To quantify the clinical impact of methicillin-resistance in Staphylococcus aureus causing infection complicated by bacteremia in adult patients, while controlling for the severity of patients' underlying illnesses.

Retrospective cohort study from October 1, 1995, through December 31, 2003.

A total of 438 patients with S. aureus infection complicated by bacteremia from a single Veterans Affairs healthcare system.

We found that 193 (44%) of the 438 patients had methicillin-resistant S. aureus (MRSA) infection and 114 (26%) died of causes attributable to S. aureus infection within 90 days after the infection was identified. Patients with MRSA infection had a higher mortality risk, compared with patients with methicillin-susceptible S. aureus (MSSA) infections (relative risk, 1.7 [95% confidence interval, 1.3-2.4]; P < .01), except for patients with pneumonia (relative risk, 0.7 [95% confidence interval, 0.4-1.3]). Patients with MRSA infections were significantly older (P < .01), had more underlying diseases (P = .02), and were more likely to have severe sepsis in response to their infection (P < .01) compared with patients with MSSA bacteremia. Patients who died within 90 days after S. aureus infection was identified were significantly older (P < .01) and more likely to have severe sepsis (P < .01) and pneumonia (P = .01), compared with patients who survived. After adjusting for age as a confounder, comorbidities, and pneumonia as an effect modifier, S. aureus infection-related mortality remained significantly higher in patients with MRSA infection than in those with MSSA infection, among those without pneumonia (hazard ratio, 1.8 [95% confidence interval, 1.2-3.0]); P < .01.

The results of this study suggest that patients with MRSA infections other than pneumonia have a higher mortality risk than patients with MSSA infections other than pneumonia, independent of the severity of patients' underlying illnesses.

To develop a predictive system for the occurrence of nosocomial pneumonia in patients who had cardiac surgery performed.

Retrospective cohort study.

Two cardiologic tertiary care hospitals in Rio de Janeiro, Brazil.

Between June 2000 and August 2002, there were 1,158 consecutive patients who had complex heart surgery performed. Patients older than 18 years who survived the first 48 postoperative hours were included in the study. The occurrence of pneumonia was diagnosed through active surveillance by an infectious diseases specialist according to the following criteria: the presence of new infiltrate on a radiograph in association with purulent sputum and either fever or leukocytosis until day 10 after cardiac surgery. Predictive models were built on the basis of logistic regression analysis and classification and regression tree (CART) analysis. The original data set was divided randomly into 2 parts, one used to construct the models (ie, “test sample”) and the other used for validation (ie, “validation sample”).

The area under the receiver–operating characteristic (ROC) curve was 69% for the logistic regression model and 76% for the CART model. Considering a probability greater than 7% to be predictive of pneumonia for both models, sensitivity was higher for the logistic regression models, compared with the CART models (64% vs 56%). However, the CART models had a higher specificity (92% vs 70%) and global accuracy (90% vs 70%) than the logistic regression models. Both models showed good performance, based on the 2-graph ROC, considering that 84.6% and 84.3% of the predictions obtained by regression and CART analyses were regarded as valid.

Although our findings are preliminary, the predictive models we created showed fairly good specificity and fair sensitivity.

Recent outbreaks of severe cases of Clostridium difficile-associated diarrhea (CDAD) reported in North America, the United Kingdom, and The Netherlands have emphasized the importance of an ongoing epidemiological surveillance of CDAD.

To determine the epidemiology of CDAD over the years 2000-2004 and the rate of nosocomial transmission of C. difficile.

Retrospective survey of inpatients with CDAD and molecular characterization of the strains isolated.

A 760-bed teaching hospital.

