Built-in software in children on long-term ventilation in real life practice

Information gathered with built-in software (BIS) on new ventilators allow clinicians to access long-term noninvasive ventilation (LTNIV) data. Nevertheless, few evidence are available in literature that highlight potential strengths and disadvantages of using BIS in pediatrics. We aim to evaluate the use of BIS in a cohort of 90 children on LTNIV in our unit, focusing mainly on adherence, air leaks, and residual sleep events. We found that caregivers' perception of ventilator use is independent from objective adherence (P = .137). Furthermore, we failed to find any predictors of adherence. As regards air leaks, we found that pre-scholars' (0-6 years old) total air leaks are lower than teenagers' (more than 12 years old) (P < .05). Multiple regressive analysis showed that age at the beginning of therapy is a predictor of total air leaks: prescholars are associated with lower values (P < .05), while scholars (6-12 years old) are associated with higher values (P < .05). Finally, we explored the validity of BIS automatic scoring of sleep events (AHI_BIS ) as compared with the manual scoring of polygraphy (AHI_PG ). AHI_BIS is within a range of 3.98 from AHI_PG in 95% of cases, with a 64% of sensitivity and a 67% of specificity in identifying a pathological state. The disagreement between the two methods seems to increase for high AHI values. In conclusion, data gathered by BIS are a useful support tool for the clinician in assessing the course of LTNIV. However, clinicians must be aware of the several limitations of built-in software, especially in pediatrics.