To investigate trends in ventilator-associated pneumonia (VAP) rates during participation in the German nosocomial infection surveillance system (Krankenhaus-Infektions-Surveillance-System [KISS]).

A total of 71 ICUs that began participating in KISS in 1999 or later and continued participation for at least 36 months were selected. Beginning with the first month of participation, the pooled mean rate of VAP in the ICUs was calculated for each year of participation. The incidence densities for the 3 years of participation were compared using the Pearson x2 test. Relative risks (RRs) and 95% confidence intervals (CIs) were calculated. VAP rates were calculated for each ICU and year of participation, and rates for years 1 and 3 were compared using the Wilcoxon test for paired samples.

Twenty-nine medical-surgical, 18 medical, 20 surgical, 2 neurosurgical, and 2 pediatric ICUs met the selection criteria. Surveillance data were available on 181,275 patients, for whom there were 613,098 patient-days and 224,138 ventilator-days. A total of 2,043 cases of VAP were reported. The ICUs had a pooled VAP rate of 10.5 cases per 1,000 ventilator-days during year 1 of KISS surveillance. In year 2, the rate decreased by 19%, to 8.7 cases per 1,000 ventilator-days (RR, 0.81 [95% CI, 0.73-0.90]). In year 3, the rate decreased by 24% from year 1, to 8.0 cases per 1,000 ventilator-days (RR, 0.76; 95% CI, 0.68-0.85). Both results were significant (P < .001 by the Pearson x2 test). Comparison of the VAP rates of the ICUs did not show a significant difference between years 1 and 3 of KISS participation.

Surveillance was associated with a significant reduction in the pooled rate of VAP during years 1-3 of KISS participation.

To assess the duration of shedding of influenza A virus detected by polymerase chain reaction (PCR) and cell culture among patients hospitalized with influenza A virus infection.

Mayo Clinic (Rochester, Minnesota) hospitals that cater to both the community and referral populations.

Patients 18 years old and older who were hospitalized between December 1, 2004, and March 15, 2005, with a laboratory-confirmed (ie, PCR-based) diagnosis of influenza A virus infection were consecutively enrolled. Additional throat swab specimens were collected at 2, 3, 5, and 7 days after the initial specimen (if the patient was still hospitalized). All specimens were tested by PCR and culture (both conventional tube culture and shell vial assay). Information on demographic characteristics, date of symptom onset, comorbidities, immunosuppression, influenza vaccination status, and receipt of antiviral treatment was obtained by interview and medical record review. Patients were excluded if informed consent could not be obtained or if the date of symptom onset could not be ascertained.

Of 149 patients hospitalized with influenza A virus infection, 50 patients were enrolled in the study. Most patients were older (median age, 76 years), and almost all (96%) had underlying chronic medical conditions. Of 41 patients included in the final analysis, influenza A virus was detected in 22 (54%) by PCR and in 12 (29%) by culture methods at or beyond 7 days after symptom onset. All 12 patients identified by culture also had PCR results positive for influenza A virus.

Hospitalized patients with influenza A virus infection can shed detectable virus beyond the 5- to 7-day period traditionally considered the duration of infectivity. Additional research is needed to assess whether prolonging the duration of patient isolation is warranted to prevent nosocomial outbreaks during the influenza season.

Intravascular catheters are indispensable tools in modern medical therapy. In spite of their great benefits, however, the widespread use of catheters leads to several complications, including infections that cause significant morbidity, mortality, and economic losses for hospitalized patients.

This study was conducted at Farabi Hospital, a 495-bed facility at Karadeniz Technical University Medical School in Trabzon, Turkey, and involved 3 separate periods: preeducation, education, and posteducation. Patients with intravascular catheters were monitored daily, as were the results of their physical examinations. The information acquired was recorded in a questionnaire.

