To assess nosocomial transmission of imipenem-resistant Pseudomonas aeruginosa (IRPA) following bronchoscopy during August through October 1998.

Traditional and molecular epidemiological investigation of a case series.

University-affiliated community hospital.

18 patients with IRPA bronchial-wash isolates.

We reviewed clinical data, performed environmental cultures and molecular analysis of all IRPA isolates, and observed disinfection of bronchoscopes.

Of 18 patients who had IRPA isolated from bronchoscopic or postbronchoscopic specimens, 13 underwent bronchoscopy for possible malignancy or undiagnosed pulmonary infiltrates. Following bronchoscopy, 3 patients continued to have IRPA isolated from sputum and demonstrated clinical evidence of infection requiring specific antimicrobial therapy. The remaining 15 patients had no further IRPA isolated and remained clinically well 3 months following bronchoscopy. Pulsed-field gel electrophoresis revealed that all strains except one were >95% related. STERIS SYSTEM 1 had been implemented in July 1998 as an automatic endoscope reprocessor (AER) for all endoscopes and bronchoscopes. Inspection of bronchoscope sterilization cycles revealed incorrect connectors joining the bronchoscope suction channel to the STERIS SYSTEM 1 processor, obstructing peracetic acid flow through the bronchoscope lumen. No malfunction warning was received, and spore strips remained negative.

The similarity of diverse connectors and limited training by the manufacturer regarding AER for bronchoscopes were the two factors responsible for the outbreak. Appropriate connections were implemented, and there was no further bronchoscope contamination. We suggest active surveillance of all bronchoscopy specimen cultures, standardization of connectors of various scopes and automated processors, and systematic education of staff by manufacturers with periodic on-site observation.

To determine whether reprocessed single-use devices would meet regulatory standards for sterility and meet the same materials standards as a new device.

The study included single-use and reusable biopsy forceps and papillotomes and a reusable stone retrieval basket. The suitability of these devices for cleaning and disinfection or sterilization was examined.

Testing of cleanability was conducted on devices contaminated with technetium 99-radiolabeled human blood. Instruments were cleaned using hospital recommended practices for manual cleaning. Gamma counts per second were determined before and after cleaning to localize contaminants, which were additionally visualized using light and scanning electron microscopy. X-ray phc-toelectron spectroscopy was used to quantify contamination elements on the materials tested. Residual bioburden testing on instruments contaminated with microorganisms suspended in coagulable sheep blood was carried out to establish the efficacy of disinfection and sterilization.

All devices remained contaminated after cleaning, but single-use devices and the stone basket tended to be more heavily contaminated than reusable forceps and papillotomes. Cleaning procedures facilitated distribution of contaminants further into the lumens of the disposable forceps. Decreased concentrations of silicon and increased concentrations of carbon and nitrogen suggested that layers of silicon lubricant had been removed and contaminants were organic material. Reusable devices were effectively disinfected, but single-use devices were not Sterilization could not eliminate the challenge microorganisms completely.

None of the reprocessed single-use instruments were effectively cleaned, disinfected, or sterilized. This condition may provide an opportunity for the viability of non-resistant or nosocomial organisms and viruses. Additionally, reprocessing procedures may result in material destruction of fragile devices. Cost-saving initiatives that have inspired reprocessing of single-use devices, despite the absence of data establishing the efficacy of decontamination and the durability of materials throughout reprocessing, should be pushed into the background.

To evaluate an unusual number of rapidly growing acid-fast bacilli, later identified as Mycobacterium chelonae, and pink bacteria, later identified as Methylo-bacterium mesophilicum, from fungal cultures obtained by bronchoscopy.

Outbreak investigation.

An academic medical center performing approximately 500 bronchoscopies and 4,000 gastrointestinal endoscopies in 1998.

Patients undergoing bronchoscopy July 21 to October 2, 1998.

The infection control department reviewed patient charts and bronchoscopy logs; obtained cultures of source water, faucets, washers, unopened glutaraldehyde, glutaraldehyde from the washers, and endoscopes; observed endoscope and bronchoscope cleaning and disinfecting procedures; reviewed glutaraldehyde monitoring records; and sent M chelonae isolates for DNA fingerprinting.

M chelonae, M mesophilicum, gram-negative bacteria, and various molds grew from endoscopes, automated washers, and glutaraldehyde from the washers but not from unopened glutaraldehyde. The endoscopy unit regularly monitored the pH of glutaraldehyde, and the logs contained no deficiencies. The above sources remained positive for the same organisms after a glutaraldehyde cleaning cycle of the automated washers. DNA fingerprinting of the M chelonae revealed that they were clonally related.

