Skip to main content
. 2021 Jan 3;116(7):1709–1715. doi: 10.1111/add.15352

Table 1.

Interrupted time–series analysis: estimated effect on beer per capita consumption (BPC) per week (LALs) of the first‐ and second‐wave COVID‐19 restrictions in Australia

n On‐premises BPC Off‐premises BPC
Date (week beginning) LALs P‐value 95% CI LALs P‐value 95% CI
Starting level (pre‐intervention intercept) 27 July 2015 0.0139 < 0.001 0.013, 0.014 0.0480 < 0.001 0.047, 0.050
Trend before first‐wave restrictions 27 July 2015–16 March 2020 247 < −0.0001 0.463 < −0.001, < −0.001 < −0.0001 0.001 < −0.001, < 0.001
Intervention 1
Level change after first‐wave restrictions 23 March 2020 −0.0129 < 0.001 −0.013, −0.012 +0.0057 0.610 −0.016, 0.028
Trend change after first‐wave restrictions 23 March–11 May 2020 8 < 0.0001 0.770 −0.001, < 0.002 −0.0009 0.673 −0.005, 0.003
Intervention 2
Level change after partial removal of first‐wave restrictions 18 May 2020 +0.0026 0.006 0.001, 0.004 +0.0101 0.198 −0.005, 0.025
Trend change after partial removal of first‐wave restrictions 18 May–29 June 2020 7 +0.0015 < 0.001 0.001, 0.002 +0.0002 0.926 −0.004, 0.005
Intervention 3
Level change after second‐wave restrictions 6 July 2020 −0.0041 0.001 −0.006, −0.002 +0.0049 0.384 −0.006, 0.016
Trend change after second‐wave restrictions 6 July–5 October 2020 14 −0.0015 < 0.001 −0.002, −0.001 +0.0006 0.287 < 0.001, 0.002

Total n = 276 (weeks beginning 27 July 2015–5 October 2020, inclusive).

BPC = beer per capita consumption;

LALs: litres of pure alcohol.

Per capita is based on the Australian population aged 15+ years.

On‐premises: beer supplied from containers > 48 litres (e.g. kegs). Off‐premises: beer supplied from containers < 48 litres (e.g. cans, bottles).

Models adjusted for autocorrelation with lag length based on Cumby–Huizinga test: on‐premises BPC (lag 3, P‐value = 0.0003); off‐premises BPC (lag 1, P‐value = 0.0032).