Quality of poly(somno)graphy recordings in children.

The aim of the study was to assess the scorability of the signals of four poly(somno)graphy devices and transcutaneous carbon dioxide tracings (PtcCO2 ) of one device in children. The presence (0%, < 25%, 25%-50%, 50%-75%, 75%-99%, 100% of recording time) and quality (bad, average, good) of the signal of each sensor were analysed. During a 5-month period, 364 poly(somno)graphies were performed in 12 different hospital units. Forty-one children had poor/bad cooperation, and 13 severe behaviour disorders. Seventy-one and 293 poly(somno)graphies were performed in children aged ≤ 2 and > 2 years, respectively; nine poly(somno)graphies failed. For the four poly(somno)graphy devices, the signal was present during 99% of recording time for the electroencephalogram, 99% for thoracic belt, 97% for abdominal belt, 97% for body position, 95% for the microphone, 92% for pulse oximetry, 87% for tracheal sound, 71% for oronasal thermistor, 52% (41% for ≤ 2 years, 55% for > 2 years old) for nasal pressure and 86% for PtcCO2 . The signal was of good quality in 98% of poly(somno)graphies for body position, 96% for microphone, 96% for thoracic belt, 95% for pulse oximetry, 91% for abdominal belt, 91% for tracheal sound, 82% for oronasal thermistor, 78% for electroencephalogram, 73% for nasal pressure and 46% of PtcCO2 recordings. The scorability was comparable between devices. Nasal pressure and oronasal thermistor had the lowest scorability, especially in children aged ≤ 2 years. This underlines the necessity of the development or improvement of alternative, ideally face-free, sensors, or miniaturized devices adapted for infants and children.