AB031. Validation of claims approach to identify asthma and COPD overlap syndrome patients in the United States

Background: With the increasing recognition of asthma and chronic obstructive pulmonary disease (COPD) overlap syndrome (ACOS) and its significant disease and economic burden, knowledge about real-world patient characteristics and how they are treated is needed to aid in better understanding this phenotype. However, ACOS patient identification is challenging. Currently, there is no validated claims-based ACOS patient selection algorithm available to assist this type of research. To assess the validity of a retrospective claims-based algorithm to identify high likelihood ACOS patients using patient medical records as the criterion.

Methods: Patients were identified from the US based HealthCore Integrated Research Database (HIRD) between January 1, 2006 and October 31, 2014. A claims-based algorithm to identify high likelihood ACOS patients was based on the following inclusion criteria: age ≥40 years, ≥2 ICD-9 diagnoses (≥30 days apart) for asthma, (ICD-9 CM: 493.xx), ≥2 diagnoses (≥30 days apart) for COPD (ICD-9 CM: 491.xx, 492.xx, and 496.xx), ≥2 ICD-9 procedure, Current Procedural Terminology (CPT), or Healthcare Common Procedure Coding System (HCPCS) codes (≥30 days apart) for COPD-related procedures, ≥3 prescription fills (≥30 days apart) for asthma/COPD medication, and ≥2 CPT codes for spirometry test. Patients were indexed on the earliest date when all of the inclusion criteria were met. Patients diagnosed with cancer (ICD-9 CM: 140.xx–209.3x, 230.xx–234.xx) during the 12-month pre-index period were excluded. Outpatient medical records closest to (before or after) the index date were abstracted. Based on the chart review, at least 2 features (positive history of allergic rhinitis, chronic sinusitis, eczema, positive skin test to environmental allergens or desensitization to environmental allergens, diagnosis of asthma or past medical history for asthma before 40 years of age, or family history of asthma) were needed to confirm the asthma component; a post-bronchodilator FEV1/FVC ratio of <70% confirmed persistent airflow limitation and thus the COPD component. ACOS was confirmed when both asthma and COPD criteria were met.

Results: A total of 20,459 high likelihood ACOS patients were identified from the HIRD. Among 5,000 patients from this population were randomly sampled and submitted for medical chart abstraction. Of those, 3,038 could not be obtained because they no longer existed, the provider could not be located, or the provider refused. Among the 1,962 records that were obtained, 1,181 were unusable, primarily because they contained no spirometry results or failed to provide FEV1/FVC values. Thus, 781 medical records were received and reviewed. Among those, 391 (50.1%) were confirmed as ACOS through chart review. Non-confirmed patients included 206 (26.4%) asthma only (showing no evidence for COPD), 106 (13.6%) COPD only (showing no evidence for asthma), and 78 (10.0%) showing no chart evidence for either asthma or COPD.

Conclusions: A patient chart review was able to corroborate only 50% of the patients identified by the retrospective claims-based algorithm as potentially diagnosable with ACOS. It appears that retrospective research on ACOS patients still requires improvement in patient identification methods. Improvement in this area will likely require longitudinal studies using a battery of claims data and electronic records, patient charts, and patient and physician surveys.

Keywords: Asthma and chronic obstructive pulmonary disease overlap syndrome; identification; claims approach