All CDAD cases diagnosed from January 1, 2000, to December 31, 2004, were reviewed. A CDAD case was defined as diarrhea in a hospitalized patient who had a stool specimen that tested positive for C. difficile cytotoxin or had a positive toxigenic culture result. CDAD was considered to be severe if a patient fulfilled at least 1 of the following 3 criteria: (1) presence of a fever (defined as temperature higher than 38.5°C), abdominal pain, and leukocyte count greater than 10,000 cells/mm3; (2) endoscopically or histologically proven pseudomembranous colitis; or (3) complications (defined as death with C. difficile infection as the primary or a contributing cause, toxic megacolon, perforation, toxic shock, and/or colectomy). CDAD was considered community-acquired if the diarrhea occurred in the patient within 72 hours after admission and if the patient had no history of hospitalization in the previous month; otherwise, CDAD was considered healthcare-associated. All the strains isolated were serogrouped and were characterized by toxinotyping and PCR ribotyping. Detection of toxin A, toxin B, and binary toxin was performed by PCR.

One hundred fifty-one cases of CDAD were diagnosed; 147 clinical records could be reviewed, and 131 strains were studied. The overall incidence of CDAD was 1.1 cases per 1,000 patients admitted, but incidence rates were higher in 2003-2004, compared with 2000-2002 (P = .017). Diarrhea was community acquired in 28 patients (19%). For patients with healthcare-associated CDAD, transmission of the strain from patient to patient (ie, infection with a strain of the same serogroup and PCR ribotype as the strain isolated from another patient hospitalized in the same ward or in a linked ward in the previous 2 months) was demonstrated in 12 cases (10.1%). Eleven percent of strains were positive for binary toxin. Binary toxin-positive strains were associated with more-severe diarrhea (P = .01 ) and with a higher case-fatality rate (P = .03). A specific clone of C. difficile (serogroup H, PCR ribotype sa026) accounted for 35 (26.7%) of all the strains isolated, but this clone was found both in healthcare-associated and community-acquired cases. Three strains belonged to toxinotype III, but only 1 was related to the hypervirulent clone involved in recent outbreaks.

The incidence of CDAD is low in our hospital, and cross-infection is limited. These results also suggest that strains with binary toxin might be more virulent.

Concern has recently arisen that the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization among populations at low risk for acquiring this pathogen has been increasing. Although previous studies have reported on the prevalence of MRSA colonization, most of these studies have been performed in healthcare settings, with few studies performed in the community among healthy individuals in nonhealthcare locations.

To determine the prevalence of community-acquired (CA) MRSA colonization in a random sample of healthy individuals in 4 nonhealthcare locations and to compare the genetic similarity between these CA isolates and nosocomial MRSA isolates.

Prospective, observational study conducted from October 2004 through February 2005.

Two colleges, 1 church, and 1 restaurant in Newark and Wilmington, Delaware.

A total of 295 healthy volunteers.

The rate of S. aureus colonization was 26.8%. Of the 79 S. aureus isolates analyzed, 3 (4%) were resistant to methicillin. Ribotyping of the 3 MRSA isolates showed that they were genetically distinct from each other. Each of the 3 CA isolates were genetically indistinguishable from 3 different nondominant clusters of MRSA isolates found in the medical and surgical intensive care units of Christiana Care Health System (Newark) and Wilmington Hospital (Wilmington).

The overall prevalence of CA-MRSA colonization in a random sample of healthy individuals in 4 nonhealthcare locations in Newark and Wilmington was 1.0%. Despite concerns that the prevalence of MRSA colonization is increasing, it still remains relatively low in our community, perhaps because CA-MRSA colonization tends to occur in clusters.

We describe the investigation of a tuberculosis (TB) exposure in which a neonatal intensive care unit (NICU) respiratory therapist was the index patient, as well as the rationale by which exposed infants were managed and possible explanations for the lack of transmission to these patients.

Description of an exposure investigation.

Academic, level IV NICU of a tertiary care children's hospital.

Contacts of a respiratory therapist with pulmonary TB disease, including household members, healthcare coworkers, and infant patients.

