Serratia marcescens is a well recognized nosocomial pathogen. We report an outbreak with this organism in 8 neonates in a neonatal intensive care unit (NICU). Seven cases were treated successfully with meropenem after the failure of imipenem treatment. Although they have similar anti-microbial effects, meropenem can effectively treat the S. marcescens sepsis resistant to imipenem.

Keywords: Imipenem, Meropenem, Newborn, Sepsis, Serratia marcescens.

Our hospital is a 150-bed, tertiary referral care hospital. The NICU contains 17 beds: 10 level 3 NICU beds and 7 intermediate-intensive beds. In December 2006, an outbreak of systemic infections caused by Serratia marcescens occurred in the NICU. Clinical details were recorded. After the first case of S. marcescens septicemia was identified, subsequent surveillance cultures such as pharyngeal swab, rectal swab, sputum and stool cultures were obtained in all other neonates to trace reservoir/colonized neonates. Environmental cultures were taken and processed by standard laboratory methods, and susceptibility to antimicrobial agents was performed per guidelines of the Clinical and Laboratory Standards Institute(6).

Eight neonates developed clinical sepsis with positive blood cultures and one of them died. Three patients (case 1,4,8) had indwelling devices (mechanical ventilation for 11 to 23 days, central lines and total parenteral nutrition for 8-20 days). The length of stay in the NICU before S. marcescens varied from 4 to 72 days for infected patients. The clinical characteristics, demography and associated underlying disorders in these cases are shown in Table I.

TABLE I



Clinical Characteristics of Neonates Infected with S. marcescens

HIE: hypoxic ischemic encephalopathy; RDS: respiratory distress syndrome; 

CHD: Congenital heart disease; *Preterm low birth weight; 

I:Imipenem; M: Meropenem; G:Gentamicin; TTN:Transient tachypnea of newborn. 

The index case (case 1) of S. marcescens had received imipenem for seven days due to E. coli sepsis. The next day, three infected neonates (case 2, 3, and 4) were identified in the NICU. Standard infection control measures were reinforced. Despite efforts, four additional cases (case 5, 6, 7, and 8) of clinical septicemia in neonates occurred within ten days. Cultures were obtained weekly until sterile.

Infected neonates were treated empirically with imipenem for 5-10 days before culture results were available. Four neonates (cases 1, 5, 6 and 7) received intravenous imipenem (60 mg/kg per day divided into three doses) before S. marcescens sepsis occurred. After the culture results were available, gentamicin was added to the treatment regimen because of culture-antibiogram susceptibility. Lumbar puncture was not done due to severe thrombocytopenia and worstening in these babies. C-reactive protein(CRP) and blood counts were repeated frequently and cultures were repeated 72 hrs after the initiation of the treatment. As none of the neonates showed recovery signs, both clinically and with respect to the laboratory findings(CRP, thrombocyte, immature/mature ratio white blood cell and white cell count), this treatment was discontinued after 7-10 days. Case 2 died despite antibiotic therapy including imipenem and gentamicin. We ascertained that this resulted from in vivo resistance to imipenem, so we switched to intravenous meropenem (60 mg/kg per day divided into three doses) instead of imipenem for 3 weeks. A complete cure was obtained in seven patients with 21 days of therapy with this antibiotic regimen.

Cultures of samples from incubators, antiseptics, humidifiers, soap, distilled water, inhalation therapy equipment, hands of all healthcare workers and stethoscopes were negative for S. marcescens. Culture swabs from inanimate surfaces were all negative for S. marcescens. The S. marcescens isolates during the outbreak were uniformly susceptible to all of the antimicrobial agents tested, except for ampicillin, amoxycillin-clavulonate and cefazolin.

We thank Dr Mustafa Gul and Dr Pinar Ciragil who helped with laboratory processing in this study.

Contributors: All authors contributed to data search, concept, design, and reporting.

Funding: None.

Competing interests: None stated.