Outcome of Cancer Patients Considered for Intensive Care Unit Admission: A Hospital-Wide Prospective Study
http://www.100md.com
《临床肿瘤学》
the Saint Louis Teaching Hospital
Paris 7 University
Assistance Publique–Hpitaux de Paris, Paris, France
ABSTRACT
PURPOSE: To evaluate the outcome of cancer patients considered for admission to the intensive care unit (ICU).
PATIENTS AND METHODS: Prospective, one-year hospital-wide study of all cancer and hematology patients, including bone marrow transplantation patients, for whom admission to the ICU was requested.
RESULTS: Of the 206 patients considered for ICU admission, 105 patients (51%) were admitted. Of the 101 patients who were not admitted, 54 patients (26.2%) were considered too sick to benefit, and 47 patients (22.8%) were considered to be too well to benefit from the ICU. Of these 47 patients, 13 patients were admitted later. Survival rates after 30 and 180 days were significantly associated with admission status (P < .0001). Remission of the malignancy (odds ratio [OR], 3.37; 95% CI, 1.25 to 9.07) was independently associated with ICU admission, whereas poor chronic health status (OR, 0.38; 95% CI, 0.16 to 0.74) and solid tumor (OR, 0.43; 95% CI, 0.24 to 0.78) were associated with ICU refusal. In admitted patients, 30-day and 6-month survival rates were 54.3% and 32.4%, respectively. Of the patients considered too sick to benefit from ICU admission, 26% were alive on day 30 and 16.7% on day 180. Among patients considered too well to benefit, the 30-day survival rate was a worrisome 78.7%. Calibration of the Mortality Probability Model (the only score available at triage) was of limited value for predicting 30-day survival (area under the curve, 0.62).
CONCLUSION: Both the excess mortality in too-well patients later admitted to the ICU and the relatively good survival in too-sick patients suggest the need for a broader admission policy.
INTRODUCTION
Over the last decade, survival rates in critically ill cancer patients have improved dramatically.1-3 Three factors have contributed to this welcome trend. First, advances in the treatment of solid tumors4,5 and hematologic malignancies6-11 produced a 20% decrease in overall mortality in cancer patients from 1978 to 1998.12 Second, earlier admission to the intensive care unit (ICU) has resulted in better survival rates for critically ill cancer patients.13 Interventions that have been used successfully include bronchoscopy and bronchoalveolar lavage14,15 and noninvasive mechanical ventilation16,17 in patients with pulmonary infiltrates, aggressive management of septic shock,13 and investigations of specific organ failures.18-22 Third, progress has been made in selecting patients likely to benefit from ICU admission.1,16
Because the decision to recommend ICU admission is based on complex criteria, and to ensure appropriate utilization of ICU resources without depriving critically ill patients of a chance to recover, the North American and European Societies of Critical Care Medicine have developed recommendations for ICU admission, triage, and discharge.23,24 In a prospective longitudinal study performed in 26 ICUs in 2001,25 we evaluated factors associated with intensivists' refusal to admit patients to the ICU, for whom other physicians had requested ICU admission. We found that cancer patients were at high risk to be refused by the ICU, in accordance with current recommendations and with numerous articles discouraging ICU admission26,27 or prolonged intensive care28,29 for cancer patients. With good reason, oncologists and intensivists reserve ICU admission for select cancer patients with potentially reversible medical conditions and a "reasonable prospect of substantial recovery."23,24 However, the procedures used to select cancer patients for ICU admission and the outcomes associated with these procedures have not been evaluated.
The objective of this study was to determine the outcomes of cancer patients for whom ICU admission was requested. To achieve this objective, we performed a prospective longitudinal study over a 1-year period in a teaching hospital where about half the beds were in oncology and hematology wards. We included all patients for whom admission to the hospital's medical ICU was requested, regardless of whether admission was granted or not.
PATIENTS AND METHODS
The institutional review board of the French Society for Critical Care approved the study in June 2001. No information allowing identification of the patients was recorded in this prospective observational cohort study. We included all patients with solid tumors or hematologic malignancies for whom admission to the medical ICU of the Saint Louis Teaching Hospital (Paris, France) was requested by ward physicians in the same hospital between March 1, 2002, and March 1, 2003.
Description of the Saint-Louis Teaching Hospital
The Saint-Louis Teaching Hospital is a 650-bed university hospital with 330 beds in the hematology and oncology wards, where about 600 patients with newly diagnosed cancer are admitted each year. There are five hematology wards (acute leukemia, non-Hodgkin's lymphoma, myeloma, autologous stem-cell transplantation, and allogeneic stem cell or bone marrow transplantation) and three oncology wards (general oncology, lung cancer, and radiation therapy). Hospital mortality rate in the eight oncology wards is approximately 5%. The medical ICU is a closed 12-bed unit that admits 550 to 600 patients per year, including about 130 cancer patients, most of whom have hematologic malignancies. In accordance with legislation for ICUs in Europe, the daytime staff consists of five critical care specialists (intensivists) and three residents, and the night-time staff consists of one senior intensivist and one resident. In addition, a senior hematologist and a hematology resident are on call 24 hours a day for the hematology and solid-tumor wards.
Criteria for ICU Admission Used Routinely at the Saint-Louis Teaching Hospital
ICU admission is considered when a cancer patient has at least one organ failure. First, the patient is evaluated by a senior intensivist and the ward oncologist and hematologist in charge of the patient. The final decision regarding ICU admission is then made by the senior intensivist, who records the decision in the patient's chart as admission, refusal because the patient is too sick to benefit, or refusal because the patient is too well to benefit. When no beds are available, the patient is always transferred to another ICU. To admit a patient to the ICU, intensivists require patient consent and availability of life span–extending treatment options for the malignancy (ie, patients cannot be in palliative care). No patients had advance directives, which were not then recognized by French law.
Data Collection
The ICU physician who made the decision to accept or refuse admission to the ICU collected the following data for each study patient: age, sex, type and stage of the underlying malignancy, chronic health status and comorbidities, reason for considering ICU admission, and circumstances surrounding the decision (day or night and number of physicians involved in the decision). In addition, the Mortality Probability Model (MPM0), the only prognostic score available for patients not yet admitted to the ICU, was used to compute prognostic scores for each patient.32 The MPM0 score is based on physiologic items (coma, respiratory, and heart rates), chronic diseases (renal failure, cirrhosis, and metastatic cancer), admission diagnosis, age, and admission type. Reasons for considering ICU admission were categorized according to the nature of organ failures. "Sepsis" was defined as previously described.33 Poor chronic health status was defined as Knaus scale C (severe limitation of autonomy) or D (bedridden).30
For each patient who was refused ICU admission, the intensivist estimated the patient's expected survival time at the time of triage (< 2 days, 2 to 7 days, 7 to 28 days, more than 28 days, does not know). Patients who were transferred to another ICU because no beds were available in the Saint-Louis ICU were counted as admitted. Patients who were refused initially but admitted later on were counted as refused. Rates for 30-day and 6-month survival was recorded for all patients considered for ICU admission, whether they were finally admitted or not.
