Table 3.

Impact of the intervention on the percentage reduction in each of the quality indicators across follow-up (median 19 months)

ESAC Quality Indicators [25]	Absolute Effect of intervention (95% CI)*	% reduction in intervention group relative to control group (95% CI)*	p-value*
1. DID of antibiotics for systemic use	−0.63 (− 0.78, − 0.48)	− 4.23 (− 5.26 to − 3.21)	< 0.0001
2. DID of penicillins	−0.52 (− 0.63, − 0.42)	−6.51 (− 7.92 to − 5.22)	< 0.0001
3. DID of cephalosporins	−0.07 (− 0.11, − 0.03)	− 3.89 (− 6.18 to − 1.65)	0.0002
4. DID of macrolides, lincosamides and streptogramins	−0.06 (− 0.09, − 0.03)	−3.45 (− 5.23 to − 1.70)	< 0.0001
5. DID of quinolones	−0.01 (− 0.05, 0.02)	−0.47 (− 2.37 to 0.93)	0.3736
6. PCT of beta-lactamase-sensitive penicillinsa	−0.01 (− 0.06 to 0.05)	− 2.38 (− 15 to 11.63)	0.8223
7. PCT of combinations of penicillins with beta-lactamase inhibitors a	−1.09 (− 1.42, − 0.76)	− 2.79 (− 3.65 to − 1.94)	< 0.0001
8. PCT of 3rd- and 4th-generationcephalosporins a	−0.24 (− 0.36, − 0.12)	−5.69 (− 8.63 to − 2.80)	0.0001
9. PCT of quinolones	0.13 (− 0.06 to 0.33)	0.91 (− 0.42 to 2.30)	0.1709
10. Ratio of consumption of broad- to narrow-spectrum penicillins, cephalosporins and macrolides	- 5.37 (−8.23 to − 2.51)	−8.97 (− 13.99 to − 4.12)	0.0002

Abbreviations. PCT Percentage, DID Defined daily doses per 1000 inhabitants per day, ESAC European Surveillance of Antimicrobial Consumption

* Calculated from a mixed-effects model. The estimated value for intervention measures is an interaction between the variable “group” (intervention vs. control) and the variable “period” (post-intervention vs. baseline), adjusted for the secular trend of the indicator. The models were adjusted for the covariates of Table 1 and the seasonality. The percentage reduction was calculated by using the baseline values of the control group as reference values

^a Percentage of total consumption of antibiotics for systemic use in DID