A Risk Prediction Model and Risk Score of SARS-CoV-2 Infection Following Healthcare-Related Exposure

Abstract

Hospital workers are at high risk of contact with COVID-19 patients. Currently, there is no evidence-based, comprehensive risk assessment tool for healthcare-related exposure; so, we aimed to identify independent factors related to COVID-19 infection in hospital workers following workplace exposure(s) and construct a risk prediction model. We analyzed the COVID-19 contact tracing dataset from 15 July to 31 December 2021 using multiple logistic regression analysis, considering exposure details, demographics, and vaccination history. Of 7146 included exposures to confirmed COVID-19 patients, 229 (4.2%) had subsequently tested positive via RT-PCR. Independent risk factors for a positive test were having symptoms (adjusted odds ratio 4.94, 95%CI 3.83−6.39), participating in an unprotected aerosol-generating procedure (aOR 2.87, 1.66−4.96), duration of exposure >15 min (aOR 2.52, 1.82−3.49), personnel who did not wear a mask (aOR 2.49, 1.75−3.54), exposure to aerodigestive secretion (aOR 1.5, 1.03−2.17), index patient not wearing a mask (aOR 1.44, 1.01−2.07), and exposure distance <1 m without eye protection (aOR 1.39, 1.02−1.89). High-potency vaccines and high levels of education protected against infection. A risk model and scoring system with good discrimination power were built. Having symptoms, unprotected exposure, lower education level, and receiving low potency vaccines increased the risk of laboratory-confirmed COVID-19 following healthcare-related exposure events.

Keywords: COVID-19; SARS-CoV-2; occupational exposure; personal protective equipment; risk factors.

Figure 1

Consort type study flow diagram.

Figure 2

Areas under the ROC curve for the final logistic model (A) and the risk scoring system (B).