To determine the incidence and evaluate the antimicrobial-susceptibility patterns of bacterial infections in our neonatal units.

Retrospective surveillance study.

The neonatal units of the Hadassah University Hospitals, Jerusalem, Israel.

All newborns admitted from January 1994 through February 1999.

The records of all patients with positive blood and cerebrospinal fluid cultures were reviewed. Bacteremia was considered early-onset (vertical) when occurring within the first 72 hours of life and late-onset (nosocomial) when occurring later. The prevalence and antibiotic-resistance patterns of vertically transmitted and nosocomially acquired strains were compared and studied over time.

219 of 35,691 newborn infants had at least one episode of bacteremia (6.13/1,000 live births). There were 305 identified organisms, of which 21% (1.29/1,000 live births) were considered vertically transmitted and 79% nosocomially acquired. The most common organism causing early-onset disease (29.2%) was group B streptococcus (0.38/1,000 live births), whereas coagulase-negative staphylococci (51%) were the most prevalent in late-onset disease. All gram-positive bacteria were susceptible to vancomycin. Most gram-positive organisms other than staphylococci were susceptible to ampicillin. Gram-negative organisms represented 31% of all isolates. Generally, there was a trend of increasing resistance to commonly used antibiotics among nosocomially acquired gram-negative organisms, compared to those vertically transmitted, with statistically significant differences for ampicillin and mezlocillin (P<.05 and P<.01, respectively). Over the years, a trend toward an increasing resistance to antibiotics was observed among gram-negative organisms.

The trend of increasing bacterial resistance to commonly used antibiotics necessitates the implementation of a rational empirical treatment strategy, based on local susceptibility data, reserving certain agents for emerging resistant pathogens.

To describe hospital practices and policies relating to bloodborne pathogens and current rates of occupational exposure among healthcare workers.

Hospitals in Iowa and Virginia were surveyed in 1996 and 1997 about Standard Precautions training programs and compliance. The primary outcome measures were rates of percutaneous injuries and mucocutaneous exposures.

153 (64%) of 240 hospitals responded. New employee training was offered no more than twice per year by nearly one third. Most (79%-80%) facilities monitored compliance of nurses, housekeepers, and laboratory technicians; physicians rarely were trained or monitored. Implementation of needlestick prevention devices was the most common action taken to decrease sharps injuries. Over one half of hospitals used needleless intravenous systems; larger hospitals used these significantly more often. Protected devices for phlebotomy or intravenous placement were purchased by only one third. Most (89% of large and 80% of small) hospitals met the recommended infection control personnel-to-bed ratio of 1:250. Eleven percent did not have access to postexposure care during all working hours. Percutaneous injury surveillance relied on incident reports (99% of facilities) and employee health records (61%). The annual reported percutaneous injury incidence rate from 106 hospitals was 5.3 injuries per 100 personnel. Compared to single tertiary-referral institution rates determined more than 5 years previously, current injury rates remain elevated in community hospitals.

Healthcare institutions need to commit sufficient resources to Standard Precautions training and monitoring and to infection control programs to meet the needs of all workers, including physicians. Healthcare workers clearly remain at risk for injury. Further effective interventions are needed for employee training, improving adherence, and providing needlestick prevention devices.

Varicella-zoster virus (VZV) vaccine is recommended to protect susceptible healthcare workers (HCWs) from serious disease and to prevent nosocomial spread of VZV. We evaluated clinical outcomes and serological responses in HCWs after immunization with live attenuated VZV vaccine.

Vaccinees were immunized from 1979 to 1998 during VZV vaccine trials, as well as after licensure, and followed prospectively for 1 month to 20.6 (mean 4.6) years after vaccination. Sera were tested by fluorescent antibody to membrane antigen (FAMA), latex agglutination (LA), and enzyme-linked immunoassay (EIA) to detect VZV-specific antibodies.

The median age of the 120 HCWs was 26 years; 51 (42%) were males.

