Early onset pneumonia in neurosurgical intensive care unit patients

To investigate early onset pneumonia in a neurosurgical intensive care unit , we studied a cohort of patients over a 13-month period and compared neuro trauma (T) with non-neurotrauma (NT) patients. Data were abstracted from th e infection surveillance database. Five hundred and sixty-five adults were hospitalized in the neurosurgical intensive care unit. 57.9% had trauma and 129 patients developed 152 episodes of pneumonia. Incidence rates, restric ted to the 129 first episodes of pneumonia, were 20.1 versus 15.7/1000 pati ent days and 34.2 versus 27.9/1000 ventilation days, in the T and NT groups respectively. In both groups, the distribution of risk stratified by hospi tal days was bimodal, being highest during the first three days. However, t he risk was higher for T patients (at day 3, 20/1000 ventilation days versu s 10.2/1000 ventilation days). The daily risk peaked again at days 5 and 6, and thereafter remained low. Pneumonia occurring within the first three da ys, or early onset pneumonia (EOP), was associated with trauma (P=0.036) an d, in the NT group only, with a Glasgow coma scale score lower than 9 (P=0. 062). EOP was caused by Staphylococcus aureus (33%), Haemophilus spp. (23%) , other Gram-positive cocci (22%), and other Gram-negative bacilli (GNB) (1 9%); whereas after the third day GNB other than Haemophilus spp. accounted for 45.4% of isolates (P=0.11). This large series confirms the high inciden ce of EOP in neurosurgical intensive care units, particularly among trauma patients, in relation to risk factors different from those seen in other in tensive care patients. Further studies are needed to elaborate specific pre ventive measures during early care.