Statistical Analysis
Results are reported as medians and quartiles (interquartile ranges) or percentages. To evaluate patient characteristics, categoric variables were compared using the 2 test or Fisher's exact test, as appropriate, and continuous variables were compared using the nonparametric Wilcoxon test or the Kruskal-Wallis test. Status of vital statistics for 30-day and 6-month survival after ICU-proposal was recorded for all the study patients. Actuarial probabilities of survival were estimated using Kaplan-Meier log-rank tests.
To identify associations between patient characteristics and ICU admission, we first performed bivariate logistic regression analyses to identify variables that significantly influenced the likelihood of ICU admission, as measured by the estimated odds ratio (OR) and the 95% CI values. Variables yielding P values .20 in the bivariate analyses were entered into a multiple logistic regression model where ICU admission was the outcome variable of interest. The Hosmer-Lemeshow test was used to check goodness-of-fit of the logistic regression. We noted that omitting each of the selected variables did not induce a significant increase in the likelihood of ICU admission. All tests were two-sided, and P values less than .05 were considered statistically significant. Analyses were done using the SAS 6.12 software package (SAS/STAT User's Guide, Version 6, 1990; SAS Institute, Cary, NC).
RESULTS
Patient Characteristics
During the study period, 206 patients with cancer had one or more organ failures and were evaluated by intensivists at the request of ward oncologists. Figure 1 illustrates the patient flow chart. ICU admission was requested for the following reasons: acute respiratory failure in 120 patients (58%), shock in 60 patients (30%), coma in 43 patients (21%), sepsis in 35 patients (17%), and acute renal failure in 25 patients (12%). After being evaluated by the intensivist, 105 patients (51%) were admitted to the ICU. Of the 101 patients who were not admitted, 54 patients (26.2%) were considered too sick to benefit, and 47 patients (22.8%) were considered too well to benefit. Patients were categorized as too sick to benefit because of an absence of life span–extending treatment options for the malignancy (51 patients) and moribund status (18 patients, including 15 with no life span–extending treatment options for the malignancy). Of the 47 patients considered too well to benefit, 13 were admitted to the ICU later on. Table 1 lists patient characteristics at the time they were evaluated for ICU admission, and Table 2 details the patient groups defined by admission status. One third of the patients had solid tumors and two thirds had hematologic malignancies (mainly acute leukemia and lymphoma). One third of the patients had not yet received cancer treatment or were receiving first-line chemotherapy, and nearly one fifth of the patients (37 patients; 18%) were in remission. Chronic health status was poor in about half of the patients. For 23 patients with poor chronic health status, information was lacking about treatment options, and ICU admission for a trial of intensive care was decided; of the 23 patients, 14 patients survived at least 30 days and 11 patients for at least 6 months.
As listed in Table 1, the ward oncologist in charge of the patient was involved in two thirds of the triage decisions; the rest of the triage decisions were taken on by the intensivist alone. In 31 patients (15%), of whom 11 were admitted and 20 refused, the oncologist and the intensivist disagreed about the triage decision. Of the 20 patients who were admitted to the ICU after an initial refusal, 15 patients had been refused because they were too sick to benefit from treatment in the ICU, and 5 patients had been refused because they were too well to benefit from ICU admission.
Factors Associated With ICU Admission
Table 1 lists the results of the logistic regression with ICU admission as the outcome variable of interest. Three factors were independently associated with ICU admission: the association was positive for remission of the malignancy (OR, 3.37; 95% CI, 1.25 to 9.07) and negative for poor chronic health status (OR, 0.38; 95% CI, 0.16 to 0.74) and solid tumor (OR, 0.43; 95% CI, 0.24 to 0.78).
Survival Rates
As illustrated in Figure 2, survival rates were significantly different across admission categories. The 30-day mortality rate was 45.7% (48 deaths) in patients admitted to the ICU, 74% (40 deaths) in patients considered too sick to benefit from ICU admission, and 21.3% (10 deaths) in patients considered too well to benefit from ICU admission. Of the 13 patients initially considered too well to benefit from ICU admission but who were later admitted, eight patients (61.5%) died, whereas of the 34 patients considered too well to benefit from ICU admission, only two patients (5.9%) died. After 6 months, the mortality rate was 61.8% in patients admitted to the ICU, 83.3% in patients considered too sick to benefit from ICU admission, and 42.2% among patients considered too well to benefit from ICU admission.
Of the 20 patients refused by the ICU, about whom the intensivist and the oncologist disagreed about the triage decision, 14 patients died, including four of the five patients refused because they were considered too well to benefit from ICU admission.
Figure 3 illustrates mortality rates according to expected survival time as estimated by the intensivist who made the triage decision. Median observed survival was 2 days (range, 1 to 3 days) when expected survival was less than 2 days, 5.5 days (range, 2 to –84 days) when expected survival was less than 7 days, 16.5 days (range, 6 to –212 days) when expected survival was less than 28 days, and 161 days (range, 22 to 283 days) when expected survival was more than 28 days (P < .0001).
The Mortality Prediction Model (MPM0) scores for all patients were available at the time of the triage decision. As shown by the receiver-operating characteristic (ROC) curve in Figure 4, the Mortality Prediction Model (MPM0) score was of limited value for predicting hospital mortality (area under the curve, 0.62).
DISCUSSION
Cancer patients may experience acute life-threatening complications consisting of infections related to immunodeficiency induced by the malignancy or the treatment, organ involvement by malignant cells, or treatment toxicities. In this article, we report the results of a 1-year prospective hospital-wide study of 206 cancer patients for whom ICU admission was requested. About half of the patients were not admitted to the ICU. To the best of our knowledge, this is the first study designed to evaluate outcomes in critically ill cancer patients who were denied ICU admission.
ICU admission has been considered inadvisable for cancer patients,23,24 based on studies published in the 1990s that reported extremely low survival rates for cancer patients who were neutropenic, had received bone marrow transplantation, or required life-sustaining treatments.26,29,34-39 At the time, recommendations for triage to the ICU stated that ICU admission of cancer patients was futile.27,29,40 However, a critical reappraisal of these recommendations is in order, as many studies have reported improved survival in critically ill cancer patients over the last few years.1-3,16,41 Moreover, our results suggest that errors in judgment by the intensivist may translate into increased mortality for the patients: (1) 30-day mortality in patients considered too well to benefit from ICU admission was 21.3%; (2) 30-day mortality in patients considered too sick to benefit from ICU admission was 74%, suggesting inaccurate clinical judgment by the intensivists; (3) although expected survival as estimated by the intensivist was roughly correlated with actual overall survival, the broad CIs show that the intensivists' estimates were inaccurate for individual decision making; (4) the Mortality Prediction Model, the only prognostic score available at the time of ICU admission, showed poor calibration for predicting survival.
