To determine the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization among patients presenting for hospital admission and to identify risk factors for MRSA colonization.

Surveillance cultures were performed at the time of hospital admission to identify patients colonized with S. aureus. A case-control study was performed to identify risk factors for MRSA colonization.

A tertiary-care academic medical center.

Adults presenting for hospital admission (N = 974).

S. aureus was isolated from 205 (21%) of the patients for whom cultures were performed. Methicillin-sensitive S. aureus was isolated from 179 (18.4%) of the patients, and MRSA was isolated from 26 (2.7%) of the patients. All 26 MRSA-colonized patients had been admitted to a healthcare facility in the preceding year, had at least one chronic illness, or both. In multivariate analyses comparing MRSA-colonized patients with control-patients, admission to a nursing home (odds ratio [OR], 16.5; 95% confidence interval [CI95], 1.4 to 192.1; P = .025) or a hospitalization of 5 days or longer during the preceding year (OR, 3.91; CI95, 1.1 to 13.9; P = .035) were independent predictors of MRSA colonization.

Patients colonized with MRSA admitted to this hospital likely acquired the organism during previous encounters with healthcare facilities. There was no evidence that MRSA colonization occurs commonly among low-risk individuals in this community. These data suggest that evaluation of recent healthcare exposures is essential if true community acquisition of MRSA is to be confirmed.

To assess the impact of an automated intraoperative alert to redose prophylactic antibiotics in prolonged cardiac operations.

Randomized, controlled, evaluator-blinded trial.

University-affiliated hospital.

Patients undergoing cardiac surgery that lasted more than 4 hours after the preoperative administration of cefazolin, unless they were receiving therapeutic antibiotics at the time of surgery.

Randomization to an audible and visual reminder on the operating room computer console at 225 minutes after the administration of preoperative antibiotics (reminder group, n = 137) or control (n = 136). After another 30 minutes, the circulating nurse was required to indicate whether a follow-up dose of antibiotics had been administered.

Intraoperative redosing was significantly more frequent in the reminder group (93 of 137; 68%) than in the control group (55 of 136; 40%) (adjusted odds ratio, 3.31; 95% confidence interval, 1.97 to 5.56; P < .0001). The impact of the reminder was even greater when compared with the 6 months preceding the study period (129 of 480; 27%; P < .001), suggesting some spillover effect on the control group. Redosing was formally declined for 19 of the 44 patients in the reminder group without redosing. The rate of surgical-site infection in the reminder group (5 of 137; 4%) was similar to that in the control group (8 of 136; 6%; P = .42), but significantly lower than that in the pre-study period (48 of 480; 10%; P = .02).

The use of an automatic reminder system in the operating room improved compliance with guidelines on perioperative antibiotic prophylaxis.

Infliximab is a chimeric monoclonal antibody against tumor necrosis factor (TNF)-alpha, used for the treatment of Crohn's disease and rheumatoid arthritis. Recently, an increased risk of infection due to Mycobacterium tuberculosis and rare cases of invasive fungal disease have been reported following infliximab therapy.

A 73-year-old woman with chronic rheumatoid arthritis who had been treated with methotrexate, leflunomide, and prednisone was given the first of three doses of infliximab in June 2001. In July 2001, she presented with cough, and in August, she had a right upper lobe infiltrate that was treated with levofloxacin without improvement. In October, the patient had right upper and middle lobe infiltrates on a chest x-ray and computed tomography scan. At bronchoscopy, an endobronchial mass was biopsied, which demonstrated Aspergillus fumigatus. Our patient had frequently accompanied her daughter on visits to another medical center following a stem cell transplant, where her daughter was instructed to wear a mask during all visits because of extensive building construction. We postulate that our patient may have acquired pulmonary aspergillosis during this period. Literature reviews on granulomatous diseases following infliximab therapy and hospital-acquired aspergillosis are presented.

The temporal relationship between the administration of infliximab and A. fumigatus infection in this patient suggests a causal relationship and possible healthcare-associated acquisition. These data underscore the importance of both patient and family education on prevention strategies when potent immune-modulating medications such as infliximab have been prescribed (Infect Control Hosp Epidemiol 2003;24:477-482)

To examine a comprehensive approach for preventing percutaneous injuries associated with phlebotomy procedures.

From 1993 through 1995, personnel at 10 university-affiliated hospitals enhanced surveillance and assessed underreporting of percutaneous injuries; selected, implemented, and evaluated the efficacy of phlebotomy devices with safety features (ie, engineered sharps injury prevention devices [ESIPDs]); and assessed healthcare worker satisfaction with ESIPDs. Investigators also evaluated the preventability of a subset of percutaneous injuries and conducted an audit of sharps disposal containers to quantify activation rates for devices with safety features.

