Unravelling telemonitoring data to predict good NIV quality: the E-QualiNIV study.

Background and objective: High treatment quality, defined by mean adherence >4 hours per day, unintentional leaks <24 L/min and a residual Apnoea-Hypopnoea Index <5 events per hour, is associated with better outcomes. Adherence variance may reflect behaviour linked to better treatment quality. This study aimed to assess whether monthly adherence variance is associated with improved treatment quality in patients treated with non-invasive ventilation (NIV) for more than 4 months.

Methods: E-QualiNIV is a retrospective study evaluating treatment quality in 511 telemonitored patients with chronic respiratory failure, observed from 15 April to 31 October 2023. The study followed three steps: (1) hierarchical clustering based on individual adherence variance; (2) assessing whether monthly adherence variance in the preceding month predicted the proportion of alerts in the subsequent month and (3) evaluating treatment quality based on the number of months with low adherence variance.

Results: Cluster 1, consisting of patients with adherence variance below 3, had a significantly higher proportion (57.93%) of patients achieving high-quality treatment compared with other clusters (43.1% for cluster 2 and 46.4% for cluster 3) (p=0.035). Patients with a low adherence variance in the preceding month were more likely to achieve high-quality treatment in the following month (except for May, significant differences every month from p=0.04 to p<0.01). Those with 6 or more months of low adherence variance had a significantly higher probability of receiving high-quality treatment over the entire period (coefficient: 0.2649, p value: 0.0028) compared with those who did not (non-significant).

Conclusion: The E-QualiNIV study demonstrates that low adherence variance is associated with high-quality treatment and serves as a prognostic indicator of treatment stability and alert occurrence in the subsequent month.