Advancing evidence to enable optimal communicable disease control

The COVID-19 pandemic brought epidemiology into public focus globally. Understanding patterns and determinants of disease spread was central to risk assessment and the modelling of drivers of transmission to forecast outcomes under different intervention scenarios. Epidemiological analytics, including the reproduction number, were being discussed by the media and the public in ways epidemiologists and biostatisticians could not have foreseen. Yet the statistics being reported were largely confined to two ends of the evidence spectrum - at one end, raw case counts, hospitalisations and deaths, and at the other, sophisticated statistical modelling based on disease dynamics averaged at the whole-of-population level. Other core epidemiological analytic methods that add a more nuanced understanding of variation in disease transmission within and across populations, and risk of infection, were underrepresented. In Australia, for example, the purposeful collection of data to estimate subpopulation-specific case rates, generate relative risks across subpopulations and allow meaningful interpretation within and across populations was limited. This also hampered the real-world evaluation of specific health interventions, including vaccination, and the generation of updated population-specific estimates for statistical model parameters. This was a global phenomenon, though some countries did better than others. What was fundamentally missing was a clear investment in, and coordinated approach to, the quality of surveillance data needed for (a) tracking disease transmission and the degree of control achieved, both of which changed over time, and (b) public communication. The independent inquiry into the Australian Government's COVID-19 Response had evidence generation as a central theme, and investment in evidence synthesis capability and data sharing as clear recommendations for the way forward. The importance of evidence was also raised in discussions informing the draft global Pandemic Agreement. This remains a worrying gap in pandemic readiness, including in well-resourced countries such as Australia where the nuance in public health policy was constrained by the reliance on basic descriptive epidemiology, urban-focused population-level modelling and data insights imported from other countries.