Third external quality assurance scheme for Salmonella typing : European food- and waterborne diseases and zoonoses network

Thirty-five laboratories of the Food- and Waterborne Diseases and Zoonoses Network (FWD-Net), 29 of which located in EU/EEA countries, participated in the third international external quality assurance (EQA) scheme for the typing of Salmonella. Six laboratories were located in non-EU/EEA countries. All participating laboratories (including those in the EU/EEA) categorised 98% of the O antigens of the samples in agreement with the reference method and interpretive criteria. The participating EU/EEA laboratories categorised 91% of the H antigens of the samples in agreement with the reference method and interpretive criteria, compared with 92% for all participants. The EU/EEA laboratories were able to assign the correct serovar names for 90% of the samples (overall correct rate: 91%). In this third EQA scheme, fewer laboratories have deviating results compared with the second EQA. The results are comparable with those obtained in the first EQA scheme. The phage-typing results show that EU/EEA laboratories correctly phage-typed 80% of the S. Enteritidis strains (all laboratories: 82%). For S. Typhimurium, 79% of the strains were phage-typed correctly by the EU/EEA laboratories (overall correct rate 81%). Overall, the phage-typing results in the third EQA scheme were good, but when compared to the results of the second EQA scheme there were more deviations in the results for both S. Enteritidis and S. Typhimurium. Compared to the first EQA, results in this study were better. Antibiotic susceptibility testing (AST) results show a high level of performance: twenty of the 23 EU/EEA laboratories produced results that were more than 95% correct (all laboratories: 87%). Overall, 79% of all participating laboratories produced ≤ 5% deviations. If a threshold of 90% accuracy was applied, all but two participating laboratories would have been approved. Overall, in the third EQA scheme, 97% of 2448 evaluated tests were typed correctly, compared with 96% of 2443 evaluated tests in the second scheme, and 95% of a total of 2849 evaluated tests in the first EQA scheme. The third international external quality assurance (EQA) scheme for the typing of Salmonella spp. was launched in November 2010. The study included the laboratories of the Food- and Waterborne Diseases and Zoonoses surveillance network and was organised by the Laboratory for Zoonoses and Environmental Microbiology (LZO) of the National Institute for Public Health and the Environment (RIVM, Bilthoven, Netherlands) in collaboration with the Salmonella Reference Unit of the Laboratory of Gastrointestinal Pathogens (LGP) of the Health Protection Agency (HPA) in London and the Department of Bacteriology and TSEs at the Central Veterinary Institute (CVI) of Wageningen University (Lelystad, Netherlands). Three procedures for typing Salmonella spp. were evaluated in this EQA scheme: serotyping, phage typing and AST. The main objective of the EQA scheme was to assess whether typing of Salmonella strains by different laboratories within and outside the European Union was carried out uniformly and whether comparable results could be obtained. Thirty-five laboratories took part in this study, of which two did not return their results. Twenty-nine EU/EEA and six non-EU/EEA laboratories participated. Twenty strains of the species Salmonella enterica subspecies enterica were selected by RIVM for serotyping. Thirty-two participants performed serotyping of the strains, 26 of these were EU/EEA laboratories. Under the rules of the scheme, strains had to be typed with the method routinely used in each laboratory. The detected H and O antigens and serovar names (according to the White-Kauffmann-Le Minor scheme) had to be reported. Most problems were encountered in typing the H antigens. EU/EEA laboratories as well as all other participating laboratories typed O antigens correctly in 98% of the samples. The EU/EEA laboratories typed 91% of the samples correctly for H antigens (overall correct rate: 92%). 90% of the samples were identified with the correct serovar names by EU/EEA laboratories, compared with 91% for all laboratories. Fifteen of the 26 (58%) of the EU/EEA laboratories, and nineteen of the 32 (59%) participating laboratories correctly identified all 20 serovars. One serovar (S. Agona) was correctly typed by all participants. The HPA selected 20 strains for phage typing. Ten strains belonged to the serovar Salmonella Enteritidis and ten strains to the serovar Salmonella Typhimurium. Nineteen participants performed phage typing of the S. Enteritidis strains, 15 of which were EU/EEA laboratories. Two laboratories did not carry out phage typing on S. Typhimurium strains, therefore only 17 participants provided results on phage typing of S. Typhimurium (13 EU/EEA countries). Overall, 80% of the S. Enteritidis strains were phage-typed correctly by the EU/EEA laboratories (overall correct rate: 82%). For S. Typhimurium, 79% of the strains were phage-typed correctly by the EU/EEA laboratories (overall correct rate: 81%). Ten strains of various Salmonella serovars were selected by CVI for AST. These strains were tested by the participants for susceptibility to a panel of ten antibiotics. Twenty-eight laboratories participated in the AST of the strains, 23 of which were EU/EEA laboratories. Eleven of the participating laboratories employed a quantitative method producing minimum inhibitory concentrations (MIC values), and 17 laboratories employed a qualitative disk diffusion test producing zone diameters. The laboratories categorised the results as susceptible (S), intermediate (I) or resistant (R), based on their own interpretive criteria. Minor deviations in the interpretation of the AST results were found in 1.9% of all 2448 evaluated test results, and 1.1% of these results showed major deviations. Errors mostly occurred when inadequate interpretive criteria were applied. With respect to AST results, all but two participating laboratories would have been approved, assuming a threshold of 90% accuracy. Twenty of the 23 EU/EEA laboratories (87%) and 23 of the 28 overall participants (79%) produced ≤ 5% deviations, which points towards a very high standard of antimicrobial susceptibility testing