized definitions and processes, retained at the institution, analysed locally or using federated machine learning, and reported in an aggregated manner while preserving privacy and overcoming regulatory requirements.⁸ However, based on our experience, building a federated data model requires sites to obtain appropriate local regulatory approvals and have the necessary data infrastructure including data scientists and trained personnel to support this approach.

Another approach to enhancing site research capabilities is to enhance the research training and education of the personnel in each participating ED. We are currently collaborating with the Development Implementation, and Assessment of Novel Training in Domain-based competencies (DIAMOND), which is a web-based, curated research education platform, developed by the Clinical & Translational Awards (CTSA) mechanism in the United States.⁹ This novel and scalable platform allows research personnel to evaluate their knowledge gaps and build highly customizable, on-demand, web-based research education modules for training in research methods and procedures. The enhanced research methodology training will increase the site support and thereby sustain engagement in global research participation. Sharing of data-related resources across sites will further enhance the success of the federated data model.

LIMITATIONS

The barriers encountered and solutions proposed are based on our experience, and it is possible that research networks in other specialities or those that have more robust support may have different barriers and challenges. Our experience is that with limited resources, there is a risk of over-burdening the sites as well as the central data-coordinating center. Some of these barriers can be potentially circumvented by site commitment of time and resources to the network research goals and aligning priorities.¹⁰. EMERGE has collected information from participating sites through an ED demographics study that will allow us to better delineate each site’s patient population and research capabilities to participate in studies and thereby facilitate decisions on the type and number of active studies at any one point in time for the network.¹¹

CONCLUSION

We identified challenges in data-sharing and variation in research infrastructure among sites. Immediate next steps include the need to create regulatory-compliant federated data models, enhance research education and training, develop relevant research priorities, and identify research questions that require global participation yet can be performed at sites with limited resources.

Footnotes

Section Editor: James A. Meltzer, MD, MS

Full text available through open access at http://escholarship.org/uc/uciem_westjem

Address for Correspondence: Prashant Mahajan, MD, University of Michigan Medical School, Department of Emergency Medicine, 1540 East Hospital Drive CW 2-737, Ann Arbor, Michigan 48109-4260. Email: pmahajan@med.umich.edu. 11 / 2022; 23:947 – 951

Submission history: Revision received February 8, 2022; Submitted May 2, 2022; Accepted July 2, 2022

Conflicts of Interest: By the WestJEM article submission agreement, all authors are required to disclose all affiliations, funding sources and financial or management relationships that could be perceived as potential sources of bias. No author has professional or financial relationships with any companies that are relevant to this study. There are no conflicts of interest or sources of funding to declare.

REFERENCES

1. Klassen TP, Acworth J, Bialy L, et al. Pediatric emergency research networks: a global initiative in pediatric emergency medicine. Pediatr Res. 2010;17:224-7.

2. Papa L, Kuppermann N, Lamond K, et al. Structure and function of emergency care research networks: strengths, weaknesses, and challenges. Acad Emerg Med. 2009;16:995-1004.

3. Rincon F, Lee K. Ethical considerations in consenting critically ill patients for bedside clinical care and research. J Intensive Care Med. 2015;30:141-50.

4. Emergency Medicine Education & Research by Global Experts. 2021. Available at: https://www.emergenetwork.org. Accessed September 15, 2021.

5. Ong MEH, DoShin S, Tanaka H, et al. Pan-Asian Resuscitation Outcomes Study (PAROS): rationale, methodology, and implementation. Acad Emerg Med. 2011;18:890-7.

6. Wright D, Crouch R, Clancy M. Role of networks in supporting emergency medicine research: findings from the Wessex Emergency Care Research Network (WECReN). Emerg Med J. 2005;22:80-3.

7. Mahajan P, Shu-Ling C, Gutierrez C, et al. A global survey of emergency department responses to the COVID-19 pandemic. West J Emerg Med. 2021;22(5):1037-44.

8. Rieke N, Hancox J, Li W, et al. The future of digital health with federated learning. NPJ Digit Med. 2020;3:119.

9. New DIAMOND portal allows CTSA members to access, share training and assessment materials. 2020. Available at: https://michr.umich.edu/news/2018/10/15/new-diamond-portal-allows-ctsa-members-to-access-share-training-and-assessment-materials. Accessed September 15, 2021.

10. Stoner MJ, Mahajan P, Bressan S, et al. Pediatric emergency care research networks: a research agenda. Acad Emerg Med. 2018:1336-44.

11. Schmitzberger F, Belle A, Mahajan P. Characteristics of emergency departments participating in the Emergency Medicine Education and Research by Global Experts (EMERGE) Network. Consortium of Universities for Global Health (CUGH). 2020:104-5.