In addition to 5 household contacts, 248 healthcare coworkers and 180 infant patients were identified as possibly exposed during the 24 days that the index patient worked between December 3, 2004, and January 30, 2005. Tuberculin skin tests (TSTs) were performed for 233 of the 235 contacts with the greatest degree of exposure (household and coworker contacts) who had a previously documented negative TST result or whose TST status was unknown prior to the investigation. No cases of latent tuberculosis infection or TB disease were identified. Because of characteristics of the index case, the exposure duration and setting, the infants' small lung volumes, and lack of evidence of transmission to higher-risk contacts, infants were not clinically evaluated or empirically treated for TB disease. Surveillance for subsequent illness was carried out by primary healthcare providers and parents. No TB disease or unexplained illness in these infants was reported in the 20 months following the exposure.

After limited hospital exposure to a healthcare worker with pulmonary TB disease who is not highly contagious, neonates can be safely managed without specific evaluation for TB disease or empirical treatment.

The Dutch Working Party on Infection Prevention (Werkgroep Infectiepreventie [WIP]) aimed to determine whether certain policies on endotracheal suctioning are better than others in terms of prevention of ventilator-associated pneumonia (VAP) in patients receiving mechanical ventilation in the intensive care unit.

Publications were retrieved by a systematic search of Medline and the Cochrane Library for literature published before February 2006. Additionally, the reference lists of all identified trials were examined. All randomized trials, quasi-randomized trials, and systematic reviews or meta-analyses of randomized or quasi-randomized trials that compared different policies on endotracheal suctioning for patients receiving mechanical ventilation in the intensive care unit were selected. Two reviewers independently assessed trial quality and extracted data. Disagreements were resolved by discussion with a third reviewer. Data from the original publications were used to calculate the relative risk of VAP. Data for VAP were combined in the analysis where appropriate, by use of a random-effects model.

Ten trials were included in the review. The quality of the trials and the way they were reported were generally unsatisfactory. Eight low-quality trials indicate that use of closed instead of open suction systems has no effect on the incidence of VAP. Two moderate-quality trials indicate that changing in-line suction catheters less frequently than every 24 hours does not increase the incidence of VAP.

The WIP recommends that there be no preferential use of either open or closed endotracheal suction systems to reduce the rate of VAP, but it elucidates that the quality of the evidence is low. Considerations other than prevention of VAP should determine the choice of the suction system. When closed systems are used, the WIP recommends changing the in-line suction catheters every 48 hours. In case of mechanical failure or soiling of the suction system, they may be changed more frequently.

To evaluate safety-engineered devices (SEDs) with respect to their effectiveness in preventing needlestick injuries (NSIs) in healthcare settings and their importance among other preventive measures.

Multicenter prospective survey with a 1-year follow-up period during which all incident NSIs and their circumstances were reported. Data were prospectively collected during a 12-month period from April 1999 through March 2000. The procedures for which the risk of NSI was high were also reported 1 week per quarter to estimate procedure-specific NSI rates. Device types were documented. Because SEDs were not in use when a similar survey was conducted in 1990, their impact was also evaluated by comparing findings from the recent and previous surveys.

A total of 102 medical units from 32 hospitals in France.

A total of 1,506 nurses in medical or intensive care units.

A total of 110 NSIs occurring during at-risk procedures performed by nurses were documented. According to data from the 2000 survey, use of SEDs during phlebotomy procedures was associated with a 74% lower risk (P < .01 ). The mean NSI rate for all relevant nursing procedures was estimated to be 4.72 cases per 100,000 procedures, for a 75% decrease since 1990 (P < .01); however, the decrease in NSI rates varied considerably according to procedure type. Between 1990 and 2000, decreases in the NSI rates for each procedure were strongly correlated with increases in the frequency of SED use (r = 0.88; P < .02).

In this French hospital network, the use of SEDs was associated with a significantly lower NSI rate and was probably the most important preventive factor.

To compare the effects of hospital cleaning agents and germicides on the survival of epidemic Clostridium difficile strains.

We compared the activity of and effects of exposure to 5 cleaning agents and/or germicides (3 containing chlorine, 1 containing only detergent, and 1 containing hydrogen peroxide) on vegetative and spore forms of epidemic and non-epidemic C. difficile strains (3 of each). We carried out in vitro exposure experiments using a human fecal emulsion to mimic conditions found in situ.

