Population and Risk Modelling

Population health impact models

Computational models enable the projection of population-level effects by assessing the potential for specific outcomes. These models use mathematical methodologies to simulate the potential effects of different exposure and risk scenarios within a specified population, allowing the input of data such as product use behaviour transition probabilities (e.g. initiation, quitting, switching) and a risk estimate dependent on product use status. Customarily, such models use publicly available demographic data and published studies to inform model input parameters. Where data are not available, informed assumptions and consensus estimates can be used. The outputs of population models are often evaluated comparatively, calculating the potential health outcome of the base case scenario (e.g. cigarettes only) and comparing it to a scenario where a Smokeless Product has been introduced to a market. This comparison enables the calculation of the difference in health outcomes/ mortality (Figure 1).

Prospective modelling: predicting the population-level effect of Smokeless Products

Over the past decade, various mathematical models have been used to evaluate the potential public health impact of switching smokers to Smokeless Products. The results of these evaluations support the proposition that such products have the potential to reduce the prevalence of smoking and the related death toll across the globe.

A recent study using a vaping model by Levy et al. evaluated the potential impact on public health of relaxing restrictions on vaping products in Australia, predicting that as many as 100,000 deaths could be averted during the period 2017–2080 if Australia were to embrace Tobacco Harm Reduction by increasing access to these products.^[1] This finding aligns with similar studies on Smokeless Products' potential to reduce smoking prevalence in Germany^[2] and Russia.^[3]

Retrospective modelling: applying the Swedish Experience across the EU

An example of retrospective modelling is the analysis performed by group of physicians and academics regarding the population health impact of the Oral Tobacco Product, snus.^[4] At the time of the analysis in 2017, the total tobacco consumption level in Sweden was similar to that of other European Union (EU) countries, but Sweden consistently had much lower smoking prevalence, the reduction of which occurred alongside a rise in the use of snus. However, snus products have been banned in the European Union, aside from Sweden, since 1992.^[5]

Utilising data from the World Health Organization,^[6] which calculated the tobacco-related mortality rate of each of the relevant countries as well as the prevalence of smoking, the researchers extrapolated the impact of introducing snus in the other European countries. The resulting report predicted that had other EU Member States adopted Sweden's tobacco consumption patterns, more than 2.8 million potential deaths could have been averted over a 20-year period (Table 1).^[4]