Clinical outcomes of COVID-19 in long-term care facilities for people with epilepsy

At STE, TM and YE, early preventative measures were implemented to different degrees, but no on-site testing was available initially, with individuals only tested when admitted to hospital. Individuals were isolated within their rooms whilst presenting with COVID-19 like-symptoms, and/or transferred to dedicated units upon return from hospital, if the diagnosis was confirmed. Testing for caregivers with symptoms became available at testing stations from mid-April 2020, and on-site testing for symptomatic individuals since early May. At STE, all 146 asymptomatic individuals were tested between 29 May and 05 June, and again once since then, with weekly testing of a random sample of 50 people, either residents or staff.

Data availability

The authors confirm that the data supporting our findings are available from the corresponding author, upon reasonable request from bona-fide researchers.

3. Results

We report the outcomes in 2047 people living and working in four different long-term care-facilities, home for 404 residents with an age range of 8-91 years.

3.1 CCE

3.1.1 Testing of residents with symptoms suggestive of COVID-19

The first (#1-1) tested positive on 03 April and was an individual in their 60s, living in a large nursing home consisting of two units with 9-10 people each. This person had severe epilepsy and multiple comorbidities, including dysphagia with percutaneous endoscopic gastrostomy (PEG) in situ. They became symptomatic on the evening of 02 April, with vomiting and subsequently pyrexia possibly related to aspiration, rapid and severe clinical deterioration with reduced oxygen saturation at ∼70%, persistent high temperature not responsive to paracetamol, reduced conscious level (Glasgow Coma Scale <5). Transfer to hospital was promptly arranged and the person tested positive on 03 April; following further deterioration, death occurred six days after symptom onset.

The second (#1-2) was an individual in their 60s, with a genetic epilepsy and co-morbidities who lived in a large unit of 19 people with four self-contained flats each housing 4-5 people. On 09 April, they became pyrexial (38.7°C) and were promptly isolated in a single room in SWGC, tested and confirmed positive. They remained clinically stable until day 3, when oxygen saturation dropped to ∼85% leading to a transfer to our linked hospital facility (UCLH), given the risk of further deterioration. They tested positive again on days 7, 14 and 18, but remained clinically asymptomatic following admission, without pyrexia, and discharged back to CCE on day 40, after testing negative on two consecutive occasions.

As of 14 September, ten other individuals were promptly isolated due to the development of temperature above 37.8°C, with or without respiratory symptoms: all have repeatedly (minimum twice) tested negative and were discharged back to their residences and de-isolated 48 hours after symptom resolution.

3.1.2 Testing of asymptomatic residents

On 17 April 2020, CCE started regular weekly surveillance of residents. Of the remaining 96 people, seven were not tested in the first round as five declined and two had temporarily moved out. Of the 89 tested, four were positive (4.5%) and were immediately isolated.

On 22 April, in the second surveillance round, 95/96 were tested as only one declined Three who previously tested negative were now positive but remained asymptomatic throughout.

On 27 April, in the third surveillance round, all 96 people tested negative, including one of the asymptomatic individuals who had twice tested positive previously.

No furher positive individuals were identified in 13 further surveillance rounds, up to 14 September.

3.1.3 Contact tracing and surveillance of care staff

Following confirmation of a positive result, testing of caregivers who had been in contact over the previous two weeks with positive individuals was performed within three days. A total of 150 caregivers accepted testing; only one symptomatic caregiver tested positive on 11 April, before enhanced surveillance of residents started on 17 April. From 30 April onwards, weekly surveillance of all asymptomatic 275 caregivers has been implemented: only one tested positive, on 04 June, with two negative re-tests on 08 and 10 June. The symptomatic caregiver positive on 11 April fully recovered and tested negative on 17 April and repeatedly until 19 June, when although completely asymptomatic tested positive again. This individual has repeatedly tested negative since. On 15 May, this carergiver had positive antibody titres, suggestive of a previous infection with SARS-CoV-2. When antibodies were re-tested on 22 June, titres for Nucleocapsid, receptor binding domain and full trimeric spike were raised, suggestive of an acute re-infection.

3.2 STE:

3.2.1 Testing of residents with symptoms suggestive of COVID-19

The first (#2-1) was a young adult with epilepsy following encephalitis aged 2, dysphagia with PEG in situ and severe intellectual disability who lived in a unit with eight other people. They were admitted to hospital on 05 March, with aspiration pneumonia following an episode of vomiting, tested then negative, and was discharged 09 March. Two weeks later, on 23 March, he presented with a new cough and pyrexia, was transferred back to the hospital the same day, and then tested positive. Ventilation became necessary. Death occurred 11 days after symptom onset.

