To investigate an outbreak of multidrug-resistant Pseudomonas aeruginosa in an intensive care unit (ICU).

Epidemiologic investigation, environmental assessment, and ambidirectional cohort study.

A secondary-care university hospital with a 10-bed ICU.

All patients admitted to the ICU receiving ventilator treatment from December 1,1999, to September 1, 2000.

An outbreak in an ICU with multidrug-resistant isolates of P. aeruginosa belonging to one amplified fragment-length polymorphism (AFLP)–defined genetic cluster was identified, characterized, and cleared. Molecular typing of bacterial isolates with AFLP made it possible to identify the outbreak and make rational decisions during the outbreak period. The outbreak included 19 patients during the study period. Infection with bacterial isolates belonging to the AFLP cluster was associated with reduced survival (odds ratio, 5.26; 95% confidence interval, 1.14 to 24.26). Enhanced barrier and hygiene precautions, cohorting of patients, and altered antibiotic policy were not sufficient to eliminate the outbreak. At the end of the study period (in July), there was a change in the outbreak pattern from long (December to June) to short Quly) incubation times before colonization and from primarily tracheal colonization (December to June) to primarily gastric or enteral Quly) colonization. In this period, the bacterium was also isolated from water taps.

Complete elimination of the outbreak was achieved after weekly pasteurization of the water taps of the ICU and use of sterile water as a solvent in the gastric tubes.

To determine the duration of colonization with vancomycin-resistant Enterococcus (VRE) and the adequacy of 3 consecutive negative cultures to determine clearance.

Retrospective cohort study.

A university hospital.

Patients identified by perirectal cultures as VRE carriers who had follow-up cultures.

Follow-up perirectal cultures were collected in inpatient and outpatient settings, at least 1 week apart, when patients were not receiving antibiotics with activity against VRE. The likelihood of culture positivity was analyzed given prior culture results and time from the initial positive culture.

A total of 116 patients colonized with VRE had 423 follow-up cultures, a mean of 204 days (range, 4 to 709 days) after their initial isolate. The first follow-up culture, collected a mean of 125 days after the initial positive isolate, was negative in 64%. After 1 negative follow-up culture, the next one was negative in 92% of the patients. After 2 negative cultures, 95% remained culture-negative. After 3 sequential negative cultures, 35 (95%) of 37 patients remained culture-negative. As the interval between the initial and the follow-up isolates increased, the probability that a subsequent culture would be positive decreased (P < .001, chi square for trend). Prolonged hospitalization, intensive care, and antibiotic use each decreased the likelihood of clearing VRE.

These data support the Centers for Disease Control and Prevention criterion of 3 sequential negative cultures, at least 1 week apart, to remove patients from VRE isolation. Nevertheless, this may reflect a decrease in the quantity of VRE to an undetectable level and these patients should be observed for relapse, especially when re-treated with antibiotics.

We investigated cases of acute hepatitis B in The Netherlands that were linked to the same general surgeon who was infected with hepatitis B virus (HBV).

A retrospective cohort study was conducted of 1,564 patients operated on by the surgeon. Patients were tested for serologic HBV markers. A case–control study was performed to identify risk factors.

The surgeon tested positive for hepatitis B surface antigen (HBsAg) and hepatitis B e antigen (HBeAg) with a high viral load. He was a known nonresponder after HBV vaccination and had apparently been infected for more than 10 years. Forty-nine patients (3.1%) were positive for HBV markers. Transmission of HBV from the surgeon was confirmed in 8 patients, probable in 2, and possible in 18. In the remaining 21 patients, the surgeon was not implicated. Two patients had a chronic HBV infection. One case of secondary transmission from a patient to his wife was identified. HBV DNA sequences from the surgeon were completely identical to sequences from 7 of the 28 patients and from the case of secondary transmission. The duration of the operation and the occurrence of complications during or after surgery were identified as independent risk factors. Although the risk of HBV infection during high-risk procedures was 7 times higher than that during low-risk procedures, at least 8 (28.6%) of the 28 patients were infected during low-risk procedures.