As expected, survival was strongly associated with admission status. Our finding that 30-day mortality was less than 50% in a population of critically ill cancer patients that included patients needing mechanical ventilation, vasopressors, and/or renal replacement therapy is consistent with earlier data and highlights the recent improvements in survival among carefully selected cancer patients.1-3,16,41 In contrast, survival rates for patients who were not admitted differed from expected survival rates. Of the 54 patients considered too sick to benefit from ICU admission, 14 patients (26%) were still alive 30 days later. This suggests that many of the decisions to refuse ICU admission on the grounds of too-sick-to-benefit status were based on the absence of cancer treatment options, which indicates that hope for cancer control is lost, but not necessarily that the patient is going to die in the very short term. However, of the 47 patients considered too well to benefit from ICU admission, 10 patients died; eight of these 10 patients died after a later admission to the ICU, suggesting that ICU admission might have been inappropriately delayed. This finding should be viewed as possibly identifying an area for improvement. Our data indicate that refusing patients with cancer admission to the ICU because of too-well-to-benefit status should be viewed with great caution and that alternative strategies may deserve evaluation in this population.
Clearly, moribund or chronically bedridden cancer patients and patients who refuse ICU admission should be managed by the wards. However, full ICU management should be offered to patients with newly diagnosed cancer and acute life-threatening cancer-related events such as bulky mediastinal disease, tumor lysis syndrome, pulmonary leukemic infiltration, or leukostasis at the onset of acute leukemia. Admission decisions should be made on a case-by-case basis, somewhere between the two ends of the decision-making spectrum, based on the clinical evaluation and after discussions with the oncologist and patient. In addition, decisions should not be delayed until night-time, when the incidence of ICUs denying admission to patients is higher, the intensivist must make the decision alone, and ICU admission could be lifesaving. When uncertainty or disagreement exists about the criteria for ICU admission, a trial of ICU management should be offered to ensure that no patients are deprived of a chance for recovering from their acute complication. After 3 days of full life-support management (including transfusions, antibiotics, mechanical ventilation, vasopressors, and dialysis), a reduction in the number of organ failures indicates that additional life-sustaining treatment is in order, whereas absence of a response or an increase in the number of organ failures should prompt a discussion of the appropriateness of continuing aggressive treatment.13,31,42
Four arguments support a trial of ICU management, rather than ICU refusal for cancer patients. First, as shown by the current study, criteria for ICU admission are far from infallible and delayed admission is associated with a high mortality rate, which is in keeping with previous reports.43,44 In addition, severity scores computed before or at ICU admission are of limited value for predicting which patients will survive.45,46 Second, recent evidence of increasing survival rates among critically ill cancer patients indicates a need for reappraising the criteria used to predict outcomes in this population. In particular, neutropenia, autologous bone marrow transplantation, and the characteristics of the underlying malignancy45,47,48 now seem to have little influence on the outcomes of ICU management.1,2,16,49 Third, the number of organ failures at ICU admission is correlated with survival,13,45,49 but an even stronger correlation could exist with the evolution of organ failures during the patient's first 3 days in the ICU.13,31 Fourth, failure of 3 days of noninvasive mechanical ventilation is almost consistently followed by death in the ICU,42 suggesting that a period of early noninvasive supportive care in the ICU may give intensivists the time they need to accurately evaluate the prognosis and to identify the appropriate goals of treatment.
Our study has a number of limitations. First, it was conducted at a single institution. Admission policies may vary across hospitals. However, the large number of cancer patients managed in our hospital and the standardized nature of our triage policies25 suggest that our results could be applied to other settings. Second, we did not evaluate the criteria used to request ICU admission. Conceivably, the selection decision could have occurred before the request for ICU admission was made, as oncologists and hematologists might be reluctant to consider ICU admission in the sickest patients. Third, in about one third of the ICU admission decisions made, there was no time for the admitting intensivist to discuss the decision with the oncologist. Moreover, none of the cancer patients had advance directives about ICU admission. Routine discussions with patients about the outcomes of ICU management and their wishes regarding ICU admission should they experience acute life-threatening events might decrease the uncertainty associated with triage decisions. This approach is being evaluated in our hospital. Fourth, in this study, a majority of the cancer patients presented with hematologic malignancies rather than solid tumors. This majority could differ across cohorts of critically ill cancer patients. Fifth, our assumption that ICU admission benefits survival remains theoretical and needs to be evaluated. Last, expanding the cancer-patient population in the ICU would limit ICU access for other patient groups, unless the number of ICU beds is increased. This would not necessarily have to translate into denial of ICU admission in another patient population; instead, the result might be earlier ICU discharge for patients who would otherwise have been kept in the ICU.
In summary, half of the critically ill cancer patients for whom ICU management was considered were denied ICU admission because they had poor chronic health status and/or no cancer treatment options. Among the patients who were denied ICU admission because they were felt to be too well to benefit from admission, one fourth were subsequently admitted, and mortality was high in this subgroup. Because triage criteria are inaccurate, further studies are needed to evaluate the potential benefits of a broad ICU admission policy involving a trial of ICU therapy followed by re-evaluation of the patient. In this policy, the trial should be offered to all patients for whom a benefit from ICU management cannot be convincingly ruled out.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported by grants from the French Society for Critical Care Medicine and the Assistance Publique–Hpitaux de Paris.
Presented in part at the Congress of the French Society of Intensive Care (SRLF), January 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Azoulay E, Recher C, Alberti C, et al: Changing use of intensive care for hematological patients: The example of multiple myeloma. Intensive Care Med 25:1395-1401, 1999
Kress JP, Christenson J, Pohlman AS, et al: Outcomes of critically ill cancer patients in a university hospital setting. Am J Respir Crit Care Med 160:1957-1961, 1999
Staudinger T, Stoiser B, Mullner M, et al: Outcome and prognostic factors in critically ill cancer patients admitted to the intensive care unit. Crit Care Med 28:1322-1328, 2000
Bergh J, Wiklund T, Erikstein B, et al: Tailored fluorouracil, epirubicin, and cyclophosphamide compared with marrow-supported high-dose chemotherapy as adjuvant treatment for high-risk breast cancer: A randomised trial—Scandinavian Breast Group 9401 study. Lancet 356:1384-1991, 2000
