Gains in life expectancy in the Australian population due to reductions in smoking: comparisons between interventions targeting the population versus interventions in a specific high risk group

BACKGROUND: Four decades of population-based tobacco control strategies have contributed to substantial reduction in smoking prevalence in Australia. However, smoking prevalence is still double in socially disadvantaged groups compared to those that are not. But not all tobacco control strategies successfully used in the general population is effective in specific high-risk population groups. Hence, an effective way to reduce smoking in high risk population groups may include targeting them specifically to identify and support smokers to quit. In this backdrop, we examined whether tobacco control interventions at the population-level are more effective in increasing life expectancy among Australians compared to interventions targeting a high risk group or a combination of the two when smoking prevalence is reduced to 10 and 0% respectively. METHODS: Using the risk percentiles approach, analyses were performed separately for men and women using data from various sources such as the 2014–15 National Health Survey linked to death registry, simulated data for high risk groups, and the Australian population and deaths data from the census. Indigenous status was simulated by preferentially assigning those who are indigenous to lower SES quintiles. The age-sex distribution of mental disorder status was simulated using its distribution from 2016 National Drug Strategy Household Survey with 25.9% of mentally ill being assigned to current smoking category and the rest to non-smoking category. The age-sex distribution of prisoners was simulated based on 2014 ABS Prisoners Australia survey with 74% of prisoners being assigned to current smoker category and the rest to non-smoker category. Homelessness status was simulated according to age, sex and indigenous status for 2011 census with all homeless being allocated to the lowest SES category. The age-sex distribution of total cholesterol level was simulated based on 2011–13 Australian Health Survey. RESULTS: The results showed that the combined approach for reducing smoking is most effective for improving life expectancy of Australians particularly for the socially disadvantaged and mentally ill groups both of which have high fraction of smokers in the population. For those who were mentally ill the gain in ALE due to reduction of smoking to 10% was 0.53 years for males and 0.36 years for females which were around 51 and 42% respectively of the maximal gains in ALE that could be achieved through complete cessation. CONCLUSIONS: Targeting high-risk population groups having substantial fraction of smokers in the population can strongly complement the existing population-based smoking reduction strategies. As population and high risk approaches are both important, the national prevention policies should make judicious use of both to maximize health gain.