Transmission of HBV from surgeons to patients at a low rate can remain unnoticed for a long period of time. Prevention requires a more stringent strategy for vaccination and testing of surgeons and optimization of infectious disease surveillance. Policies allowing HBV-infected surgeons to perform presumably low-risk procedures should be reconsidered.

To evaluate the usefulness of an assay using two polymerase chain reaction-based genotyping methods in the practical surveillance of methicillin-resistant Staphylococcus aureus (MRSA).

Nosocomial infection and colonization were surveyed monthly in a university hospital in Japan for 20 months. Genotyping with mec-HVR is based on the size of the mec-associated hypervariable region amplified by polymerase chain reaction. Toxin genotyping uses a multiplex polymerase chain reaction method to amplify eight staphylococcal toxin genes.

Eight hundred nine MRSA isolates were classified into 49 genotypes. We observed differing prevalences of genotypes for different hospital wards, and could rapidly demonstrate the similarity of genotype for outbreak isolates. The incidence of genotype D: SEC/TSST1 was significantly higher in isolates causing nosocomial infections (49.5%; 48 of 97) than in nasal isolates (31.4%; 54 of 172) (P = .004), suggesting that this genotype may represent the nosocomial strains.

The combined use of these two genotyping methods resulted in improved discriminatory ability and should be further investigated.

Catheter-related phlebitis is a frequent problem in the clinical setting. Risk factors for catheter-related phlebitis were assessed at a single tertiary-care institution where no routine change policy for peripheral intravenous catheters is in place.

In a nonrandomized, observational trial, peripheral intravenous Teflon catheters were inserted in patients with a diagnosis of leukemia, lymphoma, solid tumor, acquired immunodeficiency syndrome, other serious infection, or autoimmune disorder. Underlying disease, age, white blood cell count at the time of insertion, physician placing the catheter, catheter bore, duration of cannulation, reason for removal of the catheter, and visual inspection of the insertion site were recorded.

Four hundred twelve catheters were inserted in 175 patients. The number of catheterizations per episode varied between 1 and 7. Three hundred sixty-four (88.3%) catheter placements were evaluable. The mean duration of cannulation was 4.2 days. The overall incidence of phlebitis was 12.9%. Catheters in leukopenic patients showed a longer duration of cannulation compared with catheters in nonleukopenic patients, but no difference regarding the phlebitis rate.

Findings in this study partly contrast with data reported in the literature. In particular, leukopenia, female gender, prolonged duration of cannulation, antibiotics, and choice of insertion site could not be shown to be risk factors.

To identify antibiotic resistance trends and risk factors for resistance of Serratia species to third-generation cephalosporins.

Retrospective survey of medical records.

A 2,200-bed, tertiary-care hospital.

One hundred twenty-two patients with Serratia bacteremia between January 1991 and June 2001.

Infectious disease physicians collected data from medical records regarding patient demographics, underlying disease or condition, portal of entry, microorganism, antibiogram, complications, antibiotics received, and outcome.

Among 122 Serratia isolates, 117 (95.9%) were Serratia marcescens and 110 (90.2%) were of nosocomial origin. During the study period, the 122 isolates showed a high rate of resistance to third-generation cephalosporins (45.9%) and extended-spectrum penicillins (56.6%). The resistance rate to ciprofloxacin was 32.0%. The resistance rate to third-generation cephalosporins increased from 31.7% for 1991 to 1995 to 54.9% for 1996 to 1998 and 50.0% for 1999 to 2001. In the multivariate analysis, prior use of a second-generation cephalosporin (adjusted odds ratio [OR], 5.90; 95% confidence interval [CI95], 1.41 to 24.6; P = .015) or a third-generation cephalosporin (OR, 3.26; CI95, 1.20 to 8.87; P = .020) was a strong independent risk factor for resistance to third-generation cephalosporins. The overall case-fatality rate was 25.4% (Serratia bacteremia-related case-fatality rate, 13.1%).

