Validation of an Artificial Intelligence driven framework to automatically detect red flag symptoms in screening for rare diseases in electronic health records: hereditary transthyretin amyloidosis polyneuropathy as a key example.

Rare life-threatening conditions, such as multisystemic hereditary transthyretin amyloidosis (ATTRv) polyneuropathy, are often underdiagnosed or diagnosed late in the disease course, although early diagnosis is crucial for treatment success. Red flag symptoms have been identified, but manual screening of multidisciplinary medical records on this set of symptoms is time-consuming. This study aimed to validate a Natural Language Processing (NLP) algorithm to perform such a search in an automated manner, in order to improve early diagnosis and treatment. A novel state-of-the-art NLP procedure was applied to extract red flag symptoms from patients' electronic medical records and to select patients at risk for ATTRv polyneuropathy for further clinical review. Accuracy of the algorithm was assessed through comparison with a manual standard on a random sample of 300 patients. Out of a retrospective sample of 1015 patients, the NLP algorithm yielded 128 patients with three or more red flag symptoms of which 69 patients were considered eligible for genetic testing after clinical review. High accuracy was found in the detection of red flag symptoms, with F1 scores between 0.88 and 0.98. A relative increase of 48.6% in genetic testing, to identify patients with a rare disease earlier, was demonstrated. An NLP algorithm, after clinical validation, offers a valid and accurate tool to detect red flag symptoms in medical records across multiple disciplines, supporting better screening for patients with rare diseases. This opens the door to further NLP applications, facilitating rapid diagnosis and early treatment of rare diseases.

Detection of transthyretin amyloid cardiomyopathy by automated data extraction from electronic health records.

AIMS: Transthyretin amyloid cardiomyopathy (ATTR-CM), a progressive and fatal cardiomyopathy, is frequently misdiagnosed or entails diagnostic delays, hindering patients from timely treatment. This study aimed to generate a systematic framework based on data from electronic health records (EHRs) to assess patients with ATTR-CM in a real-world population of heart failure (HF) patients. Predictive factors or combinations of predictive factors related to ATTR-CM in a European population were also assessed. METHODS AND RESULTS: Retrospective unstructured and semi-structured data from EHRs of patients from OLV Hospital Aalst, Belgium (2012-20), were processed using natural language processing (NLP) to generate an Observational Medical Outcomes Partnership Common Data Model database. NLP model performance was assessed on a random subset of EHRs by comparing algorithm outputs to a physician-generated standard (using precision, recall, and their harmonic mean, or F1-score). Of the 3127 HF patients, 103 potentially had ATTR-CM (age 78 ± 9 years; male 55%; ejection fraction of 48% ± 16). The mean diagnostic delay between HF and ATTR-CM diagnosis was 1.8 years. Besides HF and cardiomyopathy-related phenotypes, the strongest cardiac predictor was atrial fibrillation (AF; 72% in ATTR-CM vs. 60% in non-ATTR-CM, P = 0.02), whereas the strongest non-cardiac predictor was carpal tunnel syndrome (21% in ATTR-CM vs. 3% in non-ATTR-CM, P < 0.001). The strongest combination predictor was AF, joint disorders, and HF with preserved ejection fraction (29% in ATTR-CM vs. 18% in non-ATTR-CM: odds ratio = 2.03, 95% confidence interval = 1.28-3.22). CONCLUSIONS: Not only well-known variables associated with ATTR-CM but also unique combinations of cardiac and non-cardiac phenotypes are able to predict ATTR-CM in a real-world HF population, aiding in early identification of ATTR-CM patients.