During the preeducation period (October 2003 through March 2004), 405 intravascular catheters inserted into 241 patients were observed for 5,445 catheter-days. Seventy-one cases of intravascular catheter-related infection (CRI) were identified, giving a CRI rate of 13.04 infections per 1,000 catheter-days. The catheter-related bloodstream infection (CRBSI) rate was 8.3 infections per 1,000 catheter-days, and the exit-site infection (ESI) rate was 3.5 infections per 1,000 catheter-days. During the posteducation period (June through November 2004), 365 intravascular catheters inserted into 193 patients were observed for 5,940 catheter-days. Forty-five cases of CRI were identified, giving a rate of 7.6 infections per 1,000 catheter-days. The CRBSI rate was 4.7 infections per 1,000 catheter-days, and the ESI rate was 2.2 infections per 1,000 catheter-days. When findings from the 2 periods were compared, it was determined that education reduced CRI incidence by 41.7%.

CRI can be prevented when hospital personnel are well informed about these infections. We compared the knowledge levels of the relevant personnel in our hospital before and after theoretical and practical training and identified a significant increase in knowledge after training (P < .0001 ). Parallel to this, although still below ideal levels, we identified a significant improvement in the incidence of CRI during the posteducation period (P = .004). The rate was low for the first 3 months of this period but increased 2.08 times after the third month. In conclusion, regular training for the residents in charge of inserting intravascular catheters and the nurses and interns who maintain the catheters is highly effective in reducing the rate of CRI in large teaching hospitals.

To describe the incidence and etiology of nosocomial bloodstream infections in children at a general hospital.

Review of nosocomial bloodstream infections detected in children during 1991-2005. Data were prospectively gathered through active surveillance. Annual rates of infection were compared.

A public general hospital in San Luis Potosi, Mexico.

Children younger than 15 years of age admitted to pediatric wards and subjected to prospective surveillance for nosocomial infection.

Measures instituted to decrease the incidence of hospital-acquired infection during the 15-year study period included establishing active surveillance for hospital-acquired infection, reinforcing compliance with handwashing recommendations, decreasing the degree of crowding on wards, establishing guidelines for the management of intravenous catheters and solutions, preparing parenteral nutrition and intravenous solutions under a laminar air-flow hood, and increasing nursing personnel.

There were 868 nosocomial bloodstream infections detected in 29,273 subjects (overall rate, 2.94 episodes per 100 discharges). Infection rates were greatest among children admitted to the neonatal intensive care unit and lowest for those admitted to the school-age ward and the infectious diseases ward. There was a significant decrease in rates of nosocomial bacteremia in all of the wards. The organisms isolated most commonly were Klebsiella pneumoniae, Candida species, and coagulase-negative staphylococci. Mortality rates were higher for children with a gram-negative bacterial bloodstream infection (45.2%) and lower for children with a gram-positive bacterial infection (19.2%).

Rates of nosocomial bloodstream infection decreased over the past 15 years at our hospital but continue to cause significant mortality. Continuing efforts to decrease the frequency of and mortality due to bloodstream infection are warranted.

Use of an alcohol-based hand rub for hand hygiene has recently been recommended by the Centers for Disease Control and Prevention. However, the proper technique for using hand rub has not been well described and is not routinely taught in hospitals.

To evaluate the impact of training on proper technique as outlined by the European Standard for testing alcohol-based hand rubs (European Norm 1500) in a clinical study.

Prospective study including 180 healthcare workers (HCWs) in a 450-bed, university-affiliated geriatric hospital where alcohol-based hand rub was introduced in the late 1970s.

Structured training program in hand hygiene with alcohol-based hand rub. Technique for using hand rub was tested by the addition of a fluorescent dye to the disinfectant and the number of areas missed was quantified by a validated visual assessment method. In addition, the number of bacteria eradicated was estimated by calculating the difference between the log10 number of colony-forming units (cfu) of bacteria on the fingertips before and after the procedure, and reported as reduction factor (RF).

Log10 cfu bacterial counts on fingertips before and after training in the appropriate technique for using hand rub.

At baseline, only 31% of HCWs used proper technique, yielding a low RF of 1.4 log10 cfu bacterial count. Training improved HCW compliance to 74% and increased the RF to 2.2 log10 cfu bacterial count, an increase of almost 50% (P < .001). Several factors, such as applying the proper amount of hand rub, were significantly associated with the increased RF.

These results demonstrate that education on the proper technique for using hand rub, as outlined in EN 1500, can significantly increase the degree of bacterial killing.

