Identifying older people most vulnerable to COVID-19

As the COVID-19 pandemic has progressed from the first reported UK cases in late January and first UK death in February 2020, governments across the UK have implemented various policies to prevent spread and limit impact of the infection. Early on, decisions were made to try to protect people believed to be at greatest risk of the most adverse COVID-19 outcomes. Government advice has focused on three different population groups:

1. Clinically vulnerable: identified by doctors and expert groups, largely focused on those who normally receive flu-vaccine, who were advised to follow strict social distancing measures;
2. Clinically extremely vulnerable: a sub-group of those who were clinically vulnerable who were advised to ‘shield’ because it was believed their medical condition may put them at even higher risk of becoming very unwell or dying with COVID-19 if they became infected. This meant staying at home at all times and avoiding all face-to-face contact;
3. The rest of the population who were advised to follow mandatory social distancing measures during lockdown.

As we have learned more about COVID-19, it has become clear that some people are more likely than others to become very unwell, requiring hospitalisation, or to die if they become infected. Factors such advanced age, sex, ethnicity, weight, occupation, having some medical conditions and being on certain medications may increase the risk of getting the virus, of being hospitalised or dying if infected.

In response to this, England’s Chief Medical Officer (CMO) has asked an expert group to consider the latest evidence on risk factors for severe COVID-19 outcomes and death, and assess whether it is possible to create a risk calculator that takes into account these various risk factors.^1,2

Three BGS members,³ together with other organisations concerned about people with greater vulnerability to COVID-19 infection, have discussed these proposals in July 2020 with England’s Deputy Chief Medical Officer (DCMO). BGS has specifically raised concerns that the framing of this work is focused on health conditions and does not consider functional needs or the balancing risks of isolation for older people who will be captured by a risk model based purely on multimorbidity.

This paper aims to summarise the concerns raised by the BGS and make recommendations for how this should be taken forward.

BGS position on risk modelling for COVID-19 shielding

BGS recommends that factors over and above disease alone should be taken into account when developing national isolation and shielding policy, given the particular vulnerabilities of people living with multimorbidity and/or frailty, who in the main will be predominantly older-and that an approach predicted on age alone risks inadvertently selecting people not at significant risk from hospitalisation or death due to COVID (or other illness) yet who may be at equal or greater risk of harm from isolation itself. To assist with this BGS recommend that consideration be given to phenotypic and/or cumulative deficit frailty for which in England there are data now readily available.

Frailty and adverse health outcomes

Frailty is defined as a distinctive health state related to the ageing process in which multiple body systems gradually lose their in-built reserves. At least 10% of people aged over 65 years have frailty, rising to between a quarter and a half of those aged over 85.⁴

Frailty is a strong predictor of adverse outcomes for older people hospitalised because of acute illness.⁵ There are a number of validated scoring mechanisms for identifying those at risk of poor outcomes based on prior frailty for example the Hospital One Year Mortality Score and Hospital Frailty Risk Score.^6, 7

The World Health Organisation (WHO) defines frailty as a clustering of multi-system dysregulations, leading to a loss of dynamic homeostasis, reduced physiological reserve and greater vulnerability to subsequent morbidity and mortality. This is often manifested by maladaptive response to stressors, leading to a vicious cycle that results in functional decline and other serious adverse health outcomes. WHO have by consensus recommended moving away from diseases focused models when designing health strategies focused on population ageing, and moving towards strategies which focus on enhancing intrinsic capacity and functional ability.⁸

In 2019 the task force of the International Conference of Frailty and Sarcopenia Research (ICFSR) developed clinical practice guidelines to overview the current evidence-base and to provide recommendations for the identification and management of frailty in older adults.⁹ ICFSR made multiple evidence based recommendations including the routine identification of frailty among people aged 65 and over using a validated instrument.

