Impact of Observability Period on the Classification of COPD Diagnosis Timing among Medicare Beneficiaries with Lung Cancer

Eman Metwally, Sarah E. Soppe, Jennifer L. Lund, Sharon Peacock Hinton, Caroline A. Thompson

Abstract

Investigators often use claims data to estimate the diagnosis timing of chronic conditions. However, misclassification of chronic conditions is common due to variability in healthcare utilization and in claims history across patients.

Introduction

Investigators often want to use administrative healthcare data to estimate the timing of initial diagnoses of chronic conditions to examine its association with health outcomes. However, healthcare utilization (HCU) data (electronic health records (EHRs) and administrative claims) can be discontinuous based on several factors such as insurance coverage, healthcare access, and severity of the underlying condition, which could prevent achieving this goal. To estimate the time of first diagnosis of a chronic condition, the current recommendations suggest ensuring at least one or two years of observable lookback period prior to the appearance of the chronic condition diagnosis in the claims [1]. 

The observability period is frequently approximated by the continuous enrollment period (i.e., membership period in coverage through a particular insurance provider), while the look back period (LBP) is the claim search period for indicators of case definition before the index date (e.g., the baseline date or date of diagnosis of a second condition).

Methods
Data source
We used the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) cancer registry data linked to Medicare enrollment and insurance claims data from the Centers for Medicaid and Medicare Services (CMS). The SEER-Medicare data reflect linkage of two large population-based sources to provide detailed information about Medicare beneficiaries with cancer. The SEER database includes data from 21 cancer registries and represents approximately 40% of the US population [14]. The SEER registry file includes clinical, sociodemographic and cause of death information for patients with cancer.

Results

The baseline study population included 185,405 older adults with lung cancer with at least one year of continuous Medicare enrollment, of which 70.8% had COPD based on one-year LBP. The mean age at lung cancer diagnosis was 76.4 years old, 50.9% were female, 84.3% were Non-Hispanic White, 7.2% were Non-Hispanic Black, and 4% were Hispanic. Requiring longer continuous enrollment reduced the distribution of patients younger than 70 years old (from 19.3% to 0%), increased prevalence of COPD diagnosis (from 70.8% to 74.6%), without notable changes in other sociodemographic or clinical characteristics. 

Discussion

In this study, we explored several alternate approaches for searching Medicare claims to identify the best lookback and observability (i.e., continuous enrollment) periods to ascertain diagnosis timing of a chronic condition such as COPD and provide a framework for future studies that would utilize one of these approaches in their methods. Using a large population dataset (SEER-Medicare), we observed that all available LBP with at least one year of continuous enrollment is the most efficient approach to mitigate COPD misclassification while minimizing losses in sample size.

Conclusion

Exposure misclassification of a chronic condition is common using shorter versus longer LBPs in administrative claims data. The length of optimum LBP and continuous enrollment (i.e., observability) period depends on the context of the research question and the data generating mechanisms. In Medicare FFS claims, we estimated that all available LBP with one year of required continuous enrollment was the best approach to tradeoff between COPD diagnosis misclassification and loss in sample size.

Citation: Metwally E, Soppe SE, Lund JL, Hinton SP, Thompson CA (2024) Impact of observability period on the classification of COPD diagnosis timing among Medicare beneficiaries with lung cancer. PLOS Digit Health 3(10): e0000633. https://doi.org/10.1371/journal.pdig.0000633

Editor: Catherine G. Bielick, Beth Israel Deaconess Medical Center, UNITED STATES OF AMERICA

Received: March 27, 2024; Accepted: September 5, 2024; Published: October 22, 2024

Copyright: © 2024 Metwally et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data used in this study can’t be shared because it is a sample of SEER-Medicare data, which is owned by the National Cancer Institute. To obtain a data use agreement (DUA), please contact the National Cancer institute at: https://healthcaredelivery.cancer.gov/seermedicare/obtain/seerdua.pdf.

Funding: This work was supported by the National Institute of Environmental Health Sciences: 2T32ES007018 to EM and the National Cancer Institute: R01CA264176 to CAT. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

 

 

Source: https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0000633#sec022