Serratia marcescens can cause serious infections in patients in neonatal intensive care units (NICUs), including sepsis, pneumonia, urinary tract infection, and conjunctivitis. We report the utility of genetic fingerprinting to identify, investigate, and control two distinct outbreaks of S. marcescens.

An epidemiologic investigation was performed to control two clusters of S. marcescens infections and to determine possible routes of transmission. Molecular typing by pulsed-field gel electrophoresis determined the relatedness of S. marcescens strains recovered from neonates, the environment, and the hands of healthcare workers (HCWs).

Two geographically distinct level III-IV NICUs (NICU A and NICU B) in two university-affiliated teaching hospitals in New York City.

In NICU A, one major clone, “F,” was detected among isolates recovered from four neonates and the hands of one HCW. A second predominant clone, “A,” was recovered from four sink drains and one rectal surveillance culture from an asymptomatic neonate. In NICU B, four neonates were infected with clone “D,” and three sink drains harbored clone “H.” The attributable mortality rate from bloodstream infections was 60% (3 of 5 infants). The antimicrobial susceptibilities of clone F strains varied for amikacin, cefepime, and piperacillin/tazobactam.

S. marcescens causes significant morbidity and mortality in preterm neonates. Cross-transmission via transient hand carriage of a HCW appeared to be the probable route of transmission in NICU A. Sinks did not harbor the outbreak strains. Antimicrobial susceptibility patterns did not prove to be an accurate predictor of strain relatedness for S. marcescens.

To study clusters of infections caused by Serratia marcescens in a neonatal intensive care unit (NICU) and to determine risk factors for S. marcescens infection or colonization.

Genotyping of S. marcescens isolates was performed by pulsed-field gel electrophoresis (PFGE). A retrospective case-control study was conducted.

A tertiary-care pediatric hospital with a 16-bed NICU.

All neonates with at least one culture positive for S. marcescens in the NICU during December 1999 to July 2002. Case-patients (n = 11) treated in the NICU during December 1999 to February 2000 were included in the case-control study. Neonates treated in the NICU for at least 72 hours during the same period with cultures negative for S. marcescens were used as control-patients (n = 27).

S. marcescens was cultured from 19 neonates; 9 were infected and 10 were colonized. PFGE analysis identified three epidemic strains; each cluster consisted of identical isolates, except one isolate in the first cluster that was different. The risk factors identified were low birth weight, prematurity, prolonged respiratory therapy, prolonged use of antibiotics, and maternal infection prior to delivery. Overcrowding and understaffing were recorded simultaneously with the clusters.

PFGE analysis showed three independent clusters. Several factors contributed to spread of the epidemic strains: (1) there were many severely premature and susceptible neonates, (2) the NICU was overcrowded during the clusters, and (3) transmission was likely to occur via the hands of staff. Cohorting and improvement of routine infection control measures led to the cessation of each cluster.

To investigate a cluster of patients infected and colonized with Serratia marcescens in a neonatal intensive care unit (NICU).

In June 2001, two neonates in the NICU had clinical infections with S. marcescens and one died. Infection control surveillance data for the NICU revealed that S. marcescens was rarely isolated from clinical specimens. Surveillance and environmental cultures were performed and isolates were typed using pulsed-field gel electrophoresis. Staff and neonates were cohorted and a waterless, alcohol-based handwashing agent was introduced. A case-control study was performed.

From June 2 through August 20, 2001, 11 neonates with S. marcescens infection and colonization were identified. The incidence of S. marcescens infections increased from 0.19 per 1,000 patient-days in 2000 to 0.52 per 1,000 patient-days in 2001 (P < .0001). In the first 3 weeks of the investigation, there were 2 sets of patients and sinks with indistinguishable strains; however, in subsequent weeks, all isolates were of unique strains, signifying no further transmission of the two initial predominant strains. Neonates with S. marcescens were more likely to have a lower gestational age and birth weight. There was no association between cases and healthcare workers (HCWs).