Thatcher N, Girling DJ, Hopwood P, et al: Improving survival without reducing quality of life in small-cell lung cancer patients by increasing the dose-intensity of chemotherapy with granulocyte colony-stimulating factor support: Results of a British Medical Research Council Multicenter Randomized Trial—Medical Research Council Lung Cancer Working Party. J Clin Oncol 18:395-404, 2000
Boogaerts MA, Demuynck H: The changing face of stem cell transplantation by the use of recombinant human granulocyte colony stimulating factor. Eur J Cancer 30A:S34-S39, 1994 (suppl 3)
Nachman JB, Sather HN, Sensel MG, et al: Augmented post-induction therapy for children with high-risk acute lymphoblastic leukemia and a slow response to initial therapy. N Engl J Med 338:1663-1671, 1998
Linker C, Damon L, Ries C, et al: Intensified and shortened cyclical chemotherapy for adult acute lymphoblastic leukemia. J Clin Oncol 20:2464-2471, 2002
Kantarjian H, Sawyers C, Hochhaus A, et al: Hematologic and cytogenetic responses to imatinib mesylate in chronic myelogenous leukemia. N Engl J Med 346:645-652, 2002
O'Brien SG, Guilhot F, Larson RA, et al: Imatinib compared with interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med 348:994-1004, 2003
Coiffier B, Lepage E, Briere J, et al: CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med 346:235-242, 2002
Brenner H: Long-term survival rates of cancer patients achieved by the end of the 20th century: A period analysis. Lancet 360:1131-1135, 2002
Larche J, Azoulay E, Fieux F, et al: Improved survival of critically ill cancer patients with septic shock. Intensive Care Med 29:1688-1695, 2003
Mayaud C, Cadranel J: A persistent challenge: The diagnosis of respiratory disease in the non-AIDS immunocompromised host. Thorax 55:511-517, 2000
Gruson D, Hilbert G, Valentino R, et al: Utility of fiberoptic bronchoscopy in neutropenic patients admitted to the intensive care unit with pulmonary infiltrates. Crit Care Med 28:2224-2230, 2000
Azoulay E, Alberti C, Bornstain C, et al: Improved survival in cancer patients requiring mechanical ventilatory support: Impact of noninvasive mechanical ventilatory support. Crit Care Med 29:519-525, 2001
Hilbert G, Gruson D, Vargas F, et al: Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure. N Engl J Med 344:481-487, 2001
Azoulay E, Darmon M, Delclaux C, et al: Deterioration of previous acute lung injury during neutropenia recovery. Crit Care Med 30:781-786, 2002
Azoulay E, Fieux F, Moreau D, et al: Acute monocytic leukemia presenting as acute respiratory failure. Am J Respir Crit Care Med 167:1329-1333, 2003
Azoulay E, Attalah H, Yang K, et al: Exacerbation with granulocyte colony-stimulating factor of prior acute lung injury during neutropenia recovery in rats. Crit Care Med 31:157-165, 2003
Coiffier B, Mounier N, Bologna S, et al: Efficacy and safety of rasburicase (recombinant urate oxidase) for the prevention and treatment of hyperuricemia during induction chemotherapy of aggressive non-Hodgkin's lymphoma: Results of the GRAAL1 (Groupe d'Etude des Lymphomes de l'Adulte Trial on Rasburicase Activity in Adult Lymphoma) study. J Clin Oncol 21:4402-4406, 2003
Gauvin F, Toledano B, Champagne J, et al: Reactive hemophagocytic syndrome presenting as a component of multiple organ dysfunction syndrome. Crit Care Med 28:3341-3345, 2000
Society of Critical Care Medicine: Guidelines for intensive care unit admission, discharge, and triage: Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Crit Care Med 27:633-638, 1999
Society of Critical Care Medicine: Consensus statement on the triage of critically ill patients: Society of Critical Care Medicine Ethics Committee. JAMA 271:1200-1203, 1994
Azoulay E, Pochard F, Chevret S, et al: Compliance with triage to intensive care recommendations. Crit Care Med 29:2132-2136, 2001
Brunet F, Lanore JJ, Dhainaut JF, et al: Is intensive care justified for patients with haematological malignancies Intensive Care Med 16:291-297, 1990
Carlon GC: Admitting cancer patients to the intensive care unit. Crit Care Clin 4:183-191, 1988
Schuster DP, Marion JM: Precedents for meaningful recovery during treatment in a medical intensive care unit: Outcome in patients with hematologic malignancy. Am J Med 75:402-408, 1983
Rubenfeld GD, Crawford SW: Withdrawing life support from mechanically ventilated recipients of bone marrow transplants: A case for evidence-based guidelines. Ann Intern Med 125:625-633, 1996
Knaus WA, Zimmerman JE, Wagner DP, et al: APACHE-acute physiology and chronic health evaluation: A physiologically based classification system. Crit Care Med 9:591-597, 1981
Guiguet M, Blot F, Escudier B, et al: Severity-of-illness scores for neutropenic cancer patients in an intensive care unit: Which is the best predictor Do multiple assessment times improve the predictive value Crit Care Med 26:488-493, 1998
Lemeshow S, Teres D, Klar J, et al: Mortality Probability Models (MPM II) based on an international cohort of intensive care unit patients. JAMA 270:2478-2486, 1993
Alberti C, Brun-Buisson C, Burchardi H, et al: Epidemiology of sepsis and infection in ICU patients from an international multicentre cohort study. Intensive Care Med 28:108-121, 2002
Huaringa AJ, Leyva FJ, Giralt SA, et al: Outcome of bone marrow transplantation patients requiring mechanical ventilation. Crit Care Med 28:1014-1017, 2000
Lloyd-Thomas AR, Dhaliwal HS, Lister TA, et al: Intensive therapy for life-threatening medical complications of haematological malignancy. Intensive Care Med 12:317-324, 1986
Lloyd-Thomas AR, Wright I, Lister TA, et al: Prognosis of patients receiving intensive care for lifethreatening medical complications of haematological malignancy. Br Med J (Clin Res Ed) 296:1025-1029, 1988
Jackson SR, Tweeddale MG, Barnett MJ, et al: Admission of bone marrow transplant recipients to the intensive care unit: Outcome, survival and prognostic factors. Bone Marrow Transplant 21:697-704, 1998
Afessa B, Tefferi A, Hoagland HC, et al: Outcome of recipients of bone marrow transplants who require intensive-care unit support. Mayo Clin Proc 67:117-122, 1992
Crawford SW, Petersen FB: Long-term survival from respiratory failure after marrow transplantation for malignancy. Am Rev Respir Dis 145:510-514, 1992
Schuster DP: Everything that should be done—not everything that can be done. Am Rev Respir Dis 145:508-509, 1992
Darmon M, Azoulay E, Alberti C, et al: Impact of neutropenia duration on short-term mortality in neutropenic critically ill cancer patients. Intensive Care Med 28:1775-1780, 2002
Azoulay E, Thiery G, Chevret S, et al: The prognosis of acute respiratory failure in critically ill cancer patients. Medicine (Baltimore) 83:360-370, 2004
Sprung CL, Geber D, Eidelman LA, et al: Evaluation of triage decisions for intensive care admission. Crit Care Med 27:1073-1079, 1999
Metcalfe MA, Sloggett A, McPherson K: Mortality among appropriately referred patients refused admission to intensive-care units. Lancet 350:7-11, 1997
Blot F, Guiguet M, Nitenberg G, et al: Prognostic factors for neutropenic patients in an intensive care unit: Respective roles of underlying malignancies and acute organ failures. Eur J Cancer 33:1031-1037, 1997
Sculier JP, Paesmans M, Markiewicz E, et al: Scoring systems in cancer patients admitted for an acute complication in a medical intensive care unit. Crit Care Med 28:2786-2792, 2000
Azoulay E, Moreau D, Alberti C, et al: Predictors of short-term mortality in critically ill patients with solid malignancies. Intensive Care Med 26:1817-1823, 2000
Massion PB, Dive AM, Doyen C, et al: Prognosis of hematologic malignancies does not predict intensive care unit mortality. Crit Care Med 30:2260-2270, 2002
Khassawneh BY, White P Jr, Anaissie EJ, et al: Outcome from mechanical ventilation after autologous peripheral blood stem cell transplantation. Chest 121:185-188, 2002(Guillaume Thiery, elie Az)
Paris 7 University
Assistance Publique–Hpitaux de Paris, Paris, France
ABSTRACT
PURPOSE: To evaluate the outcome of cancer patients considered for admission to the intensive care unit (ICU).
