Background/Objectives: During the SARS-CoV-2 pandemic currently ongoing worldwide, several challenges were encountered in managing patients in out-of-hospital residences. The aim of this study is to establish a risk score indicating the probability of SARS-CoV-2 infection and estimate the seroprevalence in nursing homes.

Design: This is a retrospective cross-sectional observational study.

Participants and Setting: 231 patients (median age 86 years, min 53 max 100 years) were enrolled from three nursing homes of Pavia and its surroundings.

Measurements: Medical history, clinical and instrumental data were correlated to the results of nasopharyngeal swab and serology.

Results: Patients with positive nasopharyngeal swab and/or serology were 170 (74%, 95%CI: 67%-79%) and seroprevalence was 64%. Variables associated with COVID-19 infection used to build the clinical score were: anosmia and ageusia, pulse oximetry <90%, conjunctivitis, rhinorrhea, myalgia. The probability of COVID-19 positivity increased linearly over the clinical score values (score 0: 55%; score 1: 83%: score 2+: 95%). By adding lung ultrasound to the score, a Clinical & LUS Risk Score was created, which allowed further stratification. The area under the ROC Curve for the two models was 0∙73 and 0∙77, respectively. Of note, fever, one of the pivotal signs in COVID-19 patients, was not a common manifestation among nursing home residents with SARS-CoV-2 infection, therefore it was not included in the scores.

Conclusions: Given the high SARS-CoV-2 seroprevalence in nursing homes, the Clinical San Matteo Risk Score, implemented by lung ultrasound when available, can help for an early identification, isolation and treatment of possible positive cases.