Augmented Curation of Unstructured Clinical Notes from a Massive EHR System Reveals Specific Phenotypic Signature of Impending COVID-19 Diagnosis

Understanding the temporal dynamics of COVID-19 patient phenotypes is necessary to derive fine-grained resolution of the pathophysiology. Here we use state-of-the-art deep neural networks over an institution-wide machine intelligence platform for the augmented curation of 8.2 million clinical notes from 14,967 patients subjected to COVID-19 PCR diagnostic testing. By contrasting the Electronic Health Record (EHR)-derived clinical phenotypes of COVID-19-positive (COVIDpos, n=272) versus COVID-19-negative (COVIDneg, n=14,695) patients over each day of the week preceding the PCR testing date, we identify diarrhea (2.8-fold), change in appetite (2-fold), anosmia/dysgeusia (28.6-fold), and respiratory failure (2.1-fold) as significantly amplified in COVIDpos over COVIDneg patients. The specific combination of cough and diarrhea has a 4-fold amplification in COVIDpos patients during the week prior to PCR testing, and along with anosmia/dysgeusia, constitutes the earliest EHR-derived signature of COVID-19 (4-7 days prior to typical PCR testing date). This study introduces an Augmented Intelligence platform for the real-time synthesis of institutional knowledge captured in EHRs. The platform holds tremendous potential for scaling up curation throughput, with minimal need for training underlying neural networks, thus promising EHR-powered early diagnosis for a broad spectrum of diseases.