PATIENTS AND METHODS: Prospective, one-year hospital-wide study of all cancer and hematology patients, including bone marrow transplantation patients, for whom admission to the ICU was requested.
RESULTS: Of the 206 patients considered for ICU admission, 105 patients (51%) were admitted. Of the 101 patients who were not admitted, 54 patients (26.2%) were considered too sick to benefit, and 47 patients (22.8%) were considered to be too well to benefit from the ICU. Of these 47 patients, 13 patients were admitted later. Survival rates after 30 and 180 days were significantly associated with admission status (P < .0001). Remission of the malignancy (odds ratio [OR], 3.37; 95% CI, 1.25 to 9.07) was independently associated with ICU admission, whereas poor chronic health status (OR, 0.38; 95% CI, 0.16 to 0.74) and solid tumor (OR, 0.43; 95% CI, 0.24 to 0.78) were associated with ICU refusal. In admitted patients, 30-day and 6-month survival rates were 54.3% and 32.4%, respectively. Of the patients considered too sick to benefit from ICU admission, 26% were alive on day 30 and 16.7% on day 180. Among patients considered too well to benefit, the 30-day survival rate was a worrisome 78.7%. Calibration of the Mortality Probability Model (the only score available at triage) was of limited value for predicting 30-day survival (area under the curve, 0.62).
CONCLUSION: Both the excess mortality in too-well patients later admitted to the ICU and the relatively good survival in too-sick patients suggest the need for a broader admission policy.
INTRODUCTION
Over the last decade, survival rates in critically ill cancer patients have improved dramatically.1-3 Three factors have contributed to this welcome trend. First, advances in the treatment of solid tumors4,5 and hematologic malignancies6-11 produced a 20% decrease in overall mortality in cancer patients from 1978 to 1998.12 Second, earlier admission to the intensive care unit (ICU) has resulted in better survival rates for critically ill cancer patients.13 Interventions that have been used successfully include bronchoscopy and bronchoalveolar lavage14,15 and noninvasive mechanical ventilation16,17 in patients with pulmonary infiltrates, aggressive management of septic shock,13 and investigations of specific organ failures.18-22 Third, progress has been made in selecting patients likely to benefit from ICU admission.1,16
Because the decision to recommend ICU admission is based on complex criteria, and to ensure appropriate utilization of ICU resources without depriving critically ill patients of a chance to recover, the North American and European Societies of Critical Care Medicine have developed recommendations for ICU admission, triage, and discharge.23,24 In a prospective longitudinal study performed in 26 ICUs in 2001,25 we evaluated factors associated with intensivists' refusal to admit patients to the ICU, for whom other physicians had requested ICU admission. We found that cancer patients were at high risk to be refused by the ICU, in accordance with current recommendations and with numerous articles discouraging ICU admission26,27 or prolonged intensive care28,29 for cancer patients. With good reason, oncologists and intensivists reserve ICU admission for select cancer patients with potentially reversible medical conditions and a "reasonable prospect of substantial recovery."23,24 However, the procedures used to select cancer patients for ICU admission and the outcomes associated with these procedures have not been evaluated.
The objective of this study was to determine the outcomes of cancer patients for whom ICU admission was requested. To achieve this objective, we performed a prospective longitudinal study over a 1-year period in a teaching hospital where about half the beds were in oncology and hematology wards. We included all patients for whom admission to the hospital's medical ICU was requested, regardless of whether admission was granted or not.
PATIENTS AND METHODS
The institutional review board of the French Society for Critical Care approved the study in June 2001. No information allowing identification of the patients was recorded in this prospective observational cohort study. We included all patients with solid tumors or hematologic malignancies for whom admission to the medical ICU of the Saint Louis Teaching Hospital (Paris, France) was requested by ward physicians in the same hospital between March 1, 2002, and March 1, 2003.
Description of the Saint-Louis Teaching Hospital
The Saint-Louis Teaching Hospital is a 650-bed university hospital with 330 beds in the hematology and oncology wards, where about 600 patients with newly diagnosed cancer are admitted each year. There are five hematology wards (acute leukemia, non-Hodgkin's lymphoma, myeloma, autologous stem-cell transplantation, and allogeneic stem cell or bone marrow transplantation) and three oncology wards (general oncology, lung cancer, and radiation therapy). Hospital mortality rate in the eight oncology wards is approximately 5%. The medical ICU is a closed 12-bed unit that admits 550 to 600 patients per year, including about 130 cancer patients, most of whom have hematologic malignancies. In accordance with legislation for ICUs in Europe, the daytime staff consists of five critical care specialists (intensivists) and three residents, and the night-time staff consists of one senior intensivist and one resident. In addition, a senior hematologist and a hematology resident are on call 24 hours a day for the hematology and solid-tumor wards.
Criteria for ICU Admission Used Routinely at the Saint-Louis Teaching Hospital
ICU admission is considered when a cancer patient has at least one organ failure. First, the patient is evaluated by a senior intensivist and the ward oncologist and hematologist in charge of the patient. The final decision regarding ICU admission is then made by the senior intensivist, who records the decision in the patient's chart as admission, refusal because the patient is too sick to benefit, or refusal because the patient is too well to benefit. When no beds are available, the patient is always transferred to another ICU. To admit a patient to the ICU, intensivists require patient consent and availability of life span–extending treatment options for the malignancy (ie, patients cannot be in palliative care). No patients had advance directives, which were not then recognized by French law.
Data Collection
The ICU physician who made the decision to accept or refuse admission to the ICU collected the following data for each study patient: age, sex, type and stage of the underlying malignancy, chronic health status and comorbidities, reason for considering ICU admission, and circumstances surrounding the decision (day or night and number of physicians involved in the decision). In addition, the Mortality Probability Model (MPM0), the only prognostic score available for patients not yet admitted to the ICU, was used to compute prognostic scores for each patient.32 The MPM0 score is based on physiologic items (coma, respiratory, and heart rates), chronic diseases (renal failure, cirrhosis, and metastatic cancer), admission diagnosis, age, and admission type. Reasons for considering ICU admission were categorized according to the nature of organ failures. "Sepsis" was defined as previously described.33 Poor chronic health status was defined as Knaus scale C (severe limitation of autonomy) or D (bedridden).30
For each patient who was refused ICU admission, the intensivist estimated the patient's expected survival time at the time of triage (< 2 days, 2 to 7 days, 7 to 28 days, more than 28 days, does not know). Patients who were transferred to another ICU because no beds were available in the Saint-Louis ICU were counted as admitted. Patients who were refused initially but admitted later on were counted as refused. Rates for 30-day and 6-month survival was recorded for all patients considered for ICU admission, whether they were finally admitted or not.
