The aim of the study is to test the hypothesis that an automated algorithm for desired mask pressure improves breathing pattern and sleep quality in patients with hypercapnic ventilatory failure. For this purpose, The investigators will study different groups of patients, including those with obstructive and restrictive ventilatory defect, and obstructive sleep apnoea, non-naive to conventional bi-level positive airways pressure therapy.

Persisting ventilatory failure associated with chronic obstructive pulmonary disease (COPD), obesity-hypoventilation-syndrome, sleep apnoea or neuromuscular disease is increasingly managed with domiciliary non-invasive positive pressure ventilation (NIPPV).

Optimal settings of non-invasive ventilation are usually titrated manually and require time and expertise. The development of systems lead to automated analysis and development of algorithms to adjust ventilators. However, there is a paucity of optimal algorithms, particularly the problem of upper airway obstruction. Therefore, the central aim of this study is to develop the automated setting of an end-expiratory positive airway pressure (EPAP), because upper airway obstruction is relatively common in this group of patients. We hypothesise that an automated end-expiratory airway pressure (AutoEEP) adjusting algorithm could overcome these problems and further optimise and adjust ventilator settings. Using non-invasive ventilation in patients with hypercapnic ventilatory failure, awake and asleep, we will measure physiological outcome parameters and apply an AutoEEP algorithm, comparing it against usual care.

Inclusion Criteria:

• Subjects will be patients not naive to noninvasive ventilation, and being so treated for any form of hypercapnic ventilatory failure.
• Previously stabilised on bilevel noninvasive pressure support ventilation.
• Both genders, age <75years.
• Previously shown to have a requirement for an EEP above cm H2O in order to maintain upper airway patency, or those in whom such a raised EEP would be expected, e.g. obese patients.
• Patients also known to have adequate airway patency at an EEP of 4 to 5 cm H2O will be included to ensure specificity of the algorithm.

Exclusion Criteria:

• Acute critical illness (e.g. acute coronary syndrome, stroke)
• Serious anatomical variations of nose, sinuses, pharynx or oesophagus.
• Any condition at risk of oesophageal bleeding (e.g. oesophageal varices, gastric ulcer, etc.)
• Age >75 years
• Pregnancy
• Epilepsy
• Psychiatric disorders that could possibly influence the study
• Any kind of addiction
• Insufficient knowledge of the language
• Noninvasive ventilation otherwise contraindicated

The intervention took place over 2 nights of polysomnography. No adverse events occurred for any patient on any night polysomnography.

Exclusion of patients whose fundamental disorder was sleep-disordered breathing with an incidental diagnosis of COPD or of OHS and only moderate obesity, and without other supporting evidence, turned out not to be practicable for all cases.