Ninety eight (82%) of these study subjects received vaccine prepared by Merck and 22 (18%) by SmithKline Beecham; 25, 81, and 14 vaccinees received one dose, two doses, and three doses, respectively.

The crude attack rate was 10%; 12 of 120 HCWs developed chickenpox 6 months to 8.4 years after vaccination. The attack rates following household and hospital exposures were 18% (4/22) and 8% (6/72), respectively. All resulting illness was mild to moderate (mean 40 vesicles). Seroconversion after vaccination was documented by FAMA in 96% of HCWs, although 31% lost detectable antibodies. Compared with FAMA, LA and EIA were 82% and 74% sensitive and 94% and 89% specific, respectively.

The VZV vaccine effectively protected HCWs from varicella, particularly from serious disease. Currently available serological tests are not optimal, and improved assays are needed.

To study the relation between Staphylococcus aureus nasal and stool colonization, stool carriage of gram-negative bacilli resistant to third-generation cephalosporins (CephR), and subsequent infections during hospitalization.

Prospective study.

551 cirrhotic patients with 589 consecutive hospital stays. All patients were screened within 48 hours of admission; 589 nasal swabs, 417 stool specimens, and 589 urine samples were analyzed.

Carriage rates were 18.8% for methicillin-sensitive S aureus (MSSA), 16.3% for methicillin-resistant S aureus (MRSA), and 13.7% for CephR. We observed 87 episodes of spontaneous bacterial peritonitis, 63 cases of bacteremia, and 167 urinary tract infections occurred. Only 1 case of bacteremia and 4 urinary tract infections due to CephR occurred in patients carrying the same organism in their stools. The risk of MRSA ascitic fluid infections, bacteremia, and urinary tract infections was 3.1% versus 1% (not significant), 8.3% versus 0.8% (P<.001), and 11.4% versus 0.6% (P<.001) in carriers and noncarriers, respectively. Pulsed-field gel electrophoresis (PFGE) of isolates from 16 patients infected by MSSA (3 cases) and MRSA (13 cases) demonstrated that the colonizing strains matched the invasive strains in the 3 MSSA cases and in 8 of 13 MRSA cases.

Carriage of CephR strains is not associated with subsequent infection by these organisms in hospitalized cirrhotic patients. In contrast, MRSA carriage was an important risk factor for MRSA bacteremia and urinary tract infection.

To compare three laboratory screening protocols for the detection of methicillin-resistant Staphylococcus aureus (MRSA) from surveillance specimens (mannitol-salt agar containing 2 μg/mL of oxacillin [MSA-2], mannitol-salt agar containing 4 μg/mL of oxacillin [MSA-4], and a broth-containing protocol as recommended by the American Society for Microbiology [M-ASM]).

Blinded comparative laboratory study and cost analysis.

University-affiliated microbiology laboratory.

Outcome measurements included rate of detection of MRSA-positive specimens and patients, turnaround time, and media and technologist costs. All MRSA culture swabs obtained from any patient site from November 1998 to April 1999 were included.

The M-ASM protocol detected between 19.1% and 32.0% more MRSA-positive specimens and between 13.3% and 23.3% more MRSA-positive patients per surveillance event than the MSA-4 and MSA-2 protocols, respectively. There was no difference in positive-culture reporting time between the M-ASM and MSA-4 protocols. The broth-containing protocol was 2- to 2.5-fold more expensive than the simpler protocols, taking into account media and laboratory personnel costs.

It remains to be determined whether it is cost beneficial for a hospital to adopt the M-ASM, as the potential cost of MRSA transmission from unidentified MRSA-colonized patients is unknown. A broth-containing protocol should be considered the gold standard in future studies examining newer MRSA screening protocols.

To investigate the safety of the hospital water supply following a major flood.

Surveillance was conducted of the hospital water supply as it entered the hospital and at randomly selected water faucets throughout the facility.

A newly constructed surgical critical-care unit in a 265-bed community hospital that had to be evacuated and was out of operation for 6 weeks following a major flood of the city.