To assess the surveillance coverage obtained with Centers for Disease Control and Prevention (CDC)-recommended surveillance of healthcare-associated infections (HAIs), which is focused on intensive care units (ICUs) and emphasizes device-related infections (ie, those associated with central venous catheters, ventilators, and/or urinary catheters), compared with the surveillance coverage achieved by comprehensive hospitalwide surveillance. In addition, we assessed whether the infection rates in step-down units more resemble those in wards or ICUs.

Review of prospectively obtained, comprehensive hospitalwide surveillance data from 2004 through 2005 for an acute care tertiary care hospital with approximately 700 beds. Surveillance data was obtained by trained infection control professionals using standard CDC criteria for HAIs.

CDC-recommended ICU surveillance for catheter-related bloodstream infection (BSI) and ventilator-associated pneumonia would have detected only 87 (21.4%) of 407 catheter-related BSIs and only 66 (37.9%) of 174 respiratory tract infections that occurred in the medical and surgical services. Only 31 (34.8%) of 89 infections caused by methicillin-resistant Staphylococcus aureus and 7 (31.8%) of 22 infections caused by vancomycin-resistant Enterococcus occurred in our adult ICUs.

Rates of HAIs were highest in the ICUs, intermediate in step-down units, and lowest in the wards. The rates of infections in the step-down units were more similar to those in the wards than to those in the ICUs. To prevent HAIs, more comprehensive surveillance may be indicated.

To quantify the different behavioral components of healthcare workers' motivation to comply with hand hygiene in a healthcare institution with a 10-year history of hand hygiene campaigning.

Cross-sectional study, by use of an anonymous, self-administered questionnaire.

A 2,200-bed university teaching hospital.

A stratified random sample of 2,961 medical and nursing staff.

A total of 1,042 questionnaires (35.2%) were returned. Of the respondents, 271 (26.0%) were physicians, 629 (60.4%) were nurses, and 141 (13.5%) were nursing assistants. Overall, 1,008 respondents provided information about sex; 718 (71.2%) of these were women. Respondents provided demographic information and data about various behavioral, normative, and control beliefs that determined their intentions with respect to performing hand hygiene. Among behavioral beliefs, the perception that healthcare-associated infections are severe for patients was highly ranked as a determinant of behavior by 331 (32.1%) of the respondents, and the perception that hand hygiene is effective at preventing these infections was ranked highly by 891 respondents (86.0%). Among normative beliefs, perceived social pressure from patients to perform hand hygiene was ranked highly by 760 respondents (73.7%), pressure from superiors was ranked highly by 687 (66.8%), pressure from colleagues was ranked highly by 596 (57.9%), and pressure from the person perceived to be most influential was ranked highly by 687 (68.8%). Among control beliefs, the perception that hand hygiene is relatively easy to perform was rated highly by 670 respondents (65.1%). High self-reported rates of adherence to hand hygiene (defined as performance of proper hand hygiene during 80% or more of hand hygiene opportunities) was independently associated with female sex, receipt of training in hand hygiene, participation in a previous hand hygiene campaign, peer pressure from colleagues, perceived good adherence by colleagues, and the perception that hand hygiene is relatively easy to perform.

In a setting with a long tradition of hand hygiene campaigns, behavioral beliefs are strongly in favor of hand hygiene, but adherence is driven by peer pressure and the perception of high self-efficacy, rather than by reasoning about the impact of hand hygiene on patient safety. Female sex, training, and campaign exposure increased the likelihood of compliance with hand hygiene. This additional insight can help to shape future promotional activity.

The rate of influenza vaccination among healthcare workers (HCWs) is approximately 40%. Differences in vaccination rates among HCW groups and reasons for accepting or rejecting vaccination are poorly understood.

To determine vaccination rates and motivators among different HCW groups during the 2004-2005 influenza season.

Cross-sectional survey conducted between July 10 and September 30, 2005.

Two tertiary care teaching hospitals in an urban center.

Physicians, nurses, nursing aides, and other staff. Surveys were collected from 1,042 HCWs (response rate, 42%).