The second (#2-2) was an individual in their 50s, with a genetic epilepsy who lived in the same unit as #3. On 09 April, this individual became symptomatic with fever, lethargy and cough for 1 week, after which they rapidly deteriorated with respiration rate >32 per minute and oxygen saturation <88%. They were promptly isolated and confirmed positive on 20 April, and remained in isolation until 05 May. One caregiver at the same unit showed symptoms on the same day as #2-2, and tested positive. Another caregiver was asymptomatic and tested positive on 08 May.

The third (#2-3) was an individual in their late 50s, with refractory epilepsy of unknown cause and moderate intellectual disability, who lived in a different unit to #2-1 and #2-2. The individual became symptomatic on 22 April, with mild fever and cough, but would not consent to isolation in room and so was moved to an unused area of another building. Supplemental oxygen was used for the first few days as his oxygen saturation fell <90%, but, overall, symptoms remained mild. A positive result for COVID-19 testing was received on 01 May. The fourth (#2-4) was an individual in their early 20s, with a genetic epilepsy who was a boarder in college. The individual became symptomatic on 28 May, with mild fever. They were admitted to A&E with oxygen saturation <88% on 30 May, discharged that evening and transferred to isolation unit. A positive test result was received on 30 May; a re-swab on 03 June returned a negative result. One caregiver working in the college, but also in a hospital, was symptomatic and tested positive on 30 May.

Prior to 29 May, eight further resident were promptly isolated as they become symptomatic but only six were tested (testing was not available for the other two), and all were negative. All eight individuals were discharged back to their residences and de-isolated 24-48 hours after symptom resolution.

3.2.2 Testing of asymptomatic residents

A fifth (#2-5) individual tested positive on 07 May, during one of their frequent hospital admissions for recurrent urinary tract infections, but was considered asymptomatic for COVID-19 as malaise was attributed to the other health conditions, and was tested negative prior to discharge on 13 May. This individual in their late 40s lives in a different unit than the three symptomatic individuals tested positive. One caregiver from the same unit became symptomatic on 11 April and another on 08 May: both tested positive.

Between 29 May and 05 June 2020, all asymptomatic individuals living on-site were tested. Of the 146 tested, one young adult living as a boarder attending college was found to be positive. This individual attended class together with the symptomatic boarder #2-4. One caregiver working in the college also become symptomatic a few days earlier and tested positive on 30 May. Since early June, a random sample of 50 residents or caregivers were tested weekly, so that all have been tested twice since June. One further asymptomatic individual tested positive and isolated.

3.2.3 Contact tracing and surveillance of care staff

From 06 April onwards, testing was available for symptomatic caregivers and those needing to self-isolate for 14 days if a household member had symptoms. Contact tracing was implemented from 02 May, with testing of all caregivers who had contact with positive individuals. Of the 601 workforce, 105 were tested once, 13 symptomatic caregivers tested positive. Enhanced surveillance was implemented at the end of May, with 50 random samples from caregivers, so that all staff members have tested twice. An additional four asymptomatic caregivers were found positive after introducing contact tracing and enhanced surveillance.

3.3 TM:

3.3.1 Testing of residents with symptoms suggestive of COVID-19

By 14 September 2020, eight symptomatic residents were identified amongst the 80 people living on-site (10%). There was no access to viral testing, but they were promptly isolated for at least 48 hours after complete resolution of the symptoms

3.3.2 Testing of asymptomatic residents

There was no routine asymptomatic screening.

3.3.3 Contact tracing and surveillance of care staff

Up until 05 June, 26 of 250 staff were symptomatic, and have been tested, with two positive results.

3.4 YE:

3.4.1 Testing of residents with symptoms suggestive of COVID-19

By 14 September 2020, eight symptomatic individuals were identified amongst the 80 people living on-site (10%). All tested negative; seven were tested once, one individual twice for persistent COVID-like symptoms.

3.4.2 Testing of asymptomatic residents

There was no routine asymptomatic screening at TM or YE.

3.4.3. Contact tracing and surveillance of care staff

There was no systematic testing, but at least 22 of 517 student-facing caregivers are known to have attended communal testing centres throughout this period. Only one agency nurse who had worked also at other facilities was severely ill and tested positive.

