The purpose of the study was to determine the incidence and risk factors for the acquisition of methicillin-resistant Staphylococcus aureus (MRSA) in our community.

This study used a cross-sectional design to assess patients colonized or infected with MRSA.

The study population consisted of residents of London, Ontario, Canada, who were identified as MRSA-positive for the first time in 1997.

All acute- and chronic-care hospitals, long-term healthcare facilities, and community physicians' offices in the city of London participated in the study.

Incidence of MRSA in the community, risk factors for acquisition, especially previous hospitalization over a defined period, and strain type were evaluated.

In 1997, 331 residents of London were newly identified as MRSA-positive, representing an annual incidence of 100/100,000 persons (95% confidence interval, 88.8-110.7). Thirty-one (9.4%) individuals were not healthcare-facility patients in the previous month, and 11 (3.3%), 10 (3.0%), and 6 (1.8%) individuals had no such contact in the previous 3, 6, and 12 months, respectively. One hundred seventy-seven strains, including five of the isolates from patients with no healthcare-facility contact in the previous year, were typed. One hundred sixty (90.3%) of these isolates, including all typed strains from patients with no healthcare facility contact, belonged to a single clone.

These findings demonstrate that the incidence of MRSA is higher than previously reported and that hospital contact is the single most important risk factor for the acquisition of MRSA in our community. Screening for MRSA in previously hospitalized patients at the time of hospitalization may reduce nosocomial spread and indirectly reduce the incidence of MRSA in the community.

To determine the comparative virulence of methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-sensitive S aureus (MSSA) by consideration of predisposing factors and outcomes in patients infected with these organisms in the healthcare setting.

Analysis of an historical cohort of 504 bacteremic patients (316 MSSA and 188 MRSA), examining factors associated with mortality.

A 916-bed, university-affiliated, tertiary referral hospital.

Risk factors for the development of MRSA include male gender, admission due to trauma, immunosuppression, presence of a central vascular line or an indwelling urinary catheter, and a past history of MRSA infection. Overall mortality was 22%. Death due to bacteremia was significantly greater in the MRSA group (risk ratio, 1.68; P<.05), although these patients were not found to be more likely to die due to underlying disease during treatment of bacteremia. In those patients who recovered from bacteremia, no significant differences for the outcome of death could be determined between the MRSA and MSSA groups.

There is a general consensus in the published literature that MRSA bacteremia is more likely to be associated with death, and we confirm this conclusion. However, in contrast to other studies, our MRSA cohort does not appear to be more at risk of death due to underlying disease during treatment for bacteremia. Similarly, the general consensus that MRSA patients have an increased overall mortality was not confirmed in our study. Differences in comorbidities of patients may provide some explanation of these conflicting results, while an alternate explanation is that MRSA strains are more virulent than MSSA in some centers. Perhaps the most plausible explanation is that treatment is provided earlier and in a more aggressive fashion in some centers, leading to an overall lower mortality rate in all staphylococcal bacteremias in these institutions.

To investigate and control an apparent outbreak of lower respiratory tract infections due to Aureobasidium species.

Outbreak investigation.

University-affiliated medical center.

Nine patients who underwent bronchoscopy between June and August 1998.

Ten bronchoalveolar lavage (BAL) fluid cultures from nine patients grew Aureobasidium species during the outbreak period; whereas, respiratory specimens from only two patients grew Aureobasidium species during the preceding 6 years. No patient was judged to have true infection due to Aureobasidium species either before or after bronchoscopy. Nine of the 10 bronchoscopies that yielded Aureobasidium species were performed in the outpatient bronchoscopy suite. The Aureobasidium isolates were not associated with any one bronchoscope. Observation of bronchoscopy procedure revealed that plastic stopcocks labeled for single use were reused on different patients during BAL. There was no record of how many times each stopcock was being reused. After each use, the stopcocks were placed in an automated disinfection machine designed for bronchoscopes. Culture of the stopcocks after they had been “disinfected” yielded a heavy growth of Aureobasidium species, while culture of fluid from the automated disinfection machine was negative. Reuse of the stopcocks was halted, and, during the following 6-month period, Aureobasidium species were not isolated from any BAL specimen.