Sixty-nine percent of physicians (n = 282) and 63% of medical students (n = 145) were vaccinated, compared with 46% of nurses (n = 336), 42% of nursing aides (n = 135), and 29% of administrative personnel (n = 144). Physicians and medical students were significantly more likely to be vaccinated than all other groups (P < .0001). Pediatricians (84%) were more likely than internists (69%) and surgeons (43%) to be vaccinated (P < .0001). Among the HCWs who were vaccinated, 33.4% received the live attenuated influenza vaccine (LAIV) and 66.6% received trivalent inactivated influenza vaccine (TIV). Vaccinated HCWs were less likely than unvaccinated HCWs to report an influenza-like illness (P = .03). Vaccination with LAIV resulted in fewer episodes of influenza-like illness than did receiving no vaccine (P = .03). The most common reason for rejecting vaccination was a concern about availability. Understanding that HCWs may transmit the virus to patients correlated with vaccine acceptance (P = .0004).

Significant differences in vaccination exist among physician specialties and employee groups, and there are inadequate vaccination rates among those with the greatest amount of patient contact, potentially providing a basis for group-specific interventions.

In studies of the association between antibiotic-resistant infection and mortality, the importance of controlling for the underlying severity of illness is well recognized. However, it is unclear when the severity of illness should be assessed. Controlling for severity of illness on the day the culture specimen is obtained may underestimate the true association between resistance and mortality.

TO assess the impact of calculating the Acute Physiology and Chronic Health Evaluation (APACHE) II score at different time points on the association between antimicrobial resistance and mortality.

We used an existing data set from a study that investigated the association between fluoroquinolone resistance and mortality. The APACHE II score was calculated at 3 time points: the day the culture specimen was obtained, 1 day before the culture specimen was obtained, and 2 days before the culture specimen was obtained. Separate multivariable models were constructed using the 3 different APACHE II scores. These models were compared qualitatively.

Of 91 total subjects, 51 were infected with a fluoroquinolone-resistant strain and 40 with a fluoroquinolone-susceptible strain. The median APACHE II score for all subjects was 13 (95% confidence interval [CI], 11-15) when calculated on the day the culture specimen was obtained, 12 (95% CI, 11-13) when calculated 1 day before, and 11 (95% CI, 10-13) when calculated 2 days before the culture specimen was obtained. Of 91 subjects, 12 (13.2%) died. The 3 multivariable models (each with the APACHE II score calculated on a different day) were not substantively different; the adjusted odds ratio for the association between fluoroquinolone-resistant infection and mortality varied only from 1.38 to 1.65 in the 3 models.

APACHE II scores calculated at different time points relative to obtainment of the culture specimen did not differ substantively. Furthermore, when the adjusted association between fluoroquinolone resistance and mortality was assessed, there were no substantive differences across multivariable models that incorporated APACHE II scores calculated at different time points.

To present a hypothetical model of the change in economic costs and health benefits to society that result from nosocomial infection control programs.

We use a modeling framework to represent how 2 types of costs change with nosocomial infection control programs: costs incurred by the hospital sector and community health services, as well as the private costs to patients. We also demonstrate how to value the health benefits of nosocomial infection control programs, using quality-adjusted life years.

Hypothetical modeling to incorporate the societal perspective.

A cohort of 50,000 simulated patients at risk of surgical site infection following total hip replacement.

A total of 8 hypothetical interventions that change costs and health outcomes among the cohort by preventing cases of surgical site infection following total hip replacement.

We demonstrate that when infection control interventions reduce economic costs and increase health benefits, they should be adopted without further question. If, however, interventions increase economic costs and increase health benefits, then the trade—off between costs and benefits should be examined. Decision-makers should assess the cost per unit of health benefit from infection control programs, consider the impact on health budgets, and compare infection control with alternative uses of scarce healthcare resources.

To produce an accurate estimate of the association between catheter-associated urinary tract infection (UTI) and intensive care unit (ICU) and hospital mortality, controlling for major confounding factors.

Nested case-control study in a multicenter cohort (the OutcomeRea database).

Twelve French medical or surgical ICUs.