Cleaning agent and germicide exposure experiments yielded very different results for C. difficile vegetative cells, compared with those for spores. Working-strength concentrations of all of the agents inhibited the growth of C. difficile in culture. However, when used at recommended working concentrations, only chlorine-based germicides were able to inactivate C. difficile spores. C. difficile epidemic strains had a greater sporulation rate than nonepidemic strains. The mean sporulation rate, expressed as the proportion of a cell population that is in spore form, was 13% for all strains not exposed to any cleaning agent or germicide, and it was significantly increased by exposure to cleaning agents or germicides containing detergent alone (34%), a combination of detergent and hypochlorite (24%), or hydrogen peroxide (33%). By contrast, the mean sporulation rate did not change substantially after exposure to germicides containing either a combination of detergent and dichloroisocyanurate (9%) or dichloroisocyanurate alone (15%).

These results highlight differences in the activity of cleaning agents and germicides against C. difficile spores and the potential for some of these products to promote sporulation.

To compare the type of pathogens isolated from patients with early-onset intensive care unit (ICU)-acquired pneumonia with those isolated from patients with late-onset ICU-acquired pneumonia and to study risk factors for the isolation of pathogens that are potentially resistant to multiple drugs.

Prospective cohort study.

Patients admitted to the ICU of a 677-bed, university-affiliated teaching hospital in Belgium during 1997-2002.

ICU-acquired pneumonia was defined as a case of pneumonia that occurred 2 days or more after admission to the ICU in combination with a positive results of radiologic analysis, clinical signs and symptoms, and a positive culture result. All cases of pneumonia were categorized as either early onset (within 7 days after admission) and late onset (7 days or more after admission), with or without previous antibiotic treatment, and the corresponding pathogens were analyzed. Risk factors for the isolation of pathogens potentially resistant to multiple drugs (ie, Pseudomonas aeruginosa, Serratia marcescens, Enterobacter species, Morganella morganii, methicillin-resistant Stapylococcus aureus, Citrobacter species, Acinetobacter species, Burkholderia species, extended-spectrum β-lactamase–producing pathogens, and Stenotrophomonas maltophilia) were analyzed using logistic regression analysis.

A total of 4,200 patients stayed at the ICU for 2 or more days, 298 of whom developed ICU-acquired pneumonia, for an overall incidence of 13 cases (95% confidence interval [CI], 11-14 cases) per 1,000 ICU-days. Pathogens potentially resistant to multiple drugs were isolated from 52% of patients with early-onset pneumonia. Risk factors for the isolation of these pathogens were greater age and previous receipt of antibiotic prophylaxis (adjusted odds ratio [aOR], 4.6 [95% CI, 1.6-13.0]) or antibiotic therapy (aOR, 8.2 [95% CI, 2.8-23.8]). The length of ICU admission and hospital stay were weaker risk factors for the isolation of these pathogens.

Pathogens potentially resistant to multiple drugs were isolated in 52% of cases of early-onset ICU-acquired pneumonia. Previous antibiotic use (both prophylactic and therapeutic) is the main risk factor for the isolation of these pathogens.

The occurrence of nosocomial infections due to third-generation cephalosporin–resistant gram-negative bacteria is increasing. Gastrointestinal colonization is an important reservoir for antibiotic-resistant bacteria, and it often precedes clinical infection.

To estimate the prevalence of gastrointestinal colonization with ceftazidime-resistant gram-negative bacteria among intensive care unit (ICU) patients at a university-affiliated tertiary-care hospital during 2 distinct periods and to assess whether, at any time during the index hospitalization, colonized patients had a clinical culture positive for the same organism that was recovered from surveillance culture.

Two ICUs at the University of Maryland Medical Center, a 656-bed tertiary-care hospital located in Baltimore, Maryland. Both ICUs provide care to adult patients.

We performed a cross-sectional study of adult patients admitted to the medical ICU or the surgical ICU from June 14 to July 14, 2003, and from June 14 to July 14, 2006. Perirectal swab samples were obtained for surveillance culture on admission to the intensive care unit, weekly thereafter, and at discharge. Each culture sample was plated onto MacConkey agar supplemented with ceftazidime.

