CRP, duration of fever, the "standardized clinical impression score", a history of diarrhoea and focal signs of infection were the independent, most powerful predictors of SBI in febrile infants, identified by logistic regression analysis. Although the predictive model is not validated for direct clinical use, it illustrates the clinical potential of the used technique. This technique offers the advantage of assess the probability of SBI in every individual infant. This probability will form the best basis for well-founded decisions in the management of the individual febrile infant.

SUMMARY Faeces from 24 neonates with proved necrotising enterocolitis (NEC), from 12 with clinically suspected NEC, and from 41 control infants were quantitatively cultured under aerobic and anaerobic conditions. An important difference in colonisation with Klebsiella was found between the NEC groups and the control group. Although the cause of NEC is unknown, colonisation with Klebsiella seems to increase the risk.Necrotising enterocolitis (NEC) is found especially among preterm infants. Its principal clinical signs are distended abdomen, retention of food, and faecal blood loss. A typical radiological appearance is gas in the intestinal wall (pneumatosis intestinalis) and evidence suggests that this gas originates from bacteria.1 NEC is usually found in the terminal ileum, but it may extend to other parts of the intestine. It is associated with necrosis of the intestinal wall and in many cases leads to perforation and peritonitis. The incidence is usually reported as 10-1-5/1000 live born infants, with a mortality of about 30 %. Published figures show that the incidence of NEC is higher and the prognosis more unfavourable in infants of low birthweight and young gestational age.2 Although NEC is an important problem in neonatal intensive care units (ICU), its cause is unknown. Several factors probably play a part in its aetiology-intestinal mucosal injury (caused by ischaemia, mechanical damage, or toxic substances), enteral feeding, and the presence of bacteria.3 Whether certain bacterial species increase the risk of NEC is uncertain and a prospective controlled study of the faecal flora in NEC was undertaken to investigate this. MethodsPatients. Three groups of patients were studied. Group 1 comprised 24 children in whom NEC was diagnosed radiologically or at operation, or both, or at necropsy. Group 2 comprised 12 children who showed clinical signs of NEC-distended abdomen, retention of food, and facael blood loss-but in whom the diagnosis could nct be proved. Group 3 comprised 41 control infants, matched for birthweight and gestational age, who were cared for in the same ICU during the study period. A subgroup of the controls was considered separately in the analysis of results. This was group 4, comprising 19 children whose faeces were cultured within 7 days before or after NEC was diagnosed in another child in the ICU.Collection of material and culture. Between 20 September 1980 and 1 January 1982, faeces were collected from children who were at high risk from NEC. These were infants aged less than 1 month, born at a gestational age of less than 34 weeks, and with a birthweight of less than 2000 g, who were admitted to the ICU. Faecal cultures were made on the first day of illness in groups 1 and 2. In group 3 this was done on about day 10, which is the usual age of onset of NEC.After collection, a small amount of faeces was transferred to a bottle containing 3 ml glycerol broth, stored at 4°C, and used within 24 hours for quantitative cultures under aerobic and anaerobic conditions.4 The amount of faeces...

Chequerboard titrations carried out with modified serial dilutions of antibiotics such that consecutive concentrations in these series were four times smaller than those in two-fold serial dilutions enable MICs and MBCs to be determined with greater accuracy. Interaction indices calculated by this method can differ markedly from those calculated on the basis of two-fold serial dilutions. The differences calculated in this study ranged from -0.30 to +1.06.

Summary. The occurrence of various Pseudomonas aeruginosa strains in the sputum of 15 patients with cystic fibrosis (CF) was monitored over periods ranging from 2 to 60 months. Isolates of P . aeruginosa were typed by four different techniques, namely serotyping, active and passive pyocin typing, and phage typing. The maximum number of different serotypes found in the patients was three (one serotype in nine patients; two serotypes in five patients; three serotypes in one patient). Pyocin and phage typing showed no marked differences between strains of the same serotype in individual patients. Exacerbations of chronic respiratory infection were not associated with changes in the sputum flora, the composition of P . aeruginosa strains in which remains constant over long periods in patients with CF.