Validating the Data Completeness and Accuracy of the Canadian Cystic Fibrosis Registry.

Abstract

Introduction: The Canadian Cystic Fibrosis Registry (CCFR) was developed in the 1970s and has longitudinal demographic and clinical data on persons living with cystic fibrosis (CF) attending accredited clinics in Canada. We aimed to validate the data collection and identify potential limitations of the CCFR. Methods: Of 40 accredited CF clinics in Canada invited and based on an a priori sample size calculation, eight clinics were included. 15% of each CF clinic's population in 2019 were randomly selected. Data variables were selected based on their importance to care, epidemiologic trends, and data related to demography, clinic visits, and hospitalizations. The accuracy of the registry data was compared to the medical records as the gold standard. Each data element was categorized as correct, incorrect, or not able to be validated. The accuracy rate was calculated as the percent correct out of all records validated. Results: A total of 4382 individuals had data entered into the CCFR in 2019. The validation cohort consisted of 208 individuals from 8 clinics, which were representative across location, size of clinic (small/medium/large), and type of clinic (adult, pediatric, and combined). The 208 individuals were 52% male and 95% White, and with a median age of 26.3 years (IQR: 15.2-36.6). Approximately 95% of CCFR data on clinical measurements, infections, treatments, and hospitalizations validated were accurate as compared to the medical record. For demography, sex and date of birth had 100% accuracy. Conclusion: Our validation of the CCFR demonstrated high accuracy for clinical and demographic variables used in clinical research.