In 2003, a total of 33 (18.8%) of 176 patients were colonized with ceftazidime-resistant gram-negative bacilli; in 2006, 60 (31.4%) of 191 patients were (P<.01). This increase was largely driven by an increase in ceftazidime-resistant Klebsiella isolates (which accounted for 6.4% of isolates in 2003 and for 22.8% in 2006; P<.01). In 2003, a total of 16 (48.5%) of 33 colonized patients had a clinical culture positive for the same organism that was recovered from the perirectal surveillance culture, compared with 22 (36.6%) of 60 colonized patients in 2006 (P = .28).

Our data suggest that gastrointestinal colonization with ceftazidime-resistant gram-negative bacilli is common, that its prevalence is increasing, and that colonization may result in clinical cultures positive for these bacilli.

Pertussis outbreaks in healthcare settings result in resource-intensive control activities, but studies have rarely evaluated the associated costs. We describe and estimate costs associated with 2 nosocomial pertussis outbreaks in King County, Washington, during the period from July 25 to September 15, 2004. One outbreak occurred at a 500-bed tertiary care hospital (hospital A), and the other occurred at a 250-bed pediatric hospital (hospital B).

We estimated the costs of each outbreak from the hospitals' perspective through standardized interviews with hospital staff and review of contact tracing logs. Direct costs included personnel time and laboratory and medication costs, whereas indirect costs were those resulting from hospital staff furloughs.

Hospital A incurred direct costs of $195,342 and indirect costs of $68,015; hospital B incurred direct costs of $71,130 and indirect costs of $50,000. Cost differences resulted primarily from higher personnel costs at hospital A ($134,536), compared with hospital B ($21,645). Total cost per pertussis case was $43,893 for hospital A (6 cases) and $30,282 for hospital B (4 cases). Total cost per person exposed to a pertussis patient were $357 for hospital A (738 exposures) and $164 for hospital B (737 exposures).

Nosocomial pertussis outbreaks result in substantial costs to hospitals, even when the number of pertussis cases is low. The cost-effectiveness of strategies to prevent nosocomial pertussis outbreaks, including vaccination of healthcare workers, should be evaluated.

To determine whether the use of chlorhexidine bathing and intranasal mupirocin therapy among patients colonized with methicillin-resistant Staphylococcus aureus (MRSA) would decrease the incidence of MRSA colonization and infection among intensive care unit (ICU) patients.

After a 9-month baseline period (January 13, 2003, through October 12, 2003) during which all incident cases of MRSA colonization or infection were identified through the use of active-surveillance cultures in a combined medical-coronary ICU, all patients colonized with MRSA were treated with intranasal mupirocin and underwent daily chlorhexidine bathing.

After the intervention, incident cases of MRSA colonization or infection decreased 52% (incidence density, 8.45 vs 4.05 cases per 1,000 patient-days; P = .048). All MRSA isolates remained susceptible to chlorhexidine; the overall rate of mupirocin resistance was low (4.4%) among isolates identified by surveillance cultures and did not increase during the intervention period.

We conclude that the selective use of intranasal mupirocin and daily chlorhexidine bathing for patients colonized with MRSA reduced the incidence of MRSA colonization and infection and contributed to reductions identified by active-surveillance cultures. This finding suggests that additional strategies to reduce the incidence of MRSA infection and colonization—beyond expanded surveillance—may be needed.

To use multivariate analysis to determine risk factors for death among pediatric patients with candidemia and a central venous catheter in place.

Retrospective cohort study conducted at Santa Casa Complexo Hospitalar, a 1,200-bed teaching hospital in southern Brazil.

All cases of candidemia in pediatric patients (age, ≤ 13 years) at our medical center over a 9-year period were reviewed. A diagnosis of sepsis was required for inclusion in the study. Severity of illness was confirmed by the presence of hypotension requiring inotropes and according to the following scores: the Pediatric Risk of Mortality (PRISM) II score, the PRISM III score, and the Pediatric Logistic Organic Dysfunction score. The following 2 outcomes were evaluated: early death, defined as death occurring within 7 days after candidemia was diagnosed, and late death, defined as death 8-30 days after candidemia was diagnosed.

