Winter of discontent

Adele Groyer and John O’Brien assess the impact of the coronavirus pandemic on claims in 2020, and how this might play out

This article is based on a presentation recorded for the Life Conference on 3 November 2020. At the time we were aware that England would enter a second national lockdown on 5 November and, quoting the Game of Thrones character Jon Snow, we observed that “winter is coming”. We write this article in January 2021, amid a third national lockdown and in midwinter.

Another John Snow has become synonymous with COVID-19. One of the founders of modern epidemiology, he devoted much of his life to improving public health; in 1854 he famously persuaded authorities to remove the handle of a water pump in Soho, London, to try to curtail transmission of cholera. The John Snow Memorandum, named after him, was originally published in The Lancet on 14 October 2020 to counter proposals that a herd immunity approach should be adopted among the low-risk population. The authors of the memorandum state that “controlling community spread of COVID-19 is the best way to protect our societies and economies until safe and effective vaccines and therapeutics arrive within the coming months.”

Vaccines

At the time of the 2020 Life Conference, vaccines were in various stages of clinical trials but none had been approved. All train the body to target the spike proteins on the surface of the coronavirus. The conventional approach of using inactivated viruses has been applied in a number of vaccines developed in China. However, the approved Pfizer/BioNTech, Moderna, and Oxford/AstraZeneca vaccines all use novel approaches.

Pfizer and Moderna vaccines use mRNA, or messenger RNA, which instructs the body itself to make the viral protein, which in turn triggers an antibody response. These vaccines have to be stored at extremely cold temperatures, as mRNA materials are otherwise very unstable. The Oxford/AstraZeneca vaccine, meanwhile, uses an adenovirus vector. This is a genetically modified virus that cannot replicate in the human body but produces the coronavirus spike protein once inside the body to trigger an immune response. This vaccine can be stored in an ordinary refrigerator.  

Vaccines are not 100% effective – a typical flu vaccine is 60%-80% effective – but they are still a vast improvement on no vaccine. The approved mRNA vaccines have demonstrated efficacy in excess of 90% with a two-dose protocol. The political debate at present is whether a single dose with reduced efficacy should be prioritised to achieve some protection across a wider population as quickly as possible, given supply and distribution constraints.

Therapeutics

Medical care has improved as clinicians have gained a greater understanding of the virus.  It has been found that high flow nasal oxygen can be effective for many patients and avoids complications associated with invasive mechanical ventilation. The risk of clots is now well understood, and anti-coagulation drugs are used to good effect. Dexamethasone, a cheap steroid, is effective at reducing 28-day mortality among patients with severe COVID-19 symptoms. In January 2021 it was reported that two other repurposed drugs, tocilizumab and sarilumab, may further reduce mortality among critically ill COVID-19 patients, with clinical trial results now in peer review. However, other promising candidate treatments have disappointed, with no evidence of improved outcomes.

Long COVID

An emerging medical challenge is the management of post-COVID-19 syndrome, or ‘long COVID’. NICE defines this as signs and symptoms that develop during or following an infection consistent with COVID-19, continue for more than 12 weeks, and are not explained by an alternative diagnosis. Post-viral fatigue syndrome occurs with other viruses that, like COVID-19, can affect multiple organs in the body. COVID-19 studies suggest cardiac injury or myocarditis even among young, previously healthy patients, some of whom only had COVID-19 mildly. This has implications for medical care costs, disability and future mortality for the country in general, but also for insurers’ financials and underwriting approaches across all protection products.

Mortality impact of first wave on insured lives in the UK

On 4 January 2021, there were 30,451 confirmed COVID-19 patients in UK hospitals. This significantly exceeds the peak in the first wave, which was 21,684 on 12 April 2020. The dynamics of in-hospital survival, and also which patients are admitted to hospital, are different between these points in time. What is clear is that the insurance industry once again needs to estimate the impact of a challenging period on claims. Data collected during the first wave can provide some clues, but there are still many unknowns.

