Three mRCTs evaluated NIV in the treatment of hypercapnic respiratory failure.

In the first, Brochard et al. enrolled 85 patients with COPD exacerbations in five hospitals in three countries (France, Italy, and Spain). Patients were randomized to standard oxygen therapy or NPPV (at least 6 h/day). Hospital mortality was 29 % in the control group vs 9 % in the NIV group (p = 0.02), thanks to the lower rate of intubation in the NIV group [5].

Plant et al. conducted a mRCT in 14 hospitals in UK, enrolling 236 patients with mild to moderate respiratory acidosis during COPD exacerbations. NPPV was compared to oxygen therapy. Noninvasive ventilation was applied for as long as tolerated on the first day and then progressively suspended on day 4. In the NIV group, the mortality rate was half that of the standard group (12/118 vs 24/118) [6].

More recently, Nava et al. evaluated NIV efficacy in patients with chronic pulmonary disease and acute hypercapnic respiratory failure aged over 75 years. The study enrolled 82 patients in three respiratory intensive care units in Italy and Switzerland. Noninvasive ventilation (as NPPV) was compared to standard treatment. Survival was significantly better in the NIV group at hospital discharge (1/41 vs 6/41 deaths), after 6 and after 12 months [7].

Another nine single-center RCTs evaluated NIV efficacy on mortality for exacerbation of COPD [8–16]. Three noteworthy trials were conducted on respiratory or general wards [12, 13, 15]; only one trial randomized severely ill patients comparing NIV to tracheal intubation [16]. Meta-analysis of the results found a marked beneficial effect on mortality [17].

One mRCT evaluated NIV in hypoxemic patients.

Ferrer et al. enrolled 105 patients with severe hypoxemia (pO₂ <60 mmHg with Venturi mask at 50 % of oxygen) in three ICUs in Spain. Noninvasive ventilation (such as NPPV), applied as long as tolerated, was compared to standard oxygen therapy. Intensive care unit (18 % vs 39 %) and 90-day mortality were lower in the NIV group; the difference was prominent if pneumonia was the cause of ARF, while ARDS was a predictor of 90-day decreased survival. Only two patients in the standard group received NIV as rescue treatment [19].

Hypoxemic ARF can have various etiologies, whose responsiveness to NIV can markedly differ [3, 18, 20–22]. Several single-center RCTs [23–38] demonstrated that NIV significantly reduces mortality in cardiogenic pulmonary edema, and it is currently considered a first-line, grade-of-evidence 1A intervention. The benefit was present both for CPAP and NPPV and also for prehospital use. Noninvasive ventilation also proved effective in reducing mortality in RCTs conducted in hypoxemic ARF in immunocompromised patients [39] and chest trauma patients [3, 18, 40]. On the contrary, the advantage on survival is controversial in the case of pneumonia or ARDS, due to a high failure rate [3, 18, 41]. In this setting, some authors found NIV potentially dangerous (i.e., associated with worse survival) when applied for too long despite its failure, as it delays tracheal intubation [42]. Finally, three single-center RCTs evaluated NIV in asthma, and no death was reported in any of the studies [43–45].

Esteban et al. conducted a multicenter trial in 37 centers in eight countries (mainly in Europe and North and South America). The authors enrolled 221 patients who were electively extubated after at least 48 h of mechanical ventilation and who developed ARF within the subsequent 48 h. Noninvasive ventilation (such as NPPV, applied continuously for at least four hours) was compared to standard therapy, which included supplemental oxygen, bronchodilators, respiratory physiotherapy, and any other indicated therapy. Rescue NIV was applied in 28 patients in the control group (three died). ICU mortality rate was higher in the NIV group (25 % vs 14 %). The difference appeared to be due to a different rate of death (38 % in the NIV group vs 22 %) among reintubated patients (whose rate was not different between the two groups); moreover, the interval between the development of ARF and reintubation was significantly longer in the NIV group. A potential logical explanation proposed by the authors was that the delay in reintubation negatively affected survival, by various mechanisms like cardiac ischemia, muscle fatigue, aspiration pneumonia, and complications of emergency reintubation. A trend toward better outcomes was observed for COPD patients treated with NIV [62].