A total of 61 patients were included in the study, including 14 neonates. Most (63.9%) of these patients were girls, and the median age was 0.3 years. A total of 80.3% of candidemia cases were due to species other than Candida albicans, primarily Candida parapsilosis (32.8% of cases) and Candida tropicalis (24.6% of cases). Using multivariate analysis, we demonstrated that failure to remove the central venous catheter was an independent risk factor for early death among pediatric patients with candidemia. However, patients whose catheters were retained were sicker than patients whose catheters were removed, and catheter removal had no impact on late death. Instead, severity of illness determined using the PRISM III score was also an independent predictor of late death.

Results from this study suggest that systematic removal of catheters from pediatric patients with candidemia does not reduce the occurrence of late death.

To determine the prevalence and risk factors for bloodborne exposure and infection in correctional healthcare workers (CHCWs).

Cross-sectional risk assessment study with a confidential questionnaire and serological testing performed during 1999-2000.

Correctional systems in 3 states.

Among 310 participating CHCWs, the rate of percutaneous injury (PI) was 32 Pis per 100 person-years overall and 42 Pis per 100 person-years for CHCWs with clinical job duties. Underreporting was common, with only 25 (49%) of 51 Pis formally reported to the administration. Independent risk factors for experiencing PI included being age 45 or older (adjusted odds ratio [aOR], 2.41 (95% confidence interval (CI), 1.31-4.46]) and having job duties that involved needle contact (aOR, 3.70 [95% CI, 1.28-10.63]) or blood contact (aOR, 5.05 [95% CI, 1.45-17.54]). Overall, 222 CHCWs (72%) reported having received a primary hepatitis В vaccination series; of these, 150 (68%) tested positive for anti-hepatitis B surface antigen, with negative results significantly associated with receipt of last dose more than 5 years previously. Serologic markers of hepatitis В virus infection were identified in 31 individuals (10%), and the prevalence of hepatitis B virus infection was 2% (n = 7). The high hepatitis B vaccination rate limited the ability to identify risk factors for infection, but hepatitis C virus infection correlated with community risk factors only.

Although the wide coverage with hepatitis B vaccination and the decreasing rate of hepatitis C virus infection in the general population are encouraging, the high rate of exposure in CHCWs and the lack of exposure documentation are concerns. Continued efforts to develop interventions to reduce exposures and encourage reporting should be implemented and evaluated in correctional healthcare settings. These interventions should address infection control barriers unique to the correctional setting.

To characterize the epidemiology of severe (ie, nonsuperficial) surgical site infection (SSI) in community hospitals.

SSI data were collected prospectively at 26 community hospitals in the southeastern United States. Two analyses were performed: (1) a study of the overall prevalence rates of SSI and the prevalence rates of SSI due to specific pathogens in 2005 at all participating hospitals and (2) a prospective study of consecutive surgical procedures at 9 of the 26 community hospitals from 2000 through 2005.

In 2005, a total of 1,010 SSIs occurred after 89,302 procedures (prevalence rate, 1.13 infections per 100 procedures). Methicillin-resistant S. aureus (MRSA) was the pathogen most commonly recovered (from 175 SSIs). Trend data from 2000 through 2005 demonstrated that the prevalence rate of MRSA SSI almost doubled during this period, increasing from 0.12 infections per 100 procedures (95% confidence interval [CI], 0.12-0.13) to 0.23 infections per 100 procedures (95% CI, 0.22-0.24) (P<.0001). In adjusted analysis, MRSA SSI was significantly more prevalent at the end of the study period than at the beginning (prevalence rate ratio, 1.48 [95% CI, 1.36-1.61]; P<.0001).

MRSA was the pathogen that most commonly caused SSI in our network of community hospitals during 2005. The prevalence of MRSA SSI has increased significantly over the past 6 years.

