Increased care at discharge from COVID-19: The association between pre-admission frailty and increased care needs after hospital discharge; a multicentre European observational cohort study

COVID-19 has resulted in large numbers of people being admitted to hospital worldwide [1‐3]. The UK government guidance published on 25 August 2020 estimates that 95% of patients admitted with COVID-19 may be discharged home, but 50% would require voluntary and community support and 45% health and social care services, and 4% would be discharged to rehabilitation and 1% into long-term care facilities [4]. However, these figures do not take into account those that already have these services in place pre-admission. Patients may experience functional deterioration requiring temporary or permanent support from community social care and rehabilitation services on discharge from hospital [5‐7]. A proportion of this group may require a change in living situation, for example, a long-term care facility such as residential or nursing home.

People at higher risk of requiring hospital admission due to COVID-19 are older with greater levels of multimorbidity and frailty [8, 9]. Frailty represents increased vulnerability to stressors due to declined physiological systems and loss of homeostasis [10, 11]. In a wide range of conditions, even after adjusting for age and comorbidity, frailty has been reported as an independent predictor of mortality, prolonged hospital stay, and increased care needs following hospital discharge [12, 13]. Similar effects due to COVID-19 have been reported including mortality and increased length of hospital stay [9].

To date, most studies in older people with COVID-19 have focussed on mortality with little attention to functional outcomes. However, in older adults, mortality is not always the most important outcome. Previous studies assessing the impact of frailty on quality of life in other conditions have demonstrated that for older adults, independent living is a more important outcome than death. A higher value is placed on continuing day-to-day societal roles, reducing the risk of isolation and loneliness and avoiding poor future health outcomes [14‐16]. Currently, there is no literature describing the factors associated with a loss of independence, or associated increase in care needs, after hospital admission post-acute COVID-19. The aim of this study was to investigate the association between pre-admission frailty and change in the level of care needs on discharge from hospital in patients admitted with COVID-19.

As per the CONSORT flow diagram (Fig. 1), a total of 1671 patient records were entered into the COPE-Discharge database from 13 sites. Eight hundred and forty patients (840) were excluded: 112 patients due to missing care level at admission and discharge, 202 patients due to being alive still in hospital, and 526 had died in hospital.

Of the 831 patients meeting the inclusion criteria (Table 1), the median patient age was 71 years (58–81 [IQR], range 19 to 100) and 369 (44.4%) were women. The median length of hospital stay was 12 days (6–24 [IQR], 0–145 [min–max]), and 388 (46.7%) patients were considered to be living with frailty (CFS 5 to 9). Of the included patients, 30 (3.6%) were under 65 years old and had an increased level of care. Of the 30 patients, 9 went from living independently at admission to requiring a carer, and 13 required intermediate care. There were 36 patients with a missing care level at discharge imputed as not experiencing an increased care.

Before hospital admission, 596 patients (71.7%) lived independently at home without a carer and a further 74 (8.9%) lived at home with carer support. The remaining 19.1% were living in either a sheltered, residential, or nursing care setting (Table 2). Across the 13 sites, 193 (23.2%) patients were discharged with an increase in care level compared with that documented on admission. Fewer patients under 65 years old (30 out of 315, 9.5%) required an increase in care level compared with those of aged 80 years or older (87/242, 35.9%).

As shown in Table 2, of the 596 patients who were living at home without formal care before admission, 143 (24.0%) required an increased level of care at discharge. Of the 74 patients who were living at home with carers before admission, 17 (23.0%) required an increased level of care. Of the 158 patients who came from a sheltered accommodation, residential, or nursing home, 30 (19.0%) required an increased level of care. An increased care level occurred in 10.2%, 12.6%, and 13.1% for the least frail categories of CFS 1, 2, and 3, respectively (Table 1). This compared to the following: CFS 4, 20.0%; CFS 5, 36.9%; CFS 6, 38.5%; CFS 7, 26.1%; and CFS 8, 36.4%.

In the crude logistic regression analysis, CFS was associated with an increase in care at discharge (Table 3). Compared to CFS 1–3, CFS 4 OR = 2.71 (95% CI 1.38–5.34, p = 0.004), CFS 5 OR = 5.63 (95% CI 3.08–10.27, p < 0.0001), CFS 6 OR = 5.90 (95% CI 3.24–10.75, p < 0.0001), CFS 7 OR = 3.50 (95% CI 1.90–6.47, p < 0.0001), and CFS 8 OR = 4.30 (95%CI 1.53–12.09, p = 0.006). The covariates associated with an increased level of care at discharge were age (compared to < 65 years: 65–79 OR = 3.71 (95% CI 2.18–6.30, p < 0.0001); ≥ 80 OR = 5.79 (95% CI 3.35–9.98, p < 0.0001)) and elevated CRP (≥ 40 mg/L) OR = 0.65 (95% CI 0.45–0.94, p = 0.022).