A cluster of S. marcescens was quickly terminated after the introduction of preventive measures including cohorting of infected and colonized neonates and HCWs, contact precautions, surveillance cultures, and a waterless, alcohol-based hand antiseptic. Chromosomal typing determined that strains with an indistinguishable pattern were no longer present in the unit after control measures were implemented.

To evaluate antimicrobial use and the influence of inadequate empiric antimicrobial therapy on the outcomes of nosocomial bloodstream infections (BSIs).

Prospective cohort study with nested case-control analysis.

Neonatal intensive care unit (NICU).

All patients weighing 2,000 g or less were enrolled. Data collection included risk factors for nosocomial BSI, admission severity of illness, microbiology, antimicrobial therapy, and outcomes. Inadequate empiric antimicrobial therapy was defined as the use of antibiotics for more than 48 hours after the day that blood cultures were performed that did not cover the microorganisms causing the bacteremia or administration of antibiotics that failed to cover resistant microorganisms.

Two hundred twenty-nine patients were enrolled. Forty-five developed nosocomial BSIs. The BSI rates were 11.2, 2.8, and 0 per 1,000 catheter-days for patients weighing 1,000 g or less, between 1,001 and 1,500 g, and between 1,501 and 2,000 g, respectively. After adjustment for severity of illness, the mortality in patients with nosocomial BSI receiving inadequate empiric antimicrobial therapy was higher than in those receiving adequate therapy (adjusted odds ratio [AOR], 5.3; 95% confidence interval [CI95], 1.2-23.2). By multivariate analysis, nosocomial BSI attributed to Candida species (AOR, 6.3; CI95, 1.4-28.0) and invasive procedure prior to onset of BSI (AOR, 6.4; CI95, 1.0-39.0) were associated with administration of inadequate empiric antimicrobial therapy.

Administration of inadequate empiric antimicrobial therapy among NICU patients with nosocomial BSI was associated with higher mortality. Additional studies on the role of inadequate empiric antimicrobial therapy and the outcomes of BSIs among NICU patients are needed.

To evaluate the effects of a hand hygiene program on compliance with hand hygiene and the rate of nosocomial infections in a neonatal intensive care unit (NICU).

Open trial.

A level-III NICU in a teaching hospital.

Nurses, physicians, and other healthcare workers in the NICU.

A multimodal campaign for hand hygiene promotion was conducted beginning in September 1998. This program consisted of formal lectures, written instructions and posted reminders regarding hand hygiene and proper handwashing techniques, covert observation, financial incentives, and regular group feedback on compliance. Surveillance of handwashing compliance and nosocomial infections before and during the program was analyzed.

Overall compliance with hand hygiene improved from 43% at baseline to 80% during the promotion program. The rate of nosocomial infections decreased from 15.13 to 10.69 per 1,000 patient-days (P = .003) with improved handwashing compliance. In particular, respiratory tract infections decreased from 3.35 to 1.06 per 1,000 patient-days during the handwashing campaign (P = .002). Furthermore, the correlation between nosocomial infection of the respiratory tract and handwashing compliance also reached statistical significance (r = -0.385; P = .014).

Improved compliance with handwashing was associated with a significant decrease in overall rates of nosocomial infection and respiratory infections in particular. Washing hands is a simple, economical, and effective method for preventing nosocomial infections in the NICU

To assess the knowledge, beliefs, and practices of neonatal intensive care unit (NICU) healthcare workers (HCWs).

Self-administered survey.

A 55-bed NICU.

NICU HCWs (N = 215).

The response rate was 68%. Ninety-two percent knew central venous catheters (CVCs) should be capped, clamped, or connected to running fluids at all times. Ninety-five percent knew when to change gloves. Thirty-one percent knew the recommended duration for handwashing. Most HCWs believed sterile technique in CVC care (96%), gloves (91%), and handwashing (99%) prevent nosocomial infection (NI). Sixty-seven percent used sterile barriers to insert CVCs, 76% reported wearing gloves, 81% reported routine handwashing, 35% knew that bacterial hand counts are higher with rings, 30% knew that long fingernails are associated with higher gram-negative bacterial hand contamination, and 35% knew that artificial fingernails are associated with higher gram-negative bacterial hand contamination. Most (93%) believed HCWs can affect outcomes of patients with NIs. Fewer believed rings (40%), artificial fingernails (61%), and long fingernails (48%) play a role in NIs, or that policies concerning number of rings (50%), cutting fingernails (35%), or prohibiting artificial fingernails (47%) would prevent NIs. Sixty-one percent of HCWs regularly wore at least one ring to work, 56% wore their fingernails shorter than the fingertip, and 8% wore artificial fingernails.