Prior use of a second- or third-generation cephalosporin was the most important risk factor for bacteremia with Serratia resistant to third-generation cephalosporins, suggesting the need for antibiotic control. The potential role of patient-to-patient spread could not be fully evaluated in this retrospective study.

To study the humoral immune responses, safety, and tolerability of intradermal recombinant hepatitis B vaccination in healthcare workers (HCWs) nonresponsive to previous repeated intramuscular vaccination.

An open, prospective, before–after trial.

A tertiary referral hospital and surrounding district health service in Queensland, Australia.

Hospital and community HCWs nonresponsive to previous intramuscular hepatitis B vaccination.

Intradermal recombinant hepatitis B vaccine was administered every second week for a maximum of 4 doses. Hepatitis B surface antibody (anti-HBs) responses were assessed 2 weeks after each dose.

Protective anti-HBs levels developed in 17 (94%) of 18 study subjects. Three doses resulted in seroconversion of all responding subjects and the highest geometric mean antibody concentration. The vaccine was well tolerated.

More than 90% of previously nonresponsive HCWs responded to intradermal recombinant hepatitis B vaccine with protective anti-HBs levels.

To determine the prevalence of and the risk factors for methicillin-resistant Staphylococcus aureus (MRSA) carriage in nursing home residents in the Rhine-Neckar region of southern Germany.

Point-prevalence survey.

Forty-seven nursing homes in the region.

All residents of the approached nursing homes who agreed to participate.

After informed consent was obtained, all participants had their nares swabbed, some personal data collected, or both. All swabs were examined for growth of MRSA All S. aureus isolates underwent oxacillin susceptibility testing and polymerase chain reaction for demonstration of the meek gene. All MRSA isolates were typed using pulsed-field gel electrophoresis after digestion with SmaI.

Swabs from 3,236 nursing home residents yielded 36 MRSA strains, contributing to a prevalence rate of 1.1%. Significant risk factors for MRSA carriage in the multivariate analysis were the presence of wounds or urinary catheters, limited mobility, admission to a hospital during the preceding 3 months, or stay in a medium-size nursing home. One predominant MRSA strain could be detected in 30 of the 36 MRSA carriers.

The prevalence of MRSA in German nursing homes is still low. These residents seemed to acquire their MRSA in the hospital and transfer it to their nursing home. Apart from well-known risk factors for the acquisition of MRSA we identified the size of the nursing home as an independent risk factor. This might be due to an increased use of antimicrobials in nursing homes of a certain size.

The incidence of extended-spectrum β-lactamase (ESβL)–mediated resistance has increased markedly during the past decade. Risk factors for colonization with ESβL-producing Escherichia coli and Klebsiella species (ESβL-EK) remain unclear, as do methods to control their further emergence.

Case–control study.

Two hospitals within a large academic health system: a 725-bed academic tertiary-care medical center and a 344-bed urban community hospital.

Thirteen patients with ESβL-EK fecal colonization were compared with 46 randomly selected noncolonized controls.

Duration of hospitalization was the only independent risk factor for ESβL-EK colonization (odds ratio, 1.11; 95% confidence interval, 1.02 to 1.21). Of note, 8 (62%) of the patients had been admitted from another healthcare facility. In addition, there was evidence for dissemination of a single K. oxytoca clone. Finally, the prevalence of ESβL-EK colonization decreased from 7.9% to 5.7% following restriction of third-generation cephalosporins (P = .51).

ESβL-EK colonization was associated only with duration of hospitalization and there was no significant reduction following antimicrobial formulary interventions. The evidence for nosocomial spread and the high percentage of patients with ESβL-EK admitted from other sites suggest that greater emphasis must be placed on controlling the spread of such organisms within and between institutions.

To determine the occupational risk of Mycobacterium tuberculosis infection among healthcare workers (HCWs) and to examine the utility of tuberculin skin testing in a developing country with a high prevalence of bacille Calmette-Guerin vaccination.

Tuberculin skin test (TST) survey.

A tertiary-care referral center and a teaching hospital in Kuala Lumpur, Malaysia.

HCWs from medical, surgical, and orthopedic wards.