All patients admitted between January 1997 and August 2005 who required the insertion of an indwelling urinary catheter. Patients who developed catheter-associated UTI (ie, case patients) were matched to control patients on the basis of the following criteria: sex, age ( ± 10 years), SAPS (Simplified Acute Physiology Score) II score ( ± 10 points), duration of urinary tract catheterization, and presence or absence of diabetes mellitus. The association of catheter-associated UTI with ICU and hospital mortality was assessed by use of conditional logistic regression.

Of the 3,281 patients who had an indwelling urinary catheter, 298 (9%) developed at least 1 episode of catheter-associated UTI. The incidence density of catheter-associated UTI was 12.9 infections per 1,000 catheterization-days. Crude ICU mortality rates were higher among patients with catheter-associated UTI, compared with those without catheter-associated UTI (32% vs 25%, P = .02); the same was true for crude hospital mortality rates (43% vs 30%, P>.01). After matching and adjustment, catheter-associated UTI was no longer associated with increased mortality (ICU mortality: odds ratio [OR], 0.846 [95% confidence interval {CI}, 0.659-1.086]; P = .19 and hospital mortality: OR, 0.949 [95% CI, 0.763-1.181]; P = .64).

After carefully controlling for confounding factors, catheter-associated UTI was not found to be associated with excess mortality among our population of critically ill patients in either the ICU or the hospital.

To investigate the potential reservoir and mode of transmission of pandrug-resistant (PDR) Acinetobacter baumannii in a 7-day-old neonate who developed PDR A. baumannii bacteremia that was presumed to be the iceberg of a potential outbreak.

Outbreak investigation based on a program of prospective hospital-wide surveillance for nosocomial infection.

A 24-bed neonatal intensive care unit in a 2,200-bed major teaching hospital in Taiwan that provides care for critically ill neonates born in this hospital and those transferred from other hospitals.

Samples from 33 healthcare workers' hands and 40 samples from the environment were cultured. Surveillance cultures of anal swab specimens and sputum samples were performed for neonates on admission to the neonatal intensive care unit and every 2 weeks until discharge. The PDR A. baumannii isolates, defined as isolates resistant to all currently available systemic antimicrobials except polymyxin B, were analyzed by pulsed-field gel electrophoresis. Control measures consisted of implementing contact isolation, reinforcing hand hygiene adherence, cohorting of nurses, and environmental cleaning.

One culture of an environmental sample and no cultures of samples from healthcare workers' hands grew PDR A. baumannii. The positive culture result involved a sample obtained from a ventilation tube used by the index patient. During the following 2 months, active surveillance identified PDR A. baumannii in 8 additional neonates, and isolates from 7 had the same electrokaryotype. Of the 9 neonates colonized or infected with PDR A. baumannii, 1 died from an unrelated condition. Reinforcement of infection control measures resulted in 100% adherence to proper hand hygiene protocol. The outbreak was stopped without compromising patient care.

In the absence of environmental contamination, transient hand carriage by personnel who cared for neonates colonized or infected with PDR A. baumannii was suspected to be the mode of transmission. Vigilance, prompt intervention and strict adherence to hand hygiene protocol were the key factors that led to the successful control of this outbreak. Active surveillance appears to be an effective measure to identify potential transmitters and reservoirs of PDR A. baumannii.

The Burkholderia cepacia complex is associated with colonization or disease in patients with cystic fibrosis (CF). For patients without CF, this complex is poorly understood apart from its presence in occasional point source outbreaks.

To investigate risk factors for B. cepacia bacteremia in hospitalized, intensive care unit patients without CF.

We identified patients with 1 or more blood cultures positive for B. cepacia between May 1, 1996, and March 31, 2002, excluding those with CF. Control patients were matched to case patients by ward, duration of hospitalization, and onset date of bacteremia. Matched analyses were used to identify risk factors for B. cepacia bacteremia.

We enrolled 40 patients with B. cepacia bacteremia into the study. No environmental or other point source for B. cepacia complex was identified, although horizontal spread was suspected. Implementation of contact precautions was effective in decreasing the incidence of B. cepacia bacteremia. We selected 119 matched controls. Age, sex, and race were similar between cases and controls. In multivariable analysis, renal failure that required dialysis, recent abdominal surgery, 2 or more bronchoscopic procedures before detection of B. cepacia bacteremia, tracheostomy, and presence of a central line before detection of B. cepacia bacteremia were independently associated with development of B. cepacia bacteremia, whereas presence of a percutaneous feeding tube was associated with a lower risk of disease.

