LAUSR

Study Objectives: Healthcare workers (HCWs) in acute-care hospitals are inherently more likely than the general population to be exposed to the SARS-Cov-2 virus. There is great diversity among HCWs in degree of exposure based on intensity and duration of patient contact. The use of personal protective equipment and other infection prevention measures would be expected to significantly modify the risk of acquiring COVID-19. We hypothesized that patient-facing HCWs (PF) are more likely to acquire COVID-19 illness over time than non-patient facing HCWs (non-PF). Methods: All HCWs who were either employed or part of the medical staff at six acute-care hospitals in Phoenix, Arizona in June 2019 were invited to participate. A cohort of 1358 HCWs provided informed consent, filled out a questionnaire regarding their healthcare role, and had blood drawn between June 15th and August 15th, 2020 (Draw 1). The questionnaire and blood draws were repeated in October 2020 (Draw 2), January 2021 (Draw 3), and April 2021 (Draw 4). 881 physicians/APPs and nurses with direct patient care responsibilities, respiratory therapists, phlebotomists, and patient care technicians were categorized as PF, 477 other participants considered non-PF. SARS-CoV-2 anti-nucleocapsid IgG was measured using the Abbott Architect platform, using a cut-off of greater than 1.4 arbitrary units as a positive result. This assay does not detect anti- spike IgG and is therefore insensitive to Covid vaccination status. Because previous studies suggest that anti- nucleocapsid IgG levels decay over time, participants were treated as seropositive for all draws following their first positive draw regardless of the index result. Participants who missed a draw were treated as negative on that draw. Differences in seroprevalence were tested with a Z-score test for differences in proportion. Proportions were expressed as percentages +/- 95% confidence intervals. Results: Overall seroprevalence increased from 8.8% +/- 1.5% on Draw 1 to 11.3% +/- 1.7% on Draw 2, 19.9% +/- 2.1% on Draw 3, and 20.8% +/- 2.2% on Draw 4. There were no significant between-group differences in seroprevalence on Draw 1 (PF 9.7% +/- 2.0% vs non-PF 7.3% +/- 2.3%, p=0.136), but PF HCWs were significantly more likely to be seropositive on Draw 2 (12.6% +/- 2.2% vs. 9.0% +/- 2.6%, p=0.046), Draw 3 (22.3% +/- 2.7% vs 15.5% +/- 3.2%, p=0.0027), and Draw 4 (23.0% +/- 2.8% vs 16.6% +/- 3.3%, p=0.0049). See Figure 1 with CDC cumulative COVID-19 case rate for Arizona presented for reference. Subgroup analysis within the PF group shows that physicians/APPs were less likely than other PF groups to be seropositive at all time points (Draw 1: 6.2% +/- 2.8% vs. 11.3% +/- 2.5%, p=.018;Draw 2: 8.7% +/- 3.3% vs. 14.4% +/- 2.8%;p=.018;Draw 3: 17.4% +/- 4.4% vs. 24.5% +/- 3.4%, p=.019;Draw 4: 17.8% +/- 4.4% vs. 25.5% +/- 3.4%, p=0.0016) Conclusion: PF HCWs were more likely that non-PF HCWs to seroconvert acquire COVID-19. Among PF HCWs, physicians and APPs were the least likely to seroconvert, and their seroconversion rate was similar to the non-PF HCWs. [Formula presented]