COVID-19 prevalence and mortality in longer-term care facilities

This essay considers the factors that have contributed to very high COVID-19 mortality in longer-term care facilities (LTCFs). We compare the demographic characteristics of LTCF residents with those of community-dwelling older adults, and then we review the evidence regarding prevalence and infection fatality rates (IFRs), including links to frailty and some comorbidities. Finally, we discuss policy measures that could foster the physical and mental health and well-being of LTCF residents in the present context and in potential future pandemics.

falls per person-year (a key indicator of frailty) is nearly twice as high in Sweden compared to Spain.
• The proportion of older adults receiving longer-term care at home differs substantially across countries, ranging from 8% in the USA to 12% in Sweden and the UK. These outcomes reflect individual preferences as well as public policies aimed at encouraging residence at home rather than in LTCFs whenever possible. [26,27]. • Older adults face an array of residential options depending on income and health status, including retirement communities and group homes for those who can live independently, assisted living for those with specific needs or disabilities, care homes for the frail and cognitively impaired, skilled nursing facilities for those with severe illness, and hospices to provide end-of-life care. [21,28]. • The prevalence and degree of frailty is strongly linked to age, with an incidence of severe frailty among community-dwelling adults of about 4% at ages 65-69, 10% at ages [75][76][77][78][79], and 27% at ages 85+. [29] A recent metaanalysis concluded that the overall incidence of frailty was about 50% for ambulatory individuals (mean age 81) and about 70% for LTCF residents (mean age 86), but severe frailty was much more frequent among the latter group. [30].
Generally speaking, health and mortality data for LTCFs encompasses elder care homes and skilled nursing facilities, whereas data regarding other residential arrangements has a lower degree of consistency and comparability. For example, the U.S. Center for Medicare and Medicaid Services oversees all U.S. nursing homes, while state agencies may regulate other types of facilities for older adults (such as assisted living and group homes). The U.S. National Center for Health Statistics also reports on long-term care provided by adult day service centers, home health agencies, and hospices. [31]. The duration of care in LTCFs varies widely depending on the individual's health status and other factors. In the USA, for example, 14% of individuals in nursing homes are covered by Medicare, which reimburses up to 100 days of skilled nursing care following hospitalization. [31] More than one-third of U.S. nursing home residents have been diagnosed with heart disease, which can greatly foreshorten life expectancy. Nearly half of U.S. nursing home residents have been diagnosed with Alzheimer's disease or other types of dementia, for which the post-diagnosis 5-year survival probability is about 50%. [32,33] A study of U.K. care home residents found that the median length of stay was 1·6 years, with a 75th percentile of 3·6 years and a 90th percentile of 6·2 years. [34]. programs and regulations. Indeed, LTCFs are exceedingly rare in developing countries, where older adults generally live with their children or other relatives. For high-income countries, the OECD defines LTCFs as residential facilities that provide health services at the level of nursing care, as distinct from hospitals (which can provide much higher levels of medical care) or other residential arrangements (which may offer a mix of personal care and other services). [23] It should be noted that the OECD uses the term "elderly" in describing adults ages 65+, whereas we simply refer to this demographic group as "older adults." [24,25].
As shown in Table 2, only a small fraction of older adults reside in LTCFs, but the proportion in Sweden is notably higher than in the UK and nearly twice that of Spain and the USA. These outcomes reflect a number of distinct factors: • In all four countries, life expectancy (as of age 65) extends into the mid-80s. However, healthy average life expectancy -defined as the absence of any major disease -is typically about 5 years shorter than that, with varying degrees of healthcare and personal assistance that may be needed over the intervening period. • The proportion of older adults (ages 65 + years) who live together with their children varies markedly across countries, ranging from 1·5% in Sweden to more than 30% in Spain. Such residential arrangements are likely to be associated with differences in physical and social activity that may be consequential for health and cognition. [15,26] For example, the reported incidence of  [5], Sweden PHA [6], Scotland NRS [7], UK ONS [8,9], US CDC [10] and CMS [11]; see also France INED [12]. YLL reflects age and sex but not comorbidities. [3,13,14] Studies conducted in Scotland and in Ontario, Canada found that there were outbreaks in about 40% of LTCFs. [42,43] A systematic review and meta-analysis found that about 45% of residents were infected in facilities experiencing an outbreak, and in the majority of such instances it could not be determined how the virus was introduced into the facility. [41].

