Insights from 30 years of Salmonella proficiency testing
Implications for participants
In this case, the same strain had been used in earlier PT rounds with similar outcomes, reinforcing that performance is best explained by the strain’s atypical phenotype (lactose-positive, H 2 S-negative) rather than method incompatibility. Targeted analyst training and closer adherence to Standard Operating Procedures (SOPs) should improve results.
Recommended actions:
• Use two selective or differential agars, where your SOP allows, to mitigate medium-specific inhibition. • Do not triage by appearance alone: at low-levels, confirm atypical colonies rather than discarding non-classic morphologies (e.g. lactose-positive, H 2 S- negative, small or pale). • Maintain strict pre-enrichment conditions (ratio, time, temperature); small deviations at low inoculum levels markedly reduce sensitivity. • Run parallel selective enrichments to avoid reliance on a single pathway; use dual enrichments where permitted. • Include a low-level positive control periodically to verify recovery under marginal conditions, and reinforce recognition of weak or atypical colony forms. • Review performance trends using PORTAL to confirm that SOP refinements deliver sustained gains.
Discussion
Our long-term dataset aligns closely with wider PT experience: laboratories perform strongly under routine conditions, but performance can decline when challenged with atypical phenotypes. This mirrors findings from major PT programmes in the US ( Edson et al .; Nemser et al.), where atypical colony morphology or biochemical behaviour is consistently associated with higher false-negative rates. For the S. Bredeney case study, correct detection increased from 83% in 2015 to 93% in 2023, indicating that repeated exposure improves recognition of atypical characteristics. Atypical strains pose specific interpretive challenges. Colonies that are lactose- positive, weakly H 2 S-negative or difficult to recognise on agar – for example small, pale or poorly developed colonies – can resemble background flora or fall outside analyst expectations. Such presentation issues remain a well-documented cause of false negatives across both our dataset and external PT studies . Together, these examples illustrate how biological variation interacts with analyst decision-making at the bench, and why performance tends to improve as laboratories become more familiar with these characteristics. A second, equally well-established driver of variability is inoculum level, particularly in matrices with low water activity, high fat content or intrinsic antimicrobial components . Very low inoculum levels often produce sublethally injured cells , making recovery highly sensitive to enrichment parameters, incubation conditions and analyst technique. PT experience consistently reflects these constraints, with low-level contamination samples generating more variable detection outcomes .
10
Made with FlippingBook Online newsletter maker