Tuberculin purified protein derivative RT-23 (State Serum Institute, Copenhagen, Denmark) was used for the TST (Mantoux method).

One hundred thirty-seven (52.1%) and 69 (26.2%) of the HCWs tested had indurations of 10 mm or greater and 15 mm or greater, respectively. Medical ward HCWs were at significantly higher risk of a positive TST reaction than were surgical or orthopedic ward HCWs (odds ratio, 2.18; 95% confidence interval, 1.33 to 3.57; P = .002 for TST positivity at 10 mm or greater) (odds ratio, 2.61; 95% confidence interval, 1.44 to 4.70; P = .002 for TST positivity at 15 mm or greater). A previous TST was a significant risk factor for a positive TST reaction at either 10 mm or greater or 15 mm or greater, but a duration of employment of more than 1 year and being a nurse were only significantly associated with a positive TST reaction at a cut-off point of 15 mm or greater.

HCWs at the University of Malaya Medical Centre had an increased risk for M. tuberculosis infection that was significantly associated with the level of occupational tuberculosis exposure. A TST cut-off point of 15 mm or greater may correlate better with M. tuberculosis infection than a cut-off point of 10 mm or greater in settings with a high prevalence of bacille Calmette-Guerin vaccination (Infect Control Hosp Epidemiol 2002;23:584-590).

To determine the prevalence of stool colonization with vancomycin-resistant enterococci (VRE) among healthcare workers (HCWs) and their families.

Prospective assessment of fecal colonization with VRE.

A 603-bed, tertiary-care teaching hospital.

Healthy volunteers recruited from hospital employees and their households were screened to exclude pregnancy, diabetes mellitus, immunosuppressive disorders, and recent use of antimicrobials.

Self-obtained stool swabs were used to obtain cultures. Isolated enterococci were screened for vancomycin resistance and species were identified. Intra-household isolates were genotyped using pulsed-field gel electrophoresis (PFGE).

The participants (n = 228; age range, 28 days to 80 years) were from 137 households with and 91 without employees who had contact with patients. Enterococcus species were isolated from 127 stool specimens (55.7%). VRE were detected in 12 individuals, representing 6 E. casseliflavus, 5 E. faecium, and 1 E. gallinarum. VRE were more commonly isolated in employees who had contact with patients (5 of 52 vs 0 of 40; relative risk [RR], 1.9; 95% confidence interval [CI95], 1.5 to 2.2; P = .07) and their household members (10 of 137 vs 2 of 91; RR, 3.3; CI95, 0.7 to 14.8; P = .13). In 2 households (2 adults in a physician's household and an adult plus a child in a nurse's household) PFGE analysis demonstrated identical intra-household strains of vancomycin-resistant E. faecium.

VRE colonization was found in 5.3% of screened stools and was more prevalent in HCWs who had contact with patients and their households. Identical PFGE patterns between 2 employees who had contact with patients and their household members demonstrated probable intra-household spread. Although the mode of acquisition was uncertain, the association with employees who had contact with patients suggests possible occupational sources. These findings demonstrate the spread of VRE within the household and implicate occupational risk for its acquisition.

The Centers for Disease Control and Prevention and the American Thoracic Society recommend obtaining cultures of at least three sputum specimens for acid-fast bacilli (AFB) from patients in whom tuberculosis (TB) is suspected. On the basis of this, most hospitals isolate patients with suspected TB for 3 days or more until three smear (not culture) results are negative. Our objective was to evaluate the predictive value and sensitivity of these smears.

Observational study.

Four urban medical centers.

The posttest probability of TB given sequential negative AFB smears from 274 patients isolated for suspected TB and the sensitivity of sequential AFB smears from 209 patients with positive results on culture for pulmonary TB were measured.