The epidemiology of Clostridium difficile-associated disease (CDAD) is changing, with evidence of increased incidence and severity. However, the understanding of the magnitude of and reasons for this change is currently hampered by the lack of standardized surveillance methods.

An ad hoc C. difficile surveillance working group was formed to develop interim surveillance definitions and recommendations based on existing literature and expert opinion that can help to improve CDAD surveillance and prevention efforts.

A CDAD case patient was defined as a patient with symptoms of diarrhea or toxic megacolon combined with a positive result of a laboratory assay and/or endoscopic or histopathologic evidence of pseudomembranous colitis. Recurrent CDAD was defined as repeated episodes within 8 weeks of each other. Severe CDAD was defined by CDAD-associated admission to an intensive care unit, colectomy, or death within 30 days after onset. Case patients were categorized by the setting in which C. difficile was likely acquired, to account for recent evidence that suggests that healthcare facility-associated CDAD may have its onset in the community up to 4 weeks after discharge. Tracking of healthcare facility–onset, healthcare facility–associated CDAD is the minimum surveillance required for healthcare settings; tracking of community–onset, healthcare facility–associated CDAD should be performed only in conjunction with tracking of healthcare facility–onset, healthcare facility–associated CDAD. Community–associated CDAD was defined by symptom onset more than 12 weeks after the last discharge from a healthcare facility. Rates of both healthcare facility–onset, healthcare facility–associated CDAD and community–onset, healthcare facility–associated CDAD should be expressed as case patients per 10,000 patient–days; rates of community-associated CDAD should be expressed as case patients per 100,000 person-years.

To estimate the independent effect of a single lower respiratory tract infection, urinary tract infection, or other healthcare-acquired infection on length-of-stay and variable costs and to demonstrate the bias from omitted variables that is present in previous estimates.

Prospective cohort study.

A tertiary care referral hospital and regional district hospital in southeast Queensland, Australia.

Adults aged 18 years or older with a minimum inpatient stay of 1 night who were admitted to selected clinical specialities.

Urinary tract infection was not associated with an increase in length of hospital stay or variable costs. Lower respiratory tract infection was associated with an increase of 2.58 days in the hospital and variable costs of AU$24, whereas other types of infection were associated with an increased length of stay of 2.61 days but not with variable costs. Many other factors were found to be associated with increased length of stay and variable costs alongside healthcare-acquired infection. The exclusion of these variables caused a positive bias in the estimates of the costs of healthcare-acquired infection.

The existing literature may overstate the costs of healthcare-acquired infection because of bias, and the existing estimates of excess costs may not make intuitive sense to clinicians and policy makers. Accurate estimates of the costs of healthcare-acquired infection should be made and used in appropriately designed decision-analytic economic models (ie, cost-effectiveness models) that will make valid and believable predictions of the economic value of increased infection control.

To quantify the value of performing active surveillance cultures for detection of methicillin-resistant Staphylococcus aureus (MRSA) on intensive care unit (ICU) discharge.

Prospective cohort study.

Medical ICU (MICU) and surgical ICU (SICU) of a tertiary care hospital.

We analyzed data on adult patients who were admitted to the MICU or SICU between January 17, 2001, and December 31, 2004. All participants had a length of ICU stay of at least 48 hours and had surveillance cultures of anterior nares specimens performed on ICU admission and discharge. Patients who had MRSA-positive clinical cultures in the ICU were excluded.

Of 2,918 eligible patients, 178 (6%) were colonized with MRSA on ICU admission, and 65 (2%) acquired MRSA in the ICU and were identified by results of discharge surveillance cultures. Patients with MRSA colonization confirmed by results of discharge cultures spent 853 days in non-ICU wards after ICU discharge, which represented 27% of the total number of MRSA colonization-days during hospitalization in non-ICU wards for patients discharged from the ICU.

Surveillance cultures of nares specimens collected at ICU discharge identified a large percentage of MRSA-colonized patients who would not have been identified on the basis of results of clinical cultures or admission surveillance cultures alone. Furthermore, these patients were responsible for a large percentage of the total number of MRSA colonization-days during hospitalization in non-ICU wards for patients discharged from the ICU.