Random water samples throughout the facility were analyzed for heterotrophic plate counts (HPCs), chlorine, and coliforms utilizing standard methods.

Water samples entering the hospital met appropriate standards, indicating the city water distribution system was not contaminated. Of 169 faucets tested, 13 (22%) of 59 electronic faucets exceeded the HPC threshold, and 12 (11%) of 110 manual faucets exceeded the HPC threshold (P<.14). A comparison of two brands of electronic faucets with manual faucets and with each other revealed that the HPC threshold was exceeded by 11 (32%) of 34 brand A faucets as compared to 12 (11%) of 110 manual faucets (P<.006). The HPC threshold was exceeded by 2 (8%) of 25 brand B faucets compared to 12 (11%) of 110 manual faucets (P<.94). Contamination rates of brand A and brand B faucets differed significantly (P<.003). Similar testing 2 months after hyperchlorination of the water supply indicated that the HPC threshold was exceeded by 16 (52%) of 31 brand A faucets compared to 10 (9.%) of 110 manual faucets (P<.0000003) and by 2 (18%) of 25 brand B faucets compared to 10 (9%) of 110 manual faucets (P=1.0).

A certain brand of electronic water faucet used in the hospital was associated with unacceptable levels of microbial growth in water and was a continuing source of bacteria potentially hazardous to patients.

To investigate the background of changes of resistance phenotypes in methicillin-resistant Staphylococcus aureus (MRSA) from nosocomial infections in German hospitals by molecular typing and identification of particular resistance genes.

Isolates from the network for monitoring the spread of MRSA in Germany were subjected to quantitative susceptibility testing, to molecular typing, and to polymerase chain reaction identification of resistance genes.

The network consists of 175 German clinical microbiological laboratories collaborating with the German Reference Center for Staphylococci, which performs typing of staphylococcal isolates from nosocomial infections and data analysis.

During the past 5 years, MRSA susceptible to other antibiotics such as oxytetracycline, erythromycin, and gentamicin became more frequent. The proportion of epidemic MRSA clones that had been disseminated in the past and that exhibited broad resistance phenotypes decreased, whereas the proportion of recently emerging MRSA carrying only a few other resistance determinants has increased (1994, 11.5%; 1998, 39%).

The changing pattern of resistance phenotypes of MRSA from nosocomial infections in Germany is mainly due to the spread of recently emerging epidemic strains that are less frequently resistant to antibacterials other than oxacillin. The observed changes cannot simply be attributed to overall antibiotic consumption.

To determine risk factors for colonization with vancomycin-resistant enterococci (VRE) in a hospital outbreak.

Outbreak investigation and case-control study.

A referral teaching hospital in Melbourne, Australia.

Cases were inpatients colonized (with or without clinical disease) with VRE between July 26 and November 28, 1998; controls were hospitalized patients without VRE.

Five cases of VRE were identified between July 26 and November 8, 1998, by growth of VRE from various sites. Active case finding by cultures of rectal swabs from patients surveyed in wards was commenced on July 26, after the first isolate of VRE.

There were 19 cases and 66 controls. All the VRE identified were vanB, and all were Enterococcus faecium. One molecular type predominated (18/19 cases). In a logistic-regression model, being on the same ward as a VRE case was the highest risk factor (odds ratio [OR], 82; 95% confidence interval [CI95], 5.7-1,176; P=.001). Having more than five antibiotics (OR 11.9; CI95, 1.1-129.6; P<.05), use of metronidazole (OR 10.9; CI95, 1.7-69.8; P=.01), and being a medical patient (OR 8.1; CI95,1.4-47.6; P<.05) also were significant. Intensive care unit admission was associated with decreased risk (OR, 0.1; CI95, 0.01-0.8; P<.05).

Our findings are consistent with an acute hospital outbreak. Monitoring and control of antibiotic use, particularly metronidazole, may reduce VRE in our hospital. Ongoing surveillance and staff education also are necessary.

To measure directly the rate of contamination, during routine patient examination, of gowns, gloves, and stethoscopes with vancomycin-resistant enterococci (VRE).

