Identifying the interplay between protective measures and settings on the SARS-CoV-2 transmission using a Bayesian network

Abstract

During the COVID-19 pandemic, primary healthcare nurses played a crucial role in contact tracing to minimize the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) spread. However, contact tracing data were underutilized to explore the potential ways of virus transmission. Research on community spread, the effectiveness of individual protective measures, and the impact of different exposure environments is limited. This study focused on the transmission of SARS-CoV-2, aiming to describe the interplay between prevention measures and characteristics of exposure settings, using the data collated from a national SARS-COV-2 contact tracing programme to build a Bayesian network (BN) with 11 relevant features, which could potentially determine features with the strongest influence in SARS-CoV-2 transmission. BNs allow predictions of new scenarios when hypothetical information is introduced, which makes them of special interest in epidemiological studies. In our results, ventilation of the setting and time of exposure were the main factors for SARS-CoV-2 transmission. BN analysis can help decision-makers refine public health campaigns and priorise resources for individuals at higher risk, and also provides personalized guidance in specific protective measures for different settings or environments.