Abstract
Introduction:
Sleep related respiratory events may reduce the efficacy of NIV in patients with chronic hypoventilation (CH).
Aims:
1) To measure the frequency of Apnea (A) and hypopnea (H) under NIV for CH.
2) To evaluate the ability of the built-in ventilator software to detect A and H (AHINIV).
Methods:
Overnight polygraphy (PG) was performed in 42 patients with: neuromuscular diseases (NMD, n= 23), central- (CHC, n=3) or obesity hypoventilation syndromes (OHS, n=12) and restrictive thoracic disorders (4). A and H were scored according to the rules of the American Academy of Sleep Medicine and summarized as AH-index (AHI PG). H was scored if the flow reduction was associated with a 3 % desaturation or an autonomic activation scored from the pulse waveform. H was scored as central or obstructive (OH). OH was scored in the presence of inspiratory flattening of the flow signal and/or an associated thoracoabdominal paradox during the event. AHI PG was compared with AHINIV.
Results:
Mean values of: AHIPG: 8.2/hr (±8.5); H-index: 8.0/hr (±8.5); A-index: 0,1/hr (±0.2); oxygen desaturation index: 9/hr (± 9,5).
13 patients (31%) had an AHIPG >10 (7 NMD, 2 CHC, 4 OHS), of which 12 predominantly had OH.
AHIPG compared with AHINIV had a mean bias of 0.8 (limits of agreement -11.1-12.6).
With an AHI threshold of 5/hr and 10/hr, NIV software had a sensitivity of 77.8% and 66,7% and a specificity of 78.3% and 93.1%, respectively, compared with PG.
Conclusions:
Respiratory events are frequent under NIV for CH. OH was the most frequent event when AHIPG > 10. Relying solely on data from the ventilator as a screening tool, respiratory events may remain undetected in many patients with CH treated with NIV.
- © 2014 ERS