The posttest probabilities of having TB given one, two, and three negative AFB smears were low: 1.1% (3 of 265; 95% confidence interval [CI95], 0.23% to 3.27%), 0.4% (1 of 262; CI95, 0% to 2.1%), and 0% (0 of 260; CI95, 0% to 1.4%), respectively. Among the 209 patients with positive results on culture for pulmonary TB, 169 (81%) had an expectorated sputum specimen sent, of which 91 (54%) were positive for AFB. Forty (24%) of the 169 patients had a second expectorated sputum specimen sent after the results of the first specimen were negative; only 6 (15%) of these had positive AFB smears. None of the 10 patients in whom the first two expectorated sputum samples yielded an AFB smear without an organism had a third AFB smear that was positive.

Unless there is high clinical suspicion of pulmonary TB in a specific patient, the use of three AFB smears on expectorated sputa is a rational approach to discontinuing isolation for patients with suspected TB.

To identify the timing, pattern, and determinants of colonization of neonates by Malassezia.

Prospective observational study.

A neonatal medical and surgical unit consisting of 10 special care, 10 high-dependency, 10 intensive care, and 10 surgical cots.

All neonates (≤ 28 days of age) or infants (> 28 days of age) admitted to the unit during the 20-week period from October 1995 to March 1996.

All infants or neonates were swabbed on the day of admission and every third day thereafter and risk factors were collected for every day on the unit.

During the study period, 245 neonates and 42 infants were sampled for their entire duration of stay on the unit. Of these, 41 infants (97.6%) were colonized with Malassezia on admission to the unit and thereafter, as assessed by subsequent samples. Within the neonate population, 78 (31.8%) became colonized, but none were colonized immediately after birth. Univariate analysis showed that many factors appeared to be significantly associated with colonization in the neonates, including use of ventilation, presence of central venous catheters, use of parenteral nutrition, and use of antibacterial or antifungal drugs. However, when the data were analyzed by multivariate logistic regression to control for confounding variables, only gestational age and length of stay on the unit were found to be significantly associated with colonization.

Colonization of infants is not as unusual as previously thought and many infants have established a cutaneous Malassezia commensal flora by the age of 3 to 6 months. Factors that predispose to colonization in neonates may not be the same as those that predispose to infection.

To assess the influence of postdischarge infection surveillance on risk-adjusted surgical-site infection rates for coronary artery bypass graft (CABG) procedures.

Prospective surveillance of surgical-site infections after CABG.

Tertiary-care referral hospital.

Data on surgical-site infections were collected for 1,324 CABG procedures during 27 months. They were risk adjusted and analyzed according to the surgical surveillance protocol of the National Nosocomial Infections Surveillance (NNIS) System of the Centers for Disease Control and Prevention, with and without postdischarge data.

Data were available for 96% of the patients. Of the 88 surgical-site infections, 28% were identified prior to discharge and 72% postdischarge. More chest than harvest-site infections were identified (46% vs 11%) prior to discharge, and more harvest-site than chest infections were identified in the outpatient setting (42% vs 14%). The surgical-site infection rate for patients stratified under risk index 1, calculated without postdischarge surveillance, was 2.9%; when compared with that of the NNIS System, the P value was .29. When postdischarge surveillance was included, the surgical-site infection rate was 4.9% and statistically significant when compared with that of the NNIS System (P = .007). For patients stratified under risk index 2, the rates with and without postdischarge surveillance were 11.7% and 10.0%, respectively; when compared with the NNIS System rates, the P values were .000008 and .0006, respectively.

Only 28% of the surgical-site infections would have been detected if surveillance had been limited to hospital stay. Postdischarge surveillance identified more surgical-site infections among risk index 1 patients. Hospitals with comprehensive postdischarge surveillance after CABG procedures are likely to record higher surgical-site infection rates than those that do not perform such surveillance.

To evaluate clonal dissemination of methicillin-resistant coagulase-negative staphylococci (CNS).

Neonatal intensive care unit of a 180-bed, university-affiliated general hospital.

Neonates admitted to the neonatal intensive care unit between March 1999 and October 2000, from whom CNS were isolated as a unique pathogen. Patients from other wards from whom epidemiologically unrelated staphylococci strains were obtained served as control-patients.

Conventional methods were used for phenotypic characterization of CNS. Methicillin resistance was determined by mecA polymerase chain reaction (PCR) amplification. Genotypic characterization was done by random amplification of DNA with degenerated primers (RAPD) and repetitive element sequence-based PCR (rep-PCR).

