Impact of observability period on the classification of COPD diagnosis timing among Medicare beneficiaries with lung cancer

Healthcare data such as electronic health records and health insurance claims is a rich resource to study outcomes of chronic medical conditions such as chronic obstructive pulmonary disease (COPD). However, it can be tricky to determine the optimal period of searching the claims to identify the first diagnosis time of a chronic condition. In this study, we explored three approaches to optimize classification of COPD diagnosis based on a variety of Medicare claims search periods and observability (required continuous periods of enrollment) periods. We used a large dataset of Medicare claims linked to SEER cancer registry including more than 158,000 Medicare beneficiaries to identify diagnosis timing of COPD relative to lung cancer diagnosis. We observed that the most efficient approach to classify COPD diagnosis timing is to use all available historical claims data while requiring a minimum of one year of Medicare continuous enrollment. Our findings will help researchers concerned about accuracy of diagnosis timing of chronic conditions such as COPD and loss in sample size, especially when long periods of continuous insurance enrollment isn’t feasible.