To ensure a continued robust and effective capacity for infectious disease modelling, we need to create a sustainable, independent base of scientists and modellers who can be mobilized to provide science advice when needed. To provide effective advice we must ensure that the channels of communication and relationships among members as well as those in public health agencies are sufficiently well established in advance so that quick action and consensus on scientific advice will be possible in the face of unexpected and time-sensitive public health crises.

To facilitate these goals, we envision a permanent institute, or a federated network of institutions, to support public health organizations in providing this capacity. This includes coordinating research in this space across institutions, providing funding for continued modelling and capacity building across departments and organizations to ensure that Canada has the capacity to better deal with future crises, and providing capacity to pull in researchers to work on policy activities when the next crisis does occur. In this section we outline the overall vision for future research and training in emerging infectious disease modelling (EIDM) in Canada, while in the next section we focus more specifically on the proposed institute.

It is important that the sustainable, independent base of modellers, described above, also collaborates with a complementary group of strong modellers and epidemiologists in public health institutions who bring the following essential elements:

• modelling with deep enough knowledge to identify the utility of modelling innovations and opportunities for application to support public health objectives;

• collaboration with academics (including co-supervising/training HQP);

• modelling methods to develop outputs and emerging consensus in public health institutions; and

• knowledge translation for non-specialist public health managers and policy-makers.

To have these elements in-house, within public health institutions, is an essential component of the “modelling ecosystem” needed to provide effective scientific advice to the government when needed.

An integral part of the creation of an emerging EIDM organizational structure will be the development of well-grounded advisory components that are sufficiently independent of government and that can convene a range of scientific experts allowing for a full airing of potentially opposing views. We envisage that most of the members of this network will likely be based at academic institutions and will participate in the network’s activities as part of their own, self-directed, research programs. At the same time, when acute public health issues arise where decision-makers require prolonged and ongoing input by a member or members (e.g., the emergence of a novel infectious disease) there needs to be a mechanism through which these members can be released from their normal work responsibilities.

For the structure to be effective in providing holistic and relevant advice to decision-makers, it will need to integrate a broad range of scientific and modelling expertise. This not only includes fully integrated economic modelling (something that was lacking during the initial stages of the COVID-19 pandemic), but also integrated immunological, virological, clinical, vaccinology and evolutionary expertise. A suite of other perspectives must also be integrated from behavioural and social sciences (e.g., political scientists, sociologists, psychologists). Importantly, Indigenous perspectives and knowledge must also be included wherever and whenever possible.

Administratively, we view the future EIDM structure as an apolitical authority, recognisably independent of government, that can help to provide transparency, consensus (in the context of uncertainty and possibly multiple viewpoints), and a scientific evidence base that supports the identification of “serviceable truths” increasingly sought by public health decision-makers, particularly in an emerging pandemic context, where science is rapidly changing and decisions have society-wide impacts. Ideally this structure needs to be led by a person who is dedicated to the position full time, preferably seconded from an academic position, so that they do not lose their contact with current research. The position should rotate within the community of expertise.

EIDM requires a toolset typically not found in existing Canadian discipline-based graduate programs. Thus, a central component of our approach would be to provide our emerging workforce with holistic training that includes applied mathematics, computer science, data science, epidemiology, evolution, Indigenous health and reconciliation, infectious disease, public health, statistics and visualization. In addition, there is a need to provide experiential opportunities for trainees at all levels to conduct joint research with academic mathematical and statistical modellers, as well as across disciplines (e.g., economics, behavioural sciences), and public health/policy researchers and/or agencies. Training is needed on Indigenous knowledge, Indigenous health, and building research relationships with Indigenous communities. There is also a need to provide training for the next generation to be able to provide modelling inputs during public health emergencies. Finally, there is a vital need to include training in science communication, both for researchers and those in public health. This training would be aimed at both teaching how to convey findings in a meaningful way to the public health and policy realm, but also how to convey uncertainty about some findings.

At times of crisis (e.g., an emerging pandemic), real-time access to data relevant for modelling is essential (e.g., cases, hospitalizations, deaths, and other health outcomes; genomes of sequenced samples; economic and social variables, etc.). Mechanisms should be put in place to ensure that there is robust, ongoing, real-time surveillance, and to swiftly make all relevant new data streams available to modellers via data-sharing agreements. Ideally, we will develop automated procedures that will keep databases completely up to date going forward. Distinct from traditional disease notification data, a great deal of COVID-19 data has been made public throughout the pandemic; however, not all these data continue to be accessible on public web sites. It will be important to gather all these COVID-19 data and make them permanently accessible publicly. In addition, and recognizing that there are contexts where data cannot be immediately publicly available, models, methods and international scientific approaches can be brought in to modelling teams within public health institutions via the bridges this network will maintain.

A key way to help “do better next time” is to have in place a structure that facilitates building trust and an awareness of the expertise of others. For example, we need to make it easy for journalists to find the right people to interview, for modellers to identify appropriate collaborators in public health (and vice versa), and for trainees at all levels to find suitable programs, academic advisors, or job opportunities in public health. We have an online structure already in place to help build the connections (our network graph at http://eidm-mmie.net/ ) but continuing to support its development and expansion will make it a more powerful resource. A well-maintained network graph will make it possible for anyone to identify the cadre of Canadian scientists who have modelling expertise, understand the policy process, and have connections to international networks.

Modellers within public health and in academia must have, individually or collectively, the best mathematical methods, the data, and, in one form or another, the epidemiological/ecological/biological context of the system they are exploring. This is more than bringing mathematics and data together — for this to be successful, it will require a sustained funding model.