Nosocomial colonisation and infection in a paediatric respiratory intensive care unit

A study of the prevalence of nosocomial colonisation and nosocomial infection (NI) was conducted in the paediatric respiratory intensive care unit of a large teaching hospital serving a developing community. Surveillance specimens were collected regularly from 63 consecutive patients admitted over 4 months, and also from professional staff, boarder mothers, cleaners and the unit environment. The incidence among patients of colonisation (40%) and of NI (43%) was high. The risk of dying in children with NI was appreciably increased (relative risk 2,241, confidence interval 0,591-8,503). This did not reach statistical significance, probably because so few children escaped acquiring hospital organisms. The significant risk factor for acquiring colonisation (P = 0.008) and NI (P < 0.0001) was a ward stay of more than 10 days. In addition, for acquiring NI an age of under 6 months was also predictive (P = 0.0298). The nature of the primary illness dictated the time spent in the ward; an important proportion of patients had preventable diseases, such as measles, pneumonia and tetanus, which required prolonged treatment. All children with endotracheal intubation had hospital-acquired organisms in tracheal aspirates. Eighty-two per cent of children developed positive gastric aspirates, 17% a positive urine culture and 11% a positive blood culture. Colonisation occurred rapidly; organisms initially appeared in gastric aspirates (mean 2 days), then in tracheal aspirates (mean 5 days) and urine cultures (mean 10 days). The acquired organisms, many of which were antibiotic-resistant, were almost exclusively enteric Gram-negative bacilli (GNB) and Staphylococcus aureus.(ABSTRACT TRUNCATED AT 250 WORDS)