Statistical Analysis
Results are reported as medians and quartiles (interquartile ranges) or percentages. To evaluate patient characteristics, categoric variables were compared using the 2 test or Fisher's exact test, as appropriate, and continuous variables were compared using the nonparametric Wilcoxon test or the Kruskal-Wallis test. Status of vital statistics for 30-day and 6-month survival after ICU-proposal was recorded for all the study patients. Actuarial probabilities of survival were estimated using Kaplan-Meier log-rank tests.
To identify associations between patient characteristics and ICU admission, we first performed bivariate logistic regression analyses to identify variables that significantly influenced the likelihood of ICU admission, as measured by the estimated odds ratio (OR) and the 95% CI values. Variables yielding P values .20 in the bivariate analyses were entered into a multiple logistic regression model where ICU admission was the outcome variable of interest. The Hosmer-Lemeshow test was used to check goodness-of-fit of the logistic regression. We noted that omitting each of the selected variables did not induce a significant increase in the likelihood of ICU admission. All tests were two-sided, and P values less than .05 were considered statistically significant. Analyses were done using the SAS 6.12 software package (SAS/STAT User's Guide, Version 6, 1990; SAS Institute, Cary, NC).
RESULTS
Patient Characteristics
During the study period, 206 patients with cancer had one or more organ failures and were evaluated by intensivists at the request of ward oncologists. Figure 1 illustrates the patient flow chart. ICU admission was requested for the following reasons: acute respiratory failure in 120 patients (58%), shock in 60 patients (30%), coma in 43 patients (21%), sepsis in 35 patients (17%), and acute renal failure in 25 patients (12%). After being evaluated by the intensivist, 105 patients (51%) were admitted to the ICU. Of the 101 patients who were not admitted, 54 patients (26.2%) were considered too sick to benefit, and 47 patients (22.8%) were considered too well to benefit. Patients were categorized as too sick to benefit because of an absence of life span–extending treatment options for the malignancy (51 patients) and moribund status (18 patients, including 15 with no life span–extending treatment options for the malignancy). Of the 47 patients considered too well to benefit, 13 were admitted to the ICU later on. Table 1 lists patient characteristics at the time they were evaluated for ICU admission, and Table 2 details the patient groups defined by admission status. One third of the patients had solid tumors and two thirds had hematologic malignancies (mainly acute leukemia and lymphoma). One third of the patients had not yet received cancer treatment or were receiving first-line chemotherapy, and nearly one fifth of the patients (37 patients; 18%) were in remission. Chronic health status was poor in about half of the patients. For 23 patients with poor chronic health status, information was lacking about treatment options, and ICU admission for a trial of intensive care was decided; of the 23 patients, 14 patients survived at least 30 days and 11 patients for at least 6 months.
As listed in Table 1, the ward oncologist in charge of the patient was involved in two thirds of the triage decisions; the rest of the triage decisions were taken on by the intensivist alone. In 31 patients (15%), of whom 11 were admitted and 20 refused, the oncologist and the intensivist disagreed about the triage decision. Of the 20 patients who were admitted to the ICU after an initial refusal, 15 patients had been refused because they were too sick to benefit from treatment in the ICU, and 5 patients had been refused because they were too well to benefit from ICU admission.
Factors Associated With ICU Admission
Table 1 lists the results of the logistic regression with ICU admission as the outcome variable of interest. Three factors were independently associated with ICU admission: the association was positive for remission of the malignancy (OR, 3.37; 95% CI, 1.25 to 9.07) and negative for poor chronic health status (OR, 0.38; 95% CI, 0.16 to 0.74) and solid tumor (OR, 0.43; 95% CI, 0.24 to 0.78).
Survival Rates
As illustrated in Figure 2, survival rates were significantly different across admission categories. The 30-day mortality rate was 45.7% (48 deaths) in patients admitted to the ICU, 74% (40 deaths) in patients considered too sick to benefit from ICU admission, and 21.3% (10 deaths) in patients considered too well to benefit from ICU admission. Of the 13 patients initially considered too well to benefit from ICU admission but who were later admitted, eight patients (61.5%) died, whereas of the 34 patients considered too well to benefit from ICU admission, only two patients (5.9%) died. After 6 months, the mortality rate was 61.8% in patients admitted to the ICU, 83.3% in patients considered too sick to benefit from ICU admission, and 42.2% among patients considered too well to benefit from ICU admission.
Of the 20 patients refused by the ICU, about whom the intensivist and the oncologist disagreed about the triage decision, 14 patients died, including four of the five patients refused because they were considered too well to benefit from ICU admission.
Figure 3 illustrates mortality rates according to expected survival time as estimated by the intensivist who made the triage decision. Median observed survival was 2 days (range, 1 to 3 days) when expected survival was less than 2 days, 5.5 days (range, 2 to –84 days) when expected survival was less than 7 days, 16.5 days (range, 6 to –212 days) when expected survival was less than 28 days, and 161 days (range, 22 to 283 days) when expected survival was more than 28 days (P < .0001).
The Mortality Prediction Model (MPM0) scores for all patients were available at the time of the triage decision. As shown by the receiver-operating characteristic (ROC) curve in Figure 4, the Mortality Prediction Model (MPM0) score was of limited value for predicting hospital mortality (area under the curve, 0.62).
DISCUSSION
Cancer patients may experience acute life-threatening complications consisting of infections related to immunodeficiency induced by the malignancy or the treatment, organ involvement by malignant cells, or treatment toxicities. In this article, we report the results of a 1-year prospective hospital-wide study of 206 cancer patients for whom ICU admission was requested. About half of the patients were not admitted to the ICU. To the best of our knowledge, this is the first study designed to evaluate outcomes in critically ill cancer patients who were denied ICU admission.
ICU admission has been considered inadvisable for cancer patients,23,24 based on studies published in the 1990s that reported extremely low survival rates for cancer patients who were neutropenic, had received bone marrow transplantation, or required life-sustaining treatments.26,29,34-39 At the time, recommendations for triage to the ICU stated that ICU admission of cancer patients was futile.27,29,40 However, a critical reappraisal of these recommendations is in order, as many studies have reported improved survival in critically ill cancer patients over the last few years.1-3,16,41 Moreover, our results suggest that errors in judgment by the intensivist may translate into increased mortality for the patients: (1) 30-day mortality in patients considered too well to benefit from ICU admission was 21.3%; (2) 30-day mortality in patients considered too sick to benefit from ICU admission was 74%, suggesting inaccurate clinical judgment by the intensivists; (3) although expected survival as estimated by the intensivist was roughly correlated with actual overall survival, the broad CIs show that the intensivists' estimates were inaccurate for individual decision making; (4) the Mortality Prediction Model, the only prognostic score available at the time of ICU admission, showed poor calibration for predicting survival.
