The course of 260 adults with haematological malignancies admitted to a medical intensive care unit was studied to evaluate the value of life support techniques and to research predictive factors. The overall in the medical intensive care unit (MICU) and hospital mortality rates were respectively 43% (113 patients) and 57% (148 patients). Among survivors, 64% (49 patients) were still alive after 6 months and 44% (35 patients) after 1 year. Among 34 haemodialysed patients, the MICU mortality was 67% (23 patients) and among 111 mechanically ventilated patients 85% (94 patients). Prolonged mechanical ventilation, more than seven days, was performed in 11 of the 17 survivors and did not influence long term survival. No individual predictor of mortality was found comparing survivors and non-survivors. However, SAPS, intractable sepsis and failure of more than one organ system were significantly different in non-survivors (p less than 0.001). Among the 20 patients requiring both mechanical ventilation and haemodialysis, only two left the MICU and both died soon thereafter. We conclude that life support therapy should be initiated in patients with haematological disorders and that prolonged mechanical ventilation is compatible with long term survival. However, the combination of mechanical ventilation and haemodialysis is always associated with a poor prognosis and therefore the use of both techniques simultaneously for one patient is questionable.

Circuits on mechanical ventilators with cascade humidifiers are routinely changed every day or every other day, although humidifying cascades have been considered unlikely to increase the risk of respiratory infection because they do not generate aerosols. Moreover, changing ventilator tubings every 24 rather than every 48 h increases the risk of ventilator-associated pneumonia. To study the effects of ventilator circuit changes on the rate of nosocomial pneumonia and on patient and circuit colonization, 73 consecutive patients requiring continuous mechanical ventilation for more than 48 h were randomly assigned to either ventilator circuit changes every 48 h (Group 1, n = 38) or no change (Group 2, n = 35). Patients dying or being weaned before 96 h were not analyzed (Group 1 n = 3; Group 2 n = 7; leaving Group 1 n = 35 and Group 2 n = 28; p = 0.13). Ventilator-associated pneumonia was defined as the occurrence during mechanical ventilation or within 48 h after weaning of a new and persistent infiltrate on chest X-ray, purulent tracheal secretions, and a positive culture of a protected brush specimen (greater than or equal to 10(3) cfu/ml). Bacterial colonization was assessed every 48 h by quantitative cultures of pharyngeal swab, tracheal aspirate, humidifying cascade, and expiratory tubing trap. The two groups were similar in terms of age, indication for and duration of ventilation, and severity of illness.(ABSTRACT TRUNCATED AT 250 WORDS)

The broad range of offending organisms seen in early aspiration pneumonia precludes use of any single empiric regimen, making protected specimen brush mandatory in many patients. Nevertheless, the involvement of S.pneumoniae in a notable proportion of our patients suggests that routine penicillin prophylaxis after early aspiration (at least in most patients with community-acquired aspiration) is warranted given the potential severity of pneumococcal sepsis in such patients.