Atypical symptoms, SARS-CoV-2 test results and immunisation rates in 456 residents from eight nursing homes facing a COVID-19 outbreak.

Abstract

BackgroundFrail older persons may have an atypical presentation of COVID-19. The value of rRT-PCR testing for identifying SARS-CoV-2 nursing homes (NH) residents is not known.ObjectiveTo determine whether (i) atypical symptoms may predict rRT-PCR results and (ii) rRT-PCR results may predict immunization against SARS-CoV-2 in NH residents.DesignA retrospective longitudinal study.Settingeight NHs with at least ten rRT-PCR-positive residents.Subjects456 residents.MethodsTypical and atypical symptoms recorded in residents’ files during the 14 days before and after rRT-PCR testing were analyzed. Residents underwent blood testing for IgG-SARS-CoV-2 nucleocapsid protein 6 to 8 weeks after testing. Univariate and multivariate analyses compared symptoms and immunization rates in rRT-PCR-positive and negative residents.Results161 residents had a positive rRT-PCR (35.3%), 17.4% of whom were asymptomatic before testing. Temperature > 37.8°C, oxygen saturation < 90%, unexplained anorexia, behavioural change, exhaustion, malaise, and falls before testing were independent predictors of a further positive rRT-PCR. Among the rRT-PCR-positive residents, 95.2% developed SARS-CoV-2 antibodies vs 7.6% in the rRT-PCR-negative residents. Among the residents with a negative rRT-PCR, those who developed SARS-CoV-2 antibodies more often had typical or atypical symptoms (p = 0.02 and < 0.01, respectively).ConclusionThis study supports a strategy based on (i) testing residents with typical or unexplained atypical symptoms for an early identification of the first SARS-CoV-2 cases, (ii) rT-PCR testing for identifying COVID-19 residents, (iii) repeated wide-facility testing (including asymptomatic cases) as soon as a resident is tested positive for SARS-CoV-2, and (iv) implementing SARS-CoV-2 infection control measures in rRT-PCR-negative residents when they have unexplained typical or atypical symptoms.