Switch of Noninvasive Ventilation (NIV) to Continuous Positive Airway Pressure (CPAP) in Patients With Obesity Hypoventilation Syndrome

A Pilot Study

Sarah Orfanos; Dany Jaffuel; Christophe Perrin; Nicolas Molinari; Pascal Chanez; Alain Palot


BMC Pulm Med. 2017;17(50) 

In This Article



Patients over 18 years old, suffering from OHS, defined as an association of obesity (BMI ≥ 30 kg/m2) and daytime hypercapnia (paCO2 > 45 mm Hg), who had been undergoing NIV for more than 2 months, were prospectively recruited from February 2015 to February 2016 in the pulmonology wards of the university hospital (hôpital Nord) and the military hospital (hôpital Laveran) in Marseille, France. To be included, patients had to be clinically stable for at least 4 weeks before enrollment in the study (no hospitalization or emergency admission). The exclusion criteria were: patients who presented hypercapnia secondary to other causes (obstructive pulmonary diseases (FEV1/FVC < 70), interstitial lung diseases, neuromuscular or chest wall diseases, severe hypothyroidism or congenital central hypoventilation syndrome), and patients unable to give informed consent.

This trial respected the ethical principles of the declaration of Helsinki and was approved by the local ethic committee of the Assistance Publique Hôpitaux de Marseille (AP-HM). Written informed consent was obtained from all patients. The use of data was approved by the Data Privacy Officer of the hospitals (AP-HM).

Study Protocol

Patients who had been undergoing NIV for at least the last 2 month, were prospectively included in the study by the investigators, during a follow-up outpatient clinic.

During this clinic, the following clinical information were collected: history of the diagnosis of OHS, occurrence of acute exacerbations, NIV settings: brand, mode, respiratory rate (RR), expiratory positive airway pressure (EPAP), inspiratory positive airway pressure (IPAP) and interfaces.

The settings were in accordance with the American Academy of Sleep Medicine (AASM) recommendations.[13] NIV settings were optimized in order to avoid patient-ventilator dyssynchrony and to correct remaining respiratory events during sleep, according to the recommendations of the somnoNIV group.[14]

Quality of life was evaluated using the SRI questionnaire; daytime sleepiness was evaluated using the ESS questionnaire; and quality of sleep was evaluated using the PSQI questionnaire.

Finally, an ABG analysis was performed, during the consultation, between 10 and 12 am in an awake, seated patient, at rest for at least 10 min.

During one night with NIV, at home, the patient underwent: (i) nocturnal oximetry using a pulse oximeter (Rescan® software, Resmed®, Australia, New South Wales, Bella-Vista), on which were calculated the percentage of recording time with SaO2 < 90% (RT < 90%) and the oxygen desaturation index (ODI) (defined as the number of time per hour the oxygen saturation drops by ≥ 4%); (ii) nocturnal transcutaneous capnometry using a transcutaneous capnometer (Tosca 500®, Radiometer®, Crawley, UK), on which was calculated the mean ptCO2; (iii) a report of the mean compliance, the median leak, the mean tidal volume and the mean AHI index of the last month spent with NIV was extracted (Rescan® software for Resmed®, Australia New South Wales, Bella-Vista NIV and Directview® for Philips-Respironics®, Murrysville, Pennsylvania, USA NIV).

After undergoing these procedures, the patients underwent NIV withdrawal for seven nights. At the end of this week, pulmonary function tests (PFT) were performed according to the American Thoracic Society and European Respiratory Society 2005 guidelines,[15] and an overnight polygraphy during spontaneous breathing was performed (Alice PDx®, Philips-Respironics®, Murrysville, Pennsylvania). Polygraphic recordings were analyzed by a blinded pulmonologist certified in sleep medicine, respiratory events were scored manually according to the AASM 2012 manual.[16] The AHI, RT < 90% and ODI on this polygraphy during spontaneous breathing were reported. At this stage of the trial, patients exhibiting obstructive ventilatory disorder (FEV1/FVC < 70) or mild OSA (AHI < 15/h), were not switched to CPAP and continued their treatment with NIV. All the other patients were switched to CPAP, the settings were as follow: constant pressure, EPAP 2 cm H2O above the EPAP of the previous NIV to compensate for the absence of an IPAP on the CPAP, settings were not adjusted until measurements were repeated more than a month after this switch.

After more than a month with CPAP; SRI, ESS and PSQI questionnaires were reevaluated. An ABG analysis was performed in the same conditions as previously described. During one night with CPAP, at home, the patient underwent: (i) nocturnal oximetry; (ii) nocturnal transcutaneous capnometry; (iii) a report of the mean compliance, the median leak and the mean AHI index of the last month spent with CPAP was extracted. (Fig. 1).

Figure 1.

Study protocol. Definition of abbreviations: OHS: Obesity Hypoventilation Syndrome; NIV: Non Invasive Ventilation; SRI: Severe Respiratory Insufficiency questionnaire; PSQI: Pittsburgh Sleep Quality Index; ESS: Epworth Sleepiness Scale; AHI: Apnea Hypopnea Index; ABG: Arterial Blood Gas; FEV1: Forced Expiratory Volume in 1 s; FVC: Forced Vital Capacity; EPAP: Expiratory Positive Airway Pressure; CPAP: Continuous Positive Airway Pressure

As a secondary analysis, we analyzed the characteristics of patients failing NIV and CPAP using Salord et al. criteria to define failure: more than 30% of recording time spent below 90% saturation or paCO2 > 45 mm Hg, to determine if a characteristic could be used as a prognostic factor of ventilation failure:[10]

Statistical Analysis

Continuous data were expressed in means and SDs and categorical data were expressed as percentages and absolute.

Normality was assessed using Shapiro-Wilk normality test. Intra-group changes in variables measured at first with NIV and after the switch to CPAP were assessed using paired t-tests or non-parametric Wilcoxon tests depending on normality.

When comparing two independent groups (NIV failure, NIV success), unpaired t-tests or non-parametric Mann–Whitney tests depending on normality were used for continuous variables and x2 tests for categorical variables.