B. cepacia complex is an important emerging group of nosocomial pathogens in patients with and patients without CF. Nosocomial spread is likely facilitated by cross-transmission, frequent pulmonary procedures, and central venous access. Infection control measures appear useful for limiting the spread of virulent, transmissible clones of B. cepacia complex.

To assess the risk-adjusted incidence and predictors of surgical site infections (SSIs).

Prospective, multicenter, observational cohort study.

Seven surgical departments at 3 urban academic hospitals in St. Petersburg, Russian Federation.

All patients had surgery performed between January 15 and May 12, 2000. A total of 1,453 surgical procedures were followed up. Medical records were unavailable for less than 3% of all patients; patients were not excluded for any other reason. The mean patient age was 49.3 years, 61% were female, and 34% had an American Society of Anesthesiologists physical status classification (hereafter, “ASA classification”) of at least 3. Surgery for 45% of the patients was emergent.

In all, 138 patients (9.5%) developed SSI, for a rate that was approximately 3.5 times the risk-stratified rates in the United States. Male sex (odds ratio [OR], 1.54), ASA classifications of 3 (OR, 3.7) or 4 (OR, 5.0), longer duration of surgery (OR, 2.2), and wound classes of 3 (OR, 5.5) or 4 (OR, 14.3) were associated with increased SSI risk in multivariate analysis. Endoscopic surgery was associated with a lower risk of SSI (OR, 0.23). Antibiotic prophylaxis was used in 0%-33% of operations, and 69% of uninfected patients received antibiotics after the operation.

The SSI rates are significantly higher than previously reported. Although this finding may be attributable to inadequate antibiotic prophylaxis, local infection control and surgical practices may also be contributors. Use of antibiotic prophylaxis should be encouraged and the effect of local practices further investigated. Active SSI surveillance should be expanded to other parts of the Russian Federation.

The Victorian Hospital Acquired Infection Surveillance System (VICNISS) hospital-acquired infection surveillance system was established in 2002 in Victoria, Australia, and collates surgical site infection (SSI) surveillance data from public hospitals in Australia.

To evaluate the association between the US National Nosocomial Infections Surveillance (NNIS) system's risk index and SSI rates for 7 surgical procedures.

SSI surveillance was performed with NNIS definitions and methods for surgical procedures performed between November 2002 and September 2004. Correlations were assessed using the Goodman-Kruskal γ statistic.

Data were submitted for the following numbers of procedures: appendectomy, 545; coronary artery bypass graft (CABG), 4,632; cholecystectomy, 1,001; colon surgery, 623; cesarean section, 4,857; hip arthroplasty, 3,825; and knee arthroplasty, 2,416. NNIS risk index and increasing SSI rate were moderately well correlated for appendectomy (γ = 0.55), colon surgery (γ = 0.48), and cesarean section (γ = 0.42). A fairly positive correlation was found for cholecystectomy (γ = 0.17), hip arthroplasty (γ = 0.2), and knee arthroplasty (γ = 0.16). However, for CABG surgery, a poor association was found (γ = 0.02).

The NNIS risk index was positively correlated with an increasing SSI rate for all 7 procedures; the strongest correlation was found for appendectomy, cesarean section, and colon surgery, and the poorest correlation was found for CABG surgery. We believe that risk stratification with the NNIS risk index is appropriate for comparison of data for most procedures and superior to use of no risk adjustment. However, for some procedures, particularly CABG, further studies of alternative risk indexes are needed to better stratify patients.

To assess factors associated with adherence to recommended barrier precautions among healthcare workers (HCWs) providing care to critically ill patients with severe acute respiratory syndrome (SARS).

Fifteen acute care hospitals in Ontario, Canada

Retrospective cohort study.

All patients with SARS who required intubation during the Toronto SARS outbreak in 2003.

HCWS who provided care to or entered the room of a SARS patient during the period from 24 hours before intubation until 4 hours after intubation.

