Evaluating the impact of alcohol minimum unit pricing on deaths and hospitalisations in Scotland: a controlled interrupted time series study

Summary Background Since May 1, 2018, every alcoholic drink sold in Scotland has had minimum unit pricing (MUP) of £0·50 per unit. Previous studies have indicated that the introduction of this policy reduced alcohol sales by 3%. We aimed to assess whether this has led to reductions in alcohol-attributable deaths and hospitalisations. Methods Study outcomes, wholly attributable to alcohol consumption, were defined using routinely collected data on deaths and hospitalisations. Controlled interrupted time series regression was used to assess the legislation's impact in Scotland, and any effect modification across demographic and socioeconomic deprivation groups. The pre-intervention time series ran from Jan 1, 2012, to April 30, 2018, and for 32 months after the policy was implemented (until Dec 31, 2020). Data from England, a part of the UK where the intervention was not implemented, were used to form a control group. Findings MUP in Scotland was associated with a significant 13·4% reduction (95% CI –18·4 to –8·3; p=0·0004) in deaths wholly attributable to alcohol consumption. Hospitalisations wholly attributable to alcohol consumption decreased by 4·1% (–8·3 to 0·3; p=0·064). Effects were driven by significant improvements in chronic outcomes, particularly alcoholic liver disease. Furthermore, MUP legislation was associated with a reduction in deaths and hospitalisations wholly attributable to alcohol consumption in the four most socioeconomically deprived deciles in Scotland. Interpretation The implementation of MUP legislation was associated with significant reductions in deaths, and reductions in hospitalisations, wholly attributable to alcohol consumption. The greatest improvements were in the four most socioeconomically deprived deciles, indicating that the policy is positively tackling deprivation-based inequalities in alcohol-attributable health harm. Funding Scottish Government.


Contents Page
. Study flow diagram 3 Table S1. ICD-10 code definitions of study outcomes 4 Table S2. Change in outcomes from controlled models associated with the implementation of alcohol minimum unit pricing legislation, by sex 5 Table S3. Change in outcomes from controlled models associated with the implementation of alcohol minimum unit pricing legislation, by age-group 6   Table S4. Change in death outcomes from controlled models associated with the implementation of alcohol minimum unit pricing legislation, by socio-economic deprivation decile 7   Table S5. Change in hospitalisation outcomes from controlled models associated with the implementation of alcohol minimum unit pricing legislation, by socio-economic deprivation decile 7   Table S6. Change in outcomes associated with alcohol minimum unit pricing legislation, by sex, Scotland uncontrolled model 8 Table S7. Change in outcomes associated with alcohol minimum unit pricing legislation, by sex, England uncontrolled model 9 Table S8. Change in outcomes associated with alcohol minimum unit pricing legislation, by age-group, Scotland uncontrolled model 10 Table S9. Change in outcomes associated with alcohol minimum unit pricing legislation, by age-group, England uncontrolled model 11 Table S10. Change in death outcomes associated with alcohol minimum unit pricing legislation, by socio-economic deprivation decile, Scotland uncontrolled model 12 Table S11. Change in death outcomes associated with alcohol minimum unit pricing legislation, by socio-economic deprivation decile, England uncontrolled model 12 Table S12. Change in hospitalisation outcomes associated with alcohol minimum unit pricing legislation, by socio-economic deprivation decile, Scotland uncontrolled model 13 Table S13. Change in hospitalisation outcomes associated with alcohol minimum unit pricing legislation, by socio-economic deprivation decile, England uncontrolled model

Scotland (intervention group)
A country of the United Kingdom where alcohol minimum unit pricing was implemented

England (control group)
A country of the United Kingdom where alcohol minimum unit pricing was not implemented

Pre-intervention period:
January 2012 to April 2018 Intervention implemented: May 2018 to December 2020

Non-intervention period:
January 2012 to December 2020

Pseudo intervention period:
May 2018 to December 2020

Single interrupted time series:
Assess change in level of outcome between preintervention and the period following the implementation of the intervention

Controlled interrupted time series:
Assess change in level of outcome between pre-intervention and the period following the implementation of the intervention in Scotland, controlling for those like-for-like changes in England Deaths and hospital admissions wholly attributable to alcohol consumption

Study outcomes
Deaths and hospital admissions wholly attributable to alcohol consumption

Single interrupted time series:
Assess change in level of outcome between nonintervention and the pseudo intervention period Table S1. ICD-10 code definitions of study outcomes  Type  Study outcome  ICD-10 definition   All outcomes wholly  attributable to alcohol  consumption  Table S2. Change in outcomes from controlled models associated with the implementation of alcohol minimum unit pricing legislation, by sex Study outcome Effect estimate, % (95% confidence interval)

Data permissions and Statistical Analysis Plan
All Scottish data used in this study were sourced from Public Health Scotland (PHS). English hospitalisations data were sourced from National Health Service (NHS) Digital. An application to request this data was sent to NHS Digital, and following approval, a data sharing agreement was drafted and co-signed by NHS Digital and PHS. English deaths data were sourced from the Secure Research Service (SRS) of the Office of National Statistics (ONS). An application to request this data was sent to ONS, and was subsequently approved (study reference number 1011523). GMAW, DFM, CF, JL, and LG were certified accredited researchers and were the study members with access to all study data.
A pre-specified version of the study's Statistical Analysis Plan was published by Public Health Scotland on 29 November 2022. The Statistical Analysis Plan is available at the link below: https://publichealthscotland.scot/publications/an-analysis-plan-for-the-evaluation-of-the-impact-of-alcoholminimum-unit-pricing-on-deaths-and-hospital-admissions-in-scotland/.
We report the following changes to our Statistical Analysis Plan: • We did not undertake analysis on outcomes defined using any-cause alcohol harm. In our Statistical Analysis Plan, we had indicated this definition may be problematic because there are more coded fields in the English, compared to Scottish, data sources. Upon data retrieval we found that the rate of outcomes were higher in England than in Scotland. This was considered to be due to the increased opportunity to identify outcomes for England, compared to Scotland. These outcomes were therefore not carried forward, as alcohol health harms are understood to be higher in Scotland than England. • There were several outcomes that we could not estimate effects for. This was due to the low number of outcomes observed. Where this is the case, we have not displayed any estimates. • We undertook one additional sensitivity analyses for our main deaths and hospitalisations outcomes which involved modelling the difference between the Scottish and English time series to estimate effects, as an alternative to our standard approach which was to add the English time series data as a model covariate in Scottish models. • Outcomes which are partially attributable to alcohol consumption will be reported by Public Health Scotland in an institutional report released on March 21, 2023.

Introduction
Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 1,2 See study introduction Objectives 3 State specific objectives, including any prespecified hypotheses 1,2 See study introduction

Study design 4
Present key elements of study design early in the paper 2,3 See study methods

Setting 5
Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection 2,3 See study methods Participants 6 (a) Cohort study-Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up Case-control study-Give the eligibility criteria, and the sources and methods of case ascertainment and control selection. Give the rationale for the choice of cases and controls Cross-sectional study-Give the eligibility criteria, and the sources and methods of selection of participants 2,3 See study methods (b) Cohort study-For matched studies, give matching criteria and number of exposed and unexposed