PT  - JOURNAL ARTICLE
AU  - Catia Cilloniz
AU  - Miquel Ferrer
AU  - Eva Polverino
AU  - Albert Gabarrus
AU  - Rosanel Amaro
AU  - Jorge Puig de la Bellacasa
AU  - Adamantia Liapikou
AU  - Jose Mensa
AU  - Antoni Torres
TI  - Invasive mechanical ventilation in community acquired pneumonia
DP  - 2014 Sep 01
TA  - European Respiratory Journal
PG  - P4932
VI  - 44
IP  - Suppl 58
4099  - http://erj.ersjournals.com/content/44/Suppl_58/P4932.short
4100  - http://erj.ersjournals.com/content/44/Suppl_58/P4932.full
SO  - Eur Respir J2014 Sep 01; 44
AB  - Background: Invasive mechanical ventilation (IMV) is frequently used in patients with community-acquired pneumonia (CAP) and severe respiratory failure. However, there is little information on these patients.Methods: We prospectively enrolled consecutive patients hospitalized with CAP for 12 yrs. We assessed the characteristics and outcomes of patients on IMV, those who needed non-invasive ventilation (NIV) and no ventilatory support, and determined predictors for the need of ventilation and mortality.Results: Among 3,719 patients included, 154 (4%) required IMV, 136 (4%) NIV, and 3,429 (92%) were not ventilated. Streptococcus pneumoniae was the main pathogen. Current alcohol consumption, higher levels of C-reactive protein, worse baseline oxygenation, higher Pneumonia Severity Index, and bacteremia at admission independently predicted the need for IMV, while former tobacco consumption and fever at admission were independently associated with no need for IMV. Likewise, higher levels of C-reactive protein and worse baseline oxygenation independently predicted the need for NIV, while fever at admission was independently associated with no need for NIV. The 30-day mortality was highest in the IMV group, followed by the NIV and non-ventilation groups (51, 33% vs. 22, 16% and 192, 6%, respectively, p<0.001). Both the need for NIV and IMV independently predicted 30-day mortality in the multivariate analysis.Conclusion: Even alter adjustment for higher baseline severity, the need for NIV and IMV independently predicted mortality. Identification of these predictors may help in the initial management and hospital allocation of patients.Supported: CibeRes (CB06/06/0028)-ISCiii, 2009 SGR 911, and IDIBAPS.