Standardized interviews were conducted with eligible HCWs to assess their interactions with the SARS patient, their use of barrier precautions, their practices for removing personal protective equipment, and the infection control training they received.

Of 879 eligible HCWs, 795 (90%) participated. In multivariate analysis, the following predictors of consistent adherence to recommended barrier precautions were identified: recognition of the patient as a SARS case (odds ratio [OR], 2.5 [95% confidence interval {CI}, 1.5-4.5); recent infection control training (OR for interactive training, 2.7 [95% CI, 1.7-4.4]; OR for passive training, 1.7 [95% CI, 1.0-3.0]), and working in a SARS unit (OR, 4.0 [95% CI, 1.8-8.9]) or intensive care unit (OR, 4.3 [95% CI, 2.0-9.0]). Two factors were associated with significantly lower rates of consistent adherence: the provision of care for patients with higher Acute Physiology and Chronic Health Evaluation (APACHE) II scores (OR for score APACHE II of 20 or greater, 0.4 [95% CI, 0.28-0.68]) and work on shifts that required more frequent room entry (OR for 6 or more entries per shift, 0.5 [95% CI, 0.32-0.86]).

There were significant deficits in knowledge about self-protection that were partially corrected by education programs during the SARS outbreak. HCWs' adherence to self-protection guidelines was most closely associated with whether they provided care to patients who had received a definite diagnosis of SARS.

Studies conducted in tertiary care hospitals of different European countries and the United States have shown incidence rates of candidemia ranging from 0.17 to 0.76 and 0.28 to 0.96 per 1,000 admissions, respectively. So far, only 1 study has evaluated the incidence rates of candidemia in tertiary care hospitals in Latin American countries.

To evaluate the epidemiology of candidemia in 4 tertiary care hospitals in São Paulo, Brazil.

Multicenter, laboratory-based surveillance of candidemia.

A total of 7,038 episodes of bloodstream infection were identified, and Candida species accounted for 282 cases (4%). The incidence rate of candidemia was 1.66 candidemic episodes per 1,000 hospital admissions. Candida albicans was the most frequently isolated Candida species in all hospitals, but Candida species other than C. albicans accounted for 62% of isolates, including predominantly Candida parapsilosis and Candida tropicalis. Azole resistance was restricted to only 2% of all Candida isolates (1 isolate of Candida glabrata and 4 isolates of Candida rugosa). Candidemia was mostly documented in surgical patients with long durations of hospital stay. The crude mortality rate was 61%, and advanced age and high Acute Physiology and Chronic Health Evaluation II score were both conditions independently associated with risk of death.

We observed in our series a higher incidence rate of candidemia than that reported in European countries and the United States. Advanced age and a high Acute Physiology and Chronic Health Evaluation II score were factors associated with a higher probability of death in candidemic patients. Fluconazole-resistant Candida strains are still a rare finding in our case-based study of candidemia.

To assess adverse events associated with antiretroviral regimens for human immunodeficiency virus (HIV) postexposure prophylaxis (PEP), with a particular focus on the treatment combination of zidovudine, lamivudine, and tenofovir (ZDV-3TC-TDF).

Retrospective chart review for individuals who received HIV PEP for occupational and nonoccupational exposure, and multivariate analyses to identify risk factors for noncompletion of PEP and adverse events associated with PEP.

University of Rochester Health Service Occupational Health Program and University of Rochester AIDS Center.

Healthcare workers who received HIV PEP for occupational exposure from January 1, 1999, to December 31, 2004, and individuals who received HIV PEP for nonoccupational exposure from January 1, 2002, to December 31, 2004.

We found increased rates of nausea among subjects who received treatment with ZDV-3TC-TDF and subjects who received treatment with zidovudine, lamivudine, and indinavir (ZDV-3TC-IDV). Analyses showed that female sex was a risk factor for nausea. Compared with subjects who received treatment with ZDV-3TC-TDF, subjects who received treatment with ZDV-3TC-IDV were less likely to not complete the HIV PEP for occupational exposure.

Preventive treatment of adverse events may be necessary to ensure completion of HIV PEP.