A disconnect existed between CVC knowledge and beliefs and practice. HCWs did not know the relationship between bacterial hand counts and rings and fingernails, and did not believe rings or long or artificial fingernails increased the risk of NIs.

To describe the rate, risk factors, and outcome of ventilator-associated pneumonia (VAP) in pediatric patients.

This prospective surveillance study of VAP among all patients receiving mechanical ventilation for 48 hours or more admitted to a pediatric intensive care unit (PICU) in Saudi Arabia from May 2000 to November 2002 used National Nosocomial Infections Surveillance (NNIS) System definitions.

Three hundred sixty-one eligible patients were enrolled. Most were Saudi with a mean age of 28.6 months. Thirty-seven developed VAP. The mean VAP rate was 8.87 per 1,000 ventilation-days with a ventilation utilization rate of 47%. The mean duration of mechanical ventilation was 21 days for VAP patients and 10 days for non-VAP patients. The mean PICU stay was 34 days for VAP patients and 15 days for non-VAP patients. Among VAP patients, Pseudomonas aeruginosa was the most common organism, followed by Staphylococcus aureus. Other gram-negative organisms were also encountered. There was no significant difference between VAP and non-VAP patients regarding mortality rate. Witnessed aspiration, reintubation, prior antibiotic therapy, continuous enteral feeding, and bronchoscopy were associated with VAP. On multiple logistic regression analysis, only prior antibiotic therapy, continuous enteral feeding, and bronchoscopy were independent predictors of VAP.

The mean VAP rate in this hospital was higher than that reported by NNIS System surveillance of PICUs. This study has established a benchmark for future studies of VAP in the pediatric intensive care population in Saudi Arabia. Additional studies from the region are necessary for comparison and development of preventive measures.

To determine the incidence, clinical characteristics, and risk factors associated with the first culture-proven episode of sepsis among neonates in a neonatal intensive care unit (NICU).

Level-II NICU in the United Kingdom.

Neonates with their first culture-proven sepsis between January 1, 1996, and December 31, 2000.

Demographic data were obtained from the NICU database and chart review. Sepsis was considered early (EOS; < 72 hours old) or late (LOS; > 72 hours old). Data were also collected on potential risk factors.

Among 14,767 live births, 1,612 (11%) neonates were admitted to the NICU during the study period. Nine hundred eight were screened for sepsis. One hundred twenty-four had at least one positive culture (overall sepsis rate of 8.4 per 1,000 live births [1%] or 77 per 1,000 NICU admissions). Twenty-four neonates had EOS and 100 had LOS. Coagulase-negative staphylococci (CoNS) and group B Streptococcus were the most frequent organisms causing EOS, whereas CoNS and Escherichia coli most frequently caused LOS. Birth before 30 weeks' gestation and birth weight less than 1,500 g were risk factors for sepsis. Resuscitation at birth was the leading risk factor for EOS and respiratory support prior to sepsis, presence of a central or peripheral catheter, and total parenteral nutrition were leading risk factors for LOS.

A strong inverse relationship existed between gestational age of 30 weeks or younger and birth weight of 1,500 g or less and LOS. Resuscitation and indwelling intravenous catheters were also risk factors.

To acquire data on pediatric nosocomial infections (NIs), which are associated with substantial morbidity and mortality and for which data are scarce.

Prevalence survey and evaluation of a new comorbidity index.

Seven Swiss pediatric hospitals.

Those hospitalized for at least 24 hours in a medical, surgical, intensive care, or intermediate care ward.