The three selected phlebotomy devices with safety features reduced percutaneous injury rates compared with conventional devices. Activation rates varied according to ease of use, healthcare worker preference for ESIPDs, perceived “patient adverse events,” and device-specific training.

Device-specific features and healthcare worker training and involvement in the selection of ESIPDs affect the activation rates for ESIPDs and therefore their efficacy. The implementation of ESIPDs is a useful measure in a comprehensive program to reduce percutaneous injuries associated with phlebotomy procedures.

Hand hygiene (HH) is critical to infection control, but compliance is low. Alcohol-based antiseptics may improve HH. HH practices in Russia are not well described, and facilities are often inadequate.

Four 6-bed units in a neonatal intensive care unit in St. Petersburg, Russia.

Prospective surveillance of HH compliance, nosocomial colonization, and antibiotic administration was performed from January until June 2000. In February 2000, alcohol-based hand rub was provided for routine HH use. Eight weeks later, a quality improvement intervention was implemented, consisting of review of interim data, identification of opinion leaders, posting of colonization incidence rates, and regular feedback. Means of compliance, colonization, and antibiotic use were compared for periods before and after each intervention.

A total of 1,027 events requiring HH were observed. Compliance was 44.2% before the first intervention, 42.3% between interventions, and 48% after the second intervention. Use of alcohol rose from 15.2% of HH indications to 25.2% between interventions and 41.5% after the second intervention. The incidence of nosocomial colonization (per 1,000 patient-days) with Klebsiella pneumoniae was initially 21.5, decreased to 4.7, and then was 3.2 in the final period. Rates of antibiotic and device use also decreased.

HH may have increased slightly, but the largest effect was a switch from soap and water to alcohol, which may have been associated with decreased cross-transmission of Klebsiella, although this may have been confounded by lower device use. Alcohol-based antiseptic may be an improvement over current practices, but further research is required.

To evaluate the outcome of attempted Hickman catheter salvage in neutropenic cancer patients with Staphylococcus aureus bacteremia who were not using antibiotic lock therapy.

Retrospective cohort study.

A university-affiliated, tertiary-care hospital with 1,500 beds for adult patients.

All neutropenic cancer patients who had a Hickman catheter andS. aureus bacteremia (32 episodes in 29 patients) between January 1998 and March 2002.

Salvage attempts were defined as cases where the Hickman catheter was not removed until we obtained the results of follow-up blood cultures performed 48 to 72 hours after starting treatment with antistaphylococcal antibiotics. Salvage was considered to be successful if the Hickman catheter was still in place 3 months later without recurrent bacteremia or death.

Catheter salvage was attempted in 24 (75%) of the 32 episodes. Of the salvage attempts, the success rate was 50% (12 of 24). Salvage attempts were successful in 14% (1 of 7) of the episodes with positive follow-up blood cultures, and in 65% (11 of 17) of those with negative follow-up blood cultures (P = .07). If the analysis is confined to cases with no external signs of catheter infection, salvage attempts were successful in 14% (1 of 7) of the episodes with positive follow-up blood cultures and in 80% (8 of 10) of those with negative follow-up blood cultures (P = .02).

In neutropenic cancer patients with S. aureus bacteremia, attempted catheter salvage without antibiotic lock therapy was successful in 50% of the cases.

To determine the prevalence of positive tuberculin skin tests (TSTs), incidence of TST conversion, risk factors for positive TSTs, and history of active TB among HCWs in microbiology laboratories in New York City.

Two-year survey from May 1999 to June 2001.

Nineteen microbiology laboratories.

During the first year, interviews were conducted with 345 laboratory HCWs (mean, 18 HCWs per site; range, 2 to 51) to assess the prevalence of positive TSTs, but 3 (1%) could not recall their result and were excluded from further analyses. The mean age of the remaining 342 HCWs was 48 years; 68% (n = 233) were female, 54% (n = 183) received bacille Calmette-Guérin (BCG) vaccination, and 71% (n = 244) were foreign born. The prevalence of a positive TST was 57% (n = 196), but only 20% (n = 39) of the HCWs received isoniazid. The incidence of TST conversion in the second year of the study was 1% (1 of 108). Multivariate analysis identified age (odds ratio [OR] per year, 1.05; 95% confidence interval [CI95], 1.02–1.08), foreign birth (OR, 3.80; CI95, 1.98–7.28), BCG immunization (OR, 4.89; CI95, 2.72–8.80), and employment in a mycobacteriology laboratory (OR, 2.14; CI95, 1.25–3.68) as risk factors for a positive TST. Only one HCW had been treated for active TB.

