Nosocomial pneumonia plays a role in the onset or worsening of broncho
pulmonary dysplasia (BPD) in the intubated premature infant (P). As fe
w data are available on the frequency of nosocomial pneumonia in P, we
conducted a retrospective study to determine the incidence and risk f
actors of this disease. Method : P of less than 32 weeks of gestationa
l age (GA) hospitalised in 1991 were included in the study if they wer
e intubated, ventilated and survived more than 3 days. P were classifi
ed in 3 groups. The first (G1) consisted of those presenting the nosoc
omial pneumonia clinical, biological and radiological criteria specifi
ed by the CDC as confirmed by a protected specimen brush (PSB) sample;
the second (G2) of P with suspected nosocomial pneumonia according to
CDC criteria but from whom no PSB was performed; and the third (G3) o
f P not presenting the CDC criteria. Results : 54/84P met the inclusio
n criteria(l4 in G1, 12 in G2, 28 in G3), wereas 16 were not intubated
and 14 died before day 3. Stepwise logistical regression showed only
two independent risk factors between G1 (confirmed nosocomial pneumoni
a) and G3 (no nosocomial pneumonia) : low GA and prolonged tracheal in
tubation (PTI). The age of onset of nosocomial pneumonia was also high
ly correlated with PTI (r = 0.75, p < 0.001). The relation between PTI
and pneumonia was thus more in the sense ''PTI => higher risk of noso
comial pneumonia'' than in the sense ''nosocomial pneumonia => PTI'' C
onclusion : the incidence of nosocomial pneumonia was high in this stu
dy (25 to 48 %), to the extent that GA was low and tracheal intubation
(i.e, the period of risk exposure) long.