Thirty-five NIs were observed among 520 patients (6.7%; range per hospital, 1.4% to 11.8%). Bacteremia was most frequent (2.5 per 100 patients), followed by urinary tract infection (1.3 per 100 patients) and surgical-site infection (1.1 per 100 patients; 3.2 per 100 patients undergoing surgery). The median duration until the onset of infection was 19 days. Independent risk factors for NI were age between 1 and 12 months, a comorbidity score of 2 or greater, and a urinary catheter. Among surgical patients, an American Society of Anesthesiologists (ASA) score of 2 or greater was associated with any type of NI (P = .03). Enterobacteriaceae were the most frequent cause of NI, followed by coagulase-negative staphylococci; viruses were rarely the cause.

This national prevalence survey yielded valuable information about the rate and risk factors of pediatric NI. A new comorbidity score showed promising performance. ASA score may be a predictor of NI. The season in which a prevalence survey is conducted must be considered, as this determines whether seasonal viral infections are observed. Periodic prevalence surveys are a simple and cost-effective method for assessing NI and comparing rates among pediatric hospitals.

To describe the epidemiology of healthcare-associated infections (HAIs) among neonates.

Prospective surveillance of HAIs was conducted during 2 years. Infections beginning within 48 hours of birth were defined as HAIs of maternal origin. Death occurring during an active episode of HAI was considered related to HAI.

Seven neonatal units located in three Brazilian cities.

All admitted neonates were included and observed until discharge.

Twenty-two percent of 4,878 neonates had at least one HAI. The overall incidence density was 24.9 per 1,000 patient-days, and 28.1% of all HAIs were maternally acquired. HAI rates ranged from 12.3% in the group with a birth weight (BW) of more than 2,500 g to 51.9% in the group with a BW of 1,000 g or less. The main HAIs were bloodstream infection (BSI) and pneumonia. Coagulase-negative staphylococci, Enterobacter species, Staphylococcus aureus, and Klebsiella pneumoniae were the main pathogens. Forty percent of all deaths were related to HAI. Central venous catheter (CVC)-associated BSIs per 1,000 CVC-days ranged from 17.3 (BW, 1,501 to 2,500 g; device utilization [DU], 0.11) to 34.9 (BW, ≤ 1,000 g; DU, 34.92). Ventilator-associated pneumonia per 1,000 ventilator-days ranged from 7.0 (BW, ≤ 1,000 g; DU, 0.34) to 9.2 (BW, 1,001 to 1,500 g; DU, 0.14).

The high proportion of HAIs of maternal origin highlights perinatal care issues in Brazil and the need to improve the diagnosis of neonatal HAIs. The very low BW group and device-associated infections should be priorities for prevention strategies in this population.

To determine a potential source of MRSA colonization and infection among preterm infants in a neonatal intensive care unit (NICU) using molecular analysis of breast milk samples.

Case report, outbreak investigation.

Preterm triplets were delivered at 26 weeks' gestation via cesarean section when routine active surveillance for MRSA was performed for all infants in a NICU. Surveillance consisted of swabbing the throat, nose, and umbilicus (TNU) weekly. Although infants A and B initially had negative TNU swabs, repeat cultures were positive for MRSA on day of life (DOL) 10 and DOL 18, respectively. Surveillance and clinical cultures for infant C were negative. Infant A developed sepsis, and multiple blood cultures were positive for MRSA beginning on DOL 14. Infant B developed conjunctivitis and a conjunctival exudate culture was positive for MRSA on DOL 70. Both infants were fed breast milk via nasogastric tube. Cultures of breast milk samples for infants A and B dated prior to either infant's first positive surveillance culture were positive for MRSA. All MRSA isolates had identical results on antibiotic susceptibility testing. PFGE demonstrated identical banding patterns for the MRSA isolates from the blood culture of infant A, breast milk for infants A and B, and a surveillance swab from infant B. At no time did the mother develop evidence of mastitis or other local breast infection.

MRSA can be passed from mother to preterm infant through contaminated breast milk, even in the absence of maternal infection. Colonization and clinical disease can result.