4. Discussion

We report confirmed COVID-19 outbreaks in two out of four care facilities for people with epilepsy and additional co-morbidities. Less than 3% of individuals living in the two facilities showed COVID-19 related symptoms and tested positive. Enhanced surveillance, available at CCE, showed a high rate of asymptomatic SARS-CoV-2 infected individuals (8/10 testing positive; 80%). Our case fatality rate was high (CCE: 50%, or 10% corrected for asymptomatic; STE: 25%), but the total number of deaths, one at each of the two centres, was in line with the average death rate over similar observation periods over the last five years at each facility: there was no excess of deaths.

We succeeded in containing a widespread outbreak of SARS-CoV-2 in six of seven care units at CCE with a low rate of spread, i.e. only one infected individual per individual care unit, with no established resident-to-caregiver transmission. Only one caregiver tested positive during immediate contact tracing and none of the 275 caregivers during the weekly surveillance phase. A second outbreak in two care-units was detected early through enhanced surveillance, with one asymptomatic caregiver and one asymptomatic individual living in the facility who both tested positive, without further positive test results on immediate contact tracing and weekly surveillance since. In contrast, at STE, initially without enhanced surveillance, 13 symptomatic caregivers tested positive out of 106 since testing of symptomatic caregivers became available at STE on 6 April. An additional four asymptomatic caregivers tested positive during contact tracing and enahced surveillance. Infections of individuals living in the facilities and amongst caregivers were widespread across almost all care units at CCE (6/7) and STE (11/11). Whilst the spread of infections was contained at CCE within 3 weeks, positive test results at STE were seen throughout the 26 weeks’ observation period (see Figure 2). While symptom severity was similar between the two sites, we cassume that the difference in numbers of infected staff (3/275 at CCE vs 17/601 at STE, P

Care facilities are highly vulnerable to COVID-19 outbreaks [

9

COVID-19: Care home deaths in England and Wales double in four weeks.

,

10

COVID-19: UK government is urged to publish daily care home deaths as it promises more testing.

,

18

• Xydakis M.S.
• Dehgani-Mobaraki P.
• Holbrook E.H.
• Geisthoff U.W.
• Bauer C.
• Hautefort C.
• et al.

Smell and taste dysfunction in patients with COVID-19.

], and it is crucial to identify effective strategies to prevent infection and to reduce impact. The approach reported here focused on two main strategies: (1) early on-site enhancement of preventative and infection control measures, (2) early identification and isolation of symptomatic individuals, with enhanced surveillance and isolation of asymptomatic people living and working at CCE as an additional measure. All centres were able to implement isolation of suspected and confirmed residents in empty or re-purposed units (see Figure 1 for CCE), avoiding hospital admission and allowing continuity of care by staff acquainted with the individuals. The use of PPE was enforced early during the pandemic, but to different degrees (see Table 1), mainly depending on open market sourcing rather than centralized procurement[

19

Green DA, Zucker J, Westblade LF, Whittier S, Rennert H, Velu P, et al. Clinical Performance of SARS-CoV-2 Molecular Testing. J Clin Microbiol; 58(8):e00995-20.

]. Similar early implementation of these measures in a care-facility in the US has been reported to be effective in minimizing viral spread[

20

• Meng H.
• Xiong R.
• He R.
• Lin W.
• Hao B.
• Zhang L.
• et al.

CT imaging and clinical course of asymptomatic cases with COVID-19 pneumonia at admission in Wuhan.

]. Whilst this is reassuring, suggesting that PPE and good hand hygiene can effectively prevent transmission when in contact with confirmed positive individuals, caregivers themselves must have been pre- or asymptomatic earlier and so, unknowingly, infected colleagues and individuals under their care, as happened at CCE. The initial spread of infection across the sites, very likely caused by healthcare workers from different care units sharing accommodation (see Figure 2), questions the initial advice to healthcare workers of continuing to go to work despite household members self-isolating.

Individuals in all centres had different degrees of intellectual disability, such that it was not possible to assess reliably for the presence of non-respiratory symptoms, which have been described involving various organs[

3

https://www.bgs.org.uk/blog/atypical-COVID-19-presentations-in-older-people-–-the-need-for-continued-vigilance. [accessed 19 April 2020].

,

21

• Considine J.
• Street M.
• Bucknall T.
• Rawson H.
• Hutchison A.F.
• Dunning T.
• et al.

Characteristics and outcomes of emergency interhospital transfers from subacute to acute care for clinical deterioration.

