Refuting tobacco-industry funded research: empirical data shows decline in smoking prevalence following introduction of plain packaging in Australia

Introduction:
Country legislation to introduce plain packaging of tobacco products has been vigorously attacked by the tobacco industry claiming that there is no measurable impact on smoking rates on the basis of two industry-funded working papers which examined trends in smoking prevalence in Australia. Objective: To assess the effect of plain packaging on smoking prevalence in Australia, taking into account key tobacco control measures, with the aim of investigating the findings of the industry-funded study.

Methods:
Monthly smoking prevalence and sample sizes from repeat cross-sectional surveys were reconstructed from the published working paper using an original reverse-engineering technique that achieved nearly 100% accuracy and analysed as a time series using logistic regression. Indicator variables reflecting comprehensive smoke-free policy, graphic health warnings, 25% taxation increase, and introduction of plain packaging were constructed from official information.

Results:
Smoking prevalence declined from 25% to 18% over the 13 year period – an overall 28% relative reduction or average 2.8% (95% confidence interval 2.6% - 2.9%) annual reduction. A significantly improved fit was obtained by the full model which included terms for tax increase (4.8%, 2.7% - 6.8% reduction), comprehensive smoke-free policy (4.5%, 1.7% - 7.2% reduction) and plain packaging (3.7%, 1.1% - 6.2% reduction) in addition to an adjusted average annual reduction of 1.7% (1.3% - 2.2%).

Conclusions:
A significant decline in smoking prevalence in Australia followed introduction of plain packaging after adjustment for the impact of other tobacco control measures. This conclusion is in marked contrast to that from the industry-funded analysis.