Morbidity and mortality of patients with acute respiratory failure undergoing non-invasive mechanical ventilation in the Adult Intensive Care Unit in Hospital Nacional Cayetano Heredia, 2007-2008

Abstract

Introduction: noninvasive mechanical ventilation is a technique with a fan interface - patient through a nasal or facial mask. This reduces the morbidity associated with conventional mechanical ventilation associated with endotracheal intubation and sedation of the patient relaxation. Objectives: To assess morbidity and mortality of patients undergoing non-invasive mechanical ventilation (NIMV), in the General Intensive Care Unit (GICU) of Hospital Nacional Cayetano Heredia (HNCH) from June 2007 to May 2008.

Material and method: We obtained data from the GICU files from June 2007 to May 2008. We recorded physiological values when beginning and finishing NIMV and we defined success and failure when using NIMV.

Results: During the one-year study, 30 patients required NIMV, their average age was 49.7± 2 years, and 63.3 per cent were women. When NIMV was started, APACHE II score was 12.7 ± 5.6, and the SOFA score was: 4.3 ± 2.2; 66.7% of patients presented with co-morbidities; of them, 33.3 % had heart failure; 23 (76.6 %) had type I failure respiratory, and of them, acute pulmonary edema and severe pneumonia were the most frequent causes; one patient (3.3 %) had type Ii respiratory failure; 6 patients (20.1 %) had both types I and II respiratory failure. We found significant differences at the end of NIMV in the following parameters: decrease of respiratory workload (p: 0.001), decreased respiratory rate (p: 0.004), decreased cardiac rate (p: 0.002), reduced systolic blood pressure (p: 0.016), reduced Fi02 requirement (p: 0.001) and increased Pa02/Fi02 rate (p: 0.001). Failure when using NIMV was related to having a lower Glasgow score (p: 0.03), with the use of inotropic substances during or after NIMV was started (p: 0.01), and a high Fi02 requirement when starting NIMV (p: 0.006).

Conclusions: Patients undergoing NIMV were predominantly women presenting with co-morbidities, type I respiratory failure was the most frequent reason for undergoing NIMV. We found significant differences in physiological parameters at the beginning and the end of NIMV. Lower Glasgow scores, starting or increasing inotropic use during or after NIMV, and a greater Fi02 requirement when starting NIMV were predictors for failure when using NIMV.