Table 4.

Logistic regression models: Risks of receiving in 2008 ≥ one prescription of systemic corticosteroids (model 1) and risk of receiving ≥ one prescription of antibiotics (model 2), according to ratio groups in ICS-treated patients (non ICS users excluded)

	Model 1: Risks of receiving in 2008 ≥ one prescription of systemic corticosteroids		Model 2: Risks of receiving in 2008 ≥ one prescription of antibiotics
	Adjusted OR (a)	95% CI	Adjusted OR (a)	95% CI
UK (n = 31,641)
Low ratio (0% < R < 50%)	1	-	1	-
High ratio (R ≥ 50%)	0.54	0.50-0.57	0.80	0.76-0.84
France (n = 3,411)
Low ratio (0% < R < 50%)	1	-	1	-
High ratio (R ≥ 50%)	0.78	0.67-0.91	0.87	0.75-1.01

^(a)Adjusted for age, gender, ≥ 1 dispensing of LABA-ICS fixed combination, the number of dispensed respiratory medication classes in 2008 (<6, 6+).