TY - JOUR
T1 - Bacteremia in critically ill immunocompromised patients with acute hypoxic respiratory failure
T2 - A post-hoc analysis of a prospective multicenter multinational cohort
AU - for the Efraim investigators and the Nine-I study group
AU - Van de Louw, Andry
AU - Rello, Jordi
AU - Martin-Loeches, Ignacio
AU - Mokart, Djamel
AU - Metaxa, Victoria
AU - Benoit, Dominique
AU - Barratt-Due, Andreas
AU - Soares, Marcio
AU - Pickkers, Peter
AU - Antonelli, Massimo
AU - Demoule, Alexandre
AU - Schellongowski, Peter
AU - Kouatchet, Achille
AU - Mehta, Sangeeta
AU - Balik, Martin
AU - Bauer, Philippe R.
AU - Lemiale, Virginie
AU - Walter, Vonn
AU - Azoulay, Elie
N1 - Publisher Copyright:
© 2021 Elsevier Inc.
PY - 2021/8
Y1 - 2021/8
N2 - Purpose: The characteristics and impact of bacteremia have not been widely investigated in immunocompromised patients with acute respiratory failure (ARF). Methods: We performed a secondary analysis of a prospective cohort of immunocompromised patients with ARF (EFRAIM study). After exclusion of blood cultures positive for coagulase negative Staphylococci, we compared patients with (n = 236) and without (n = 1127) bacteremia. Results: The incidence of bacteremia was 17%. Bacterial pneumonia and extra-pulmonary ARDS were the main causes of ARF in bacteremic patients. Bacteremia involved gram negative rods (48%), gram positive cocci (40%) or were polymicrobial (10%). Bacteremic patients had more hematological malignancy, higher SOFA scores and increased organ support within 7 days. Bacteremia was associated with higher crude ICU mortality (40% versus 32%, p = 0.02), but neither hospital (49% versus 44%, p = 0.17) nor 90-day mortality (60% versus 56%, p = 0.25) were different from non-bacteremic patients. After propensity score matching based on baseline characteristics, the difference in ICU mortality lost statistical significance (p = 0.06), including in a sensitivity analysis restricted to patients with pneumonia. Conclusions: We analyzed a large population of immunocompromised patients with ARF and an incidence of bacteremia of 17%. We could not demonstrate an impact of bacteremia on mortality after adjusting for baseline characteristics.
AB - Purpose: The characteristics and impact of bacteremia have not been widely investigated in immunocompromised patients with acute respiratory failure (ARF). Methods: We performed a secondary analysis of a prospective cohort of immunocompromised patients with ARF (EFRAIM study). After exclusion of blood cultures positive for coagulase negative Staphylococci, we compared patients with (n = 236) and without (n = 1127) bacteremia. Results: The incidence of bacteremia was 17%. Bacterial pneumonia and extra-pulmonary ARDS were the main causes of ARF in bacteremic patients. Bacteremia involved gram negative rods (48%), gram positive cocci (40%) or were polymicrobial (10%). Bacteremic patients had more hematological malignancy, higher SOFA scores and increased organ support within 7 days. Bacteremia was associated with higher crude ICU mortality (40% versus 32%, p = 0.02), but neither hospital (49% versus 44%, p = 0.17) nor 90-day mortality (60% versus 56%, p = 0.25) were different from non-bacteremic patients. After propensity score matching based on baseline characteristics, the difference in ICU mortality lost statistical significance (p = 0.06), including in a sensitivity analysis restricted to patients with pneumonia. Conclusions: We analyzed a large population of immunocompromised patients with ARF and an incidence of bacteremia of 17%. We could not demonstrate an impact of bacteremia on mortality after adjusting for baseline characteristics.
UR - http://www.scopus.com/inward/record.url?scp=85104683250&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85104683250&partnerID=8YFLogxK
U2 - 10.1016/j.jcrc.2021.03.014
DO - 10.1016/j.jcrc.2021.03.014
M3 - Article
C2 - 33872917
AN - SCOPUS:85104683250
SN - 0883-9441
VL - 64
SP - 114
EP - 119
JO - Journal of Critical Care
JF - Journal of Critical Care
ER -