Methods of identifying transthyretin amyloid cardiomyopathy in secondary data sources: a systematic review

Identifying patients with transthyretin amyloid cardiomyopathy (ATTR-CM) in secondary data sets is important for assessing effectiveness and safety of treatments.

We searched bibliographic databases from inception through February 15, 2025. These were augmented by Google Scholar and hand searches of references from identified studies. Studies were included if they utilized billing code and/or pharmacy record algorithms to identify ATTR-CM (wild-type, variant or both) in secondary datasets. Study characteristics and methodological attributes were summarized.

Twenty-five studies (26 analyses) were identified. Analyses were performed in the United States (US)(53.8%) and outside the US (46.2%). They assessed wild-type alone (46.2%), variant-type alone (11.5%) and mixed-type (42.3%). Sixteen analyses (61.5%) used International Classification of Diseases-10th-Revision (ICD-10) codes, 23.1% used ICD-9 and -10 codes, 11.5% analyses used country unique billing codes, and 11.5% relied on tafamidis prescriptions (with/without billing codes) to identify ATTR-CM. Of the 22 analyses (84.6%) using ICD codes, E85.82 alone was most used to identify wild-type (40.9%), and E85.0-E85.2 to identify variant-type. Fourteen studies (53.8%) required diagnosis codes for cardiac conditions. Exclusion criteria included codes for light chain amyloidosis (53.8%), blood cancers (38.5%) and cerebral amyloid angiopathy (30.8%). Five analyses (19.2%) used data from 2018 onward.

While we utilized bibliographic databases, gray literature sources and backward citation tracking, it remains possible that not all studies were captured.

Methods for identifying ATTR-CM in studies of secondary datasets were heterogeneous. Future research should focus on optimizing ATTR-CM identification algorithms and performing validation studies.