We used a natural experimental study design to look at the impact of MUP on deaths and hospital admissions caused wholly or partially by alcohol consumption in Scotland. We used data for England as the main geographical control group. Data for English regions was used in sensitivity analyses.

• Our study design allowed us to estimate the impact of minimum unit pricing (MUP) based on the difference between outcomes following the implementation of MUP compared to a best estimate of what would have been observed had MUP not been implemented. This best estimate of what would have occurred in the absence of MUP was developed by controlling for differences in the trend and level of health harms in England, where MUP has not been implemented, and in Scotland, by observing trends in health harms prior to MUP. Adjustments were also made to incorporate differences in COVID-19- associated restrictions over time in each country. This approach allowed us to estimate the impact of MUP by isolating it from other factors which might impact health harms, such as the COVID-19 pandemic.
• When we refer to a reduction, an increase, or no change in a health outcome following MUP implementation, this is our best estimate of the impact of MUP in comparison to what could have been expected in the absence of MUP.

Deaths

• After more than two and a half years of implementation, our best estimate is that MUP significantly reduced deaths wholly attributable to alcohol consumption by 13.4% in Scotland, when using a method that accounts for deaths in a geographical control area (England), where the policy was not implemented, and underlying seasonal and secular trends. We estimate that an average of 156 deaths wholly attributable to alcohol consumption were averted each year over the study period following MUP implementation. 
• The overall reduction was driven by a 14.9% significant reduction in deaths from chronic causes wholly attributable to alcohol consumption, with significant reductions observed for both alcoholic liver disease and alcohol dependence syndrome. There was some evidence to suggest an increase in deaths from acute causes wholly attributable to alcohol consumption (6.6%), although this effect was more uncertain. This is in part due to acute causes contributing a relatively small proportion of all deaths wholly attributable to alcohol consumption.
• Significant reductions in deaths wholly attributable to alcohol consumption were estimated for: males (-14.8%), females (-12.0%), 35- to 64-year-olds (-10.0%) and those aged 65 years and over (-26.7%). All changes were  driven by deaths from chronic causes, such as alcoholic liver disease. Our results suggest that any increase in deaths from acute causes wholly attributable to alcohol consumption was likely driven by males (4.4%), with little evidence of any change for females (0.2%).
• Significant reductions in deaths wholly attributable to alcohol consumption were greatest among the four most socio-economically deprived area-based deciles, suggesting that MUP acted to reduce inequalities in alcohol-attributable deaths in Scotland.
• Our main estimate, a significant reduction of 13.4% in deaths wholly attributable to alcohol consumption, was robust to a range of different conditions as tested through our sensitivity analyses, providing greater certainty in our main finding.
• Deaths partially attributable to alcohol consumption were estimated to reduce by 8.4% in the study period following the implementation of MUP, although this effect was less certain than the estimated effect for deaths wholly attributable to alcohol consumption. Significant reductions in deaths from chronic causes partially attributable to alcohol consumption (-12.7%) offset a 7.8% increase in deaths from acute causes  partially attributable to alcohol consumption, although this effect on acute deaths was more uncertain.

Hospital admissions

• After more than two and a half years of implementation, our best estimate is that MUP reduced hospital admissions wholly attributable alcohol to alcohol consumption by 4.1% in Scotland, when using a method that accounts for admissions in a geographical control area (England), where the policy was not implemented, and underlying seasonal and secular trends. We estimate that an average of 411 hospital admissions wholly attributable to alcohol consumption have been averted each year over the study period following MUP implementation. There was slightly more uncertainty surrounding this result than for the estimates of reduced deaths.
• The estimated overall reduction was driven by a significant 7.3% reduction in hospital admissions for chronic conditions wholly attributable to alcohol consumption, achieved through significant reductions in hospital admissions for alcoholic liver disease (-9.8%) and alcohol psychoses (-7.2%).
• Hospital admissions for acute conditions wholly attributable to alcohol consumption were estimated to have increased by 9.9%, although this effect was more uncertain than for chronic conditions. This was most likely driven by a significant increase among females (15.6%). There was also some evidence of an increase among males (8.5%, although the effect for males was less certain than that for females. As admissions for acute conditions were less common, there was an overall reduction in hospital admissions wholly attributable to alcohol consumption.
• Significant reductions in hospital admissions wholly attributable to alcohol consumption were estimated for males (-6.2%), and while there was some evidence of an increase among females (3.1%) this effect was more uncertain. A reduction in hospital admissions wholly attributable to alcohol consumption was estimated for those aged 35 to 64 years (-4.8%); while there was some uncertainty around this effect, evidence for changes in other age groups was weaker.
• Reductions in hospital admissions wholly attributable to alcohol consumption were greatest among the four most deprived area-based deciles, suggesting that MUP acted to reduce inequalities in alcohol-attributable hospital admissions in Scotland.
• Results from the sensitivity analyses were varied, suggesting less certainty around the impact of MUP on alcohol-attributable hospital admissions, than on alcohol-attributable deaths.
• Hospital admissions partially attributable to alcohol consumption were estimated to have reduced by 3.4% although this effect was more uncertain. Any reductions were driven by significant reductions among males (-6.9%), particularly for chronic conditions. We estimated that the implementation of MUP was associated with a significant increase in female hospital admissions for acute conditions partially attributable to alcohol consumption.

Conclusion

• We conclude that the implementation of MUP has reduced alcohol-attributable health harms. The strongest evidence was that MUP reduced deaths wholly attributable to alcohol consumption, with reductions across all alcohol-attributable harm being primarily driven by reductions in chronic outcomes. Furthermore, our study has evidenced that MUP has acted to reduce deprivation-based inequalities in alcohol-attributable health harms.