The automated washers were contaminated with a biofilm that rendered them resistant to decontamination. The washers then contaminated the endoscopes and bronchoscopes they were used to disinfect. Our institution purchased new endoscopes and a new paracetic acid sterilization system.

To compute a risk index for nosocomial infection (NI) surveillance in geriatric long-term-care facilities (LTCFs) and rehabilitation facilities.

Analysis of data collected during the French national prevalence survey on NIs conducted in 1996. Risk indices were constructed based on the patient case-mix defined according to risk factors for NIs identified in the elderly.

248 geriatric units in 77 hospitals located in northern France.

All hospital inpatients on the day of the survey were included.

Data from 11,254 patients were recorded. The overall rate of infected patients was 9.9%. Urinary tract, respiratory tract, and skin were the most common infection sites in both rehabilitation facilities and LTCFs. Eleven risk indices, categorizing patients in 3 to 7 levels of increasing NI risk, ranging from 2.7% to 36.2%, were obtained. Indices offered risk adjustment according to NI rate stratification and clinical relevance of risk factors such as indwelling devices, open bedsores, swallowing disorders, sphincter incontinence, lack of mobility, immunodeficiency, or rehabilitation activity.

The optimal index should be tailored to the strategy selected for NI surveillance in geriatric facilities in view of available financial and human resources.

To assess the importance of diabetes, diabetes control, hyperglycemia, and previously undiagnosed diabetes in the development of surgical-site infections (SSIs) among cardiothoracic surgery patients.

A 540-bed tertiary-care university-affiliated hospital.

Prospective cohort and case-control studies.

All patients having cardiothoracic surgery between November 1998 and September 1999 were eligible for participation. One thousand patients had preoperative hemoglobin Ale determinations. Seventy-four patients with SSIs were identified.

Diabetes (odd ratio [OR], 2.76; P<.001) and postoperative hyperglycemia (OR, 2.02; P=.007) were independently associated with development of SSIs. Among known diabetics, elevated hemoglobin Ale values were not associated with a statistically significantly increased risk of infection; the mean Ale value was 8.44% among those with infections compared with 7.80% for those without (P=.09). Forty-two (6%) of 700 patients without prior diabetes history had evidence of undiagnosed diabetes; their infection rate was comparable to that of known diabetics (3/42 [796] vs 17/300 [6%]; P=.72). An additional 30% of nondiabetics had elevated hemoglobin Ale determinations or perioperative hyperglycemia.

Postoperative hyperglycemia and previously undiagnosed diabetes are associated with development of SSIs among cardiothoracic surgery patients. Screening for diabetes and hyperglycemia among patients having cardiothoracic surgery may be warranted to prevent postoperative and chronic complications of this metabolic abnormality.

To determine the safety and cost-effectiveness of replacing the intravenous (TV) tubing sets in hospitalized patients at 4- to 7-day intervals instead of every 72 hours.

Prospective, randomized study of infusion-related contamination associated with changing IV tubing sets within 3 days versus within 4 to 7 days of placement.

A tertiary university cancer center.

Cancer patients requiring IV infusion therapy were randomized to have the IV tubing sets replaced within 3 days (280 patients) or within 4 to 7 days of placement (232 patients). Demographic, microbiological, and infusion-related data were collected for all participants. The main outcome measures were infusion- or catheter-related contamination or colonization of IV tubing, determined by quantitative cultures of the infusate, and infusion- or catheter-related bloodstream infection (BSI), determined by quantitative culture of the infusate in association with blood cultures in febrile patients.

The two groups were comparable in terms of patient and catheter characteristics and the agents given through the IV tubing. Intent-to-treat analysis demonstrated a higher level of tubing colonization in the 4- to 7-day group versus the 3-day group (median, 145 vs 50 colony-forming units; P=.02). In addition, there were three episodes of possible infusion-related BSIs, all of which occurred in the 4- to 7-day group (P=.09). However, when the 84 patients who received total parenteral nutrition, blood transfusions, or interleukin-2 through the IV tubing were excluded, the two groups had a comparable rate of colonization (0.4% vs 0.5%), with no catheter- or infusion-related BSIs in either group.