]. For example, acute-onset anosmia may manifest either early in the disease process or in people with mild or no constitutional symptoms[

22

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/879639/ COVID-19-adult-social-care-action-plan.pdf. [accessed 27 April 2020].

]. Similarly, due to limited compliance, the false negative rate of testing can be expected to be higher in this population than the already quoted 20-30% [

23

Covid-19: Hong Kong scientists report first confirmed case of reinfection.

]. Thus, enhanced surveillance through repeat testing of all ‘asymptomatic’ individuals is vital for case ascertainment in such settings, to identify covert transmitters and individuals at risk of rapid deterioration [

24

• Dravet C.
• Bureau M.
• Oguni H.
• Fukuyama Y.
• Çokar O.

Severe myoclonic epilepsy in infancy (Dravet syndrome).

,

25

Initial Review and Analysis of COVID-19 Host Genetics and Associated Phenotypes.

]: three of the seven asymptomatic SARS-CoV-2 positive individuals at CCE in round two tested negative during the first round of surveillance, and the first positive individual #2-1 from STE was initially tested negative on admission to hospital, but not when discharged. According to UK public health guidance, a negative test was not required prior to discharge from hospital back to a care facility [

26

https://blogs.sciencemag.org/pipeline/archives/2020/04/21/watching-for-mutations-in-the-coronavirus. [accessed 27 April 2020].

]. Such discharges may contribute to the risk of infection spreading within care-facilities. We also describe a case of re-infection among caregivers, a phenomenon which has been recently reported in the literature, although the mechanisms of immunity, or its loss, underlying re-infection have not been yet established [

27

The molecular biology of coronaviruses.

].

Not surprisingly, contact tracing at CCE proved difficult, not only for asymptomatic individuals testing positive without data on when the infection might have occurred, but also due to caregivers sharing accommodation (contacts of contacts, see Figure 2), large numbers of agency workers, in particular in CCE-Unit 2, and delay in obtaining test results (up to 5 days after testing). Testing of symptomatic caregivers at STE (13 positive out of 105 tested) and TM (2/26) returned similar numbers of positive tests in symptomatic people compared to the general UK population (as of 13 September 2020: 368,504 people/19,293,329 tests), with the official numbers not accounting for multiple tests in hospitals for the same individual (two negative tests prior to discharge). Together with a low rate of infected individuals, this is re-assuring as it suggests that early implementation of preventative and infection control measures in all four long-term care-facilities (see Table 1) can reduce the infection risk in high-risk environments [

11

• Gandhi M.
• Yokoe D.S.
• Havlir D.V.

Asymptomatic Transmission, the Achilles’ Heel of Current Strategies to Control COVID-19.

], be it for vulnerable individuals living in long-term care facilities or their caregivers, to a level similar to that observed in the general population. We also show, however, that these measures alone, without identification of asymptomatic people through enhanced surveillance, do not contain the spread of infection.

Despite the frailty and multiple co-morbidities of our population, the impact to date of SARS-CoV-2 in all the facilities has been limited. Children and young adults appear to have lower infection rates, although access to testing, even of symptomatic individuals, was limited in this age group. Enhanced surveillance, as at CCE, is required to determine the true infection rate in the younger age groups. Three of the confirmed positive individuals at CCE/STE and one of the suspected individuals at YE have an underlying genetic condition frequently observed in people with severe epilepsy, with mutation in the SCN1A gene, which is known to be associated with fever sensitivity and elevated risk of early mortality [28] Host genetic predictors of outcome in SARS-CoV-2 infections are yet to be established [29]. SARS-CoV-2 RNA mutations and additional molecular mechanisms may explain variability in clinical presentation [30-31].

5. Conclusions

We provide evidence of the need for enhanced surveillance for SARS-CoV-2 of asymptomatic people in high-risk environments. We recognize that CCE was fortunate to have extensive collaboration between basic science repurposed for high-throughput viral testing (the Francis Crick Institute), high-level virological and clinical input (from UCLH), and the ability to redeploy clinical academics (from UCL), to support dynamic and purposeful care teams. All centres benefit from close integration between health and social care with close reviews by epilepsy consultants from UCLH and/or GOSH. Such multidisciplinary input is not available to all care facilities, but the strategies outlined here may provide generally applicable guidance for other facilities facing similar challenges, in particular in preparation for a potential second wave of infection. We hope that such integration between science, healthcare and social care can also generate a new model for the care of the most vulnerable in society in the future. We must learn that there are better ways to be a civil society, to ensure that those living in care-facilities are not excluded from the expertise and interventions available for the wider population.