(B-227) The Impact of COPD Severity Misclassification on Patients with Different Comorbidities Using a Validated Process: A Nationwide Cohort Study

Background: The absence of the modified Medical Research Council (mMRC) dyspnea scale and the COPD Assessment Test (CAT) in Taiwan's claims database makes accurate classification of chronic obstructive pulmonary disease (COPD) severity challenging. Misclassification may lead to inaccurate patient outcome assessments, particularly for those with comorbidities like cardiovascular disease, diabetes mellitus (DM), or chronic kidney disease (CKD), which might affect COPD diagnosis and management.

Objectives: We aimed to evaluate the impact of COPD severity misclassification in a nationwide claims database on patients with varying comorbidity profiles.

Methods: We identified a cohort from the COPD pay-for-performance (P4P) program between 2017 and 2022, using the enrollment date as the index date. The GOLD classification (Groups A-D) served as the gold standard. Patients were excluded if they had a follow-up period of less than one year or a gap exceeding 90 days between their first re-visit after enrollment. Our validation study assessed severity based on acute exacerbation (AE) hospitalizations and the use of medications such as antibiotics, steroids, SABAs, and SAMAs. Cases were considered severe if there was at least one hospitalization with medication use; otherwise, they were classified as mild if no related records existed in the year before the index date. This classification was based on the Taiwan National Health Insurance (NHI) claims database, with an accuracy rate of 78.8%. We validated severity classification by linking P4P records with the NHI database, categorizing cases as true or false positives at two severity levels: mild (GOLD A+B) and severe (GOLD C+D). We then assessed the one-year risk of COPD AE between true and false positives within each severity group. Subgroup analyses were performed based on comorbidities, including cardiovascular disease, stroke, DM, and CKD. Covariates were adjusted, including age, lung function tests, comorbidities, and COPD-related medications.

Results: A total of 22,828 patients met the study criteria. In the mild group, the adjusted hazard ratio (aHR) for AE in false positives was 1.07 (95% CI: 0.96–1.18) compared to true positives. However, misclassified patients with CKD, depression, and DM had an increased risk of AE, particularly those with depression (aHR 2.84, 95% CI: 1.41–5.71).
In the severe group, misclassification was associated with a significantly lower risk of AE (aHR 0.85, 95% CI: 0.77–0.94). This effect was more pronounced in patients with stroke (aHR 0.43, 95% CI: 0.27–0.70) and DM (aHR 0.78, 95% CI: 0.63–0.95).

Conclusions: The misclassification of COPD severity in claims databases may affect outcome assessments for COPD patients, particularly those with stroke, CKD, depression, and DM.