High unrecognized SARS-CoV-2 exposure of newly admitted and hospitalized psychiatric patients

Abstract: Background: Patients with pre-existing mental disorders are at higher risk for SARS-CoV-2 infection and adverse outcomes, and severe mental illness, including mood and psychosis spectrum disorders, is associated with increased mortality risk. Despite their increased risk profile, patients with severe mental illness have been understudied during the pandemic, with limited estimates of exposure in inpatient settings. Objective: The aim of this study was to describe the SARS-CoV-2 seroprevalence and antibody titers, and pro-inflammatory cytokine concentrations of newly admitted or hospitalized psychiatric inpatients without known history of COVID-19 infection, using robust quantitative multi-antigen assessments, and compare patients' exposure to that of hospital staff. Methods: This multi-centric, cross-sectional study compared SARS-CoV-2 seroprevalence and titers of 285 patients (University Psychiatric Centre Duffel [UPCD] N = 194; Assistance-Publique-Hopitaux de Paris [AP-HP] N = 91), and 192 hospital caregivers (UPCD N = 130; AP-HP N = 62) at two large psychiatric care facilities between January 1st and the May 30th 2021. Serum levels of SARS-CoV-2 antibodies against Spike proteins (full length), spike subunit 1 (S1), spike subunit 2 (S2), spike subunit 1 receptor binding domain (S1-RBD) and Nucleocapsid proteins were quantitatively determined using an advanced capillary Western Blot technique. To assess the robustness of the between-group seroprevalence differences, we performed sensitivity analyses with stringent cut-offs for seropositivity. We also assessed peripheral concentrations of IL-6, IL-8 and TNF-a using ELLA assays. Secondary analyses included comparisons of SARS-CoV-2 seroprevalence and titers between patient diagnostic subgroups, and between newly admitted (hospitalization 647 days) and hospitalized patients (hospitalization >7 days) and correlations between serological and cytokines. Results: Patients had a significantly higher SARS-CoV-2 seroprevalence (67.85 % [95% CI 62.20-73.02]) than hospital caregivers (27.08% [95% CI 21.29-33.77]), and had significantly higher global SARS-CoV-2 titers (F=29.40, df=2, p<0.0001). Moreover, patients had a 2.51-fold (95% CI 1.95-3.20) higher SARS-CoV-2 exposure risk compared to hospital caregivers (Fisher's exact test, P <0.0001). No difference was found in SARS-CoV-2 seroprevalence and titers between patient subgroups. Patients could be differentiated most accurately from hospital caregivers by their higher Spike protein titers (OR 136.54 [95% CI 43.08-481.98], P < 0.0001), lower S1 (OR 0.06 [95% CI 0.02-0.15], P < 0.0001) titers and higher IL-6 (OR 3.41 [95% CI 1.73-7.24], P < 0.0001) and TNF-\u3b1 (OR 34.29 [95% CI 5.00-258.87], P < 0.0001) and lower titers of IL-8 (OR 0.13 [95% CI 0.05-0.30], P < 0.0001). Seropositive patients had significantly higher SARS-COV-2 antibody titers compared to seropositive hospital caregivers (F = 19.53, df=2, P