Symptom Relief Is Possible in Elderly Dying COVID-19 Patients: A National Register Study

Introduction

The COVID-19 pandemic is creating fear not only among the public but also in health care staff, due to the alarming reports on large groups of dying patients with acute respiratory distress syndrome (ARDS). Symptoms at admission to hospitals are frequently reported,^1–3 but little is written about symptom relief in the dying patients.

In studies so far, including a recent study on 5700 patients hospitalized in the New York City area, ² focus has mainly been on describing initial symptoms such as fever and cough, risk factors, including comorbidities, laboratory findings, and prognostic factors. With few exceptions, reports on deaths have been from hospitals and/or ICUs, where multivariable analyses have highlighted independent risk factors such as age, comorbidities, organ failure, elevated markers, for example, d-dimer. ³ These data are important, especially for the development of COVID-19 care. However, from the individual, dying patient's point of view, symptom prevalence and the possibility to relieve distressing symptoms are even more important.

In hospital and ICU studies, the mean age is relatively low. In two Chinese studies, the median age at admission to hospital and ICU were 56³ and 47 years, ¹ respectively. In the recent New York study, median age at admission was 63 years. ² However, emerging statistics show that when all deaths in COVID-19 are considered, the mean age is substantially higher. Recent data indicate that about 40–50% of deaths occur among elderly people in nursing homes, ⁴ but little is known about the way they die. In Sweden, people residing at nursing homes are frail and elderly, with several chronic comorbidities and with a high need of daily support and attendance, whereas palliative care patients with complex symptoms and substantial medical needs are cared for in specialized palliative care services. Thus, Swedish nursing homes constitute a form of long-term facilities for persons needing much help with activities of daily living functions. Moreover, 60–70% are estimated to suffer from cognitive failure or are diagnosed with dementia. For this reason, Swedish nursing homes are very different from other accommodations for retired persons, for example, “65+ accommodations” for healthy elderly people who like the possibility to have access to social activities, communal restaurant facilities, and common social areas in adjunction to their apartment.

In 2019, the median period of care from admission to death was <2 years in nursing homes and the median age at death was about 86 years. Generally, an admission to a Swedish nursing home is approved only if six to eight home help visits per day are not enough for a person to reside in his or her own home. For this reason, the population dying due to COVID-19 at hospitals and nursing homes are different at least regarding age and frailty. Moreover, Swedish hospitals are mainly staffed with physicians and registered nurses, whereas the great majority of the staff in nursing homes (>90%) constitute of assistant nurses and only to a minor part of registered nurses. Physicians are not directly employed by the nursing homes but constitute a (scarce) resource provided by the county councils.

In Swedish health care, there have been efforts to safeguard the quality of care for dying people, which have been previously described. ⁵ The Swedish Register of Palliative Care (SRPC), a nationwide quality register of end-of-life (EOL) care, encourages all county councils and municipalities in Sweden to retrospectively complete a questionnaire about EOL care with focus on last week of life. The register provides EOL data and allows comparisons across different care settings. So far, 30 peer-reviewed publications based on SRPC data have been published. Owing to a reasonable coverage (∼60% of all deaths in the country), some of the questions retrieved from the SRPC have even become national quality indicators for a good death in Sweden, adopted by the National Board of Health and Welfare. ⁶

Aims

The aim of the study was to examine symptom control data on all reported COVID-19-related deaths in hospitals and nursing homes (data set retrieved April 24, 2020), using the SRPC, a validated national quality register.^5,7 A further aim was to compare occurrence of breakthrough symptoms and relief of symptoms in hospitals and nursing homes, assuming a null hypothesis (H₀), that is, no differences between the groups.

Patients and Methods

The Methods section and the Results are reported based on the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) criteria. ⁸

Results

Discussion

Many of the reports, so far, are on COVID-19 patients with ARDS, struggling for their lives in respirators at ICUs. To our knowledge, this is the first systematic report on symptom prevalence and symptom relief focusing on dying patients outside ICUs.

Compared with the reference data, patients dying from COVID-19 were elderly, suffered more often from breathlessness that was difficult to completely relieve, whereas breakthrough pain was less common. Also, in general, there were more breakthrough symptoms in the COVID-19 population and the chance to symptom relief was lower.

However, the data from nursing homes with surprisingly low prevalence of symptoms—symptoms that in most cases were possible to alleviate completely or at least partially—are in contrast with anecdotal reports on agonized patients dying in ICUs. The core symptom, breathlessness, was reported in 35% in nursing homes, in great contrast to 73% at hospitals and almost in line with figures (25%) for the total cohort of all patients dying in 2019. Meanwhile, the possibility to complete relief of breathlessness was higher in nursing homes. Thus, our null hypothesis (comparable symptoms and symptom relief in dying persons regardless of services) was rejected.

A possible explanation is that these frail elderly people die early in the course, in most cases before ARDS develops. This is strongly supported by local data from nursing homes in the Stockholm region, were the median duration from first symptom to death is only six days (mean 6.2) (personal communication), which is much shorter than for patients dying in ICUs. In comparison, Guan et al. report that the median hospital stay (which was preceded by days with initial symptoms) for the severely ill patients in Wuhan, China, was 13 days. ¹ If our tentative explanation holds true, admission of dying frail elderly to hospitals/ICUs for high-flow oxygen and/or for mechanic ventilator treatment will possibly prolong the dying process, eventually with more severe symptoms at the very EOL and increased suffering.

From a clinical point of view, the possibility to relieve symptoms is as central as the very occurrence of symptoms. In Swedish health care, the specialized palliative care units have propagated for evidence-based use of “four essential palliative drugs,” needed for quality care of the dying. ⁹ Such palliative guidelines have been disseminated both to nursing home services and to acute hospitals. As seen from the data, these drugs are prescribed to a high extent, both in nursing homes and in acute hospitals.

The proportion of “I do not know” responses is of certain interest (Table 3). In this study, for all comparisons, the proportions were higher in hospitals. This may be due to several reasons. The ELQ is supposed to be retrospectively completed, based on data from the medical records. Insufficient documentation may result in several “I do not know.” Another reason might be that the person completing the questionnaire was not highly involved in the actual patient care. However, it cannot be out ruled that the differences also may reflect the actual interest in symptoms and symptom relief. If symptom control and support is the main focus of care in people dying from COVID-19, the proportions of “I do not know” should be low for central symptoms such as breathlessness and anxiety.

Conclusions

When studying all deaths related to COVID-19 and reported to the SRPC, a picture emerges of persons in a relatively high age, certainly different from the reports on patients dying at ICUs. The central and feared symptom, breathlessness, is much less frequent in nursing homes and when present, it is easier to relieve. Our data indicate that these differences are due to differences between the groups. Persons residing in Swedish nursing homes are elderly and frail and die much earlier in the course, before severe respiratory distress develops.