A large, academic, tertiary-care hospital.

Between January 1997 and December 1998, 49 patients colonized or infected with VRE were entered in the study.

After routine examination, the examiner's glove fingertips, gown (the umbilical region and the cuffs), and stethoscope diaphragm were pressed onto Columbia colistin-nalidixic acid (CNA) agar plates with 5% sheep blood plus vancomycin 6 ug/mL. The stethoscope diaphragm was sampled again after cleaning with a 70% isopropanol wipe.

VRE were isolated from at least 1 examiner site (gloves, gowns, or stethoscope) in 33 (67%) of 49 cases. Gloves were contaminated in 63%, gowns in 37%, and stethoscopes in 31%. All three items were positive for VRE in 24%. One case each had stethoscope and gown contamination without glove contamination. Only 1 (2%) of 49 stethoscopes was positive after wiping with an alcohol swab. Contamination at any site was more likely when the patient had a colostomy or ileostomy. Patients identified by rectal-swab culture alone were as likely to contaminate their examiners as were those identified by clinical specimens.

Our study revealed a high rate of examiner contamination with VRE. The similar risk of contamination identified by surveillance and clinical cases reinforces concerns that patients not known to be colonized with VRE could serve as sources for dissemination. Wiping with alcohol is effective in decontaminating stethoscopes.

To evaluate the incidence of nosocomial infection (NI) in pediatric patients who received cardiothoracic surgery and to identify possible associated risk factors.

Prospective observational study.

The cardiac surgery and cardiac intensive care units at the Regina Margherita Children's Hospital, Turin, Italy.

All patients who underwent surgery from July 20,1998, to July 19,1999, were enrolled, except patients with operative catheterization only.

Clinical data were collected daily from July 20, 1998, to July 19, 1999. NIs were diagnosed according to US Centers for Disease Control and Prevention criteria.

104 patients were included in the present study, 80 (76.9%) of whom underwent extracorporeal circulation. The NI ratio was 48.1% (50/104); the percentage of patients with NI was 30.8% (32/104): 23.1% developed one infection, 7.7% two or more. The rate of NI was 2.17 per 100 days of hospitalization (50/2,304). The most common pathogen was Pseudomonas aeruginosa, Important risk factors were length of preoperative admission >5 days, total length of admission >10 days, open chest during postoperative phase, and cyanotic heart disease. There was a significant association between sepsis and central venous catheterization for 3 days or more. Rate of sepsis was 19 per 1,000 catheter days (16/852).

NIs represent a frequent complication for children who undergo heart surgery. Based on our data, we suggest decreasing the preoperative stay as much as possible. The higher NI incidence in patients with an open chest postoperatively suggests that an alternative antibiotic strategy should be considered for these patients.

To determine whether surveillance and infection control interventions decrease the incidence of catheter-related (CR) bloodstream infections (BSIs) in Korea.

A medical-surgical intensive care unit (ICU) of a university hospital in Korea.

The CR infection rate of the intervention period was compared to that of historical controls for a 4-month period.

All patients with a central venous catheter in the intensive care unit (ICU) from October 1998 to January 1999.

Active infection control programs were initiated during the intervention period. Data collected included patient characteristics, risk factors of CR infection, and the microbiology laboratory results. Laboratory-proven CR infection rates were compared between the intervention group and control group.

304 catheters were inserted into 248 patients. The intervention group and the control group showed similar characteristics, but more patients in the intervention group received steroid therapy, and subclavian insertion was more common in the intervention group. CR BSI occurred in 1.3 per 1,000 catheter-days in the intervention group and 4.2 in the control group (binomial test, P=.14). CR infections were associated with the duration of ICU admission by multivariate logistic regression.

The data suggested that an active infection surveillance and control program could reduce the rate of CR BSI in an ICU.

To determine the impact of methicillin-resis-tant Staphylococcus aureus (MRSA) colonization on the occurrence of S aureus infections (methicillin-resistant and methicillin-suscep-tible), the use of glycopeptides, and outcome among intensive care unit (ICU) patients.

