![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Department of Clinical Microbiology, St. Elisabeth Hospital, Tilburg, The Netherlands,1 Department of Intensive Care, St. Elisabeth Hospital, Tilburg, The Netherlands,2 Department of Clinical Microbiology, Amphia Hospital, Breda, The Netherlands3
Received 17 February 2005/ Returned for modification 11 April 2005/ Accepted 27 June 2005
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hospital outbreaks due to multiresistant K. pneumoniae (MRKP) strains have been described throughout the world (7, 11, 13, 14, 15, 20). Outbreaks of K. pneumoniae infections have been associated with a wide variety of sources and reservoirs, including sinks (22), ultrasonography gel (6), an intravenous dextrose solution (9), cockroaches (3), food blenders (8), contaminated breast milk (5), bath soap (23), and water baths (1). However, the majority of outbreaks could not be associated with an environmental source. Patient-to-patient transmission via the hands of hospital staff was considered the main route of transmission (14).
The scope of the present study was to investigate a nosocomial outbreak at the intensive care unit (ICU) of St. Elisabeth Hospital, Tilburg, The Netherlands, caused by a K. pneumoniae strain that was resistant to gentamicin, tobramycin, and trimethoprim-sulfamethoxazole and that produced an ESBL. The incidence of multiple-resistant Klebsiella spp. at our hospital is low and corresponds to the low incidence of ESBL-producing members of the family Enterobacteriaceae reported in The Netherlands (21). The initial investigation of defining cases and possible sources was continued by a case-control study to identify the risk factors for the acquisition of an MRKP strain. In addition, molecular genotyping of clinical and environmental isolates was performed.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Identification of outbreak. In November 2000, three patients in the ICU developed fatal sepsis caused by a K. pneumoniae strain expressing resistance to extended-spectrum cephalosporins by means of ESBL production and additional resistance to gentamicin, tobramycin, and trimethoprim-sulfamethoxazole. The infection control department was alerted, and the infection control practitioner conducted a visit to the ICU. A preliminary review of the patients' characteristics and a search of the microbiology reports of the ICU patients for specimens positive by culture for MRKP for the months of October and November were performed. It was noted that all fatal cases were operated on in one of two operation rooms (ORs; OR-X and OR-Y) within 5 days before the development of sepsis. In response, the infection control practitioner visited the OR complex and took environmental samples from OR-X and OR-Y.
Microbiological methods.
In addition to the culture of samples for clinical indications, samples for surveillance cultures (respiratory material, urine, drains, and wounds) were obtained from all patients on artificial ventilation and were performed twice a week as standard procedure at the ICU. All identification and susceptibility tests were performed according to the CLSI (formerly the National Committee for Clinical Laboratory Standards). K. pneumoniae strains were identified by conventional CLSI-recommended biochemical methods, completed with an API 20E test kit (bioMerieux SA, Lyon, France). Antibiotic susceptibility testing was performed by a microdilution method in microtiter panels containing 11 antibiotics: amoxicillin, amoxicillin-clavulanate, and ceftazidime (GlaxoSmithKline CH, Zeist, The Netherlands); cefuroxime (Pharmachemie, Haarlem, The Netherlands); ceftriaxone (Roche, Mijdrecht, The Netherlands); ciprofloxacin (Bayer, Mijdrecht, The Netherlands); ofloxacin (Aventis Pharma, Hoevelaken, The Netherlands); trimethoprim-sulfamethoxazole and tobramycin (Sigma-Aldrich Chemie BV, Zwijndrecht, The Netherlands); meropenem (Astra Zeneca, Zoetermeer, The Netherlands); and gentamicin (Bufa BV, Uitgeest, The Netherlands). ESBL production was tested by disk diffusion tests (Oxoid Ltd., Basingstoke, England), as follows: cefpodoxime (10 µg), ceftazidime (30 µg), and cefotaxime (30 µg) were plated with and without clavulanate at 1 µg, 10 µg, and 10 µg, respectively. ESBL production was confirmed by a clavulanate-induced increase in the zone size of 
5 mm. Environmental specimens for culture were taken with cotton swabs and were plated on blood agar and cultured for 2 days at 35°C. The choice of sites for culture was guided by preliminary investigations of the nature of the outbreak, information presented in the literature, and previous experiences in our hospital.