Assessing the Cause of Death
At an early stage of the pandemic, the WHO established guidelines for certifying COVID-19 as a confirmed cause of death (in cases with a positive test) or as a suspected cause of death (based solely on observed symptoms). [44] During 2020-21, the ratio of confirmed to total COVID-19 fatalities was 97% in England and 96% in Sweden. [6,45,46] Nonetheless, the precise application of the WHO guidelines has differed somewhat across countries, and hence caution is needed in making international comparisons. Moreover, implementing these guidelines is particularly complex in assessing fatalities of LTCF residents, who may have been extremely frail or had a severe preexisting illness.
Autopsies can be useful in determining causes of death but are quite unusual for elderly individuals. However, several peer-reviewed studies have reported on systematic autopsies of older adults to determine the frequency of cases in which COVID-19 was not a cause of death although the deceased had a confirmed positive test. [47][48][49][50] In these studies, COVID-19 was found to be a direct or secondary cause of death in the vast majority of PCR-positive individuals, regardless of whether the death occurred at home or at a hospital or LTCF. Diffuse alveolar damage was evident in nearly all instances in which COVID-19 was a causal factor of death, while thrombosis was observed in a large proportion of those cases. [47,50] Autopsies have also documented some specific cases, such as carbon monoxide poisoning, where LTCF residents clearly died "with COVID" rather than "from COVID." [51].

Gauging Fatality Rates
The proportion of deaths among confirmed infections -commonly referred to as the case fatality rate (CFR)can be used as an approximate gauge of the severity of COVID-19 in LTCFs. [41] However, numerous studies have documented the pitfalls of using CFRs, reflecting the incidence of asymptomatic COVID-19 infections that are never detected as well as constraints on the availability of rapid tests, especially during the early stages of the pandemic. [35,38,40,52].
Importantly, as shown in Table 2, the median age of LTCF residents is markedly higher than that of the overall population of older adults: This age gap is about 9 years in the USA, 10 years in the UK, and 12 years in Spain and Sweden. Accounting for these age differences is crucial in comparing COVID-19 IFRs for LTCF residents vs. community-dwelling older adults.

COVID-19 Prevalence in LTCFs
Numerous epidemiological studies have found that the prevalence of COVID-19 in LTCFs was far higher than among community-dwelling older adults. For example, a serology study of northern Italy during spring 2020 identified anti-SARS-CoV-2 antibodies in about 50% of LTCF residents compared to about 11% of the general population. [35] The LTCF/community ratio of prevalence was even higher in other locations: 13x in a study of 179 UK care homes and about 30x in a study of LTCFs in Fulton County, USA. [36][37][38] These disparities reflect the extent to which communitydwelling older adults could effectively limit their potential exposure to the virus (getting home food deliveries, minimizing social contacts, etc.), whereas LTCF residents inevitably have frequent and direct physical contact with LTCF staff. [39] Moreover, about one-third of COVID-19 infections are asymptomatic, even in elderly adults. [40,41].
Protecting LTCF residents was particularly difficult at the early stages of the pandemic when rapid tests and protective equipment were not readily available and methods of mitigating infection risk were not well-established. In such circumstances, a single LTCF staff member might inadvertently spread the virus to a large fraction of LTCF residents.  Chudasama et al. (2021). [15][16][17][18][19][20][21][22] It should also be noted that the IFR for LTCF residents may have been extraordinarily high during the initial onset of the COVID-19 pandemic, due to incapacitation of facility staff and limited information about effective treatment methods as well as constraints on hospitalization and admission to intensive care units. [53,54,57] For example, at one small LTCF in southwestern France, an outbreak started on 14 March 2020 that infected half of the staff and twothirds of the residents, whose case fatality rate was nearly 30%. [58] Similarly, an outbreak was identified on 17 March 2020 at a small LTCF in Ontario, Canada, infecting half of the staff and nearly all of the residents, whose case fatality rate was 45%. [59] In a study of Swedish LTCF residents (median age 87), the 30-day mortality during the first wave of the pandemic was about 40% for COVID-19 cases compared to about 6% for the control group. [16] In a comparable study of Ontario, Canada, the 30-day mortality for COVID-positive LTCF residents (median age 86) was 24% higher than that of the control group and moderately above the metaregression IFR prediction of 17% (CI: 14-22%) for that age cohort. [60,61].