The prevalence of positive TSTs was high among laboratory HCWs, but the TST conversion rate was low. Higher rates of treatment for latent TB infection are desirable.

To identify factors that increase the risk of sternal surgical wound infection after cardiovascular surgery and to develop a bedside clinical risk index using these factors.

A risk index was developed using clinical data collected from a cohort of 11,508 cardiac surgery patients and validated using three independent subsets of the data. With two of these subsets, we derived a logistic regression equation and then modified the scoring algorithm to simplify the calculation of patient risk scores by clinicians. The final subset was used to validate the index. The area under the receiver operating characteristic (aROC) curve was the primary measure of goodness of fit.

Toronto General Hospital, a teaching hospital and the largest center for cardiac surgery in Ontario, Canada.

Cardiac surgery patients receiving cardiopulmonary bypass between April 1, 1990, and December 31, 1995, who survived at least 6 days after surgery.

Variables that were used to construct the risk index included reoperation due to complication (odds ratio, 4.3; range, 1.9 to 8.5), diabetes (odds ratio, 2.4; range, 1.5 to 3.7), more than 3 days in the intensive care unit (odds ratio, 5.4; range, 3.2 to 8.7), and use of the internal mammary artery for revascularization (odds ratio, 3.2; range, 1.7 to 5.8). Validation showed that the index had an aROC curve of 0.64.

The risk index described in this article allows clinicians to quickly stratify patients into four risk groups associated with an increasing risk of sternal surgical wound infection. It may be used perioperatively or as part of a wound infection surveillance system.

To evaluate the airborne viable bacterial concentrations generated during sputum induction and their reduction with exhaust ventilation, ultraviolet germicidal irradiation (UVGI), or both.

Exhaust ventilation, upper air UVGI lights, and a portable UVGI unit were operated independently or in combination while and after sputum induction was performed for 58 patients suspected of having active tuberculosis. Viable airborne bacteria were sampled with volumetric air samplers, grown on blood agar, and identified with standard techniques.

During and immediately after sputum induction, concentrations of airborne bacteria, particularly respiratory tract or oropharyngeal organisms, increased rapidly, regardless of environmental conditions. The subsequent rate of reduction of airborne bacteria was most rapid with the portable UVGI unit, followed by upper air UVGI with air mixing. Exhaust ventilation achieved high air changes per hour, but efficacy in reducing airborne bacterial concentrations was low. However, the continuous entrainment of bacteria-laden air from the hallway outside may have resulted in underestimation. The efficacy of a wall-mounted upper air UVGI fixture was significantly less if there was no air mixing. The irradiation from this fixture was of adequate germicidal intensity only in a narrow horizontal plane 2.5 m above the floor.

Sputum induction was associated with a rapid and substantial increase in airborne bacteria despite the use of exhaust ventilation providing more than 30 air changes per hour, and the adjunct use of UVGI. This emphasizes that health-care workers involved in similar cough-inducing procedures performed for patients with suspected tuberculosis must wear appropriate personal respirators (Infect Control Hosp Epidemiol 2003;24:483-489)

Determine differences in patterns of percutaneous injuries (PIs) in different types of hospitals.

Case series of injuries occurring from 1997 to 2001.

Large midwestern healthcare system with a consolidated occupational health database from 9 hospitals, including rural and urban, community and teaching (1 pediatric, 1 adult) facilities, ranging from 113 to 1,400 beds.

Healthcare workers injured between 1997 and 2001.

Annual injury rates for all hospitals decreased during the study period from 21 to 16.5/100 beds (chi-square for trend = 22.7; P = .0001). Average annual injury rates were higher at larger hospitals (22.5 vs 9.5 Pis/100 beds; P = .0001). Among small hospitals, rural hospitals had higher rates than did urban hospitals (14.87 vs 8.02 PIs/100 beds; P = .0143). At small hospitals, an increased proportion of injuries occurred in the emergency department (13.7% vs 8.6%; P = .0004), operating room (32.3% vs 25.4%; P = .0002), and ICU (12.3% vs 9.4%; P = .0225), compared with large hospitals. Rural hospitals had higher injury rates in the radiology department (7.7% vs 2%; P = .0015) versus urban hospitals. Injuries at the teaching hospitals occurred more commonly on the wards (28.8% vs 24%; P = .0021) and in ICUs (11.4% vs 7.8%; P = .0006) than at community hospitals. Injuries involving butterfly needles were more common at pediatric versus adult hospitals (15.8% vs 6.5%; P = .0001). The prevalence of source patients infected with HIV and hepatitis C was higher at large hospitals.