Prospective observational cohort survey.

A medical-surgical ICU with 10 single-bed rooms in a 460-bed, tertiary-care, university-affiliated hospital.

A total of 1,044 ICU patients were followed for the detection of MRSA colonization from July 1, 1995, to July, 1 1998.

MRSA colonization was detected using nasal samples in all patients plus wound samples in surgical patients within 48 hours of admission or within the first 48 hours of ICU stay and weekly thereafter. MRSA infections were defined using Centers for Disease Control and Prevention standard definitions, except for ventilator-associated pneumonia and catheter-related infections, which were defined by quantitative distal culture samples.

One thousand forty-four patients (70% medical patients) were included in the analysis. Mean age was 61±18 years; mean Simplified Acute Physiologic Score (SAPS) II was 36.4±20; and median ICU stay was 4 (range, 1-193) days. Two hundred thirty-one patients (22%) died in the ICU. Fifty-four patients (5.1%) were colonized with MRSA on admission, and 52 (4.9%) of 1,044 acquired MRSA colonization in the ICU. Thirty-five patients developed a total of 42 S aureus infections (32 MRSA, 10 methi-cillin-susceptible). After factors associated with the development of an S aureus infection were adjusted for in a multivariate Cox model (SAPS II >36: hazard ratio [HR], 1.64; P=.09; male gender: HR, 2.2; P=.05), MRSA colonization increased the risk of S aureus infection (HR, 3.84; P=.0003). MRSA colonization did not influence ICU mortality (HR, 1.01; P=.94). Glycopeptides were used in 11.4% of the patients (119/1,044) for a median duration of 5 days. For patients with no colonization, MRSA colonization on admission, and ICU-acquired MRSA colonization, respectively, glycopeptide use per 1,000 hospital days was 37.7, 235.2, and 118.3 days. MRSA colonization per se increased by 3.3-fold the use of glycopeptides in MRSA-colonized patients, even when an MRSA infection was not demonstrated, compared to non-colonized patients.

In our unit, MRSA colonization greatly increased the risk of S aureus infection and of glycopeptide use in colonized and non-colonized patients, without influencing ICU mortality. MRSA colonization influenced glycopeptide use even if an MRSA infection was not demonstrated; thus, an MRSA control program is warranted to decrease vancomycin use and to limit glycopeptide resistance in gram-positive cocci.

To determine factors that correlate with increased antibiotic use among adult inpatients in Taiwan.

Retrospective survey of medical records.

14 acute-care hospitals (8 regional hospitals, 6 medical centers) in Taiwan.

A systematic probability sample from each hospital, totaling 663 adult inpatients who were discharged or had died in early 1999.

Infectious disease physicians at the 14 hospitals collected data from medical records regarding patient demographics, hospitalization, discharge diagnosis, and antibiotics received.

A total of 447 (67%) patients received antibiotics for an overall rate of 813 antibiotic-days (number of days patients received each antibiotic)/1,000 patient-days. Both the proportion of beds in intensive care units ([ICUs] Pearson correlation coefficient [r], 0.67; 95% confidence interval [CI95], 0.36-0.89; P<.01) and the proportion of patients admitted to surgical services (r, 0.66; CI95, 0.20-0.88; P=.01) correlated with the mean patient rate of antibiotic-days/hospital-day (MPAUD). In contrast, we found no correlation between the proportion of patients who received antibiotics and the MPAUD. Using multiple linear regression, medical center status was the only independent predictor for increased MPAUD (regression coefficient [b], 0.15; CI95, 0.05-0.24; P<.01). There was no correlation between pooled rates of antibiotic-days/hospital-day and any hospital demographic factors. First-generation cephalosporin (39%) and aminoglycoside (24%) use accounted for the majority of antibiotic-days.

Antibiotic use is greater in medical centers than in regional hospitals and appears to be independent of surgical case mix or the proportion of ICU beds. Determination of multiple, independent measures of antibiotic use may be necessary to understand the relation between antibiotic use and resistance in hospitals.