Case-control study. A case-control study was conducted to investigate risk factors and possible transmission routes in the November 2000 outbreak in the ICU. Cases were defined as patients who were admitted to the ICU for >24 h in November 2000 and who were positive for MRKP by culture of specimens taken between 1 November 2000 and 31 December 2000. Samples for culture were taken from specific infection sites or for surveillance, and samples from both colonized and infected patients were included. Case finding was conducted by a retrospective review of all microbiology culture reports from this period. Seven cases were recorded, and subsequently, 14 unmatched controls were randomly selected from a pool of all patients admitted to the ICU for >24 h in November 2000.
Data for the cases and the controls were collected by retrospective chart review and included age; gender; hospital admission dates; ICU admission dates; Apache score at the time of admission to the ICU; attending specialty; wards attended by the patient before and after admission to the ICU; units attended during the ICU admission; mortality at end of the ICU admission; mortality at end of the hospital admission; number of days of artificial ventilation; presence of urinary catheter, artery catheter and central venous catheter; presence of drains; presence of open wounds; surgery during hospital admission; OR number; operating team; antibiotic use; number of specimens presented for culture taken during the study period; date of first MRKP-positive culture; and the MRKP-positive culture sites.
Molecular genotyping. Molecular genotyping was performed by multienzyme amplified fragment length polymorphism (ME-AFLP) analysis, as described previously (24). In short, the genomic DNA of each isolate was extracted with a QiAmp DNA minikit (QIAGEN, Hilden, Germany). After addition of 5 µl of RNase (1 U; Sigma), the concentration of DNA was determined with a spectrophotometer (260 nm; Genequant, Pharmacia, United Kingdom). The restriction-ligation reaction was performed with 500 ng of DNA in a final volume of 30 µl. The reaction mixture consisted of the restriction enzymes EcoRI, PstI, XbaI, and NheI (2 U each; Roche Molecular Biochemicals); 3 µl of T4 ligase buffer; 1 U of T4 ligase (Roche Molecular Biochemicals); and 20 pmol of oligonucleotide adaptors. The restriction-ligation mixture was incubated at 37°C for 3 h. DNA was precipitated by using 2.5 M of ammonium acetate in 100 µl chilled absolute ethanol. After the DNA was washed with 70% ethanol, it was resuspended in 100 µl of TE (Tris-EDTA) buffer and stored at 4°C. PCRs were carried out with Ready-To-Go beads (Amersham Pharmacia, Little Chalfont, United Kingdom) supplemented with 1 µl MgCl2 (25 mM). Each primer (10 pmol) and 100 ng of chromosomal template DNA were used in a final reaction volume of 25 µl. In a DNA thermocycler (GeneAmp PCR system 2400; Perkin-Elmer), the reaction mixtures were preheated for 4 min at 94°C. Thirty-three amplification cycles of 1 min at 94°C, 1 min at 60°C, and 2.5 min at 72°C were performed. Analysis of 10 µl of each PCR product was performed by agarose gel electrophoresis on a 1.5% agarose gel (MP agarose; Roche Molecular Biochemicals). Marker DNA (100-bp ladder; Gibco BRL, Life Technologies, Scotland) was loaded after every four samples. Analysis of the banding patterns on gel images was performed with the software program Gelcompar II, which is included in Bionumerics software (Applied Maths, Kortrijk, Belgium). Cluster analysis of the fingerprints was performed by the unweighted pair group method with arithmetic averages, and similarities between AFLP patterns were calculated by using the Pearson product-moment correlation coefficient (25). The patterns of the strains analyzed were considered identical when there was 90 to 100% homology.
A computer-based search for positive cultures of K. pneumoniae resistant to extended-spectrum cephalosporins and aminoglycosides at our hospital was performed for the years 1998, 1999, and 2000. Most of these isolates were stored at –80°C and could be included for molecular typing (one isolate per patient).
Data analysis. Univariate analysis was performed by Fisher's exact test for categorical variables, Student's t test, or the Mann-Whitney U test for continuous variables (EpiCalc 2000, version 1.02). Odds ratios were calculated when possible. Statistical significance was defined a priori as a P value of <0.05 (two tailed).
