Case-Volume and Mortality in Hematological Patients with Acute Respiratory Failure
L. Lecuyer 1,
S. Chevret 1,
B. Guidet 2,
P. Aegerter 2,
P. Martel 2,
B. Schlemmer 1,
E. Azoulay 1
1 Intensive Care Unit and Biostatistics Dept, Saint-Louis Teaching Hospital; Intensive Care Unit, Saint-Antoine Teaching Hospital; Biostatistics Dept, Ambroise Paré Teaching Hospital, Assistance Publique Hôpitaux de Paris; University Paris 7; and U717 Inserm, Paris, France
2 Intensive Care Unit and Biostatistics Dept, Saint-Louis Teaching Hospital; Intensive Care Unit, Saint-Antoine Teaching Hospital; Biostatistics Dept, Ambroise Paré Teaching Hospital, Assistance Publique Hôpitaux de Paris
 |
Abstract |
|---|
High case-volume is associated with improved survival in medical and surgical conditions. We sought to determine whether ICU case-volume was associated with survival of critically ill patients with hematological malignancies and acute respiratory failure (ARF). We used a regional database containing data from 1,753 hematological patients with ARF admitted to 28 medical ICUs from 1997 through 2004. Multivariate analysis using mixed models was performed to adjust for severity of illness and other confounding factors, including a propensity score that incorporates differences between ICUs with different case-volumes. The three case-volume tertiles were: low volume (<12 admissions·year-1), intermediate volume (12–30·year-1), and high volume (>30·year-1). In univariate analyses, ICU case-volume was not associated with ICU mortality. After adjusting on prognostic factors for ICU mortality and the propensity score, patients in high-volume ICUs had lower mortality than other patients. A case-volume increase of one admission per year led to a mortality reduction with an odds ratio of 0.98 [0.97–0.99] (P=0.002). Mortality was independently associated with severity of organ dysfunction. In ICUs admitting larger numbers of critically ill hematological patients with ARF, mortality was lower than in other ICUs. The mechanisms of the relationship between volume and outcome among hematological patients with ARF deserve additional studies.