To quantitate the magnitude and consistency of positive (airflow out) and negative (airflow in) hospital special-ventilation-room (SVR) airflow.

A room-pressure evaluation was conducted during two seasons on a total of 18 rooms: standard rooms, airborne infection isolation rooms, and protective environment rooms. The pressures were measured using a digital pressure gaugepiezoresistive pressure sensor that measured pressure differentials. With doors closed, the rooms were measured a minimum of 30 times each for a cooling season and a heating season.

The standard rooms showed the least amount of variability in pressure differential, with an average of −0.2 Pa (median, −0.2 Pa), and an interquartile range (IQR) of 0.4 Pa. Airborne infection isolation rooms showed more variability in pressure, with an average of −0.3 Pa (median, −0.2 Pa) and an IQR of 0.5 Pa. Protective environment rooms had the greatest fluctuation in pressure, with an average of 8.3 Pa (median, 7.7 Pa) and an IQR of 8.8 Pa. Dramatic pressure changes were observed during this evaluation, which may have been influenced by room architectural differences (sealed vs unsealed); heating, ventilation, and air-conditioning zone interactions; and stack effect.

The pressure variations noted in this study, which potentially affect containment or exclusion of contaminants, support the need for standardization of pressure requirements for SVRs. To maintain consistent pressure levels, creating an airtight seal and continuous pressure monitoring may be necessary.

We have developed and analyzed a large surgical prophylaxis database and now report the factors significantly associated with early infection, readmission due to infection, and death within 28 days of surgery. This study is intended to be a stepping-stone for further studies using our clinical database.

A computerized database of 9,016 surgical patients from a 400-bed community hospital was examined. Multivariate logistic regression and tree-based modeling were used to identify factors associated with the outcomes. Factors considered included surgical procedure, prophylactic antibiotic, age, gender, serum creatinine, and albumin.

12.6% had an early infection, 2.5% were read-mitted due to infection, and 2.5% died within 28 days. Most combination prophylactic antibiotics were associated with an increased probability of an early infection. Decreased albumin and increased age were associated with an increased probability of an early infection. Tracheostomy and amputations were associated with an increased probability of an early infection, whereas gallbladder and orthopedic procedures involving the arm were associated with a decreased probability. Factors associated with readmission due to infection included dialysis shunt, vascular repair, and an early infection. Factors associated with increased probability of death within 28 days included age, albumin, serum creatinine, and an early infection. Gallbladder procedures and obstetric-gynecologic procedures were associated with a decreased probability of death within 28 days.

Older patients and those with a decreased albumin were most likely to have an early infection. To the extent that an early infection was a significant risk factor for readmission due to infection, the impact of age and albumin on the probability of readmission due to infection is demonstrated by their effects on early infections. Interestingly, albumin and age were significantly associated with death within 28 days, in addition to early infection, showing the predictive association between these factors and early death.

To determine how many infections are missed if the postdischarge surveillance (PDS) follow-up of intensive care unit (ICU) patients that is required by the National Nosocomial Infection Surveillance System method is not done.

ICU patients were followed up and surveillance results with PDS (gold standard) and without PDS were compared.

Surgical or interdisciplinary ICUs in eight German acute-care hospitals.

All 1,857 patients within a 6-month period in the participating ICUs (a total of 9,129 ICU-patient–days).

Without PDS, 45 urinary tract infections (UTIs) were diagnosed, compared with 53 with PDS; thus, 15% of the UTIs were missed if no postdischarge follow-up was performed. Three nosocomial pneumonias (4%) and one bloodstream infection (8%) also were missed if surveillance was carried out without PDS. A total of 198 nosocomial infections (NIs) were recorded with PDS, compared to 175 NIs without PDS. Approximately 12% of all ICU-associated NIs were missed if no follow-up was done.

Since it is very time-consuming to follow patients after their transfer from the ICU, we do not perform a postdischarge follow-up of patients in the course of routine surveillance.