Investigation of an outbreak of CTX-M-15-producing Escherichia coli of sequence types 131 and 1441 in a neonatal surgical ward: comparison of typing methods
Författare och institution:
Nahid Karami (Institutionen för biomedicin, avdelningen för infektionssjukdomar); Lisa Helldal (Institutionen för biomedicin, avdelningen för infektionssjukdomar); M. Andersson (-); Christel Unosson (Institutionen för biomedicin, avdelningen för infektionssjukdomar); Leif Larsson (Institutionen för biomedicin, avdelningen för infektionssjukdomar); Christina Åhrén (Institutionen för biomedicin, avdelningen för infektionssjukdomar); Edward R.B. Moore (Institutionen för biomedicin, avdelningen för infektionssjukdomar)
20th European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), Vienna, Austria,
Objectives: In a surgical ward caring mainly for newborns spread of CTX-M-15-producing E.coli had been ongoing at least since September 2008 when we finally recognized the outbreak in late December. We have compared various typing methods with pulsed-field gel electrophoresis (PFGE) to verify the actual outbreak and subsequently to determine number of affected children.
Methods: In addition to clinical sampling 125 children hospitalized between Sept-Dec were screened for extended-spectrum beta-lactamase (ESBL)-producing bacteria in stool during Dec-Feb. From Jan-June 2009 newly admitted children were screened at admission and twice weekly. Fifty-one E coli isolates with ESBL from 27 children were found. These isolates have been typed with PFGE, multiple-locus-variable number tandem repeat analysis (MLVA), a mini multiple-locus-sequence typing (MLST) method (dnaJ, purA and fumC genes) as well as with the Phene Plate (PhP) biochemical fingerprinting system.
Results: When the outbreak was revealed five children had developed infections with ESBL-producing E. coli that were of two PFGE-types (A and B) later considered to be the outbreak strains. One or both were spread to 21 children. Six children had multiple types. Altogether 38 isolates (20 children) were of type A (ST 131), 7 isolates (5 children) of type B (ST 1441). In addition E coli of six distinct PFGE-types (C-?) were found in one child each. MLVA gave identical discriminatory results as PFGE for all isolates tested. Mini-MLST could not differentiate ST 131 isolates of to distinct PFGE-types (type A and C) but accurately predicted the ST-types of each PFGE-type when confirmed with standard MLST according to http://mlst.ucc.ie/mlst/dbs/Ecoli. By comparing resistance pattern we thus missed the outbreak by a month. PhP indicated that all initial isolates were singletons and there was hardly any correlation with PFGE.
Conclusion: If transmission has been ongoing for a long time several types of ESBL-producing bacteria may be found in an outbreak and all isolates including repeat and screening isolates need to be typed to identify affected patients. Only genetic typing, gave satisfactory results in this outbreak. MLVA gave identical results to PFGE and is thus attractive being faster, cheaper and easier to communicate. Our mini-MLST was somewhat less discriminatory but despite using only three house keeping genes accurately predicted the ST-types.
Ämne (baseras på Högskoleverkets indelning av forskningsämnen):
Biologiska vetenskaper ->
Biokemi och molekylärbiologi ->
Cell- och molekylärbiologi ->
Biologiska vetenskaper ->
MEDICIN OCH HÄLSOVETENSKAP ->
Klinisk medicin ->
Escherichia coli, ESBL, CTX-M-15, ST-131, ST-1441, outbreak
Visa i Endnote-format