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


Study Population

Twenty-two patients were included. Five dropped-out due to unwillingness to be hospitalized for the polygraphy prior to the switch to CPAP. Two patients were excluded prior to the switch, one had an obstructive syndrome on the PFT, and the other had an AHI < 15/h on the polygraphy. Fifteen of the 30 patients who were initially selected, were switched to CPAP (Fig. 2).

Figure 2.

Flow chart. Definition of abbreviations: OHS: Obesity Hypoventilation Syndrome; FEV1: Forced Expiratory Volume in 1 s; FVC: Forced Vital Capacity; NIV: Non Invasive Ventilation; CPAP: Continuous Positive Airway Pressure; AHI: Apnea Hypopnea Index

Anthropometric characteristics of the patients are shown in Table 1.

Seven nights after the NIV was discontinued, polygraphy of the weaning period showed a high mean AHI (AHI = 53.5/h ±26.9) and the mean RT < 90% was 21.8% (24.2) (Table 2).

Regarding settings, all NIV were in spontaneous timed (ST) mode, mean EPAP was 7.7 ± 1.8 cm H2O, mean IPAP was 18.3 ± 3 cm H2O, mean back-up RR was 13.9 ± 1.8 cycles/min. Two patients received additional oxygen therapy, flow rates were not modified after switching to CPAP. (Table 3).

Effect of the Switch to CPAP on the Measured Data

There were no significant differences, after the switch to CPAP, in mean AHI, diurnal and nocturnal alveolar gas exchanges (diurnal ABG, nocturnal oximetry and transcutaneous capnometry), median leaks and compliance (Table 4, Fig. 3).

Figure 3.

Measurement of mean AHI; diurnal and nocturnal alveolar gas exchange parameters and mean compliance, in 15 patients, with NIV and one month after the switch to CPAP. Definition of abbreviations: NIV: Non Invasive Ventilation; CPAP: continuous positive airway pressure; AHI: Apnea Hypopnea Index; RT < 90% spO2: percentage of recording time below 90% saturation; NS: no significant difference p > 0.05

Regarding ESS, SRI and PSQI scores, the only significant difference was in the ESS, which was lower after the switch to CPAP (8.21 ± 5.54 vs 4.29 ± 3.43) (p = 0.004). There were no significant differences in the SRI and all its subscales and in the PSQI and all its components between before and after the switch to CPAP. (Table 5, Fig. 4).

Figure 4.

Changes in ESS, PSQI and SRI scores one month after the switch to CPAP. Definition of abbreviations: NIV: Non Invasive Ventilation; CPAP: Continuous Positive Airway Pressure; ESS: Epworth Sleepiness Scale, PSQI: Pittsburgh Sleep Quality Index; SRI: Severe Respiratory Insufficiency questionnaire; NS: no significant difference; *: significant difference p < 0.05

At the end of the trial, through a self-administered questionnaire, the patients' preference for one mode of ventilation was determined. Twelve patients (80%) reported that they preferred CPAP over NIV and wanted to continue their treatment with CPAP. Two patients (13%) had no preference. Only one patient wanted to discontinue CPAP and go back to NIV, despite the fact that there were no objective criteria of deterioration during CPAP treatment.


None of the patients presented with clinical deterioration after more than a month of CPAP and needed to be switched back to NIV. Two patients were admitted in the intensive care unit during the trial, due to causes unrelated to the mode of ventilation. One presented with acute renal failure complicated by symptomatic hyperkalemia and the other with left heart failure secondary to non-compliance to diuretics.


When taking into account French social security reimbursement rate, NIV cost was 63.16€ per week and CPAP cost was 18€ per week. This switch may allow an economy of 2348€ per year per patient.

NIV Failure

When analyzing the characteristics of the patients failing NIV, according to Salord's criteria: more than 30% of recording time spent below 90% saturation or paCO2 > 45 mm Hg; BMI was the only factor associated with NIV failure (p = 0.014) (Table 6).