Significant differences exist in injury rates and patterns among different types of hospitals. These data can be used to target intervention strategies.

To compare the percutaneous injury rate associated with a standard versus a safety resheathable winged steel (butterfly) needle.

Before-after trial of winged steel needle injuries during a 33-month period (19-month baseline, 3-month training, and 11-month study intervention), followed by a 31-month poststudy period.

A 1,190-bed acute care referral hospital with inpatient and outpatient services in New York City.

All healthcare workers performing intravascular-access procedures with winged steel needles.

Safety resheathable winged steel needle.

The injury rate associated with winged steel needles declined from 13.41 to 6.41 per 100,000 (relative risk [RR], 0.48; 95% confidence interval [CI95], 0.31 to 0.73) following implementation of the safety device. Injuries occurring during or after disposal were reduced most substantially (RR 0.15; CI95, 0.06 to 0.43 Safety winged steel needle injuries occurred most often before activation of the safety mechanism was appropriate (39%); 32% were due to the user choosing not to activate the device, 21% occurred during activation, and 4% were due to improper activation. Preference for the safety winged steel needle over the standard device was 63%. The safety feature was activated in 83% of the samples examined during audits of disposal containers. Following completion of the study, the safety winged steel needle injury rate (7.29 per 100,000) did not differ significantly from the winged steel needle injury rate during the study period.

Implementation of a safety resheathable winged steel needle substantially reduced injuries among healthcare workers performing vascular-access procedures. The residual risk of injury associated with this device can be reduced further with increased compliance with proper activation procedures.

To examine the clinical and epidemiologic features, excess length of stay, extra costs, and mortality attributable to bloodstream infection (BSI) in neutropenic patients with hematologic malignancies.

Prospective cohort and matched case-control study.

All adult neutropenic patients with hematologic malignancies admitted to Cologne University Hospital between May 1, 1997, and April 30, 1998, were prospectively observed. Case-patients were defined as patients with nosocomial BSI; control-patients were selected among patients without BSI.

During the study period, the BSI rate in neutropenic patients was 14.3 per 100 neutropenic episodes. Eighty-four case-patients were included. Matching was successful for 96% of the cohort; 81 matched pairs were studied. The mean total length of stay was significantly longer for patients with BSI than for control-patients (37 vs 29 days; P = .002). Extra costs attributable to the infection averaged $3,200 (U.S.) per patient. The crude mortality rates of case-patients and control-patients were 16% and 4%, respectively (P = .013), with an attributable mortality of 12% (odds ratio, 11). Eighty-seven percent of patients met the criteria for sepsis according to the American College of Chest Physicians/Society of Critical Care Medicine. Severe sepsis or septic shock occurred in 13% of patients and was correlated with mortality (55% vs 10% in patients without severe sepsis or septic shock; P =.01).

Nosocomial BSI in neutropenic patients is significantly associated with an excess length of hospital stay, extra costs, and excess mortality. Severe sepsis and septic shock are closely correlated with an adverse outcome.

To define the extent of nosocomial transmission of methicillin-resistant Staphylococcus aureus (MRSA) in patients admitted to a tertiary-care hospital.

A blinded, prospective surveillance culture study of patients admitted to the hospital to determine the transmission (acquisition) rate of MRSA Risk factors associated with the likelihood of MRSA colonization on admission were investigated.

Tertiary-care military medical facility.

All patients admitted to the medicine, surgery, and pediatric wards, and to the medical, surgical, and pediatric intensive care units were eligible for inclusion.

Five hundred thirty-five admission and 374 discharge samples were collected during the study period. One hundred forty-one patients were colonized with methicillin-susceptible S. aureus (MSSA) and 20 patients (3.7%) were colonized with MRSA on admission. Of the 354 susceptible patients, 6 acquired MRSA during the study for a transmission rate of 1.7%. Patients colonized with MRSA on admission were more likely to be older than non-colonized or MSSA-colonized patients, to have received antibiotics within the past year, to have been hospitalized within the prior 3 years, or to have a known history of MRSA. Patients acquiring MRSA had an average hospital stay of 17.7 days compared with 5.3 days for those who did not acquire MRSA. Pulsed-field gel electrophoresis of the 6 MRSA isolates from patients who acquired MRSA revealed 4 distinct band patterns.

Most patients colonized with MRSA were identified on admission samples. Surveillance cultures of patients admitted may help to prevent MRSA transmission and infection.