ESC Heart Failure | Publication | Detection of ATTR-CM by automated data extraction from EHRs
Information held in Electronic Health Records (EHRs) hold a significant opportunity to provide physicians and researchers with better and more insights to improve disease management and treatment.
Click below to read about this study together with Heart Center Aalst, in which complete EHRs of Heart Failure (HF) patients were processed and analyzed using LynxCare’s cutting-edge NLP data processing technology in search of phenotypes compatible with ATTR-CM.
Key findings include:
* Among 3127 HF patients, 103 potentially had ATTR-CM
* Average diagnostic delay between HF and ATTR-CM diagnosis: 1.8 years
* Atrial fibrillation (AF) emerged as a significant cardiac predictor
* Carpal tunnel syndrome identified as a strong non-cardiac predictor
* Strongest combination predictor: AF, joint disorders, and HF with preserved ejection fraction
The NLP model demonstrated commendable performance in identifying patients with ATTR-CM, holding great potential for future studies and clinical practice. The study also underscores the significance of early ATTR-CM diagnosis for effective disease management. Beyond established variables, novel combinations of cardiac and non-cardiac phenotypes offer valuable insights for early patient identification.
Key findings include:
* Among 3127 HF patients, 103 potentially had ATTR-CM
* Average diagnostic delay between HF and ATTR-CM diagnosis: 1.8 years
* Atrial fibrillation (AF) emerged as a significant cardiac predictor
* Carpal tunnel syndrome identified as a strong non-cardiac predictor
* Strongest combination predictor: AF, joint disorders, and HF with preserved ejection fraction
The NLP model demonstrated commendable performance in identifying patients with ATTR-CM, holding great potential for future studies and clinical practice. The study also underscores the significance of early ATTR-CM diagnosis for effective disease management. Beyond established variables, novel combinations of cardiac and non-cardiac phenotypes offer valuable insights for early patient identification.
