Introduction: Follow-up of children on chronic non-invasive ventilation (NIV) could be improved by telemonitoring, using the ventilator?s built-in software (BIS) parameters as alternative for in-hospital sleep studies to reduce costs, enhance patient independence and contribute to early detection of infections.

Aim: This study investigates whether analysis of BIS parameters can predict abnormal nocturnal transcutaneous CO2 (TcCO2) and saturation (SaO2) measurements in children on chronic NIV.

Methods: Children on chronic NIV in follow-up at the Antwerp University Hospital between Oct 2017 and Sept 2021 were retrospectively included. TcCO2 and SaO2 measurements from sleep studies were collected together with BIS parameters at three different time points: the night of the sleep study (BIS₁), mean values from 48h (BIS₂) and 72h (BIS₃) before the sleep study. Predictions were calculated for following outcome measures: % recording time TcCO2 > 46.9 mmHg (%RT TcCO2; abnormal if >2%), recording time SaO2 < 93% (RT SaO2; abnormal if >1h), abnormal TcCO2 or SaO2, mean TcCO2, mean SaO2.

Results: 69 measurements were included. %RT TcCO2 was separately predicted by reached tidal volume₂ (OR 0.97; p = 0.05; AUC = 30%) and reached IPAP₁ (OR 1.05; p = 0.05; AUC = 66%). Leakage₁ predicted RT SaO2 (OR 1.21; p = 0.03; AUC = 84%). Mean TcCO2 was correlated to reached tidal volume₂ (? = -0.31; p = 0.03).

Conclusion: Certain BIS parameters can predict nocturnal hypercapnia and desaturation in children on chronic NIV. Future studies with larger sample size are warranted to further investigate the predictive value of BIS parameters identified in this study.