Evaluation of the World Health Organization Global Invasive Bacterial Vaccine-Preventable Disease (IB-VPD) Surveillance Network’s Laboratory External Quality Assessment Programme, 2014–2019

Abstract

Introduction. In 2009, the World Health Organization (WHO) established the Global Invasive Bacterial Vaccine Preventable Disease (IB-VPD) Surveillance Network (GISN) to monitor the global burden and aetiology of bacterial meningitis, pneumonia and sepsis caused by Haemophilus influenzae (Hi), Neisseria meningitidis (Nm) and Streptococcus pneumoniae (Sp). Hypothesis/Gap Statement. The GISN established an external quality assessment (EQA) programme for the characterization of Hi, Nm and Sp by culture and diagnostic PCR. Aim. To assess the performance of sentinel site laboratories (SSLs), national laboratories (NLs) and regional reference laboratories (RRLs) between 2014 and 2019 in the EQA programme. Methodology. Test samples consisted of bacterial smears for Gram-staining, viable isolates for identification and serotyping or serogrouping (ST/SG), plus simulated cerebrospinal fluid (CSF) samples for species detection and ST/SG by PCR. SSLs and NLs were only required to analyse the slides for Gram staining and identify the species of the live isolates. RRLs, and any SLs and NLs that had the additional laboratory capacity, were also required to ST/SG the viable isolates and analyse the simulated CSF samples. Results. Across the period, 69–112 SS/NL labs and eight or nine RRLs participated in the EQA exercise. Most participants correctly identified Nm and Sp in Gram-stained smears but were less successful with Hi and other species. SSLs/NLs identified the Hi, Nm and Sp cultures well and also submitted up to 56 % of Hi, 62 % of Nm and 33 % of Sp optional ST/SG results each year. There was an increasing trend in the proportion of correct results submitted over the 6 years for Nm and Sp. Some SSLs/NLs also performed the optional detection and ST/SG of the three organisms by PCR in simulated CSF from 2015 onwards; 89–100 % of the CSF samples were correctly identified and 76–93 % of Hi-, 90–100 % of Nm- and 75–100 % of Sp-positive samples were also correctly ST/SG across the distributions. The RRLs performed all parts of the EQA to a very high standard, with very few errors across all aspects of the EQA. Conclusion. The EQA has been an important tool in maintaining high standards of laboratory testing and building of laboratory capacity in the GISN.