Progress

Prediction tools are instruments that use existing data to predict outcomes, in the short-term or long-term, for an individual or group. Historical data are used to build a mathematical model that captures important trends. Current data are put into the model to predict what will happen next, or to suggest actions to take for optimal outcomes. To improve lifecycle health, what is needed is not identifying more risk factors. Rather, we need to know how existing risk and resilience factors work together, and the critical periods of these interactions where interventions or supports could be provided. We also need to integrate environmental, biological, and social factors to develop complex models with large explanatory value and that can predict resilience and risk. In doing so, these models can help inform health practice, policy and behaviours.

We are conducting research to build a comprehensive library of early life resiliency and risk factors and their influences on developmental outcomes and disease pathways. We are also testing new approaches with machine learning to understand gene-environment interactions to predict child health outcomes. These efforts, and others, are first steps to create datasets and mathematical models that could help inform, in early life, about future health trajectories.

Health promotion and participatory decision-making requires the input of knowledge users across sectors and disciplines, including experts-by-experience. As researchers and clinicians, we know we do not have all the answers, and we cannot do this work alone. We are undertaking community engagements and using co-design approaches to brainstorm new or reframed questions that may reveal novel solutions or paths to pursue, or new ways of thinking, around promoting, predicting, and improving lifecycle health.  Through these collaborative efforts we aim to: identify and understand the value and needs of an early life health promotion and knowledge mobilisation tool; understand the evidence and experiences that bolster the case for new and improved approaches for early health prediction and decision-making; define what a successful roadmap and tool might look like and; identify next steps, ways to maintain momentum, and information flows. 

Health promotion and knowledge mobilisation tool prototyping must consider tool design, features, accessibility and functions. Our preliminary work starts upstream of tool prototyping, where we first aim to harmonise datasets and scientific literature so that our tool can leverage the available existing knowledge on biological, social and environmental determinants of health in early life, to use in the mathematical models and educational materials we are developing. We are using natural language processing (NLP) to gather and extract this scientific knowledge. We are also developing knowledge graphs (KGs) that can take this diverse and often fragmented information and better visualise the relationships that exist between health determinants. This can allow us to understand mechanisms of disease and potential areas for support and interventions. Collectively, this work will help us refine the tool’s functions.

We also know that for health technologies to effectively improve or facilitate the delivery of health information and services, these technologies need to be both accessible and user friendly, otherwise they will not meet the needs of their stakeholders and will have low uptake and impact. We are working to understand the views, values, and needs of knowledge users so that these are incorporated in the design of our tool. This work will help us develop the tool’s design, features, accessibility and functions.