Table 2.

Results growth models (2009–2015 alcohol-related hospital episodes statistics)

Parameter	Unconditional estimate (SE)	Conditional (1) estimate (SE)	Conditional (2) estimate (SE)
Baseline (intercept)*	5.013 (0.014)	4.702 (0.024)	4.717 (0.028)
	p<0.0001
Seasonal trend (quarter 1)	ref	ref	ref
July–September (2)	0.042 (0.004)	0.042 (0.004)	0.042 (0.004)
October–December (3)	0.020 (0.004)	0.020 (0.004)	0.020 (0.004)
January–March (4)	−0.007 (0.004)	−0.007 (0.004)	−0.007 (0.004)
	p<0.0001	p<0.0001	p<0.0001
Slope	−0.048 (0.021)	−0.010 (0.031)	−0.056 (0.057)
	p=0.022	p=0.003	p=0.047
Effect on intercept
No policy		ref	ref
Low policy		0.004 (0.027)	0.006 (0.027)
Medium		0.063 (0.028)	0.069 (0.029)
High policy		0.049 (0.031)	0.064 (0.033)
		p=0.210	p=0.047
Deprivation at baseline		2.159 (0.117)	2.254 (0.153)
		p<0.0001	p<0.0001
Population at baseline		−0.020 (0.009)	−0.028 (0.011)
		p=0.028	p=0.016
Alcohol-related crime at baseline		−0.014 (0.003)	−0.017 (0.004)
		p<0.0001	p<0.0001
Effect on slope
No policy		ref	ref
Low policy		0.001 (0.054)	−0.006 (0.055)
Medium policy		−0.044 (0.056)	−0.065 (0.058)
High policy		−0.183 (0.060)	−0.229 (0.067)†
		p=0.017	p=0.006
Deprivation at baseline			−0.300 (0.311)
			p=0.335
Population at baseline			0.027 (0.023)
			p=0.257
Alcohol-related crime at baseline			0.011 (0.008)
			p=0.158
Variance (intercept)	0.057	0.030	0.030
Variance (slope)	0.115	0.112	0.112
Covariance (intercept, slope)	−0.49	−0.62	−0.62
Variance (residual)	0.011	0.011	0.011
BIC	−9137.81	−9288.12	−9250.71

Conditional (1) baseline effects, and interaction policy and time, (2) baseline effects plus all interactions with time.

*Baseline population age-adjusted rate of alcohol-related (narrow definition) unique hospital episode statistics is exp(5.013)=150.4 (95% CI 146.3 to 154.5) per 100 000 people.

†Note that because of rescaling of the time axis, the annual decline in the high policy category (see text) is calculated by 1−exp(0.229/10)≈−2%.

BIC, Bayesian Information Criterion.