Using Direct and Indirect Estimates for Alcohol-Attributable Mortality: A Modelling Study Using the Example of...

Area of Science:

• Public Health
• Epidemiology
• Health Economics

Background:

• Comparative risk assessments (CRAs) for alcohol use typically rely on indirect harm estimates.
• Eastern European CRAs often use Russian cohort studies instead of global risk relations.
• A Lithuanian tax intervention provided a unique opportunity to compare direct and indirect mortality estimates.

Purpose of the Study:

• To compare direct and indirect estimates of alcohol-attributable mortality in Lithuania post-tax increase.
• To evaluate the accuracy of Russia-specific versus global risk relations in CRAs.

Main Methods:

• A statistical modeling study compared direct (time-series) and indirect (attributable-fraction) mortality estimates.
• The study analyzed data before and after a significant alcohol excise tax increase in Lithuania.
• Russia-specific and global relative risks were assessed against a time-series gold standard.

Main Results:

• Indirect estimates underestimated the reduction in alcohol-attributable mortality by at least 63% compared to direct estimates.
• Russia-specific estimates were closer to the direct estimates than global ones.
• Higher estimates for alcohol-attributable cardiovascular mortality contributed to the closer Russia-specific results.

Conclusions:

• Current CRAs using indirect estimates significantly underestimate alcohol-attributable mortality reduction.
• Global relative risks and price elasticities require re-evaluation.
• New regional estimates based on cohort studies are recommended to replace global ones.