Severe noninfectious acute lung injury in early phase of bone marrow transplantation

Predictive factors of intensive care unit admission in patients with haematological malignancies and pneumonia

OBJECTIVE

To describe early signs at the onset of pneumonia occurring in the haematology ward which could be associated with a transfer to the ICU.

DESIGN

A 13-month preliminary prospective observational cohort study.

SETTING

Department of haematology and (32-bed) medical intensive care unit (ICU).

PATIENTS

Fifty-three of 302 patients hospitalised in the haematology ward who developed presumptive clinical evidence of pneumonia were enrolled.

MEASUREMENTS AND RESULTS

At the onset of the clinical evidence of pneumonia (day 1), we compared variables between patients requiring an ICU admission and those who did not. Twenty-four patients (45%) required a transfer to the ICU. Factors associated with ICU admission were: numbers of involved quadrants: 2.3 vs 1, P=0.001 and oxygenation parameters (initial level of O(2) supplementation: 3.5 vs 0.9 l/min, P<0.05), the presence of hepatic failure (58% vs 10%, P<0.01), Gram-negative bacilli isolated in blood culture (7 vs 1, P=0.01). In the multivariate analysis, a decrease of 10% in the SaO(2) and the requirement of nasal supplementary O(2) at the onset of acute respiratory failure increased the risk of admission to MICU, respectively, by 18 and by 14. The overall 6-month mortality rate of the 53 patients was 28%.

CONCLUSION

Parameters of oxygenation and radiological score could be associated with this transfer on day 1 of the onset of pneumonia occurrence. A further study should evaluate an earlier selection of this type of patient, followed by an "early" admission to the MICU, in order to improve ICU outcome.

Noninvasive continuous positive airway pressure in neutropenic patients with acute respiratory failure requiring intensive care unit admission

OBJECTIVE

To evaluate the tolerance and the efficacy of noninvasive continuous positive airway pressure (CPAP) in severe acute respiratory failure occurring in intensive care unit (ICU) neutropenic patients with hematologic malignancies, and to establish predictive variables of efficacy of this method.

DESIGN

Prospective study over a 5-yr period.

SETTING

Hematologic and medical intensive care unit of a teaching hospital.

METHODS

Among 129 neutropenic patients admitted to the ICU, 64 patients presented with febrile acute hypoxemic normocapnic respiratory failure (PaO2/FIO2 ratio <200) and were enrolled. In addition to standard therapy, patients received CPAP with a facial mask. The initial settings of the CPAP were 6 cm H2O positive end-expiratory pressure and FIO2 0.8 (80%). Physiologic measurements were performed at the end of 45 mins of ventilation with first adjustments. CPAP was used with a sequential mode (45 mins/3 hrs). CPAP was efficient if intubation was avoided.

RESULTS

The setting of CPAP, after adjustments, was as follows: positive end-expiratory pressure 7 +/- 1 cm H2O and FIO2 0.7 +/- 0.1 (70% +/- 10%). For the 64 patients, CPAP was administered for a total of 6 +/- 2 hrs during the first 24 hrs. The mean duration of CPAP was 7 +/- 3 days. A reduction in respiratory rate to less than 25 breaths/min was achieved in 53% of patients. PaO2/FIO2 ratio increased from 128 +/- 32 to 218 +/- 28. CPAP was successful in avoiding endotracheal intubation in 16/64 patients. A total of 16 responders and four nonresponders survived. Hepatic failure was a criterion indicating the failure of CPAP: 1/16 vs. 26/48 (p = .001). In multivariate analysis, two variables were predictive of failure of CPAP: Simplified Acute Physiology Score II (58 +/- 14 vs. 41 +/- 11) and a hepatic failure at the entry into the study.

CONCLUSION

CPAP was efficient in 25% of cases. All the responders survived. This noninvasive method was used as a way to avoid mechanical ventilation, which is well correlated with a poor prognosis in neutropenic ICU patients. Further controlled studies are needed to confirm the efficacy of noninvasive CPAP and to evaluate the most appropriate selection of immunocompromised patients.