[Study of correlations between antibiotics prescriptions and guidelines in a neonatal intensive care unit]

The increasing use of wide spectrum antibiotics has been reported to be associated with a greater prevalence of multi-resistant bacteria.

The aims of this study were to survey the use of antibiotics and to evaluate the correlations between patterns of prescription of antibiotics and prescription guidelines in a neonatal intensive care unit.

In this 6-month study, all newborns admitted to the NICU and treated with antibiotics were included. Data regarding criteria of antibiotic prescription, length of treatment, and criteria of withdrawing treatment were collected. The correlation between prescriptions and guidelines was evaluated a posteriori by a non-prescriber physician. One hundred and sixteen newborns were included, of whom nine had received antibiotics on more than one occasion. Mean gestational age was 33.5 weeks. In 82% of cases, the reason for hospitalisation was respiratory distress syndrome.

Patients received systemic antibiotics for primary infection (78%), nosocomial infection (17%) and postsurgical prophylaxis (5%). Suspected foeto-maternal infections (SFMI) were the dominant features of primary infection (96%). In 49% of cases, suspected infection was not proven and justified withdrawal of treatment within 3 days. Sixty percent of nosocomial infections occurred in newborns with gestational ages of less than 28 weeks. Bacterial criteria were decision-making factors only in nosocomial infections. An absence of observance of guidelines occurred in 9% of treated newborns, and in most cases involved excessive length of treatment.

These results show: (1) the majority of antibiotic prescriptions were for not proven SMFI; (2) a low rate of nosocomial infections; (3) the predominance of nosocomial infections in premature newborns; (4) less than 10% of non-observance of guidelines. It appears necessary to develop more precise guidelines to limit antibiotic use and to evaluate them regularly.