Ventilator data: An adjunct to cardiorespiratory sleep study (CRSS) in monitoring children on long term non-invasive ventilation (LT-NIV)

Annual CRSS is recommended for monitoring children on LT-NIV, but limitations include poor data quality and inability to assess adherence. Ventilator trend data can include minute ventilation (MV), leak and compliance. We compare such data with CRSS in children on LT-NIV. Method: Analysis of CRSS (clinician) and [ResMed] ventilator (physiologist) reports and breath-by-breath data from the same night using Cohen’s Kappa (κ). A summary of the trend [ResMed] data from 30 days prior were also reviewed. <bold>Results:</bold> 40 datasets (38 patients): 23 on bilevel NIV, 15 on continuous positive airway pressure. Overall median age of 11.1 yrs (60% male). 74% had compliance >80%, 15% had <50% compliance. Interrater reliability for overall impression was poor (κ=0.19, 95% CI 0.03-0.34). However, 38% of ventilator reports noted a stable study, but possible mask issues that CRSS did not identify; Adjusting ‘review-mask issues’ to ‘stable’ in these cases (as none led to mask changes) κ improved to moderate (κ=0.55, 95% CI 0.23-0.74). The data were unbalanced (~70% cases were in 2 of 5 categories), which limited the maximal value of κ. There was very good agreement (κ= 0.86) for back-up-rate and good agreement (κ= 0.63) for desaturation assessments. Deviations in CRSS from trend parameters (e.g. MV, leak) were seen in 10%. <bold>Conclusion:</bold> Ventilator data complements CRSS by providing trend and supplemental data; identifying opportunities for ventilator optimisation even in those with stable CRSS (leading to a lower than expected κ for the overall impression). It has potential for use as a screening tool e.g. to identify mask/ compliance issues to be addressed prior to CRSS.