Assessing Smoking Status and Risk of SARS-CoV-2 Infection: A Machine Learning Approach among Veterans

Area of Science:

• Epidemiology
• Public Health
• Infectious Diseases

Background:

• The association between smoking and SARS-CoV-2 infection risk remains incompletely understood.
• Electronic health records offer a valuable resource for investigating population health trends.

Purpose of the Study:

• To evaluate the impact of smoking status on the risk of SARS-CoV-2 infection.
• To identify factors associated with SARS-CoV-2 positivity in a large veteran cohort.

Main Methods:

• Retrospective cohort study utilizing electronic medical records of veterans tested for SARS-CoV-2 between February 2020 and February 2021.
• Veterans were categorized into Never Smokers (NS), Former Smokers (FS), and Current Smokers (CS).
• A cascade machine learning algorithm was employed to calculate adjusted odds ratios (aOR) for potential confounders.

Main Results:

• A total of 1,176,306 veterans were tested, with an overall positivity rate of 19.6%.
• Current Smokers (CS) exhibited the lowest positivity proportion (11.5%), followed by Former Smokers (FS) (19.2%) and Never Smokers (NS) (22.0%).
• Adjusted odds of testing positive were significantly lower for CS (aOR: 0.51) and FS (aOR: 0.89) compared to NS. Pre-existing conditions like dementia and pneumonia increased risk, while phenylephrine use and cancer history decreased risk.

Conclusions:

• Current and former smoking are associated with a reduced risk of SARS-CoV-2 infection.
• Smoking status is an important factor influencing COVID-19 risk, necessitating further research into the underlying mechanisms.