Molecular epidemiology of fluoroquinolone-resistant ST131 Escherichia coli producing CTX-M extended-spectrum beta-lactamases in nursing homes in Belfast, UK.

OBJECTIVES: Between January 2004 and May 2006 Escherichia coli producing extended-spectrum ß-lactamases (ESBLs) were isolated from the faeces of 118/294 residents from 16 nursing homes in Belfast. Of these, 58 isolates belonged to UK strain A, a variant of the international ST131 clone. Here we investigated the remaining 60 ESBL producers. METHODS: MICs were determined and interpreted using BSAC methodology. Isolates were characterized by phylogenetic typing, real-time PCR and PFGE. Plasmids were rep typed by PCR and their similarity to IncI1 reference plasmid pEK204 was investigated by restriction fragment length polymorphism analysis. The molecular environments surrounding bla(CTX-M) were determined by DNA sequencing and PCR. RESULTS: Fifty-nine of 60 isolates belonged to the B2, ST131 lineage; of these 28 belonged to the previously defined UK strain C, while the other 31 were clustered into five groups by PFGE. Forty-nine isolates harboured bla(CTX-M-3) on plasmids of five different rep types (I1, FIA, FIA-FIB, N and Y) and 11 harboured bla(CTX-M-15) on F-type plasmids (FIA and FIA-FIB). All CTX-M-3 ESBL producers and three with CTX-M-15 ESBL had an intact copy of ISEcp1 immediately upstream of bla(CTX-M); the remaining eight with CTX-M-15 ESBL had a truncated ISEcp1. CONCLUSIONS: Gut colonization among nursing home residents in Belfast with ciprofloxacin-resistant E. coli producing ESBLs almost entirely involves clonal spread of ST131 variants, with similar genetic environments for bla(CTX-M-3) or bla(CTX-M-15) as in pEK204 and pEK499. Such diversity indicates dissemination of both plasmids and ESBL genes among a single commonly multiresistant clone.

Nursing homes as a reservoir of extended-spectrum beta-lactamase (ESBL)-producing ciprofloxacin-resistant Escherichia coli.

BACKGROUND: To assess the prevalence and risk factors for faecal carriage of fluoroquinolone-resistant, extended-spectrum beta-lactamase (ESBL)-producing, Escherichia coli (MDR E. coli) among residents in nursing homes in Northern Ireland. METHODS: Between January 2004 and May 2006, retrospective histories of hospital admissions, antimicrobial treatment and co-morbidities were collected. Faecal samples were cultured for MDR E. coli. These isolates and their ESBL genes were typed by a reference laboratory. RESULTS: Of the 294 patients included in the study, faecal samples from 119 (40.5%) grew MDR E. coli. The proportion of carriers in the different homes ranged from 0% to 75%. Epidemic strain A belonging to the ST131, O25:H4 lineage with the CTX-M-15 enzyme accounted for 58 (49%) of all isolates; its proportion varied from 0% to 100% among homes. Fifty-one percent of carriers had no history of recent hospital admission and only 13.5% had a known history of ESBL E. coli colonization or infection. In a multivariate logistic regression model, days of fluoroquinolone use [odds ratio (OR) = 1.33, 95% confidence interval (CI) 1.04-1.69, P = 0.02] and a history of urinary tract infection (OR = 2.56, 95% CI 1.37-4.78, P = 0.003) were the only variables independently associated with the risk of carrying MDR E. coli. CONCLUSIONS: The high level of faecal carriage of MDR E. coli in nursing home residents demonstrates their importance as a reservoir population. Public health measures to combat spread of these organisms should address the needs of this group.

Clinico-epidemiological features of infections caused by CTX-M type extended spectrum beta lactamase-producing Escherichia coli in hospitalised patients.

A review of medical records of 45 of 53 hospitalised patients with positive cultures for CTX-M type ESBL-producing Escherichia coli between 01 January and 31 May 2004 was conducted. The mean age of the population studied was 73.1 (+/-14.6) years and the majority (55.6%) had been under the care of the internal medicine or elderly care service. In the majority (77.8%) of instances the isolate was attributed to a clinical infection rather than colonisation and the commonest clinical specimen to yield the organism was urine, which was positive in 57.8% of patients. Acquisition of the organism was categorised as nosocomial in 68.9% of patients; in this subgroup, the median duration of inpatient stay prior to recovery of the organism was 24 (range 3-240) days. Haemodialysis-dependence was the most common of the comorbidities evaluated. The mean number of antibiotics prescribed per patient in the 30 days prior to first isolation of the organism was 1.7 (range 0-4). Furthermore, the mean number of antibiotic-days exposure per patient during this period was 13.9 (range 0-48). The most frequently received class of antibiotic was beta-lactam/beta-lactamase inhibitor combinations. Of 35 infections, 26 (74.2%) were successfully treated. Overall 12 patients with infection died (34.3%); attributable mortality was presumed in seven (20%).