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
Case 3 was a patient with multiple traumas admitted to ICU-A on 7 November. He received laporotomies three times: immediately after admission and on 9 and 12 November in OR-Y, OR-Y, and OR-X, respectively. An MRKP strain was isolated from blood sampled on 11 November and an abdominal discharge sampled on 12 November. Cefuroxime and tobramycin treatment was started as empirical therapy on 10 November and was changed to ofloxacin on 12 November. Case 3 died postoperatively of sepsis with multiorgan failure on 13 November.
Case 4, who had pancreatitis, was admitted to the internal medicine ward on 27 October and transferred to ICU-A on 30 October, where he stayed until his death on 21 November. A laporotomy was performed on 3 November in OR-X, on 17 November in OR-X, and on 20 November in OR-Z. Case 4 was treated with cefuroxime for an infected pancreas and was in stable condition. An MRKP isolate was first isolated from discharge from the ear sampled on 18 November and from abdominal rinsing fluid and abdominal drains sampled on 19 November. On November 21 he developed a fatal sepsis, and an MRKP strain was isolated from blood sampled on that day.
Case 5 was admitted to the surgery ward on 17 October; operated upon on 18 October (esophageal resection, OR-Z); and transferred to ICU-C, then to ICU-B (from 25 October), and then to ICU-A (from 7 November). On 14 November an explorative thoracotomy was performed in OR-Y. Case 5 received trimethoprim-sulfamethoxazole and vancomycin from 15 November. An MRKP strain was first isolated from discharge from a thorax drain sampled on 18 November, and 2 days later multiple thoracic drainage sites were positive for MRKP. On 21 November the patient developed a fatal sepsis and an MRKP strain was isolated from the blood sampled on that day.
Case 6 was admitted to ICU-C on 12 October after a thoracic operation in a different hospital. On 18 October he was transferred to the surgery department of St. Elisabeth Hospital but was readmitted to ICU-C on 7 November after an explorative laparotomy in OR-X. On 11 November, case 6 was transferred to ICU-A, where he stayed until 14 December. Since it was a surveillance culture and there were no indications for infection, no antibiotic treatment for the MRKP infection was given. Case 6 was discharged from the hospital on 20 December.
Case 7, who had subarachnoidal bleeding, was admitted to ICU-A from 24 to 28 November, to the neurosurgery ward from 28 November until 20 December, and to ICU-C from 20 December until death on 24 December. An MRKP strain was first isolated from cerebrospinal fluid sampled on 19 December. Case 7 was not operated upon but received a spinal catheterization for pain relief 6 days prior to the positive culture. This procedure was performed in the recovery ward of the OR complex by an anesthetist who had not been involved in any of the operations on the other cases. Case 7 was treated with meropenem but died as a result of infection with the MRKP strain.
The index patient was a trauma patient repatriated from Morocco and admitted to ICU-A on 18 September. The patient was not operated upon in our hospital. Respiratory specimens, blood, and a thorax drain sampled at admission were culture positive for an MRKP strain. Treatment was started with meropenem. This patient recovered and was transferred to the surgery ward on October 9. Barrier isolation was imposed at both the ICU and the surgery ward during admission of the index patient. Identification of the index case was inferred from molecular genotyping of the MRKP isolates from this patient. The typing patterns of the MRKP isolates from this patient were identical to the typing patterns of the outbreak strains.
Case-control study. Seven cases and 14 unmatched controls were defined, and the data for these individuals are summarized in Table 2. Comparable values of the baseline characteristics were found between the cases and the controls; these included gender, age, and Apache score. The mean length of stay in the ICU, in total (17.0 ± 14.33 days) and prior to the acquisition of an MRKP strain (12.0 ± 11.18 days), was longer for the cases than for the controls (7.3 ± 1.39 days). Also, the cases received ventilatory support, both in total and prior to the acquisition of MRKP, for longer periods than controls. None of these differences was statistically significant (Table 2). An operation performed in OR-X or OR-Y was found to have the strongest association with the acquisition of an MRKP strain, with an odds ratio of 36.0 (95% confidence interval, 2.7 to 481.2; P = 0.003). All six cases had undergone an operation before the acquisition of an MRKP strain (Table 1). These six operations were performed by 17 operation team members in total, 12 of which attended only one operation and 5 of which attended two operations. The average number of days between the operation and the subsequent culturing of MRKP was 4.8 (standard deviation, ±5.19; Table 1).
