Prognostic factors associated with mortality in patients with cancer and febrile neutropenia. Factores pronósticos relacionados con la mortalidad en pacientes con cáncer y neutropenia febril.

Abstract

Introduction: Febrile neutropenia (FN) patients represent a heterogeneous population having a variable risk of developing complications, including death. Objectives: Describing clinical characteristics and outcomes in a cohort of FN patients, as well as the variables influencing mortality. Materials and methods: An ambidirectional study was made to include information concerning 270 patients and 412 episodes of FN which was collected from the 1st May 2001 to the 31st March 2008 at the Instituto Nacional de Cancerología (INC) in Bogotá, Colombia. All the patients included in the study had confirmed neoplasia, were over 18 years old and had been diagnosed with a FN. Demographic, microbiological and clinical variables concerning the patients’ treatment and final outcome were evaluated. The variables modifying overall survival were submitted to univariate and multivariate analysis. Results: The patients’ average age was 41.6 years; bacteremia incidence was 26.5%, microbiological findings being most frequent in Gram-negative germ haemocultures. The FN mortality rate was 7.3 per 100,000 patients per day and univariate analysis of variables significantly associated with mortality revealed the following hazard ratios (HR): mental state (HR 13.34, 6.34-32.82 95%CI), hypotension (HR 5.4, 2.77-10.83 95%CI), renal failure (HR 5.40, 2.40-12.12 95%CI), dehydration (HR 6.67, 3.40-13.08 95%CI), respiratory failure (HR 9.82, 8.48-20.01 95%CI) and fungemia (HR 4.88,4.16-16.30 95%CI). Both univariate and multivariate analysis rejected bacteremia being associated with mortality. Mortality was only modified by respiratory failure in multivariate analysis (HR 4.23, 1.87-9.59 95%CI), compared to altered mental state (HR 3.96, 1.61-9.75 95%CI) and dehydration (HR 2.49, 1.10-5.65 95%CI). Conclusions: The mortality rate due to FN was comparable to that reported in the medical literature. Bacteremia did not modify mortality; however, it did modify other factors which could support identifying high risk patients.