(185) Descriptive Characteristics and Validation of Fibrotic Interstitial Lung Disease Patients (ILD) with and without progression Using a Claims Based Algorithm
Associate Director Carelon Research Wilmington, United States
Background: Interstitial lung disease (ILD) results from approximately 200 heterogeneous etiologies of pulmonary parenchymal disorders; patients with certain etiologies may be more likely to develop progressive fibrosing-ILD (PF-ILD). Currently, there is a limited understanding of how to identify patients with PF-ILD using claims codes. Additionally, there is a need for a greater understanding of the characteristics of PF-ILD patients and how to best categorize PF-ILD within administrative claims databases.
Objectives: We aim to present the results of a validation of a PF-ILD claims based algorithm using medical records. In the first step towards validation, we will describe the demographic and clinical characteristics among fibrosing ILD patients who progressed to a PF-ILD as compared to those who did not, according to a claims-based algorithm.
Methods: Previous studies have used a clinically informed claims-based algorithm to identify patients with PF-ILD. This algorithm requires two or more diagnoses of fibrotic ILD within 365 days, progression is defined when the patient meets one of several criteria (e.g., ≥1 claim for oxygen therapy). We applied this algorithm to identify patients who developed PF-ILD in the HealthCore Integrated Research Database (HIRD) from October 2016 until October 2020. We reported demographic and clinical characteristics at ILD diagnosis among those with incident PF-ILD based on our algorithm compared to those without. Additionally, we will use available medical records and high-resolution CT imaging data, which is currently being collected and adjudicated, to assess the validity of our algorithm.
Results: We identified 41,621 ILD patients in the HIRD, 30,153 (72%) of whom the disease was later characterized as PF-ILD. The average age at ILD diagnosis of those with subsequent PF-ILD was 69 years (SD:14.5) as compared to 65 years (SD: 14.8) for those who did not have progressive disease. Individuals who met the criteria for PF-ILD were more likely to have pulmonary hypertension (15% vs 7%), pulmonary embolism (7% vs. 4%), lung cancer (7% vs. 3%), chronic obstructive pulmonary disease (56% vs. 38%), and pneumonia (48% vs. 26%) prior to their ILD diagnosis as compared to those who did not progress to PF-ILD.
Conclusions: Commercially insured individuals who developed PF-ILD were older and more likely to have comorbidities before their ILD diagnosis as compared to those who did not progress to PF-ILD. Additionally, we plan to present the early results of our validation of the clinically informed claims-based algorithm.
