Antimicrobial resistance surveillance in Europe: regional pooling of national data from a small number of sites can be misleading

Abstract

We read with interest the results of the European component of theRegionalResistanceSurveillancestudyreportedbyJonesandcolleaguesintheir online article on October 14, 2013 ( Jones et al., 2013). This articlepresents antimicrobial susceptibility test results from a collection ofEuropean bacterial isolates with centralised determination of MICs for alarge panel of antibiotics, followed by interpretation according to 2international breakpoint de finitions. The authors suggest that the resultscan be used to validate results from national and regional antimicrobialresistance surveillance programs, such as the European AntimicrobialResistance Surveillance Network (EARS-Net) ( ECDC, 2013).Unfortunately, 2 important factors were not taken into account bythe authors, which preclude the use of their results for validationpurposes or to properly reflect the heterogeneous resistance situationin Europe. Firstly, the results were based on pooled data from only 47hospitals in 21 countries. Secondly, the authors mentioned largevariations in resistance percentages between countries, but theseinter-country differences were neither thoroughly presented, nordiscussed. As a consequence, the pooled estimates could be subject tobias due to poor representativeness and might not reflect the truepicture of antimicrobial resistance in Europe.Results from EARS-Net, based on routine microbiological data fromapproximately 1300 hospitals in Eu rope, have repeatedly shown thatdifferencesinantimicrobialresistancepercentagesbetweencountriesaresubstantial. As an example, the national percentage of invasive Klebsiellapneumoniae isolates resistant to third-generation cephalosporins variedbetween 2% and 75% among the 30 Europe an countries that participatedin EARS-Net in 2012. Similar significant inter-country differences can befound for many antimicrobial-microorganism combinations in thepublicly available on-line EARS-Net database ( ECDC, 2014).As the authors state in their conclusions, antimicrobial resistancesurveillance programs should be more widely supported to monitoremerging resistance trends and follow the impact of structuredinterventions at national, regional, and local levels. Moreover,although European readers may be tempted to use the data providedby the European component of the Regional Resistance Surveillancestudy to update their guidelines for empiric therapy of infectedpatients, due to the aforementioned limitations, much caution shouldbe exercised in doing so. While national antimicrobial resistance dataremain essential to describe the magnitude and trends in antimicro-bial resistance in and between European countries, large variationsmay exist even within 1 single country (Carbonne et al., 2013;SWEDRES-SVARM, 2013; ECDC, 2013). It is, therefore, necessary forhospitals and physicians to be familiar with their local surveillancedataandusethemasabasisforthecreationoflocal guidelinesaswellas for empiric antimicrobial treatment of infections.The Regional Resistance Surveillance study is indeed complemen-tary to other surveillance programs such as EARS-Net and should, inprinciple, provide additional, useful information. This goal, however,is only partially achieved by this study due to the small number ofparticipating sites and the fact that the pooled data at the regionallevel provide misleading information on antimicrobial resistanceacross Europe.Conflicts of interestWe declare no conflict of interest.Liselotte Diaz HogbergAnna-Pelagia MagiorakosOle E. HeuerDominique L. MonnetEuropean Centre for Disease Prevention and ControlTomtebodavagen 11a, 171 83 Stockholm, SwedenE-mail address: liselotte.diaz-hogberg@ecdc.europa.euhttp://dx.doi.org/10.1016/j.diagmicrobio.2014.03.015References