Cardiovascular risk factors in COVID-19 and artificial intelligence-an innovative approach to current pandemic

ABSTRACT

Background: Coronavirus 2019 (COVID-19), also known as Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), infection is a pandemic that causes acute respiratory injury, hospital admission and death. Older adults are at a higher risk of serious illness and death from this pandemic. Many COVID-19 patients have a pre-existing cardiovascular disease (CVD). We aim to develop a risk factor stratification tool, using Artificial Intelligence (AI) method, to predict mortality, ICU admission, and length of hospital stay, in patients with CVD during this pandemic. Methods: This is a retrospective cohort study. An IRB approval was obtained. Patients with confirmed (SARS-Cov-2) test, age more than 60 and older, who were admitted to the Sparrow hospital between March 2020 and October 2020 were included. CV risk factors including Hypertension (HTN), Chronic Ischemic Heart Disease (CHD), Heart Failure (HF), and Cardiac Arrhythmia (CA) were used. Results: Of the 426 patients with COVID-19(mean age74.5 years), at least 1 CVD was identified in most patients. HTN being the most common (55%), followed by CHD (22%), HF (20%) and CA (3%). Multivariable logistic regression has been conducted to identify risk factors for adverse outcomes and competing risk survival analysis for mortality. Outcomes measures included hospital stay > 7 days, ICU admission, and death. Discussion: Our data suggests patients with HTN required longer hospital stay, had higher ICU admissions and death rate. Conclusion: CV risk factors are common in older adults. HTN is the commonest CVD in this population. Several CV risk factors may contribute to the severity of COVID19 and its impact on older adults. Our study suggests that CV risk factors including HTN, HF, CHD, and CA have major impact on COVID-19 infection in hospitalized geriatric populations - see graph 1. Patients with HTN, had longer hospital stay, ICU admission, and mortality. Based on this work, we suggest that a large data sample might be required to develop an AI software that can help predict outcomes and the need for certain resources for older patients.