|
Infection control measures. Because the fatal cases had all had an operation in OR-X or OR-Y less than 5 days prior to the development of sepsis caused by an MRKP strain, environmental samples were taken from these ORs. In both OR-X and OR-Y, K. pneumoniae was isolated from the roll boards used to transport patients from the bed to the operation table (Fig. 2). The multiresistant K. pneumoniae outbreak strain was defined as being resistant to aminoglycosides and trimethoprim-sulfamethoxazole and ESBL positive. Resistance to ciprofloxacin and ofloxacin varied from susceptible to resistant both within patients and between patients. The roll boards were disinfected thoroughly or replaced, if it was worn out. In addition, barrier precautions for colonized or infected patients were reinforced on the ICU. The outbreak ended after these interventions were implemented.
|
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A large number of risk factors for the acquisition of ESBL-producing K. pneumoniae strains have been reported, including the length of hospital stay (2, 10), the length of ICU stay (4), mechanical ventilation (2, 10, 17), urinary catheterization (2, 10), arterial catheterization (2), the presence of a central venous catheter (10, 19), prior antibiotic treatment (10, 16, 19), and more.
In the present study, a number of risk factors for the acquisition of the MRKP strain agreed with those that had been reported previously. We found that (i) the mean length of stay at the ICU was longer, (ii) the number of patients receiving ventilatory support was higher, and (iii) the number of central venous catheters and artery catheters was higher in cases than in the controls.
Our case-control study showed that an operation in OR-X or OR-Y was associated with a 36-fold increased risk of acquiring an MRKP strain and hence presented a significant risk factor for mortality due to infection with MRKP. Also, a short period of time (mean number of days, 4.8) between an operation in OR-X or OR-Y prior to the acquisition of MRKP was found. The contribution of OR-X and OR-Y to the outbreak was supported by the recovery of MRKP from cultures of environmental samples of the roll boards in OR-X and OR-Y. The similarity of these environmental isolates to the outbreak strain was confirmed by molecular typing. Moreover, the outbreak was contained after thorough disinfection and/or replacement of the roll boards in the ORs. Thus, the present study indicated a transmission route which differentiated this outbreak from previously reported outbreaks with multiresistent Klebsiella pneumoniae strains.
A minority of the cases could not be related to OR-X or OR-Y. One case had not been operated upon. In a second case, the time between the operation and the first isolation of MRKP took 15 days. The index case patient also did not have an operation, and other than admission to ICU-A, no clear connection could be found between this patient and any of the cases.
It is conceivable that secondary to the transmission at the ORs, cross contamination within the ICU occurred. Six of the seven cases were admitted to unit ICU-A at the time that MRKP was isolated for the first time. In the case-control study, ICU-A was found to be a risk factor for the acquisition of the MRKP strain. Unfortunately, no environmental samples for culture were taken from ICU-A. If, however, an environmental source was present in ICU-A and this source did play a significant role in the outbreak, the outbreak would probably not have been controlled after the interventions taken in the ORs.
To our knowledge, this study is the first to report on the contamination of roll boards in ORs with multiresistant Enterobacteriaceae. Roll boards usually do not come into contact with the intact mucosal barriers or sterile tissue, and these utensils are expected to impose only a small risk for the transmission of microorganisms. Prompt cleaning and decontamination are done on a regular basis, e.g., when blood or other potentially infectious material is spilled on the surface. This procedure is regarded to be sufficient and according to Dutch and international guidelines for infection control (2a).
From the results of this study it may be recommended that the construction of roll boards be improved to facilitate effective decontamination; they should have smooth surfaces instead of loosening fibers and open ball bearings. Alternatively, more frequent cleaning and disinfection procedures for roll boards in ORs might be considered to decrease the risk of unwanted infection by a pathogen from environmental sites.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
mara Almeida, B. Flannery, M. C. Almeida Lins, J. L. Mello Sampaio, L. Martins Teixeira, L. E. Vaz Miranda, L. W. Riley, and J. L. Gerberding. 2003. Extended-spectrum ß-lactamase-producing Klebsiella pneumoniae in a neonatal intensive care unit: risk factors for infection and colonization. J. Hosp. Infect. 53:198-206.[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Antimicrob. Agents Chemother. | Clin. Microbiol. Rev. |
|---|---|
| Clin. Vaccine Immunol. | ALL ASM JOURNALS |
|---|