Of the 831 included participants, 810 (97.5%) exhibited complete data and were included in the mixed-effects multivariable logistic regression analysis. After adjustment for the other comorbidities and CRP on admission, an increase in the level of care at discharge was associated with frailty. Compared to CFS 1–3, CFS 4 aOR = 1.99 (95% CI 0.97–4.11, p = 0.062), CFS 5 aOR = 3.77 (95% CI 1.94–7.32, p = 0.0001), CFS 6 aOR = 4.04 (95% CI 2.09–7.82, p < 0.0001), CFS 7 aOR = 2.16 (95% CI 1.12–4.20, p = 0.022), and CFS 8 aOR = 3.19 (95% CI 1.06–9.56, p = 0.039). Other covariates associated with an increased level of care were age (compared to < 65 years: 65–79 aOR = 2.82 (95% CI 1.57–5.06, p = 0.0005); > 80 aOR = 3.87 (95% CI 2.07–7.26, p < 0.0001)) and CAD (aOR = 0.49, 95% CI 0.29–0.82, p = 0.007).

A sensitivity analysis carried out that only included patients that received a longer-term increase in care (by removing intermediate care) at discharge found frailty was associated with an increased care needs, compared to CFS 1–3.

Our study has demonstrated that pre-admission frailty status is associated with a person’s level of independence after discharge from hospital following admission with COVID-19. Although many patients had no formal care prior to admission, almost a quarter acquired COVID-19-related functional dependence exhibited by additional care required on discharge. These findings are in keeping with other studies demonstrating worse functional outcomes associated with greater pre-hospital levels of frailty [20].

Although not synonymous with advancing age, frailty is more prevalent in the older population, and as a consequence, older people are more likely to require care. These data have shown that risk of increased care needs following hospitalisation with COVID-19 is also associated with increasing age, with those over 80 years or older significantly more likely to require increased care at discharge than younger people.

The association of frailty and age with increased functional dependence on discharge can be used to assist with early identification, at the point of hospital admission, of people at risk of needing more care. These data can aid clinicians in providing early tailored assessments and management plans to patients. One paradigm of this tailored assessment is a comprehensive geriatric assessment (CGA) effected through geriatric medicine clinicians in conjunction with the multidisciplinary team. CGA produces a personalised and integrated management plan using a multimodal, multimorbidity, and multidisciplinary approach over time. It has been shown to reduce post-acute long-term functional dependence and mortality [21‐23]. Ensuring CGA-trained clinicians are embedded in acute hospital pathways will allow early screening of vulnerable patients and the potential chance to modify post-COVID outcomes.

These data may be unique to hospital admissions during the pandemic and not pertain to non-COVID times. Hospital-based teams are being faced with unique challenges: streamlined discharge processes aimed at maintaining capacity within the acute setting, reduced access to specialist inpatient therapy resources due to high service demand, sickness-related staff absences, and the breakdown of pre-pandemic informal care arrangements due to social distancing restrictions [4]. It is difficult to interpret if these data reflect the Department of Health’s estimate that 50% of patients would be able to be discharged home without care—50.7% in this cohort returned to their own home with no support, but this figure excludes the 19% in this study admitted from sheltered accommodation or a care home—and whether the Department of Health estimate includes these groups [4]. Furthermore, these findings have implications for both health and social care funding with a large economic package required to provide adequate support for a more dependent population post COVID-19. Local providers can use these data to anticipate short-term budgets based on the frailty statuses of the hospitalised population. Future research should focus on the trajectory of patients' COVID-19-acquired functional dependence, as well as examine population-based frailty scoring with the same associations of an increased level of care, in order to improve the regional and national social care predicted spend [4].

To minimise the personal and societal burden of COVID-19-related functional decline and increased care needs, urgent strategies are required: firstly, avoiding transmission of the virus through public health techniques such as contact tracing, infection control measures, and social distancing including shielding; secondly, modification of frailty pre-admission may ameliorate the risk of post-acute deterioration [24‐28]; thirdly, post-acute rehabilitation through specialist multidisciplinary rehabilitation services needs to be targeted at those who have been affected to promote a return to pre-admission function [29‐31].

Frailty is associated with an increased level of care post hospital admission with COVID-19. In this study, 23.2% of discharged patients required an increased care level on discharge. This study suggests that future public health approaches must take into account a large number of patients with increased care needs and position adequate resources to ensure robust supported discharge schemes for those admitted to hospital. Frailty screening should become a standard practice at admission in order to identify patients early who are most likely to benefit from multidisciplinary health and social care attention.