As expected, survival was strongly associated with admission status. Our finding that 30-day mortality was less than 50% in a population of critically ill cancer patients that included patients needing mechanical ventilation, vasopressors, and/or renal replacement therapy is consistent with earlier data and highlights the recent improvements in survival among carefully selected cancer patients.1-3,16,41 In contrast, survival rates for patients who were not admitted differed from expected survival rates. Of the 54 patients considered too sick to benefit from ICU admission, 14 patients (26%) were still alive 30 days later. This suggests that many of the decisions to refuse ICU admission on the grounds of too-sick-to-benefit status were based on the absence of cancer treatment options, which indicates that hope for cancer control is lost, but not necessarily that the patient is going to die in the very short term. However, of the 47 patients considered too well to benefit from ICU admission, 10 patients died; eight of these 10 patients died after a later admission to the ICU, suggesting that ICU admission might have been inappropriately delayed. This finding should be viewed as possibly identifying an area for improvement. Our data indicate that refusing patients with cancer admission to the ICU because of too-well-to-benefit status should be viewed with great caution and that alternative strategies may deserve evaluation in this population.
Clearly, moribund or chronically bedridden cancer patients and patients who refuse ICU admission should be managed by the wards. However, full ICU management should be offered to patients with newly diagnosed cancer and acute life-threatening cancer-related events such as bulky mediastinal disease, tumor lysis syndrome, pulmonary leukemic infiltration, or leukostasis at the onset of acute leukemia. Admission decisions should be made on a case-by-case basis, somewhere between the two ends of the decision-making spectrum, based on the clinical evaluation and after discussions with the oncologist and patient. In addition, decisions should not be delayed until night-time, when the incidence of ICUs denying admission to patients is higher, the intensivist must make the decision alone, and ICU admission could be lifesaving. When uncertainty or disagreement exists about the criteria for ICU admission, a trial of ICU management should be offered to ensure that no patients are deprived of a chance for recovering from their acute complication. After 3 days of full life-support management (including transfusions, antibiotics, mechanical ventilation, vasopressors, and dialysis), a reduction in the number of organ failures indicates that additional life-sustaining treatment is in order, whereas absence of a response or an increase in the number of organ failures should prompt a discussion of the appropriateness of continuing aggressive treatment.13,31,42
Four arguments support a trial of ICU management, rather than ICU refusal for cancer patients. First, as shown by the current study, criteria for ICU admission are far from infallible and delayed admission is associated with a high mortality rate, which is in keeping with previous reports.43,44 In addition, severity scores computed before or at ICU admission are of limited value for predicting which patients will survive.45,46 Second, recent evidence of increasing survival rates among critically ill cancer patients indicates a need for reappraising the criteria used to predict outcomes in this population. In particular, neutropenia, autologous bone marrow transplantation, and the characteristics of the underlying malignancy45,47,48 now seem to have little influence on the outcomes of ICU management.1,2,16,49 Third, the number of organ failures at ICU admission is correlated with survival,13,45,49 but an even stronger correlation could exist with the evolution of organ failures during the patient's first 3 days in the ICU.13,31 Fourth, failure of 3 days of noninvasive mechanical ventilation is almost consistently followed by death in the ICU,42 suggesting that a period of early noninvasive supportive care in the ICU may give intensivists the time they need to accurately evaluate the prognosis and to identify the appropriate goals of treatment.
Our study has a number of limitations. First, it was conducted at a single institution. Admission policies may vary across hospitals. However, the large number of cancer patients managed in our hospital and the standardized nature of our triage policies25 suggest that our results could be applied to other settings. Second, we did not evaluate the criteria used to request ICU admission. Conceivably, the selection decision could have occurred before the request for ICU admission was made, as oncologists and hematologists might be reluctant to consider ICU admission in the sickest patients. Third, in about one third of the ICU admission decisions made, there was no time for the admitting intensivist to discuss the decision with the oncologist. Moreover, none of the cancer patients had advance directives about ICU admission. Routine discussions with patients about the outcomes of ICU management and their wishes regarding ICU admission should they experience acute life-threatening events might decrease the uncertainty associated with triage decisions. This approach is being evaluated in our hospital. Fourth, in this study, a majority of the cancer patients presented with hematologic malignancies rather than solid tumors. This majority could differ across cohorts of critically ill cancer patients. Fifth, our assumption that ICU admission benefits survival remains theoretical and needs to be evaluated. Last, expanding the cancer-patient population in the ICU would limit ICU access for other patient groups, unless the number of ICU beds is increased. This would not necessarily have to translate into denial of ICU admission in another patient population; instead, the result might be earlier ICU discharge for patients who would otherwise have been kept in the ICU.
In summary, half of the critically ill cancer patients for whom ICU management was considered were denied ICU admission because they had poor chronic health status and/or no cancer treatment options. Among the patients who were denied ICU admission because they were felt to be too well to benefit from admission, one fourth were subsequently admitted, and mortality was high in this subgroup. Because triage criteria are inaccurate, further studies are needed to evaluate the potential benefits of a broad ICU admission policy involving a trial of ICU therapy followed by re-evaluation of the patient. In this policy, the trial should be offered to all patients for whom a benefit from ICU management cannot be convincingly ruled out.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported by grants from the French Society for Critical Care Medicine and the Assistance Publique–Hpitaux de Paris.
Presented in part at the Congress of the French Society of Intensive Care (SRLF), January 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Azoulay E, Recher C, Alberti C, et al: Changing use of intensive care for hematological patients: The example of multiple myeloma. Intensive Care Med 25:1395-1401, 1999
Kress JP, Christenson J, Pohlman AS, et al: Outcomes of critically ill cancer patients in a university hospital setting. Am J Respir Crit Care Med 160:1957-1961, 1999
Staudinger T, Stoiser B, Mullner M, et al: Outcome and prognostic factors in critically ill cancer patients admitted to the intensive care unit. Crit Care Med 28:1322-1328, 2000
Bergh J, Wiklund T, Erikstein B, et al: Tailored fluorouracil, epirubicin, and cyclophosphamide compared with marrow-supported high-dose chemotherapy as adjuvant treatment for high-risk breast cancer: A randomised trial—Scandinavian Breast Group 9401 study. Lancet 356:1384-1991, 2000
