Factors Affecting Domiciliary Non-Invasive Ventilation Compliance

Abstract

Few international studies have investigated factors affecting domiciliary non-invasive ventilation (D-NIV) compliance, and data from the UK are limited. We assessed compliance (defined as ≥ 4 h/night for at least 70% of the time) in a retrospective UK population study, at three time points (0-1 month, 3-4 months and 11-12 months), for all patients commenced on D-NIV over a 5-year period. A total of 359 patients were included. Non-compliant vs. compliant patients were significantly younger (median age 64 (IQR 52-72) vs. 67 (58-75) years, p = 0.032) and more likely to have schizophrenia, consistent at both 3-4 months (5% vs. 1%, p = 0.033) and 11-12 months (5% vs. 2%, p = 0.049). Repeated measures ANOVA demonstrated that the minutes [median (IQR)] of D-NIV used significantly increased at the three time points (0-1 month, 3-4 months and 11-12 months) for patients with hypertension [310 (147.5-431) vs. 341 (89-450) vs. 378 (224.5-477.5), p = 0.003]; diabetes [296.5 (132.5-417.5) vs. 342.5 (94.5-438.5) vs. 382 (247.5-476.25), p = 0.002] and heart failure [293 (177-403) vs. 326 (123-398) vs. 365 (212-493), p = 0.04]. In conclusion, younger and comorbid schizophrenic patients have lower D-NIV compliance rates, and our data suggest that persistence with D-NIV over a year may improve overall use.

Keywords: Compliance; Non-invasive ventilation; Schizophrenia.

Fig. 1

a Box-plot of the number of minutes of NIV use across the three time points in different demographic sub-groups of our cohort. b Box-plot of the number of minutes of NIV use across the three time points in patients with different commodities. c Box-plot of the number of minutes of NIV use across the three time points in patients with different index indications for NIV. The length of the box represents the interquartile range (IQR) with the top of the box representing the third quartile and bottom the first quartile. The width of the box is arbitrary. The horizontal line across the box represents the median value. The whisker length has been determined by calculating the maximal whisker length (1.5 times the IQR). The 3rd quartile plus maximal whisker length denotes the upper boundary, whilst the 1st quartile minus the maximal whisker length denotes the lower boundary. Any values falling outside these boundaries have been classified as outliers and have not been shown. The minimum and maximal values shown are the smallest and largest value in the data set that falls within the set boundaries. This follows standardised nomenclature for a modified box-plot as originally described by Tunkey in 1977. [14] AF atrial fibrillation, BAME Black, Asian and Minority Ethnic, COPD chronic obstructive pulmonary disease, CWD chest wall deformity, HTN hypertension, IHD ischaemic heart disease, NMD neuromuscular disease, OHS obesity hypoventilation syndrome, T2DM type 2 diabetes. All statistical analysis was conducted using a repeat measures ANOVA with a Greenhouse–Geisser correction

Fig. 1

a Box-plot of the number of minutes of NIV use across the three time points in different demographic sub-groups of our cohort. b Box-plot of the number of minutes of NIV use across the three time points in patients with different commodities. c Box-plot of the number of minutes of NIV use across the three time points in patients with different index indications for NIV. The length of the box represents the interquartile range (IQR) with the top of the box representing the third quartile and bottom the first quartile. The width of the box is arbitrary. The horizontal line across the box represents the median value. The whisker length has been determined by calculating the maximal whisker length (1.5 times the IQR). The 3rd quartile plus maximal whisker length denotes the upper boundary, whilst the 1st quartile minus the maximal whisker length denotes the lower boundary. Any values falling outside these boundaries have been classified as outliers and have not been shown. The minimum and maximal values shown are the smallest and largest value in the data set that falls within the set boundaries. This follows standardised nomenclature for a modified box-plot as originally described by Tunkey in 1977. [14] AF atrial fibrillation, BAME Black, Asian and Minority Ethnic, COPD chronic obstructive pulmonary disease, CWD chest wall deformity, HTN hypertension, IHD ischaemic heart disease, NMD neuromuscular disease, OHS obesity hypoventilation syndrome, T2DM type 2 diabetes. All statistical analysis was conducted using a repeat measures ANOVA with a Greenhouse–Geisser correction