Forty methicillin-resistant CNS isolates obtained from neonates were characterized as Staphylococcus epidermidis (33), S. hominis (5), S. warneri (1), and S. auricularis (1). Both RAPD and rep-PCR indicated the presence of 4 different clones among the 33 S. epidermidis isolates. In turn, the 4 randomly selected, epidemiologically unrelated methicillin-resistant CNS strains obtained from control-patients showed 3 new profiles by RAPD and 2 by rep-PCR, which differed from the corresponding patterns mentioned earlier. Persistence of S. hominis in a neonate could be assessed by both genotypic techniques.

The molecular characterization of the methicillin-resistant CNS studied indicated dissemination of one particular methicillin-resistant CNS clone among the neonates in the ward studied. Although RAPD showed a superior power to discriminate among methicillin-resistant CNS isolates, both RAPD and rep-PCR detected intraspecific and interspecific genomic diversity.

We report surveillance of nosocomial diarrhea in children at our institution during the past decade and note different epidemiology of diarrhea due to viruses and Clostridium difficile.

A prospective cohort study.

A university-affiliated pediatric hospital with 180 beds serving an urban area and providing referral care for the Maritime Provinces of Canada.

Children younger than 18 years.

Surveillance was conducted from 1991 to 1999 using personal contact with personnel and review of microbiology and medical records. Nosocomial diarrhea was defined as loose stools occurring more than 48 hours after admission, with at least two loose stools in 12 hours and no likely non-infectious cause.

Nosocomial diarrhea was the third most common nosocomial infection (217 of 1,466; 15%), after bloodstream and respiratory infections, with from 0.5 to 1 episode per 1,000 patient-days. Of 217 nosocomial diarrhea episodes, 122 (56%) had identified pathogens: C. difficile (39 of 122; 32%), rotavirus (38 of 122; 31%), adenovirus (36 of 122; 30%), and other viral (9 of 122; 7%). The median age was 1.3 years (range, 11 days to 17.9 years), 0.80 year for children with viral diarrhea, 3.9 years for children with C. difficile, and 1.5 years for children with diarrhea without a causative organism identified (P< .0001). Most children with nosocomial diarrhea were incontinent (diapered) at the time of their first episode (138 of 185; 75%), but preexisting incontinence was more common in those with viral diarrhea (93%) compared with those with no organism identified (71%) or those with C. difficile-associated diarrhea (CDAD) (49%) (P< .0001).

C. difficile is the single most common cause of nosocomial diarrhea in our tertiary-care center, although all viral pathogens account for 69% of cases. Diapered status appears to be a risk factor for CDAD in children, and CDAD occurs more often in older children than viral nosocomial diarrhea. Further characterization of risk factors for, and morbidity associated with, nosocomial CDAD in children is warranted.

To identify exposures associated with acute hepatitis B virus (HBV) infection among residents with diabetes in a skilled nursing facility.

Residents from Unit 3 and other skilled nursing facility residents with diabetes were tested for serologic evidence of HBV infection. Two retrospective cohort studies were conducted. Potential routes of HBV transmission were evaluated by statistical comparison of attack rates.

A 269-bed skilled nursing facility.

All skilled nursing facility residents with diabetes and skilled nursing facility residents who lived on the same unit as the index case (Unit 3) for some time during the case's incubation period.

All 5 residents with acute HBV infection had diabetes and resided in Unit 3. The attack rate among the 12 patients with diabetes in Unit 3 was 42%, compared with 0% among 43 patients without diabetes (relative risk, 37.2; 95% confidence interval, 4.7 to ∞). Acutely infected patients with diabetes received more morning insulin doses (P = .05), and more insulin doses (P = .03) and finger sticks (P = .02) on Wednesdays than did noninfected patients with diabetes. Two chronically infected patients with diabetes in Unit 3 were positive for hepatitis B e antigen and regularly received daily insulin and finger sticks. Of the 4 acute and 3 chronically infected residents from whom HBV DNA was amplified, all were genotype F and had an identical 678-bp S region sequence. Although no component of the lancets or injection devices was shared among residents, opportunities for HBV contamination of diabetes care supplies were identified.

