Mortality of medical oncological patients admitted to intensive care.

Editor,—We were interested to read the abstract by Martin, Quintin and Hinds concerning the prognosis of patients admitted to intensive care with life-threatening complications of haematological malignancy.1 We recently conducted an audit on our intensive care unit with similar results. We audited all patients admitted to the intensive care unit from the Regional Cancer Centre over a 5-yr period (Jan 1993–Dec 1997). All admissions from medical oncology were included. We calculated the mortality in this group and compared it with the background mortality of the unit. We collected data on age, sex, tumour type, duration of stay, mechanical ventilation, extracorporeal renal support, APACHE II score and white cell count. Significant prognostic factors were identified using Fisher’s exact test (P,0.05). Twenty-two patients were identified. Mean age was 46.6 yr, with an equal number of males and females. Haematological malignancies comprised 45% of these admissions; the remainder were solid tumours. The mean APACHE II score was 25 (SD 8.1). The mortality for oncological admissions was 81% compared with a background of 28%, with a relative risk of mortality of 2.9 (95%CI 2.3, 3.6). Neutropenia, renal support and mechanical ventilation were significantly associated with increased mortality (89%, 90% and 93%, respectively, P,0.0003), with relative risks of mortality of 3.1 (95%CI 2.4, 4.0), 3.2 (95%CI 2.5, 4.0) and 3.3 (95%CI 2.8, 3.9), respectively. No patient who required IPPV and renal support survived, with a relative risk of mortality of 3.5 (95%CI 3.3, 3.8). These results confirm the findings in the abstract and in previous studies.2 In view of the very poor prognosis, should patients be given the information required to make an informed decision on whether they wish to have intensive treatment should they become critically ill as a result of their tumour and its treatment?