In patients at low risk for infection from infusion- or catheter-related infection who are not receiving total parenteral nutrition, blood transfusions, or interleukin-2, delaying the replacement of IV tubing up to 7 days may be safe, as well as cost-effective.

To compare molecular typing by amplified fragment-length polymorphism (AFLP) analysis with pulsed-field gel electrophoresis (PFGE) with respect to the ability to differentiate between epidemiologically related and unrelated isolates of common nosocomial pathogens recovered during a period of endemicity.

Retrospective laboratory analysis.

Tertiary-care institution.

17 isolates of Acinetobacter baumannii, 22 isolates of Pseudomonas aeruginosa, and 22 vancomycin-resistant Enterococcus faecium (VRE) were typed by both methods.

AFLP generated comparable results to PFGE for A baumannii and P aeruginosa isolates; both methods identified epidemiologically related and unrelated isolates. However, strain typing of VRE isolates produced discordant results between the two methods. PFGE identified 10 different strain types and differentiated between all epidemiologically related and unrelated isolates. In contrast, AFLP generated only five different strain types, three of which contained both epidemiologically related and unrelated isolates.

Molecular typing by AFLP is comparable to PFGE for A baumannii and P aeruginosa isolates. For VRE isolates, however, PFGE remains the method of choice.

To investigate an outbreak of Burkholderia cepacia.

Observational study and chart review.

Adult non-cystic fibrosis (CF) patients.

Intensive care units (ICUs) at a university-affiliated teaching hospital.

As part of the epidemiological investigation, we conducted a chart review and collected environmental samples. A review of work schedules of healthcare workers also was performed. We used B cepacia selective agar for preliminary screening for all isolates, which subsequently were confirmed as members of the B cepacia complex by polyphasic analysis employing conventional biochemical reactions and genus- and species-specific polymerase chain reaction assays. Pulsed-field gel electrophoresis, randomly amplified polymorphic DNA typing, and automated ribotyping were used to genotype the isolates. As part of the intervention, contact isolation precautions were initiated for all patients identified as having had a culture positive for B cepacia.

Between September 1997 and September 1999, B cepacia was isolated from 31 adult patients without CF in ICUs at a university-affiliated teaching hospital. Based on geographic clustering and genotypic analysis, three distinct clusters were observed involving 20 patients. Isolates from 17 of these patients were available for testing and were found to be of the same strain (outbreak strain). Further taxonomic analysis indicated that the outbreak strain was B cepacia complex genomovar III. Twelve (71%) of the 17 patients were judged to be infected, and 5 (29%) were colonized with this strain. Six of 200 environmental cultures from multiple sources in the hospital's ICUs yielded B cepacia. Two of these isolates, both recovered from rooms of colonized patients, were the same genotype as the outbreak strain recovered from patients.

Despite an extensive investigation, the source of the B cepacia clone involved in this outbreak remains unknown. The spatial and temporal pattern of cases suggests that cross-transmission of a genetically related strain contributed to clustering among patients. The initiation of contact isolation may have limited the extent of this transmission. Additional studies are needed to elucidate better the epidemiology of nosocomial B cepacia infection among non-CF adult patients.

To investigate the presence and clonal distribution of Legionella species in the water supply of 20 hospitals in Catalonia, Spain.

20 hospitals in Catalonia, an area of 32,000 km2, located in northeast Spain.

Environmental cultures of 186 points of potable water supply and 10 cooling towers were performed for the presence of Legionella species. Following filtration and acid treatment, the samples were seeded in selective MWY (modified Wadowsky Yee)-buffered charcoal yeast extract-a agar. All isolates obtained were characterized microbiologically and genotyped by Sfil pulsed-field gel electrophoresis (PFGE).

73 of 196 water samples, representing 17 of the 20 hospitals included in the study, were positive for Legionella pneumophila (serogroups 1, 2-14, or both). The degree of contamination ranged from 200 to 74,250 colony-forming units/L. Twenty-five chromosomal DNA subtypes were detected by PFGE. A single DNA subtype was identified in 10 hospitals, 2 DNA subtypes were observed in 6 hospitals, and 1 hospital exhibited 3 different DNA subtypes. Each hospital had its own Legionella DNA subtype, which was not shared with any other hospitals.

Legionella was present in the water of most of the hospitals studied; each such hospital had a unique, dominant chromosomal DNA subtype. The verification of several genomic DNA restriction profiles in such a small geographic area demonstrates the great genetic diversity of Legionella in the aquatic environment.