Reuse of medical equipment labeled for single use is potentially hazardous, especially if no quality control system is in place to monitor sterility and function after reprocessing.

To determine the cause of an outbreak of Pseudomonas aeruginosa cerebral ventriculitis among eight patients at a community hospital neurosurgical intensive care unit. All had percutaneous external ventricular catheters (EVCs) to monitor cerebrospinal fluid (CSF) pressure.

Cohort study of all patients who had EVCs placed during the epidemic period (August 8-October 22, 1997). A case-patient was any patient with P aeruginosa ventriculitis during the epidemic period. Pulsed-field gel electrophoresis (PFGE) was performed on all isolates.

P aeruginosa was significantly more likely to be isolated from CSF per EVC placed in the epidemic than pre-epidemic (January 1-August 7, 1997) periods (8/61 [13%] vs 2/131 [1.5%], P = 002). During the epidemic period, ventriculitis was significantly more likely after EVC placement in the operating room than in other units (8/24 vs 0/22, P = .004). EVC placement technique differed for EVCs placed in the operating room (little hair was removed, preventing application of an occlusive dressing) versus other hospital units (more hair was removed, and an occlusive dressing was applied). Among patients who had operating room EVC placement, contact with one healthcare worker was statistically significant (7/13 vs 0/8, P = .02). Hand cultures of this worker were negative. All isolates had closely related PFGE patterns.

These data suggest that a single healthcare worker may have contaminated EVC insertion sites, resulting in an outbreak of P aeruginosa ventriculitis. Affected patients were unlikely to have had an occlusive dressing at the EVC insertion site. Application of a sterile occlusive dressing may decrease the risk of ventriculitis in patients with EVCs.

To collect information about the safety of taking antiretroviral drugs for human immunodeficiency virus (HIV) postexposure prophylaxis (PEP).

A voluntary, confidential registry.

Hospital occupational health clinics, emergency departments, private physician offices, and health departments in the United States.

492 healthcare workers (HCWs) who had occupational exposures to HIV, were prescribed HIV PEP, and agreed to be enrolled in the registry by their healthcare providers were prospectively enrolled in the registry. Three hundred eight (63%) of 492 of the PEP regimens prescribed for these HCWs consisted of at least three antiretroviral agents. Of the 449 HCWs for whom 6-week follow-up was available, 195 (43%) completed the PEP regimen as initially prescribed. Forty-four percent (n=197) of HCWs discontinued all PEP drugs and did not complete a PEP regimen. Thirteen percent (n=57) discontinued ≥1 drug or modified drug dosage or added a drug but did complete a course of PEP. Among the 254 HCWs who modified or discontinued the PEP regimen, the two most common reasons for doing so were because of adverse effects attributed to PEP (54%) and because the source-patient turned out to be HIV-negative (38%). Overall, 340 (76%) HCWs with 6-week follow-up reported some symptoms while on PEP: nausea (57%), fatigue or malaise (38%), headache (18%), vomiting (16%), diarrhea (14%), and myalgias or arthralgias (6%). The median time from start of PEP to onset of each of the five most frequently reported symptoms was 3 to 4 days. Only 37 (8%) HCWs with 6-week follow-up were reported to have laboratory abnormalities; review of the reported abnormalities revealed that most were unremarkable. Serious adverse events were reported to the registry for 6 HCWs; all but one event resolved by the 6-month follow-up visit. Fewer side effects were reported by HCWs taking two-drug PEP regimens than by HCWs taking three-drug PEP regimens.

Side effects from HIV PEP were very common but were rarely severe or serious. The nature and frequency of HIV PEP toxicity were consistent with information already available on the use of these antiretroviral agents. Clinicians prescribing HIV PEP need to counsel HCWs about PEP side effects and should know how to manage PEP toxicity when it arises.

To determine the risk factors for acquisition of nosocomial primary bloodstream infections (BSIs), including the effect of nursing-staff levels, in surgical intensive care unit (SICU) patients.

A nested case-control study.

A 20-bed SICU in a 1,000-bed inner-city public hospital.

28 patients with BSI (case-patients) were compared to 99 randomly selected patients (controls) hospitalized ≥3 days in the same unit.

