INTRODUCTION
Infections are one of the main causes of death in neutropenic patients with onco-hematologic diseases.
Bacteremia is found in 10-25% of the febrile neutropenia (FN) episodes, especially in those with profound and sustained neutropenia (absolute neutrophil count below 100/uLfor more than 7 days), with high mortality rates per episode.
Gram-negative bacteria are the most frequent isolation, with increasing resistance patterns.
Local epidemiological bacterial isolates and resistance patterns are crucially important in determining first-choice empirical therapy, which determines the prognostic of the patients.
OBJECTIVES
Describe the bacterial isolates in blood cultures and its resistance patterns.
Identify if the initial antibiotic therapy was appropriate regarding the bacteria isolations.
Evaluate the application and compliance with the Hospital de Clinicas protocol for initial antibiotic therapy of FN in onco-hematologic diseases.
METHODS
Retrospective observational study held in Hospital de Clínicas Manuel Quintela Montevideo – Uruguay from August 1st 2011 to December 31st 2015. Bacteremia of first episode of FN in high risk hemato-oncologic patients, older than 18 years old, were analyzed.
RESULTS
A total of 200 episodes of neutropenia were observed, 82/200 (41%) FN and 20/82 (24%) bacteraemias in 19 patients.
Mean age was 37 years (range 18-78), females 12/19 (60%). Median neutropenia duration was 26 days, and profound neutropenia 14 days.
Regarding antibacterial prophylaxis, 95% received it.
The primary diseases were acute myeloid leukemia 75%, (50% in induction phase), acute lymphoblastic leukemia (10%) and non-Hodgkin's lymphoma and Hodgkin lymphoma 15%.
In 50% of the episodes of FN the origin was not found, in the rest: 15% was cutaneous, 10% pulmonary, 10% abdominal, 10% catheter associated infection and urinary 5%.
95 % were late bacteraemias with a predominance of Gram negative bacteria, 65% of them multi resistant.
More than one bacteria were isolated in 3/20 positive blood cultures.
The isolations were the following: 23 bacteria: 20 Gram-negative bacilli y 3 Gram-positive coccus. E coli 11, 4 of them produced extended spectrum b-lactamase (ESBL), K. pneumoniae ESBL (4), E. cloacae (3), S. maltophilia (1), P. aeruginosa (1), methicillin-resistant Staphylococcus aureus (MRSA) (1), S. epidermidis (1), S. mitis (1).
Regarding the application of the protocol for empirical antibiotic treatment we found that there was compliance with it in 55 % (11/20) and there was not in 45 % (9/20).
The empirical antibiotic treatment was appropriate in 65% (13/20) vs. when the protocol was followed the antibiotic choose was appropriate in 95% of the episodes, on the contrary when it was not followed the treatment was appropriate in only 25% of the episodes.
The crude mortality was 6/20 (30%) (4 acute myeloid leukemia in induction phase and 2 non-Hodgkin's lymphoma in second line therapy).
CONCLUSIONS
The majority of bacteraemia were late and the majority by Gram-negative bacteria, 65% of them multi resistant. All of them were susceptible to carbapenemes.
There was a compliance with the initial antibiotic empirical therapy protocol in 55% of the episodes, and in this group the treatment was appropriate in 95%.
The application of an initial antibiotic treatment protocol is very important and it is necessary to improve the compliance with it in the unit.