Frailty and Comorbidities
Frailty has been linked to elevated COVID-19 mortality among LTCF residents, reflecting a combination of increased incidence of infection as well as more adverse consequences-the same combination that has been observed for other infectious diseases. [62,63] Individuals with severe frailty or cognitive deficiencies require extensive assistance with daily routines and hence greater exposure if a staff person becomes infected. Moreover, frailty has a strong association with immune dysregulation that is likely to exacerbate vulnerability to COVID-19 as with other infectious diseases. [64,65] However, a systematic review of the links between frailty and COVID-19 mortality found that nearly all studies have involved hospitalized LTCF patients; such findings are subject to elevated risk of bias for the reasons noted above. [66] That review identified two other studies involving samples of LTCF residents, but those findings were hampered by limited size and diagnostic methods. [67,68] A large-scale study of UK Biobank participants found a strong association between frailty, hospitalization, and COVID-19 mortality in the overall sample but did not find a significant link between frailty and mortality within the subsample of hospitalized patients. [69].
COVID-19 mortality has also been independently associated with some specific comorbidities. Large-scale studies of LTCF residents with COVID-19 did not observe any significant association between 30-day mortality risk and certain chronic conditions (i.e., chronic obstructive pulmonary An alternative metric is the proportion of deaths among hospitalized COVID-19 patients, commonly referred to as the hospitalization fatality rate (HFR). For the general population, most COVID-19 deaths have occurred in hospitals, and the HFR has generally been much higher than the CFR, reflecting the fact that hospital admissions involve individuals who are experiencing severe symptoms. By contrast, LTCF residents with COVID-19 have only rarely been transferred to a hospital. For example, a UK study found that only 4% of the COVID-19 fatalities of LTCF residents occurred in hospitals. [42].
The low frequency of hospitalization for LTCF patients reflects several distinct factors. First, LTCFs are able to provide significant medical treatment for COVID-19 (such as anticoagulants, antibiotics, and hydration), thereby avoiding a hospital transfer that could be quite traumatic for the patient. Second, some LTCF residents (or their legal guardian) may have signed a "do not resuscitate" order, signaling their preference for avoiding the use of invasive procedures (such as a mechanical ventilator) that would prolong the individual's life. Finally, in some circumstances where hospital capacity is highly constrained, triage procedures may assign priority to younger and healthier patients while limiting the hospital admission of the aged and infirm. [53,54] All of these factors create risk of bias that hampers the use of HFRs in drawing inferences regarding LTCF residents.
The gold standard in gauging severity is the number of deaths in proportion to the number of infected individuals, commonly referred to as the infection fatality rate (IFR). [2] For example, a study in Georgia USA found an IFR of 18% for LTCFs that conducted periodic antigen tests of all residents and staff and an IFR of 15% for facilities where all residents were tested promptly after identification of an index case. [37] The Vivaldi serology study analyzed a large representative sample of care home residents in England and found seropositivity of 33% during the baseline period of summer and early fall 2020. [55] Thus, in the total population of about 425,000 care home residents, there were about 140,000 COVID-19 infections. [56] As of 30 October 2020, England reported 20,617 COVID-related fatalities of care home residents, of which more than 80% were in residents ages 80+. [8] In effect, the estimated IFR of LTCF residents in England was about 15%.
These estimates of IFR for LTCF residents are comparable in magnitude to the predictions of a prior metaregression analysis of age-specific IFRs published in this journal. [2] In particular, LTCF residents in Georgia USA have a median age of about 82 years, while LTCF residents in England have a median age of 86 years. The metaregression predicts an IFR of 15% (CI: 12-20%) for a comparable cohort with a median age of 85 years.