Case- and control-patients were similar in age, severity of illness, and type of central venous catheter (CVC) used. Case-patients were significantly more likely than controls to be hospitalized during a 5-month period that had lower regular-nurse-to-patient and higher pool-nurse-to-patient ratios than during an 8-month reference period; to be in the SICU for a longer period of time; to be mechanically ventilated longer; to receive more antimicrobials and total parenteral nutrition; to have more CVC days; or to die. Case-patients had significantly lower regular-nurse-to-patient and higher pool-nurse-to-patient ratios for the 3 days before BSI than controls. In multivariate analyses, admission during a period of higher pool-nurse-to-patient ratio (odds ratio [OR]=3.8), total parenteral nutrition (OR=1.3), and CVC days (OR=1.1) remained independent BSI risk factors.

Our data suggest that, in addition to other factors, nurse staffing composition (ie, pool-nurse-to-patient ratio) may be related to primary BSI risk. Patterns in intensive care unit nurse staffing should be monitored to assess their impact on nosocomial infection rates. This may be particularly important in an era of cost containment and healthcare reform.

To investigate and control an outbreak of colonization and infection caused by vancomycin-resistant enterococci (VRE) in a burn intensive care unit (BICU).

Epidemiological investigation, including multiple point-prevalence culture surveys of patients and environment, cultures from hands of healthcare workers (HCWs), pulsed-field gel electrophoresis (PFGE) typing of patient and environmental isolates, case-control study, and institution and monitoring of control measures.

BICU in an 800-bed university medical center in Galveston, Texas.

Between June 6, 1996, and July 14, 1997, 21 patients were colonized by VRE, and 4 of these patients developed bacteremia. Of 2,844 environmental cultures, 338 (11.9%) were positive, but all hand cultures from HCWs were negative. PFGE typing indicated that the outbreak was clonal, with VRE isolates from patients differing by ≤4 bands from the index case. Thirteen of 14 environmental isolates varied by ≤4 bands from the pattern of the index case. A case-control study analyzed by exact logistic regression identified diarrhea (odds ratio [OR], 43.9; 95% confidence interval [CI95], 5.5-infinity; P=.0001) and administration of an antacid (OR, 24.2; CI95,2.9-infinity; P=.002) as independent risk factors for acquisition of VRE. During a 5-week period in October and November 1996, all patient and 317 environmental cultures were negative for VRE. The outbreak recurred from a contaminated electrocardiogram lead that had not been identified during the prior 5 weeks. VRE were finally eradicated from the BICU in July 1997, using barrier isolation and a very aggressive environmental decontamination program.

A VRE outbreak in a BICU over 13 months was caused by a single clone. After apparent eradication of VRE from a BICU, recrudescence of the outbreak occurred, evidently from a small inapparent source of environmental contamination. Changes in gastrointestinal (GI) tract function (motility) and administration of medications, other than antibiotics, that have an effect on the GI tract may increase the risk of GI tract colonization by VRE in burn patients. Application of barrier isolation and an aggressive environmental decontamination program can eradicate VRE from a burn population.

To assess the activity against three non-enveloped viruses (an adeno-, a rhino- and a rotavirus) of a gel containing 60% ethanol, using experimentally contaminated thumb- and fingerpads of 12 panelists, as per standard procedure E-1838-96 of the American Society of Testing and Materials.

Each digit received 10 μL of the test virus suspension. The inocuLum from the thumbs was eluted immediately with 990 μL of Earle's balanced salt solution (EBSS) to assess the amount of virus on each digit (0-minute control). The inoculum on the fingers was allowed to dry (20-25 minutes), and virus was eluted from two fingerpads to determine the loss in virus infectivity upon drying (baseline titer). Then the dried inoculum on randomly selected fingers was exposed to 1 mL of the test product or standard hard water (200-ppm calcium carbonate) for 20 seconds. The virus remaining was eluted with 1 mL of EBSS, titrated to determine the amounts eliminated, and compared to the baseline titer.

Each digit received at least 104 plaque-forming units of virus in 10 μL The amounts of adeno-, rhino-, and rotaviruses surviving the drying were 30%, 75%, and 42%, respectively. The product reduced the infectivity titers of the three viruses by 3 to >4 log10 when compared to a reduction of ≤ 1 log10 for the hard-water rinse.

