We were able to gather information about signs and symptoms about NH residents with confirmed COVID-19 in the week prior to the RT-PCR SARS-CoV-2 test up to two weeks after testing; we also collected data about 30 day mortality. Of the residents with confirmed COVID-19, 44% of the residents were tested based on the presence of core symptoms (S-based) and 56% of the residents were tested based on transmission prevention (TP-based). CT-values did not differ significantly between both groups. In the 7 days prior to the test, only fever was observed as core symptom in the S-based group. We found that falling and somnolence were the most common reported other symptoms prior to the test in the TP-based group. Thus, extra attention should be paid to these symptoms since they might indicate a possible infection. This is in line with previous studies, in which symptoms such as falling and delirium were reported, but in which debuts with core symptoms were more common13–15. In the 14 days following the test, apart from the core symptoms, other symptoms that were common were decreased appetite, body weakness, dysphagia, somnolence, confusion and being anxious. Residents with confirmed COVID-19 who were tested based on transmissions prevention (TP-based) and not because of the presence of core symptoms on day of testing did develop core symptoms (fever, cough, dyspnea) in 73% of the cases, but 27% did not. To prevent transmission, we cannot base the testing policy on solely testing residents with core symptoms. The results of this study ask for repeated testing of all residents that have been in close contact with a resident with confirmed COVID-19. Furthermore all residents living together with a resident with confirmed COVID should be isolated for a quarantine period and personal protective equipment should be used in the care of these residents. Repeated testing of all care personnel of these wards is also appropriate16.
Our results thus support the approach for repeated testing, irrespective of symptoms, in skilled nursing facilities that has been advocated since May 202017.
Our study showed no difference between CT-values of S-based residents and TP based residents, similar to previous studies.18–19, 16,20
In our study, 40% of the residents with confirmed COVID-19, died within 30-days. This percentage is higher than the 34% reported by McMichael et al.,21 and the 26% reported by Arons et al2. This differences could be due to a difference in the prevalence of dementia, in our study the prevalence is much higher compared to the other studies. Moreover, we observed that residents with core symptoms on day of testing (S-based testing) were 2.5 times more likely to decease within 30 days than residents without these core symptoms on day of testing (TP-based). Thus, even though for transmission prevention all residents should be tested irrespective of symptoms, staff should still be more alert when the core symptoms are present since the mortality risk is higher for these residents as compared to residents without these symptoms. However, please note that mortality rate was still high 27% in the TP-based group and that a large part of the TP-based group did develop core symptoms after ~ 4 days.
We are the first to show day-to-day fluctuations of the oxygen saturation in COVID-19 positive NH residents; oxygen saturation decreases approximately two days before testing in all residents who tested positive. We observed that oxygen saturation was lower for the S-based group prior to test than for the TP-based group. In addition, we observed an increase in temperature ~ 3 days prior to test. Day-to-day fluctuations in temperature in residents with COVID-19 were described previously9. Yet, our day-to-day temperature measurements add to this by making a distinction between NH residents with and without core symptoms on day of testing. We observed that temperature for the NH residents without core symptoms at day of test (TP-based) also increased 2 days prior to test. These results of the daily fluctuations in temperature and oxygen saturation (shifts 1 to 2 days before the positive test) can contribute to earlier testing and earlier detection of the 'invisible' SARS-CoV-2 virus. If testing is done sooner, the residents can also be placed in a separated cohort earlier and the chance of possible spreading of the SARS-CoV-2 virus is smaller.
Furthermore, measuring the temperature and oxygen saturation on a daily basis is non-invasive and is independent from the ability to express symptoms. Many NH residents frequently have difficulty putting their symptoms into words and therefore run the risk of being missed. For example, our results demonstrated that almost all residents who were tested based on transmission prevention were living in a psychogeriatric ward and had dementia. By measuring daily changes in their temperature and oxygen saturation we may reduce the chance of transmission in this vulnerable group of NH residents.
This study had some limitations that must be acknowledged when interpreting results. The study was carried out within one organization and based on EHR data. Because of the use of EHR data we may have missed symptoms and also workload during the COVID-19 pandemic may have influenced symptom registration in the EHR negatively. As we used both care staff and physicians registration, this might have been somewhat overcome. Measuring the oxygen saturation and especially the temperature (tympanic and rectal) will not have been the same in every ward, but it is close to practice. A RT-PCR test has relatively low sensitivity (63–78%)22. Consequently we will have missed cases of COVID-19.