(114) The Use of Daily Home Spirometry to Predict Clinical Outcomes in Persons with Cystic Fibrosis

Background: Monitoring lung function with home spirometry has been increasing in popularity among persons with cystic fibrosis (pwCF) since the start of the COVID-19 pandemic. While clinic spirometry is interpreted from validated standards, expected test-to-test variation of home spirometry and how variation during baseline health may relate to clinical changes are unknown. This aim of this study was to determine variation in baseline lung function during daily home spirometry and identify associations with clinical outcomes.

Methods: Subjects were selected on the basis of available spirometry data from a cohort of pwCF enrolled in a long-term airway microbiome study. Subjects were provided with a PiKo-6 hand-held spirometer (nSpire Health, Inc. Longmont, CO, USA) and asked to perform 3 daily spirometry maneuvers. Validity of home spirometry (percent predicted forced expiratory volume in one second (ppFEV₁)) compared to clinic was measured using Bland-Altman plots. Spirometry acceptability across multiple maneuvers in the same day was assessed using the American Thoracic Society (ATS) guidelines where grade A or B (≥2 maneuvers within 150mL) was considered acceptable. Variation of ppFEV₁ was assessed by calculating a mean ppFEV₁ and coefficient of variation (CoV). The association between CoV and PEx was tested using Cox proportional hazards regression models.

Results: Thirteen subjects (62% female) were included, with a mean age of 28.7±8.3 years and mean ppFEV₁ of 59.9±8.2. The median study duration was 377 days (range, 33–730). Subjects used the home spirometer on average 51.2% of the study days (range, 15-97%). On average, 58.9% of subjects (range, 12-100%) used the home spirometer at least twice a week and 76.8% (range, 65-100%) at least once a week.

To focus on periods of baseline health, days associated with pulmonary exacerbation (PEx, i.e.: spirometry performed 2 weeks prior and during times of antibiotic therapy) were excluded. A median of 204 (range, 11–728) days of baseline spirometry readings were available for further analysis. To evaluate validity of home compared to clinic spirometry, Bland–Altman plots demonstrated overall good agreement with a slight bias (+0.042L) toward higher readings for clinic ppFEV₁ (95% limits of agreement, −0.11-0.19L). Spirometry quality was graded as acceptable on most study days (mean 90.6±4.6%) in which ≥2 maneuvers were recorded. Intra-individual variation in baseline ppFEV₁ was high with a mean variation of 17.6±5.9% day-to-day and 15.2±6.2% week-to-week. Neither rates of acceptable spirometry grades nor CoV were associated with lung disease severity. Of the 13 subjects, 10 experienced ≥1 PEx, with a total of 32 PEx during the study. CoV was not associated with either time-to-first PEx (hazards ratio [HR], 0.78; 95% confidence interval [CI], 0.51-1.21; P=0.24) or time-to-subsequent PEx (HR, 0.91; 95% CI, 0.73-1.12; P=0.28) during the study.

Conclusions: Our study suggests that while home spirometry has generally good validity and acceptability, variation of lung function during baseline health is present and often exceeded expected variation in clinic spirometry per ATS standards. Variability may represent normal physiological variation, be related to the home spirometer itself or other factors but did not portend to upcoming PEx. Recognition of variation during baseline health provides important context for interpretation of home spirometry.

Acknowledgements: This work was supported by grants from CFF (THORNT21F0, CAVERL20Y5) and NIH (5R01-HL-136647).

References: