Noninvasive positive pressure ventilation: effect on mortality in acute cardiogenic pulmonary edema: a pragmatic meta-analysis.

INTRODUCTION: In contrast to a series of recent meta-analyses (MAs), the 3CPO (Three Interventions in Cardiogenic Pulmonary Oedema) randomized controlled trial (RCT) reported in 2008 did not find a significant mortality benefit of noninvasive positive pressure ventilation (NPPV) in acute cardiogenic pulmonary edema (ACPE).
OBJECTIVES: This paper combines data collected in the 3CPO trial together with data from recent MAs and calculates a revised risk ratio for NPPV in ACPE. Reasons for the discrepancy in mortality estimates are identified and discussed through contrasting the methodology and results of the 3CPO trial with previous RCTs. PATIENTS AND METHODS: Patients included adults with ACPE secondary to a variety of insults such as hypertension, acute coronary syndromes, dietary indiscretion, arrhythmias and valvular lesions and assessed by clinical parameters (respiratory rate, crackles, oxygen saturation) and chest radiograph. Data was collected from MAs published after 2005 and their respective RCTs. As opinions regarding RCTs worthy of inclusion in the analyses were varied, 3 sets of RCTs were combined with the 3CPO data. The first set of data duplicated the RCTs chosen in the Cochrane; the second set, a comprehensive set, included all RCTs cited in any of the MAs reviewed; and the third set, a high quality RCT set, assessed data from only those RCTs included in at least 4 out of the 5 MAs reviewed. Data were analyzed with both fixed and variable effect modes using Revman software.
RESULTS: All combinations of RCTs and modes of analysis predict a significant mortality benefit. The combined data predicts a risk ratio for mortality using NPPV of 0.75 (95% CI: 0.61-0.92).
CONCLUSIONS: An analysis of the existing RCT data, inclusive of the 3CPO trial, predicts a continued and significant mortality benefit of NPPV in ACPE.