Abstract
INTRODUCTION: Prolonged gaps in care of >12-months are frequent among people with cystic fibrosis (pwCF) and are associated with reduced lung function. Comprehensive analysis of patient-level predictors of visit frequency is needed to optimize protocols for stable pwCF and identify subgroups at high risk of gaps and poor outcomes, promoting equitable treatment for all pwCF. OBJECTIVE: To determine sociodemographic and disease-related factors predictive of visit frequency in pwCF and to assess how these effects vary across the lifespan. METHODS: We conducted an observational cohort study using data from 2004-2016 for pwCF aged 6-60 years in the US Cystic Fibrosis Foundation Patient Registry. We modeled the relationship between patient-level characteristics and between-visit interval (BVI) using multivariable longitudinal semiparametric regression. BVI was defined as the number of days from the index encounter to the previously recorded visit. RESULTS: The study included 28,588 pwCF with 859,568 encounters. Overall, 55% of visits occurred within 90 days of the prior visit, adhering to national guidelines. On average, adults without common CF-complications attended clinic approximately every 4 months, with a BVI ≥ 110 days from age 23-56. Males attended clinic less frequently than females (9.8% longer BVI; 95% CI 9.1%, 10.5%; p<0.001), as did non-white individuals (3.6% longer BVI than whites; 95% CI 2.2%, 5.0%; p<0.001), with the greatest differences seen in young adults. Those with public and private insurance largely adhered to current guidelines (maximum average BVI of 90 and 95 days, respectively). In contrast, uninsured individuals over age 25 had a mean BVI ≥ 30 days longer than the insured. CONCLUSIONS: Frequent visits in those with CF-complications likely reflects higher need, while less frequent visits in male, non-white, and uninsured individuals may reflect patient-preference or structural barriers to care. Risk factors for gaps in care should inform changes to CF care recommendations going forward.