International Journal of Antimicrobial Agents, Vol. 55 No. 1 (2020) 105800
[available on the Internet at <https://www.sciencedirect.com/science/article/pii/S092485791930247X#!> (last accessed on 24 April 2020)].
🔴 To the best of authors’ knowledge, this is the first time that an econometric time series model (obtained through a methodology inspired by the well-known Box–Jenkins technique) is used for characterising EU prevalence of antimicrobial resistance by means of EARS-Net EU nosocomial resistance data. Hereby, it has been possible to statistically test if there is any change in the global behaviour of this data.
🔴 The statistically proven brakes and reversions observed in the EU antimicrobial resistance prevalence constitute noteworthy milestones in the tackling of antimicrobial resistance in Europe.
🔴 It has been proven through an AR model with interventions a growth deceleration in the global increasing tendency of Escherichia coli resistance to third-generation cephalosporins, resulting in a much softer slope. Concerning Klebsiella pneumoniaeand Escherichia coli resistanceto fluoroquinolones in the EU, a slowdown in their increasing tendency has also been proven but, in these cases, towards a stabilization into a constant value.
🔴 It has been proven via an AR econometric model with interventions a modification in the growing trend of K. pneumoniae and P. aeruginosa resistance to third-generation cephalosporins that consists in an inversion into a very slightly descending slope.
🔴 For all these data series, there is a temporal association between the statistically proven deceleration in their global tendency and the enforcement of EU referral procedures that optimize and restrict the use of different antibiotics in the treatment of diseases in human and animals.
Infections with bacteria harbouring resistance to cephalosporins or fluoroquinolones (FQ) constitute a serious hazard to human health.
To establish a methodology based on econometric analysis and the largest European Union (EU) resistance database (EARS-Net), to model nosocomial antimicrobial resistance (AMR) in the EU and to detect tendency changes, steps or peaks. The contribution of legislation based on third-generation cephalosporin (3GC) and FQ class referrals to resistance rate patterns is evaluated.
Resistance to 3GC and FQ was examined in nosocomial Escherichia coli, Klebsiella pneumoniae and Pseudomonas aeruginosa in at least 25 out of 30 EU countries (> 94% population coverage), weighted by their mean annual population, between 2006 and 2016. Autoregressive integrated moving average (ARIMA) model analysis, inspired by Box–Jenkins methodology, was prepared to adjust series to a mathematical model to detect hypothetical changes in the general behaviour. To the best of the authors’ knowledge, this is the first study to use ARIMA with interventions to model overall nosocomial AMR data compiled in EARS-Net.
Results and conclusions
Econometric ARIMA models statistically prove the occurence of slowdowns and reversions in the increasing trend of AMR prevalence in nosocomial E. coli and K. pneumoniae to 3GC and FQ, as well as resistance of P. aeruginosa to 3GC. The resistance of P. aeruginosa to FQ exhibited a descending slope. The presented decreasing trends constitute noteworthy milestones in tackling AMR in Europe.