Search

From July 1995 to November 1996, multi-resistant Pseudomonas aeruginosa O11 was isolated from 36 patients admitted to a neurosurgery intensive care unit. The strain was resistant to ticarcillin, ceftazidime, imipenem, gentamicin and ciprofloxacin, and susceptible to amikacin. Nine patients were colonized only; the remaining 27 patients had at least one infected site (17 urinary infections, 10 pneumonias and four with sinusitis). P. aeruginosa O11 with the same resistance pattern was isolated from tap water. The strain was also cultured from enteral nutrition solutions given to two infected patients. DNA macrorestriction analysis with XbaI established the similarity of the isolates from patients, tap water and solutions. The outbreak was controlled after reinforcement of isolation procedures for infected patients, changing the mode of enteral nutrition and replacement of all sinks in the unit. The sinks were presumably the main source of P. aeruginosa during this outbreak, via the hands of the nursing staff or nutrition solutions contaminated with tap water.

Seventy isolates of Klebsiella pneumoniae with extended-spectrum beta-lactamases (ESBLs) were compared. These were isolated from 51 patients on 10 separate wards in one hospital over an 18-month period between 1992 and 1994. Antibiograms were determined and the isolates were typed by pulsed-field gel electrophoresis of their DNA digestion with XbaI. The isolates were compared to three genotypically different epidemic strains responsible for previous outbreaks at the hospital between 1988 and 1991. Isolates from 84% of the present patients had closely related XbaI patterns, and most (74%) produced an ESBL with an iso-electric point (pI) of 7.0. A similar pattern was found for one of the previous epidemic strains, but it produced an ESBL with a pI of 7.8; isolates with this latter enzyme variant were found only in six of the present patients. The two other previous epidemic strains had ESBLs with a pI of 6.3 and organisms related to them were found in one and two of the present patients, respectively.

Thirty-seven isolates of extended-spectrum beta-lactamase-producing (ESBL) Klebsiella pneumoniae implicated in five nosocomial outbreaks (I-V) on three distinct wards of our hospital were compared using capsular typing, biotyping, antibiotyping, enzyme electrophoresis typing and DNA macrorestriction analysis with Xba I resolved by pulsed-field gel electrophoresis. The isolates from each outbreak had common phenotypic and genotypic characteristics indicating that they were related epidemiologically. Isolates from outbreaks I (four patients) and V (13 patients), although they occurred in two different wards (neurology and surgery) and three years apart, produced the same ESBL with a pI of 7.8 (SHV-4) and were of serotype K25. The Xba I patterns were closely related. The isolates of outbreaks II (seven patients), III (four patients) and IV (seven patients), which occurred in a single surgical intensive care unit, produced an ESBL with a pI of 6.3 (TEM-3). Isolates from outbreaks III and IV, which occurred six months apart, were of serotype K68 and had similar Xba I patterns suggesting that the two outbreaks were due to a single strain which persisted endemically in the ward. The isolates from outbreak II were of serotype K62, and had distinct characteristics from the two later outbreaks. The Xba I patterns of the isolates from outbreaks "I and V', II and "III and IV' had Dice similarity coefficients under 40% showing that the three groups were genetically distant. DNA macrorestriction analysis was a useful complement to phenotypic methods for identifying K. pneumoniae strains responsible for outbreaks harbouring a common ESBL.