Abstract
OBJECTIVE: Acute exacerbations of COPD(AECOPD) can have severe impacts on patients with the disease and a heavy burden on health care resources. Electronic health records (EHRs) are a valuable resource for identifying cases of AECOPD and research. Studies have attempted to validate case definitions of AECOPD and this review aims to summarize validated AECOPD definitions in EHRs and to provide guidance on the best algorithms to use to ensure accurate cohorts of AECOPD cases are available for researchers using EHRs. METHODS: MEDLINE and Embase were searched and studies that met the inclusion criteria were reviewed by ≥2 reviewers. Data extracted included the algorithms used to identify AECOPD, the reference standards used to compare against the algorithm, and measures of validity. The risk of bias was assessed using QUADAS-2 adapted for this review. RESULTS: Out of 2784 studies found by the search strategy, 12 met the inclusion criteria. The clinical terminology used to build algorithms to detect AECOPD included codes from the International Classification of Diseases (ICD) Ninth Revision, Clinical Modification and Tenth Revision (ICD-9-CM and ICD-10), along with the Read codes from United Kingdom general practices. AECOPD can be identified within EHRs using validated definitions, however, the validity of AECOPD definitions varies considerably depending on the algorithm used and the settings to which they are applied. CONCLUSION: Although there are validated definitions that can be used to identify AECOPD, there is no clear consensus on which provides the highest validity or the most sensitive and specific definition to use.