Real-life evaluation of NIV to CPAP switch in patients with chronic respiratory failure. A case control study
Résumé
Background Non-invasive ventilation (NIV) is a standard of care for hypercapnic chronic respiratory failure (CRF). Obstructive sleep apnea syndrome (OSA) frequently contributes to hypoventilation in CRF patients. CPAP improves hypercapnia in selected COPD and obese patients, like NIV. We aimed to describe the profile of patients switching from NIV to CPAP in a cohort of patients on long-term ventilation and to identify the factors associated with a successful switch. Methods In this case-control study, 88 consecutive patients who were candidates for a NIV-CPAP switch were compared with 266 controls among 394 ventilated patients treated at the Dijon University Hospital between 2015 and 2020. They followed a standardized protocol including a poly(somno)graphy recorded after NIV withdrawal for three nights. CPAP trial was performed if severe OSA was confirmed. Patients were checked for recurrent hypoventilation after 1 and 23[14–46] nights under CPAP. Results Patients were 53% males, median age 65 [56–74] years, and median BMI 34 [25–38.5] kg/m2. Sixty four percent of patients were safely switched and remained on long-term CPAP. In multivariate analysis, the probability of a NIV-CPAP switch was correlated to older age (OR: 1.3 [1.01–1.06]), BMI (OR: 1.7 [1.03–1.12]), CRF etiology (OR for COPD: 20.37 [4.2–98,72], OR for obesity: 7.31 [1.58–33.74]), circumstances of NIV initiation (OR for acute exacerbation: 11.64 [2.03–66.62]), lower pressure support (OR: 0.90 [0.73–0.92]), lower baseline PaCO2 (OR: 0.85 [0.80–0.91]) and lower compliance (OR: 0.76 [0.64–0.90]). Among 72 patients who went home under CPAP, pressure support level was the only factor associated with the outcome of the NIV-CPAP switch, even after adjustment for BMI and age (p = 0.01) with a non-linear correlation. Etiology of chronic respiratory failure, age, BMI, baseline PaCO2, circumstances of NIV initiation, time under home NIV or NIV compliance were not predictive of the outcome of the NIV-CPAP switch. Conclusions A NIV-CPAP switch is possible in real life conditions in stable obese and COPD patients with underlying OSA.