The CMI’s Mortality Monitor software enables comparison of the 2020 weekly death registrations in the general population in England and Wales against the corresponding figures for 2019. The peak in death registrations was in the week ending 17 April 2020, while death occurrences peaked in the previous week. The peak in the 13-week average standardised mortality ratio (SMR) for 2020 vs 2019 was 146% for registrations between 20 March and 19 June. If we limit the age range to 20-64 years, the SMR was 129% during the same period.

In August 2020, the Association of British Insurers (ABI) released results of the survey conducted among individual and group insurers to record COVID-19 death claims incurred between 1 March and 31 May 2020. The data includes a small number of critical illness and total permanent disablement claims, which do not skew the results materially. We compared these claims to one quarter of the annual claims reported by the ABI for calendar year 2019 and found that COVID-19 claims were 10.5% of these prior year claims counts and 9.3% of these prior year claims amounts. We show these figures in Table 1.  

Note that 85% of claims counts in 2019 related to low face amount whole-of-life business, which is typically guaranteed acceptance funeral cover. On an amounts basis, only 21% of claims relate to whole-of-life cover. This is noteworthy because cause of claim is less likely to be recorded for guaranteed acceptance business. Additionally, there was relatively more under-reporting of COVID-19 deaths at older ages in the general population during the first wave of the pandemic. Therefore, COVID-19 claims are likely to be understated in this survey, particularly on a counts basis.  

GRiD released claims statistics from group insurers showing the value of COVID-19 death claims paid in the first six months of 2020. It is unlikely that there was a material number of claims prior to March, and the delay between occurrence on payment means that the period should map well to the period covered by the ABI statistics. In Table 2 we show that COVID-19 death claims are 20% of one quarter of the group life claims paid in 2019.

If we assume no second order impacts on other death claims, the COVID-19 claims suggest an uptick of 20% in claims compared to the increase in general population mortality of 29% in the working age range. Group portfolios include only the working age population and most of the insured lives are actively at work, which is a positive indicator of health. Therefore, the lower increase relative to the general population is plausible.

It is surprising that the 10% uptick implied in the ABI statistics is lower than the uptick observed only for group business. We have a rationale for under-reporting on a lives basis for funeral plans, but the similar amounts figure remains mysteriously low. When the ABI releases full year claims statistics for 2020, the picture may become clearer.

Mortality impact of subsequent waves on insured lives in the UK

The claims costs incurred in 2020 and into 2021 are affected by deaths from all causes. The COVID-19 survey data collected to date therefore gives only partial information, and ideally we should compare claims from all causes against expectations. It is equally true that all-cause mortality changes should be taken into account to understand the impact on the general population.

We may overstate the impacts if we consider only COVID-19 deaths, because deaths from other causes may decrease. For example, there are likely to be fewer accidents during a lockdown, and social distancing, better hygiene and increased uptake of flu vaccines suggests there may be less flu-related mortality this winter. For those with other severe illnesses, COVID-19 will unfortunately have accelerated deaths.    

On the other hand, we may understate the short-term impact if we look only at COVID-19 deaths because of under-reporting. This no longer appears to be the case in the general population but is likely to continue to be the case for some claims, especially for guaranteed acceptance cover. Deaths from other causes may increase because urgent life-saving interventions for other conditions do not take place, either because of a lack of healthcare capacity or reluctance of patients to seek care. There are also likely to be longer-term impacts of COVID-19-related disability, delayed treatments for other conditions and other deteriorations in physical and mental health.  

No easy choices

We find ourselves in another difficult phase of the pandemic, where the choices the country has to make are not those we want to make. This brings another line from Game of Thrones’s Jon Snow to mind: “Sometimes there is no happy choice, only one less grievous than the others.” It is our hope that by the time this article is published, the choices are easier and the outcomes are happier than they are now.

Adele Groyer is head of pricing and research, UK & Ireland, at Gen Re

Dr John O’Brien is life/health chief medical officer at Gen Re

Image credit | iStock