A brief analysis of Covid-19 mortality risk by age group

Summary

• Combining recent reports and data published by Imperial College, London with standard mortality tables by age groups it is possible to make approximate estimates of the severity of Covid-19 in terms of standard mortality risk
• This appears to show that the risk to someone who has contracted the virus is approximately equal to one year’s “Annual equivalent mortality risk” – effectively doubling the chance of death for the next 12 months for any individual
• While this figure is approximately constant by age there also appears to be a spike at the 60+ age group relative to this risk which may be explainable by typical causes of death at different age groups

Findings

On March 16 2020, the Imperial College Covid-19 Response Team led by Professor Neil Ferguson published on their findings on strategies to deal with the Covid-19 (coronavirus) epidemic in the UK (1).

This paper has been immediately credited with a dramatic and swift change by the UK and US governments towards “suppression” (lock-down) style strategies to reduce the spread of the disease. This is to avoid estimated likely overall fatalities from the projected figures in the region of 500,000 in the UK if no action is taken. 

Their paper focuses on modelling of different strategies to project likely outcomes which can then be used by governments to decide their approaches in view of wider health, social and economic considerations. 

Simultaneously many research teams worldwide are naturally racing to develop vaccines and drug treatments to deal with Covid-19 since as the Imperial College teams point out suppression cannot be considered to be a viable strategy indefinitely due to its significant social and economic cost.

 The Imperial College paper provides a breakdown of fatalities by age group of Covid-19, based on early cases from China and other countries affected early on in the pandemic. This is sourced from Verity et al (2).

Table 1: Infection Fatality Ratio for Covid-19 

This table gives the percentage of identified cases of contracting Coronavirus by age group that suffer fatality due to Covid-19.

Much has been made in international news coverage of the fact that the virus affects older patients significantly more than younger individuals. This can be seen in the table above with its increasing rates of fatalities by age.  

However the author of this analysis has seen no attempt to combine any finding on Covid-19 with standard mortality tables, to attempt some “benchmarking” of the danger of the virus relative to expected higher mortality rates. This is the point of this short study. 

The average life expectancy in China is around 75 years, placing it at 5 years below the UK average. This author does not have access to detailed Chinese mortality rates, but a study published by the UK Office of National Statistics in 2005 (the last time such detailed figures could be found) and presented on the Bandolier website 3 gives detailed mortality rates by age group. 

We will therefore neglect the differences in life expectancy or mortality rates by age group between the two counties in order to combine the available data. For improved accuracy it would be possible to adjust for this slight difference.

 By approximately converting UK mortality rates by age bands to be compatible with Table 1 above and ignoring biases and errors in combining data across different countries and date points we obtain the Table 2 below. Note that there is likely to be some statistical error in Covid-19 fatality ratios given the relatively small data set from which they have been drawn.

Table 2: Comparison of Covid-19 fatalities with standard mortality risk

 For all age groups, the annual mortality equivalent risk due to contracting Covid-19 is very roughly around one year. This can be very broadly interpreted as for an individual of any age group, the risk of fatality in contracting Covid-19 is approximately equal to standard mortality risk of surviving one further year of life at that point. This gives an intuitive quantitative measure of the severity of the virus as an alternative to absolute probability levels.

Put another way, for any individual of any age, contracting the virus approximately doubles the chance that individual will die in the next year compared to all normal causes in aggregate.

This last statement does assume some independence between the cases that were fatal under Covid-19 by age group and their age adjusted mortality risk from other causes, in other words were the individual by age group that died from Covid-19 more likely to die in the next year than their peers.

Table 2 also shows that this annual equivalent risk reaches a maximum at the 60-69 age group, rising to 1.64 years.

Data from Public Health England published in 2015 4, shows that in both sexes at age group 60-69, lung cancer emerges as a serious threat. It is the number one cause of death for women and number two cause for men, having not been in the top five causes at any younger age group for either sex.

 A further reference of available medical literature shows that the connection between respiratory disease (which is how Covid-19 typically causes death) and lung cancer has been previously documented. See for example: Denholm et al 5

 Conclusion

 The main focus in the medical profession is to help understand and control the coronavirus/Covid-19 pandemic to reduce fatalities. This short paper has analysed fatalities by age group rather than overall numbers to provide further insight.

Across all age groups risk of fatality from Covid-19 is approximately equal to one year of standard mortality risk.

However the connection between respiratory disease and deteriorating lung condition resulting in lung cancer or other illnesses may help to explain the why the annual equivalent mortality rate at the 60-69 group is observed to be the highest from available Covid-19 data.

References: 

1.     Report 9: Impact of non-pharmaceutical interventions (NPIs) to reduce COVID19 mortality and healthcare demand  - https://doi.org/10.25561/77482), 2020

2.     R, Okell LC, Dorigatti I, et al. Estimates of the severity of COVID-19 disease. medRxiv 2020. https://www.medrxiv.org/content/10.1101/2020.03.09.20033357v1. 2020

3.     Mortality statistics 2005. National Statistics Series DH2 no.32 (www.statistics.gov.uk) and https://www.bandolier.org.uk/booth/Risk/dyingage.html

4.     https://www.gov.uk/government/publications/health-profile-for-england/chapter-2-major-causes-of-death-and-how-they-have-changed - Tables 2 and 3. 2015

5.     Is Previous Respiratory Disease a Risk Factor for Lung Cancer? Denholm et al, American Journal of Respiratory and Critical Care Medicine 2014 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4214084/

T.M. Mortimer PhD.

London, England

28 March 2020