The clinical and epidemiological features of an outbreak of nosocomial Moraxella catarrhalis infection in a respiratory unit are described. Six isolates from five patients were shown to be indistinguishable by immunoblotting and restriction endonuclease analysis and different from 11 other, unrelated clinical strains. (Thorax 1993;48:91-92) Central Microbiological Laboratories

Sodium dodecyl sulphate-polyacrylamide gel electrophoresis of whole cell protein, immunoblotting with normal human serum and restriction endonuclease analysis using Taq I enzyme were applied to 38 clinically significant isolates of Moraxella (Branhamella) catarrhalis obtained during a suspected outbreak of nosocomial infection. Each of 18 strains had individual profiles by at least two of the three methods (unique strains). The remaining 20 strains were assigned to five groups (A-E) on the basis of similarity by at least two of the three methods. Isolates within groups A, D and E were homologous by all three methods. Immunoblot groups B and C had two distinct whole cell protein profiles (B1 and B2) but indistinguishable restriction endonuclease profiles (group B/C). This emphasizes the need to use more than one technique in characterizing strains from suspected outbreaks of nosocomial infection. Grouped strains were more likely to originate from the same hospital ward than unique strains and were associated with a significantly longer median time from patient admission to strain isolation (14 versus 3.5 days, p less than 0.005). Furthermore, the beta-lactamase activity was homologous within the groups. The results suggest that nosocomial infection involving several distinct Moraxella catarrhalis strains persisted over a period of months, involving at least 20 patients on three different wards. Such infection is probably common in wards harbouring suitably predisposed patients. The mode of transmission remains to be elucidated, but the above three techniques possess sufficient reproducibility and discriminatory ability to constitute suitable investigative tools.

SUMMARYWe have used molecular techniques to characterize 51 group A streptococci from Scotland and 17 'serious disease' isolates from other countries, in order to establish the clonal structure of invasive Streptococcus pyogenes strains circulating between 1986 and 1993. Strains were grouped by restriction endonuclease analysis, pulsed field gel electrophoresis and ribotyping patterns, and were examined for the presence of alleles of the speA gene by polymerase chain reaction and DNA sequence analysis. Serious and fatal infections in Scotland were caused by several clones. One clone (9 of 51 strains) was M type 1 and possessed the speA gene allele 2. This was the clone previously identified as causing severe infection in the USA. Another clone (5 of 51 strains) was M type 3 and had speA gene allele 3. In view of the clear association of more than one clone with severe, invasive and fatal infections, horizontal gene exchange between genotypes merits further investigation.

Summary. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) of whole-cell proteins (WCP), immunoblot analysis and DNA restriction-endonuclease analysis (REA) were applied as potential typing methods to 31 clinically significant strains of Moraxella (Branhamella) catarrhalis, five of which came from a suspected outbreak of nosocomial infection in a respiratory-diseases ward. Twelve of 31 isolates were placed in four groups, each of which containea strains indistinguishable by the three typing techniques used. Each of a further two groups contained two strains, and they were similar by at least one technique; the remaining 15 strains were unique by all three methods. Four of five strains from the suspected outbreak were indistinguishable by SDS-PAGE of WCP, immunoblotting and REA. Results show that SDS-PAGE of WCP, immunoblotting and REA are suitable techniques for characterising M . catarrhalis and that there is a considerable degree of strain heterogeneity. Nosocomial infection with M . catarrhalis may be relatively common and further epidemiological studies with a combination of typing techniques are indicated.

No abstract