Frailty identification in England

Since 2017 frailty has been routinely identified in England as part of the General Practice General Medical Services (GMS) contractual requirements for frailty.¹⁰ This requires a two stage approach to frailty diagnosis and coding in the general practice record. It recommends use of validated tools (such as the electronic frailty index - eFI) to identify those at risk followed by clinical confirmation of frailty status using validated tools such as the clinical frailty scale, gait speed, PRISMA-7, timed up and go.¹¹

Clinical frailty is therefore already routinely identified and encoded by severity of frailty into GP electronic patient record systems for the purposes of extraction for assurance against the GMS contract, and has been since 2017. This data is easily accessible to guide personalised care discussions and planning and optimal frailty management formed part of the 2019 NHS Long Term Plan¹² and commissioning frameworks set out within the Rightcare frailty toolkit.¹³

QFrailty

QFrailty¹⁴ is derived from QMortality and QAdmissions to classify patients into four groups at risk of all cause admissions and death. It is derived from outcomes not based on the prior characteristics of individuals who might be considered at risk. It is not specifically aligned to either of the two accepted clinical constructs for frailty:

1. cumulative deficit frailty^15, 16
2. or phenotypic frailty¹⁷

QFrailty comprises:

• age
• body mass index
• Townsend deprivation score
• ethnic group
• smoking status
• alcohol intake
• unplanned hospital admissions in the past 12 months
• atrial fibrillation
• antipsychotics
• cancer
• asthma or chronic obstructive pulmonary disease
• living in a care home
• congestive heart failure
• corticosteroids
• cardiovascular disease
• dementia
• epilepsy
• learning disability
• leg ulcer
• chronic liver disease or pancreatitis
• Parkinson’s disease
• poor mobility 
• rheumatoid arthritis
• chronic kidney disease
• Type 1 diabetes
• Type 2 diabetes 
• venous thromboembolism
• anaemia
• abnormal liver function test result
• high platelet count
• visited doctor in the past year with either appetite loss, unexpected weight loss or breathlessness.

Frailty and multimorbidity

QFrailty is largely configured around single or multiple long term health conditions (multimorbidity) as a predictor of death or hospitalisation. Although separate concepts, it is apparent that there is a large overlap between frailty and multimorbidity.^18, 19 The majority of older people have multiple long term conditions. Most older people with frailty have multimorbidity, but the majority of people with multimorbidity are not phenotypically frail, despite being at greater risk for adverse health outcomes than their age peers.

Conversely the eFI is designed to include functional measures based on the cumulative deficit model of frailty. eFI is validated²⁰ and includes activity limitation, housebound, mobility and transfer problems, requirement for care and social vulnerability. This matches what Rockwood (and to an extent Fried) conceptualised as frailty. eFI is included in every GP system in England with contractual obligation to secondarily validate and code for moderate and severe frailty since 2017. This could be included into a risk model to establish whether two distinct populations (multimorbidity and frailty) are identifiable and if so their attendant risks of the primary and secondary outcomes.

Balancing risks for older people

In segmenting populations of older people at risk of adverse outcomes following onset of acute illness it is important therefore to consider both the accumulation of disease (multimorbidity) and/or the presence of frailty (based on a cumulative deficit or phenotypic construct). Risk models based solely on disease risk are likely to select older people, but will not consider their other vulnerabilities for example linked to the presence or absence of frailty.

In addition, interventions which focus on reducing the risk of adverse outcomes of hospitalisation and death through isolation must be considered for their impact on individuals, who as a result of underlying frailty or disease, may be at risk of other health outcomes, including increased risk of cardiac death and deterioration in cognitive and mental health.²¹ ICFSR recommends multi-component physical activity programmes, protein and calorific supplementation, social support and home based training for those identified with frailty or pre-frailty at risk.²²

BGS recommend therefore, that the balancing risks of isolation, particularly for people living with frailty and/or multimorbidity are considered in light of the non-COVID-19 excess mortality now evident among mainly older people living at home during the pandemic reported by ONS.²³