Fig 1.
The workflow of the modeling approach.
Consumption and resistance data were acquired from ECDC, and other parameters were found in the literature or used as free parameters. The model was fit to the data reported by ECDC to optimize the free parameters. Sensitivity analyses were performed to test the robustness of the model. Counterfactual scenarios were applied to understand the functional dependencies of the prevalence of resistance from possible drivers.
Table 1.
Free model parameters of the fit.
Fig 2.
The model was fitted to the data of the annual prevalence of resistance in Klebsiella pneumoniae reported by ECDC from 2005 to 2015. Circles represent the reported data, and solid and dotted lines represent the fit with variable between-country and uniform for all hospital transmission rates, respectively.
Fig 3.
Correlation between antibiotic consumption and prevalence of resistance.
Correlation between the consumption of different classes of antibiotics in different settings (x-axes), and the prevalence of resistance to 3rd generation cephalosporins (A), prevalence of resistance to carbapenems (B). Consumption rates are given as mean yearly consumption in the years 2006–2015 in DDD per day per 1000 inhabitants.
Fig 4.
Counterfactual scenarios corresponding to variation of hospital transmission rate.
Plots represent the dependence of change in prevalence of resistant strains between 2005 and 2015 on the level of the hospital transmission rate. Green and purple areas represent the decrease and increase in hospital transmission rate, respectively.
Fig 5.
Counterfactual scenarios corresponding to variation of antibiotic consumption.
Plots represent the dependence of change in prevalence of resistant strains between 2005 and 2015 on the level of antibiotic consumption. Green and purple areas represent the decrease and increase in antibiotic consumption, respectively. (A) Outpatient consumption of 3rd and 4th generation cephalosporins (B) Inpatient consumption of 3rd and 4th generation cephalosporins (C) Inpatient consumption of carbapenems.
Fig 6.
Counterfactual scenarios corresponding to variation of import of ESBL strain.
Plots represent the dependence of change in prevalence of resistant strains between 2005 and 2015 from the level of the import of ESBL strain. Green and purple areas represent the decrease and increase in import of ESBL strain, respectively.