Thatcher N, Girling DJ, Hopwood P, et al: Improving survival without reducing quality of life in small-cell lung cancer patients by increasing the dose-intensity of chemotherapy with granulocyte colony-stimulating factor support: Results of a British Medical Research Council Multicenter Randomized Trial—Medical Research Council Lung Cancer Working Party. J Clin Oncol 18:395-404, 2000
Boogaerts MA, Demuynck H: The changing face of stem cell transplantation by the use of recombinant human granulocyte colony stimulating factor. Eur J Cancer 30A:S34-S39, 1994 (suppl 3)
Nachman JB, Sather HN, Sensel MG, et al: Augmented post-induction therapy for children with high-risk acute lymphoblastic leukemia and a slow response to initial therapy. N Engl J Med 338:1663-1671, 1998
Linker C, Damon L, Ries C, et al: Intensified and shortened cyclical chemotherapy for adult acute lymphoblastic leukemia. J Clin Oncol 20:2464-2471, 2002
Kantarjian H, Sawyers C, Hochhaus A, et al: Hematologic and cytogenetic responses to imatinib mesylate in chronic myelogenous leukemia. N Engl J Med 346:645-652, 2002
O'Brien SG, Guilhot F, Larson RA, et al: Imatinib compared with interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med 348:994-1004, 2003
Coiffier B, Lepage E, Briere J, et al: CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma. N Engl J Med 346:235-242, 2002
Brenner H: Long-term survival rates of cancer patients achieved by the end of the 20th century: A period analysis. Lancet 360:1131-1135, 2002
Larche J, Azoulay E, Fieux F, et al: Improved survival of critically ill cancer patients with septic shock. Intensive Care Med 29:1688-1695, 2003
Mayaud C, Cadranel J: A persistent challenge: The diagnosis of respiratory disease in the non-AIDS immunocompromised host. Thorax 55:511-517, 2000
Gruson D, Hilbert G, Valentino R, et al: Utility of fiberoptic bronchoscopy in neutropenic patients admitted to the intensive care unit with pulmonary infiltrates. Crit Care Med 28:2224-2230, 2000
Azoulay E, Alberti C, Bornstain C, et al: Improved survival in cancer patients requiring mechanical ventilatory support: Impact of noninvasive mechanical ventilatory support. Crit Care Med 29:519-525, 2001
Hilbert G, Gruson D, Vargas F, et al: Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure. N Engl J Med 344:481-487, 2001
Azoulay E, Darmon M, Delclaux C, et al: Deterioration of previous acute lung injury during neutropenia recovery. Crit Care Med 30:781-786, 2002
Azoulay E, Fieux F, Moreau D, et al: Acute monocytic leukemia presenting as acute respiratory failure. Am J Respir Crit Care Med 167:1329-1333, 2003
Azoulay E, Attalah H, Yang K, et al: Exacerbation with granulocyte colony-stimulating factor of prior acute lung injury during neutropenia recovery in rats. Crit Care Med 31:157-165, 2003
Coiffier B, Mounier N, Bologna S, et al: Efficacy and safety of rasburicase (recombinant urate oxidase) for the prevention and treatment of hyperuricemia during induction chemotherapy of aggressive non-Hodgkin's lymphoma: Results of the GRAAL1 (Groupe d'Etude des Lymphomes de l'Adulte Trial on Rasburicase Activity in Adult Lymphoma) study. J Clin Oncol 21:4402-4406, 2003
Gauvin F, Toledano B, Champagne J, et al: Reactive hemophagocytic syndrome presenting as a component of multiple organ dysfunction syndrome. Crit Care Med 28:3341-3345, 2000
Society of Critical Care Medicine: Guidelines for intensive care unit admission, discharge, and triage: Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Crit Care Med 27:633-638, 1999
Society of Critical Care Medicine: Consensus statement on the triage of critically ill patients: Society of Critical Care Medicine Ethics Committee. JAMA 271:1200-1203, 1994
Azoulay E, Pochard F, Chevret S, et al: Compliance with triage to intensive care recommendations. Crit Care Med 29:2132-2136, 2001
Brunet F, Lanore JJ, Dhainaut JF, et al: Is intensive care justified for patients with haematological malignancies Intensive Care Med 16:291-297, 1990
Carlon GC: Admitting cancer patients to the intensive care unit. Crit Care Clin 4:183-191, 1988
Schuster DP, Marion JM: Precedents for meaningful recovery during treatment in a medical intensive care unit: Outcome in patients with hematologic malignancy. Am J Med 75:402-408, 1983
Rubenfeld GD, Crawford SW: Withdrawing life support from mechanically ventilated recipients of bone marrow transplants: A case for evidence-based guidelines. Ann Intern Med 125:625-633, 1996
Knaus WA, Zimmerman JE, Wagner DP, et al: APACHE-acute physiology and chronic health evaluation: A physiologically based classification system. Crit Care Med 9:591-597, 1981
Guiguet M, Blot F, Escudier B, et al: Severity-of-illness scores for neutropenic cancer patients in an intensive care unit: Which is the best predictor Do multiple assessment times improve the predictive value Crit Care Med 26:488-493, 1998
Lemeshow S, Teres D, Klar J, et al: Mortality Probability Models (MPM II) based on an international cohort of intensive care unit patients. JAMA 270:2478-2486, 1993
Alberti C, Brun-Buisson C, Burchardi H, et al: Epidemiology of sepsis and infection in ICU patients from an international multicentre cohort study. Intensive Care Med 28:108-121, 2002
Huaringa AJ, Leyva FJ, Giralt SA, et al: Outcome of bone marrow transplantation patients requiring mechanical ventilation. Crit Care Med 28:1014-1017, 2000
Lloyd-Thomas AR, Dhaliwal HS, Lister TA, et al: Intensive therapy for life-threatening medical complications of haematological malignancy. Intensive Care Med 12:317-324, 1986
Lloyd-Thomas AR, Wright I, Lister TA, et al: Prognosis of patients receiving intensive care for lifethreatening medical complications of haematological malignancy. Br Med J (Clin Res Ed) 296:1025-1029, 1988
Jackson SR, Tweeddale MG, Barnett MJ, et al: Admission of bone marrow transplant recipients to the intensive care unit: Outcome, survival and prognostic factors. Bone Marrow Transplant 21:697-704, 1998
Afessa B, Tefferi A, Hoagland HC, et al: Outcome of recipients of bone marrow transplants who require intensive-care unit support. Mayo Clin Proc 67:117-122, 1992
Crawford SW, Petersen FB: Long-term survival from respiratory failure after marrow transplantation for malignancy. Am Rev Respir Dis 145:510-514, 1992
Schuster DP: Everything that should be done—not everything that can be done. Am Rev Respir Dis 145:508-509, 1992
Darmon M, Azoulay E, Alberti C, et al: Impact of neutropenia duration on short-term mortality in neutropenic critically ill cancer patients. Intensive Care Med 28:1775-1780, 2002
Azoulay E, Thiery G, Chevret S, et al: The prognosis of acute respiratory failure in critically ill cancer patients. Medicine (Baltimore) 83:360-370, 2004
Sprung CL, Geber D, Eidelman LA, et al: Evaluation of triage decisions for intensive care admission. Crit Care Med 27:1073-1079, 1999
Metcalfe MA, Sloggett A, McPherson K: Mortality among appropriately referred patients refused admission to intensive-care units. Lancet 350:7-11, 1997
Blot F, Guiguet M, Nitenberg G, et al: Prognostic factors for neutropenic patients in an intensive care unit: Respective roles of underlying malignancies and acute organ failures. Eur J Cancer 33:1031-1037, 1997
Sculier JP, Paesmans M, Markiewicz E, et al: Scoring systems in cancer patients admitted for an acute complication in a medical intensive care unit. Crit Care Med 28:2786-2792, 2000
Azoulay E, Moreau D, Alberti C, et al: Predictors of short-term mortality in critically ill patients with solid malignancies. Intensive Care Med 26:1817-1823, 2000
Massion PB, Dive AM, Doyen C, et al: Prognosis of hematologic malignancies does not predict intensive care unit mortality. Crit Care Med 30:2260-2270, 2002
Khassawneh BY, White P Jr, Anaissie EJ, et al: Outcome from mechanical ventilation after autologous peripheral blood stem cell transplantation. Chest 121:185-188, 2002(Guillaume Thiery, elie Az)