Septicaemia in an Austrian neonatal intensive care unit: a 7-year analysis A Berger 1 , HR Salzer 1 , M Weninger 1 , B Sageder 2 and C Aspo ¨ck 3 Departments of Neonatology 1 , Medical Computer Sciences 2 and Hospital Infection 3 , University Hospital, Vienna, Austria Berger A, Salzer HR, Weninger M, Sageder B, Aspo ¨ ck C. Septicaemia in an Austrian neonatal intensive care unit: a 7-year analysis. Acta Pædiatr 1998; 87: 1066–9. Stockholm. ISSN 0803-5253 The results of blood cultures and clinical data of 101 neonates with 110 episodes of septicaemia during a 7-y study period were reviewed. The overall incidence of culture-proven sepsis within the study period was 6.0 per 100 neonatal intensive care unit admissions and the mortality rate was 14%. Three groups of pathogens accounted for 70% of all isolates: coagulase-negative staphylococci (27%), aerobic Gram-negative rods (24%) and Enterococcus faecalis (19%). Group B streptococcus was the major pathogen of very early-onset septicaemia (within 24 h of birth), whereas late-onset infections were most commonly caused by coagulase- negative staphylococci. Birthweight 1500 g, gestational age 30 weeks of gestation and early onset of symptoms within the first week of life were associated with poor prognosis. In addition, the case fatality rate of episodes caused by Gram-negative organisms was significantly higher than that of Gram-positive bacteraemia.  Microorganisms, mortality, neonatal septicaemia, risk factors A Berger, Department of Neonatology, University Hospital, Waehringerguertel 18-20, 1090 Vienna, Austria Septicaemia in neonatal intensive care medicine remains an important problem associated with high mortality and morbidity rates (1). With current advances in intensive care and subsequent increasing survival of very low birth- weight infants (VLBI), conditions for the development of septicaemia have become even more favourable. As major pathogens of neonatal sepsis vary with geo- graphic area and time, knowledge of the most common causative microorganisms of early- and late-onset sepsis in individual neonatal intensive care units (NICU) is impor- tant for the selection of infection control measures and initial empirical antibiotic therapy. The aim of the present study was to determine the causative microorganisms, frequency, risk factors and clini- cal course of neonatal septicaemia during a 7-y period in an Austrian tertiary neonatal intensive care centre. Patients and methods Blood culture results consecutively obtained from infants admitted to the NICU of the University Children’s Hospital in Vienna from Jan. 1, 1988 to Dec. 31, 1994 and the clinical records of all patients with positive results were evaluated retrospectively. Indications for blood sampling were clinical signs of infection or maternal risk factors for congenital septicaemia (premature rupture of membranes, premature labour, maternal fever). No routine blood cul- tures were taken. Blood cultures were drawn from a per- ipheral vein by a sterile technique. Blood specimens were inoculated into blood culture bottles (BACTEC 6A, 7A), which were then incubated at 37°C for a maximum of 7 d and analysed using the system BACTEC NR-760. When- ever a positive culture was detected, subcultivation and antibiotic sensitivity testing were performed. The micro- organisms were identified by routine methods (2). Infants were classified as septicaemic if they met the following three criteria: (i) at least one positive blood culture; (ii) clinical evidence of sepsis: temperature instability, increasing frequency of apnoea and/or brady- cardia, increased ventilatory requirements, feeding diffi- culties (gastric retention, abdominal distension, vomiting), rapidly deepening jaundice, sudden collapse; and (iii) treatment with specific antibiotics for at least 7 d. If the same microorganism was isolated from an infant on more than one occasion within a 7-d period this was counted as one single episode of sepsis. Corynebacterium spp., Micrococcus spp. and Bacillus spp. were regarded as con- taminants. Coagulase-negative staphylococci (CNS) were regarded as contaminants unless they were isolated from at least two blood specimens obtained at different times within one episode of infection or if isolates were obtained from a blood specimen and a second specimen from a normally sterile body fluid, from pus or from a discontin- ued vascular line in the absence of any other bacterial isolates within the same episode of infection. Catheter- related sepsis was defined as both a positive blood culture and a culture-positive intravascular catheter tip with the same microorganism, including an identical resistance pattern, in a patient with clinically suspected sepsis and no other source of infection. Patients with sepsis were divided into three groups according to age at onset of symptoms: very early (24 h), early (1–7 d) and late (7 d) onset. Data collection and frequency calculations were carried out using the computer system MONI, a database system for the surveillance of nosocomial infections (3). This computer system contains an extended classification of Acta Pædiatr 87: 1066–9. 1998  Scandinavian University Press 1998. ISSN 0803-5253