Vigilancia de la resistencia bacteriana en pediatria y su relación con el usos de antibióticos por medio de análisis de series temporales.

Abstract

Background: Time series analysis are a group of techniques aimed at adjusting a mathematical model to a series of observations taken over time. Time series analysis permits assessment of the relationships between antimicrobial use and resistance. ViResiST Project (www.viresist.org) is a surveillance system of antimicrobial resistance and use. ViResiST Project uses time series analysis to predict for the level of antimicrobial resistance for several hospitals in Spain, Scotland (U.K.), Belgium, Holland, France and United States of America. Aim: 1) Investigate the prevalence of antimicrobial resistance in pediatric population in Comunidad Valenciana (Spain) 2) Constructing autoregressive integrated moving average (ARIMA) models that analyze the temporal behaviour of bacterial resistance series 3) Assessment of the relationship between antimicrobial use and resistance by using transfer functions (Box-Jenkins) 4) Identify those situations of high level of antimicrobial resistance which may need measures of control of resistance. Methods: The data used were exported from ViResiST Project data bases. ViResiST Project collected data of antimicrobial resistance in children and antibiotic use in global population for 5-6 years from each hospital. We used the data from five hospitals of the Comunidad Valenciana (H. Dr. Peset de Valencia, H. de la Vega Baja de Orihuela (Alicante), H. Clínico de Valencia, H. General de Elche, H. General de Castellón) y two european hospitals (Aberdeen Royal Infirmary in Scotland (U.K.) y and the Academic Hospital of Rotterdam in Holland). Percentages of antimicrobial-resistant/intermediate microorganisms for the last 3 years were obtained. Data were analyzed as time series and ARIMA (Box-Jenkins) and transfer function models were built. Results: 976 antimicrobial-microorganism combinations were analyzed. We could demostrate a temporal relationship between antimicrobial use and resistance for 60 antimicrobial-microorganism combinations. Resistances and antibiotic use were higher in the spanish settings than in Aberdeen. Conclusions: ViResiST Project data are useful to know antimicrobial resistance in pediatric population and to guide empirical prescriptions of antimicrobials in pediatric infections. Widespread use of antibiotics, particularly ?-lactams in the spanish settings accounts for selection of more antibiotic resistant microorganisms.