Conventional polysomnographic techniques which identify hypopneas and apneas were compared to Transcutaneous Carbon Dioxide (PTcCO2) and End Tidal Carbon Dioxide (PeTcCO2), for non-invasively detecting Alveolar Hypoventilation in sleep. 61 patients were evaluated for apneas, hypopneas, alveolar hypoventilation using PTcCO2 > 45mmHg (based on control and elderly subject data ¹) with addition of 4mmHg (to compensate for the skin metabolic gradient ²), oximetry nadirs, End Tidal CO2 (PETCO2), Apnea and Disturbance indexes. Patients were selected if diagnostic histories included any of the following; obesity, lung disease, elevated PaCO2 (tension of arterial carbon dioxide), disorders of excessive somnolence (DOES) or AM headaches. Patients were monitored during a full night using a 16-channel Grass model 78D. Sensors included the Biochem "Lifespan100" End Tidal CO2 monitor, the "Microspan Combo TCPO2/CO2 model 5000", and a Biox II ear oximeter. Averages of the respiratory variables as well as percentage breakdowns of the patient population were obtained. Patient's questionaires were collated for subjective complaints that may be related to elevated PaCO2. Seven had arterial blood gases obtained during the nocturnal study.

15% of the 61 patients had AM headaches and 31% had symptoms of DOES. 63% of the study group had hypoventilation during wake and sleep, 14% showed hypoventilation exclusively during sleep, and 23% of the patients showed no alveolar hypoventilation. The sub-group with blood gases demonstrated a PTcCO2-PaCO2 difference of +4mmHg, consistent with a previous study ². PETCO2-PaCO2 differences were up to +/- 15.9mmHG, with greater deviations occurring at higher PaCO2. 57% of the ABG sub-group showed evidence of hypoventilation.

Normal ventilation during wake and sleep (n=15)

Hypoventilation during wake and sleep (n=37)

Hypoventilation during sleep only (n=9)

ABG sub-group with alveolar hypoventilation (n=4)

ABG sub-group with alveolar hypoventilation (n=3)

COnventional polysomnographic variables may suggest, but not diagnose, alveolar hypoventilation. PTcCO2 monitoring provided greater reliability than ETcCO2 in identifying alveolar hypoventilation. The reasons for the discrepancy in PaCO2 - PETCO2 values from previous studies ² are not apparent; however, they may reflect decresed tidal volume, increased dead space or air entrainment. PTcCO2 is a useful non-invasive method to screen patients at risk for alveolar hypoventilation in sleep.

1 Naifeh, et al. Effect of Aging on Sleep Related Changes in Respiratory Variables. Sleep, 1987, 10(2): 160-171.

2 Palmisano, et al. Clinical Accuracy of a Combined TCPO2-PCO2 Electrode. Soc. Crit. Care Med, 1984, 12: 276-277