• Non invasive ventilation
• COPD Exacerbations

The efficacy of non-invasive ventilation (NIV) for severe COPD exacerbation has been unambiguously demonstrated. The study by Struik et al recently published in Thorax1 demonstrates that pursuing nocturnal NIV after acute respiratory failure in patients with COPD exhibiting prolonged hypercapnia >48 h after termination of ventilatory support does not reduce time to readmission or rate of deaths at 1 year. They included a mixed cohort of patients with acute-on-chronic and acute respiratory failures. However, in future studies it may be interesting to select patients with persistent hypercapnia several weeks after the acute exacerbation episode. Nevertheless, long-term NIV efficacy for stable COPD remains uncertain in terms of hard outcomes. Clinical cohort studies suggest that mortality and readmission rates under NIV are dependent upon the underlying COPD subgroup.2 ,3

We recently showed that a reduction in risk of death and hospital readmission is obtained only in obese patients with COPD with NIV usage >5 h/day.2 In the Struik study,1 randomisation was stratified on body mass index although the mean body mass index of the whole cohort was 25 kgm⁻² (SD around 6 kgm⁻²). The majority of Struik's cohort had ‘respiratory COPD’ characterised by severe airflow limitation without obesity,4 a subgroup supposed to be less responsive to NIV.2

Obesity is a main risk factor of sleep apnoea, and patients with overlap syndrome (COPD+sleep apnoea) benefit from positive pressure treatment.5 We obtained positive results with NIV in obese patients with COPD after adjustment for all independent risk factors associated with prognosis, particularly sleep apnoea. This suggests that NIV, beyond correcting sleep apnoea, may also lessen the deleterious impact of nocturnal hypoventilation and improve respiratory mechanics.

In Struik's study,1 in both intention to treat and per protocol analyses (threshold for NIV compliance ≥5 h/night) NIV was not effective on time to readmission or death. In our work, we analysed the impact of daily use of NIV as a time-dependent variable without any pre-existing arbitrary threshold. We found a U-shaped curve between NIV usage and risk of readmission or death. This suggests that NIV adherence should be assessed as a continuous variable rather than with a predefined threshold.

In summary, we suggest that subsequent randomised controlled trials be conducted on specific COPD subgroups that have better likelihood of responding to NIV (ie, obese patients with comorbidities and moderate to severe airflow obstruction). This might represent a new approach to the challenging area of long-term NIV in stable patients with COPD.