Lessons for the Future
COVID-19 has had a catastrophic impact on LTCF residents. Some countries, such as France and UK, that had been hit hard by the first wave incurred fewer LTCF fatalities during the second wave, reflecting protective measures as well as the natural immunity of those previously infected. [55,82] However, the incidence of LTCF fatalities continued rising markedly during the second wave in several other countries (including Denmark, Germany, and USA), reflecting outbreaks at LTCFs that had been left unscathed during the first wave. [83,84].
More recently, mortality has been sharply reduced by the nearly universal dissemination of vaccines to LTCF residents and staff. For example, the COVID-19 mortality rate of U.S. LTCF residents in March-December 2021 was 780 per 100 K, less than one-tenth of its rate over the preceding 12-month period. [11] Nonetheless, even following booster shots, vaccinated LTCF residents may be particularly prone to so-called "breakthrough infections" due to immune dysregulation and other factors, and the frequency of such cases has increased following the emergence of the Omicron variant of the virus. [82] Thus, it would be premature to conclude that the pandemic no longer poses a substantial threat to LTCF residents.
Numerous studies have found that the risk of COVID-19 spreading within an LTCF is closely linked to the incidence of SARS-Cov-2 infections in the surrounding area. [85][86][87][88] Consequently, LTCFs have endeavored to mitigate those risks by implementing non-pharmaceutical interventions such as comprehensive testing, visitor restrictions, and use of masks. However, measures such as visitor restrictions can exacerbate LTCF residents' sense of isolation, fear, and depression, while the use of protective gear by staff can be particularly disturbing to residents with impaired cognitive or hearing abilities. Systematic reviews have found that social isolation and loneliness in LTCF residents are strongly associated with increased mortality, which is particularly concerning given that about one-third of LTCF residents were already experiencing severe loneliness prior to the pandemic. [89][90][91].
Most LTCF residents have relatively few close contacts, whereas staff members interact with many residents and with close contacts in the community and hence have a higher risk of conveying the virus into the facility. Many jobs in LTCFs are associated with low pay, and hence staff are more likely to live in high-density neighborhoods and use public transport, and many have family members who disease, hypertension, coronary artery disease, heart failure) but did identify moderately elevated odds ratios for diabetes (OR 1·2; CI: 1·1-1·4) and chronic kidney disease (OR 1·3; CI: 1·1-1·6). [16,70] Mortality risk was linked more strongly to physical and cognitive impairments, although those links might at least partially reflect increased risk of exposure for the reasons noted above. [71].

Comparison with Community-Dwelling Older Adults
Age differences are highly relevant in accounting for differences in estimated IFR between LTCF residents and community-dwelling older adults, because the metaregression indicates that IFR triples with each additional 10 years of age. [2] The study of Axfors and Ioannidis (2022), published in this journal, compute IFR estimates for community-dwelling adults that can be directly compared with the metaregression predictions. [72] Here are several notable examples: • England: The estimated IFR is 9·7% for communitydwelling adults ages 70+, which is a cohort with a median age of about 79 years. [72] At this median age, the metaregression predicts an IFR of 7·5% (CI: 6·1-9·2%). • France: The estimated IFR is 4·1% for all French community-dwelling adults ages 65+, which is a cohort with a median age of about 75 years. [73] At this median age, the metaregression predicts an IFR of 4·6% (CI: 3·8-5·6%). • USA: The estimated IFR is 2·3% for U.S. communitydwelling adults ages 65+, although that IFR is 3·6% based on COVID-19 fatalities as reported by the U.S. federal agency that oversees all skilled nursing facilities (rather than from an alternative source that encompasses group homes and assisted living facilities). [11,74] This age cohort has a median age of 72 years, and the metaregression predicts an IFR of 3·2% (CI: 2·7-3·8%).
For other national studies involving representative samples of community-dwelling older adults, the estimated IFR is about 2% for Andorra and Netherlands, 3% for Denmark, 5% for Spain, and 8% for Italy. The IFR estimate for Hungary is highly uncertain, because its seroprevalence was not distinguishable from zero in light of comprehensive data on assay characteristics. [75,76] Findings from other national studies --such as the Belgium study of residual sera, the Canada study of blood donors, the Iceland and Israel studies of healthcare patients, and the USA study of dialysis are also employed in high-contact service jobs; moreover, LTCF staff are primarily female and hence more likely to have child care responsibilities. [92,93] And if an staff member becomes ill, no substitute may be available to carry out their duties. At the early stages of the pandemic some LTCF staff took heroic steps, such as moving into the facility to limit their exposure to infection, but such measures are clearly not practical or sustainable in the vast majority of instances. [94,95]. Looking forward, perhaps the most promising approach for mitigating the risk of infectious diseases in LTCFs might be to create living spaces that are more akin to a family home. [96] As in Denmark, for example, each facility could be subdivided into a smaller set of 'houses', each of which has its own dedicated staff. [97] Implementing such an approach would necessarily be a long-term solution that would require careful consideration of funding, staffing, and other factors. Nonetheless, it is imperative for public authorities and academic researchers to engage in "lessons learned" from the COVID-19 pandemic that would have crucial benefits for protecting the health and well-being of LTCF residents.

Declarations of Interests
The authors have no financial interests nor any other conflicts of interest related to this commentary. No funding was received for producing this commentary.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.