European Gonococcal Antimicrobial Surveillance Programme external quality assessment scheme for Neisseria gonorrhoeae antimicrobial susceptibility testing, 2017

Introduction

External quality assessments (EQA) are an essential part of any laboratory-based surveillance system, allowing for the monitoring of performance and comparability of results from participating laboratories, identification of potential issues and deployment of resources and training where necessary. An EQA scheme for antimicrobial susceptibility testing in Neisseria gonorrhoeae has been available to laboratories participating in ECDC’s European Sexually Transmitted Infections (STI) surveillance network since 2010. This EQA scheme has so far shown high levels of inter-laboratory comparability in the presence of differing methodologies. Problems identified previously included reduced comparability of results determined using discs compared with those determined by agar dilution and gradient strips, media not suitably supporting gonococcal growth, and the use of gradient strips from one manufacturer. This is the third EQA to be published on antimicrobial susceptibility testing in Neisseria gonorrhoeae. 

Materials and methods

The EQA specimen panel was selected by Public Health England (PHE) and distributed by the United Kingdom National External Quality Assessment Service (UK-NEQAS). In January 2017, 28 laboratories in 27 participating countries received 10 gonococcal isolates for susceptibility testing. Of the 10 gonococcal isolates provided, one was in triplicate and two were in duplicate to test intra-laboratory concordance. The remaining isolates were all provided singularly, meaning that the N. gonorrhoeae antimicrobial susceptibility EQA panel comprised of six different strains in total. The isolates chosen by PHE were representative of a range of different antimicrobial susceptibility profiles, and consisted of the three World Health Organization (WHO) reference strains, WHO F, U and X and three UK clinical isolates from 2015 or 2016. Participating laboratories were requested to test the EQA panel using local methodology (i.e. gradient strip, agar dilution or disc diffusion) and relevant international breakpoints (i.e. CLSI, EUCAST etc.) against a range of antimicrobial agents. Results were submitted directly to UKNEQAS who issued individual laboratory reports. The results were then supplied to Public Health England (PHE) who decoded and analysed the results based on the categories of susceptibility assigned.

Results

Twenty-eight laboratories returned EQA results to UK-NEQAS. Most laboratories used gradient strips and EUCAST breakpoints. The highest level of susceptibility category concordance (other than spectinomycin; 100%) was seen with ceftriaxone (99.4%), whilst the lowest was seen with azithromycin (88.1%). Overall concordance increased for most antimicrobials in comparison with the previous distribution, except for ciprofloxacin which decreased slightly from 99.8% to 94.1%. Overall, 88.0% and 96.4% of the reported minimum inhibitory concentrations (MICs) were within one and two doubling dilutions of the modal MIC, respectively.

Discussion and conclusion

There has been further harmonisation of susceptibility testing methodologies and breakpoints used by participating laboratories; most laboratories used gradient MIC strips and all applied EUCAST breakpoints for interpretation of MIC results. Overall, the laboratories participating in EQA scheme QA17 performed well and showed good levels of competency in testing N. gonorrhoeae strains of unknown phenotype. The susceptibility category concordances increased slightly in this distribution when compared with 2016, with the exception to ciprofloxacin. The interlaboratory concordance was high (over 88% for susceptibility categories for all antimicrobials), as was the intralaboratory concordance (22/28 laboratories had ≤5% error score), demonstrating comparability between different testing methodologies and allowing confidence in decentralised testing for surveillance purposes. Most susceptibility category discrepancies were attributable to strains with MICs on or close to a breakpoint, which highlights the need to consider the actual MIC as well as susceptibility category when interpreting susceptibility results. Analysis of the individual results submitted by the participating laboratories highlighted seven centres in need of further guidance to help bring them into line with the European Gonococcal Antimicrobial Surveillance Programme (Euro-GASP) recommended target of 95% of MICs within two doubling-dilutions (four-fold) of the modal MICs and beta-lactamase assessment.