Ian G. Wilson

Vancomycin-resistant enterococci in shellfish, unchlorinated waters, and chicken

Vancomycin-resistant enterococci (VRE) have been a cause of increasing concern chiefly regarding the infection of hospital patients. There is suspicion, but limited evidence, that food and environmental spread may be important. Biomonitoring by examination of bivalve shellfish was used to assess the occurrence of VRE entering the environment. Using pre-enrichment and Lewisham and Slanetz and Bartley agars, 2/125 (1.6%) of shellfish were found to contain enterococci resistant to high levels of vancomycin. Lewisham agar allows relatively rapid identification of VRE. In a second phase of the work using pre-enrichment and Slanetz and Bartley agar, 4/151 (2.7%) shellfish and 5/27 (18.5%) raw chickens contained VRE. Using filtration and pre-enrichment, no VRE were found in 54 unchlorinated water samples. The study shows that environmental prevalence of VRE is low, and that raw chickens are frequently contaminated.

Some factors inhibiting amplification of the Staphylococcus aureus enterotoxin C1 gene (sec+) by PCR.

PCR amplification of the sec+ gene for staphylococcal enterotoxin C1 (SEC1) can be achieved from as little as 10 fg total genomic DNA (equivalent to less than 10 cells) using two nested primer pairs. The presence of bacterial cells, particularly thermonuclease-producing staphylococci, and the thermonuclease enzyme (TNase) itself, were found to be factors which individually and together reduced the sensitivity of PCR amplification.

Occurrence and characteristics of cytotoxic necrotizing factors, cytolethal distending toxins and other virulence factors in Escherichia coli from human blood and faecal samples

Escherichia coli isolates from human blood (n=266) and faecal (n=237) samples were examined for cytotoxic necrotizing factors 1 and 2 (CNF 1 and 2), cytolethal distending toxin (CDT), and putative virulence factors that have been associated with disease conditions in humans and animals. PCR showed that the chromosomally encoded, Rho-activating, CNF1 (68/544, 12.5%) was more common than the transmissible plasmid-borne CNF2 (3/544, 0.6%). The relative risk of having either CNF or CDT toxin genes in blood compared to faecal isolates was 3.88 (95% CI 2.36-6.38). This was highly significant (P<0.0001) and demonstrates the importance of these factors in bloodstream infections. Fifty-one of 65 (78%) E. coli bearing CNF1 and 11 of 21 (52%) of E. coli bearing CDT also carried the pyelonephritis-associated pilus gene, papG. The S fimbrial adhesin gene, sfa, was found in 57 blood (21%) and eight faecal samples (3%). The F17 fimbrial adhesin gene and afimbrial adhesin gene afa did not occur frequently. Haemolysin (hly) was found in all of the isolates tested. Further studies must be designed to identify the clinical significance of these genes and their role in pathogenesis.

A non-isotopic DNA hybridisation assay for the identification of Staphylococcus aureus isolated from foods

A digoxygenin-labelled total genomic DNA probe was used to identify Staphylococcus aureus. Isolation and identification of organisms was possible in less than 4 days. Identification alone could be completed in less than 2 days, compared with over 5 days for identification of isolates by multipoint inoculation. The probe showed excellent discrimination of S. aureus from other staphylococci and from a wide range of bacteria commonly associated with milk and meat. The effectiveness of this probe was tested against cultural isolation of staphylococci in milk using Baird-Parker agar followed by identification using multipoint inoculation and API Staph. The probe gave comparable results to the conventional methods and, for large sample numbers, offered lower cost and greater ease of use.

Meta‐audit of laboratory ISO accreditation inspections: measuring the old emperor's clothes

Accreditation to ISO/IEC 17025 is required for EC official food control and veterinary laboratories by Regulation (EC) No. 882/2004. Measurements in hospital laboratories and clinics are increasingly accredited to ISO/IEC 15189. Both of these management standards arose from command and control military standards for factory inspection during World War II. They rely on auditing of compliance and have not been validated internally as assessment bodies require of those they accredit. Neither have they been validated to criteria outside their own ideology such as the Cochrane principles of evidence‐based medicine which might establish whether any benefit exceeds their cost. We undertook a retrospective meta‐audit over 14 years of internal and external laboratory audits that checked compliance with ISO 17025 in a public health laboratory. Most noncompliances arose solely from clauses in the standard and would not affect users. No effect was likely from 91% of these. Fewer than 1% of noncompliances were likely to have consequences for the validity of results or quality of service. The ISO system of compliance auditing has the performance characteristics of a poor screening test. It adds substantially to costs and generates more noise (false positives) than informative signal. Ethical use of resources indicates that management standards should not be used unless proven to deliver the efficacy, effectiveness, and value required of modern healthcare interventions.