The level of virus reduction by gel was statistically significantly higher than that seen with the water control. Evidence for such activity against non-enveloped viruses supports further investigation of the benefits of this product.

To investigate an outbreak of aspergillosis in a leukemia and bone marrow transplant (BMT) unit and to improve environmental assessment strategies to detect Aspergillus.

Epidemiological investigation and detailed environmental assessment.

A tertiary-care university hospital with a 37-bed leukemia and BMT unit.

Leukemic or BMT patients with invasive aspergillosis identified through prospective surveillance and confirmed by chart review.

We verified the diagnosis of invasive fungal infection by reviewing medical charts of at-risk patients, performing a case-control study to determine risk factors for infection, instituting wet mopping to clean all floors, providing N95 masks to protect patients outside high-efficiency particulate air (HEPA)-filtered areas, altering traffic patterns into the unit, and performing molecular typing of selected Aspergillus flavus isolates. To assess the environment, we verified pressure relationships between the rooms and hallway and between buildings, and we compared the ability of large-volume (1,200 L) and small-volume (160 L) air samplers to detect Aspergillus spores.

Of 29 potential invasive aspergillosis cases, 21 were confirmed by medical chart review. Risk factors for developing invasive aspergillosis included the length of time since malignancy was diagnosed (odds ratio [OR], 1.0; P=.05) and hospitalization in a patient room located near a stairwell door (OR, 3.7; P=.05). Two of five A flavus patient isolates were identical to one of the environmental isolates. The pressure in most of the rooms was higher than in the corridors, but the pressure in the oncology unit was negative with respect to the physically adjacent hospital; consequently, the unit acted essentially as a vacuum that siphoned non-HEPA-filtered air from the main hospital. Of the 78 samples obtained with a small-volume air sampler, none grew an Aspergillus species, whereas 10 of 40 cultures obtained with a large-volume air sampler did.

During active construction, Aspergillus spores may have entered the oncology unit from the physically adjacent hospital because the air pressure differed. Guidelines that establish the minimum acceptable pressures and specify which pressure relationships to test in healthcare settings are needed. Our data show that large-volume air samples are superior to small-volume samples to assess for Aspergillus in the healthcare environment.

To determine the prevalence and determinants of fecal carriage of vancomycin-resistant enterococci (VRE) in intensive care unit (ICU), hematology-oncology, and hemodialysis patients in The Netherlands.

Descriptive, multicenter study, with yearly 1-week point-prevalence assessments between 1995 and 1998.

All patients hospitalized on the testing days in ICUs and hematology-oncology wards in nine hospitals in The Netherlands were included.

Rectal swabs obtained from 1,112 patients were screened for enterococci in a selective broth and subcultured on selective media with and without 6 mg/L vancomycin. Resistance genotypes were determined by polymerase chain reaction. Further characterization of VRE strains was done by pulsed-field gel electrophoresis (PFGE). We studied possible determinants of VRE colonization with a logistic regression analysis model. Determinants analyzed included gender, age, and log-transformed length of prior hospital stay.

The results showed that 614 (55%) of 1,112 patients were colonized with vancomycin-sensitive enterococci, and 15 (1.4%) of 1,112 carried VRE. No increase in VRE colonization was observed from 1995 to 1998. Eleven strains were identified as Enterococcus faecium and four as Enterococcus faecalis. All E faecium and one E faecalis carried the vanA gene; the other E faecalis strains harbored the vanB gene. PFGE revealed that three vanB VRE isolated from patients hospitalized in one single ICU were related, suggesting nosocomial transmission. Though higher age seemed associated with VRE colonization, exclusion of patients with the nosocomial strain from the regression analysis decreased this relation to nonsignificant. Duration of hospital stay was not associated with VRE colonization.

VRE colonization in Dutch hospitals is an infrequent phenomenon. Although nosocomial spread occurs, most observed cases were unrelated, which suggests the possibility of VRE acquisition from outside the hospital. Prolonged hospital stay, age, and gender proved unrelated to VRE colonization.