Table 1

Results of regression analysis to detect an association between smokefree legislation in England and emergency hospital admissions for asthma

Effects*	edf†	Parameter estimate	Standard error	p value‡
Parametric terms
Smokefree
Before§
After		−0.050	0.022	0.03
Region
North East§
North West		0.050	0.022	0.02
Yorkshire and Humber		0.035	0.023	0.13
East Midlands		−0.12	0.024	<0.001
West Midlands		0.051	0.023	0.02
East of England		−0.28	0.024	<0.001
London		−0.079	0.023	<0.001
South East		−0.31	0.023	<0.001
South West		−0.22	0.024	<0.001
Smooth terms
Influenza	3.58		<0.001
Temperature	5.53		0.006
Month	7.71		<0.001
Region×year
North East	1.00		0.008
North West	6.17		<0.001
Yorkshire and Humber	2.28		0.21
East Midlands	2.79		0.06
West Midlands	2.82		0.32
East of England	1.00		0.03
London	5.56		<0.001
South East	3.67		<0.001
South West	4.30		<0.001

*AR(1) term=0.22.
†edf, the effective degrees of freedom for the smooth term is a measure of the complexity of the estimated smooth term. For example, the edf for each region in the Region×Year effect represents the degree of non-linearity of the long-term trend in that region. If edf=1, the shape of the relationship between the smooth term and the outcome is linear. Values of edf increase progressively with the addition of more non-linear relationships.
‡p value from a t test on parametric regression coefficients and F test on smooth terms. p Values for smooth terms test the null hypothesis that there is no relationship between the smooth term and emergency admissions for asthma.
§Baseline category.