Study population and setting
This outbreak investigation included 394 residents at four nursing homes in central London and a convenience sample of 70 staff from three of the nursing homes. Comprehensive testing with rt-PCR for SARS-CoV-2 began on April 15, 2020, and residents who tested negative were re-tested one week later. SARS-CoV-2 genomes were extracted from nineteen samples and sequenced to evaluate genomic diversity.
Summary of Main Findings
Between March 1 and May 1, 2020, 103 (26%) of 384 residents in these four nursing homes died, approximately triple the death rate compared to the same time period in the two years prior. Just over half of the deaths (n=53, 54%) were attributed to COVID-19. Among 313 residents tested for SARS-CoV-2, 126 (40%) were initially positive. Upon re-testing (n=173), five additional residents tested positive. Many residents with COVID-19 (54/126, 43%) did not show any symptoms. Among those who had symptoms, almost one-third (22/72, 31%) did not have any symptoms considered typical of COVID-19 (i.e., cough, breathlessness, and fever). The symptom most associated with a positive SARS-CoV-2 test was new onset of anorexia. Large proportions of staff members (ranging from 11% to 22% across centers) called in sick or self-isolated during the study period. Among 70 staff members who were not exhibiting COVID-like symptoms and did not self-isolate during the study period, three tested positive for SARS-CoV-2. There were multiple distinct clusters of SARS-CoV-2 from genomic sequencing, even within single nursing homes.
This was a comprehensive outbreak investigation at four nursing homes in central London. Residents were tested at two time points. A convenience sample of nursing home staff, covering a range of staff roles, were additionally tested to assess infection prevalence and potential transmission.
COVID-related causes of death may have been under-reported in the early phases of the outbreak, especially due to the atypical presentation among the nursing home residents. Some participants (<20, though the exact number is unclear) declined or were unavailable for testing, so the point-prevalence of infection may be under-estimated. It is not clear how many participants declined to participate at the second testing point. The initial tests were conducted with combined oropharyngeal and nasopharyngeal tests, but the procedure was revised to bilateral anterior nasal swabbing on April 29, which may result in some misclassification due to different test performance during the second testing phase, but this is unlikely to affect the results given increasing evidence that the approaches are equivalent.
This was a well-conducted outbreak investigation with two point-prevalence surveys across four nursing homes in central London.
This review was posted on: 4 July 2020