Contamination of diabetes care supplies resulted in resident-to-resident transmission of HBV. In any setting in which diabetes care is performed, staff need to be educated regarding appropriate infection control practices.

To determine the additional direct costs of hospitalization attributable to catheter-associated urinary tract infection (CAUTI) in 1,497 newly catheterized patients.

Prospective observational and laboratory study.

University hospital.

Data were collected on risk factors for CAUTI (defined as > 103 colony-forming units [CFU]/mL), severity of illness, and diagnostic and therapeutic interventions in consenting newly catheterized patients. Daily urine cultures were obtained from each newly catheterized patient, but the results of these cultures were not revealed to his or her physician. During the study, one of the investigators (DGM) reviewed each patient's record and made a judgment as to which of the diagnostic tests and treatments ordered and what incremental length of stay could reasonably be ascribed to his or her CAUTI. The total hospital costs for each patient were also obtained.

Overall, 235 patients acquired CAUTIs during the study; most of the CAUTIs were completely asymptomatic, and only 52% were diagnosed by the patients' physicians using the hospital laboratory. Only 1 patient with a CAUTI had a secondary bloodstream infection. Thirty-three (13%) of the CAUTIs were caused by Escherichia coli; 63 (25%) by Klebsiella, Enterobacter, Citrobacter, Pseudomonas aeruginosa, or other antibiotic-resistant, gram-negative bacilli; 87 (35%) by enterococci or staphylococci; and 67 (27%) by Candida species. The 123 CAUTIs diagnosed by the hospital laboratory were judged to have been responsible for an additional $20,662 in extra costs of diagnostic tests and $35,872 in extra medication costs, a mean of $589 (median, $356) per CAUTI. CAUTIs caused by E. coli cost considerably less than infections caused by other gram-negative bacilli ($363.3 ± $228.2 vs $690.4 ± $783.7; P = .02) or yeasts ($821.2 ± $2,169.9). There were less striking differences in the costs per CAUTI caused by staphylococci or enterococci ($387.1 ± $434.8).

The extra direct costs associated with nosocomial CAUTI found in this prospective study, which was done in the era of managed care during the late 1990s, are substantially lower than those reported in the largest comparable studies done more than 15 years ago, most of which were retrospective, reflecting the powerful impact of cost-containment measures that are now implemented in managed care.

To determine the seroprevalence and risk factors for hepatitis C virus (HCV) infection among patients at an urban outpatient hemodialysis center.

This was a cross-sectional study of 227 patients undergoing hemodialysis at the Rogosin Kidney Center on December 15, 1998, with a response rate of 90% (227 of 253). Laboratory records were used to retrieve the total number of blood transfusions received and serologic study results. Univariate and multivariate analyses were used to examine the relationship among HCV serostatus, patient demographics, and HCV risk factors (eg, intravenous drug use [IVDU], intranasal cocaine use, multiple sexual partners, comorbidities, length of time receiving hemodialysis, and total number of blood transfusions received).

The seroprevalence of antibody to HCV (anti-HCV) was 23.3% (53 of 227) in the population. In univariate analysis, factors associated with HCV seropositivity included male gender, younger age, history of IVDU, history of intranasal cocaine use, history of multiple sexual partners, human immunodeficiency virus coinfection, increased time receiving dialysis, history of renal transplant, and positive antibody to hepatitis B core antigen. Multivariate logistic regression analysis showed that longer duration receiving dialysis and a history of IVDU were the only risk factors that remained independently associated with HCV seropositivity.

HCV is markedly more common in our urban cohort of patients receiving hemodialysis compared with patients receiving dialysis nationally and is associated with a longer duration of receiving dialysis and a history of IVDU. Stricter and more frequent enforcement of universal precautions may be required in hemodialysis centers located in areas with a high prevalence of HCV infection or IVDU among the general population.