News: COVID-19 complications can be assessed with natural language processing, study shows

Natural language processing (NLP) can be used to sort through EHR data of COVID-19 patients and “map the relationships between complications, comorbidities, and outcomes” in the hospitalized patients, according to a study published by NPJ Digital Medicine. The study aims to understand the relationship between pre-existing conditions and complications of COVID-19.

Researchers looked at the clinical notes of 1,803 COVID-19 patients to compare pre- and post-COVID-19 information. Using NLP, the study used 20 risk factors for COVID-19 severe illness and 18 COVID-19 associated complications.

The study found that the most frequent comorbidities were hypertension (27.7% of patients), Type 2 diabetes mellitus (15.4%), cancer (14.1%), and obesity (12.26%), “reflecting the most common …chronic diseases in the US.”

The most common COVID-19 complications found by the study were respiratory, cardiovascular, acute kidney injury, anemia, sepsis, and diabetic decompensation.

“In the case of pleural effusion, which remains the most frequent complication, the prevalence decreases from 4.9% (89 patients) during the early onset time period (days 1–30) to <1% (20 patients) during the later onset time periods (days 31–90),” the study says.

“In particular, patients with cardiomyopathy (2/56), chronic kidney disease (4/235), coronary artery disease (3/112), heart failure (3/138), and hypertension (5/499) appear to be more susceptible. Patients with liver disease, stroke, and Type 1 diabetes also appear to be more susceptible to complications during days 31–90 post-infection.”

The study also reviewed EHR data from patients who were negative for COVID-19 and found that hypertension is the most significant risk factor aside from deep vein thrombosis.

“Specifically, our data suggest that a recent history of hypertension is the strongest predictor of ARDS, the most significant and life-threatening complication of COVID-19, among hospitalized COVID-19 patients, similar to previous observations,” the study says.

Editor’s note: The NPJ Digital Medicine published study can be found here.