We have developed a new method to measure flow in patients with sleep-relat
ed breathing disorders (SRBD). These flow sensors are modified thermistors
located in the same sensors we use for pressure measurement in the upper ai
l-ways to find the obstructive segments during apnoeas. The aim of this stu
dy was to test if using internal thermistors as indicator of air flow has a
dvantages compared with the external thermistor method in detecting respira
tory events. A total of 50 consecutive patients with an apnoea-hypopnoea in
dex (AHI) of more than 15 were studied. A standard nocturnal polysomnograph
y (PSG) with both internal and external thermistors was performed in all pa
tients. To estimate the patients' AHI, a detailed analysis viewing all para
meters except external and internal thermistors was performed. This was fol
lowed by an analysis viewing only internal thermistors signals indicating a
irflow in the pharynx, and finally an analysis viewing only external thermi
stor signals indicating air flow at the mouth and nose. Mean AHI measured b
y the three methods showed 49.8 (SD 23.4) by the PSG, 47.8 (SD 24.9) by int
ernal thermistors alone, and 31.5 (SD 22.2) by external thermistors. There
was no statistical difference between AHI detected by PSG and internal ther
mistors, but highly significant differences between PSG and external thermi
stors (p < 0.001). The external thermistors missed an average of almost 20
respiratory events per hour.