Life and medical insurers turn to excess mortality to measure the severity of pandemic
A close study of excess mortality statistics is helping to fill in the blanks around the extent and severity of the COVID-19 pandemic. José Nicolas, Head of Strategy and Planning for Sub-Saharan Africa at global reinsurer Munich Re, believes that a clearer picture of the tragedy will emerge over time. This view emerged during his presentation to a Munich Re of Africa life insurance seminar.
The first challenge facing life and medical insurers when assessing an emerging peril is to determine the incidence rate. They must establish the parameters of loss distributions in their insured populations to determine the correct price for cover. Determining this measure is not always a simple task. “It has proven difficult to nail down the actual incidence rate of COVID-19,” said Nicolas, who added that certain measures of death, while making sense from a public health perspective, are not credible for insurance purposes.
The trouble with ‘tried and tested’ definitions
There are two definitions that analysts turn to when assessing the immediate impact of pandemic, namely the case fatality ratio (CFR) and the infection fatality ratio (IFR). The CFR expresses the number of people who go on to die from a confirmed case of COVID-19 as a percentage of total confirmed cases; whereas the IFR expresses the percentage of eventual fatalities based on estimates of the total number of people that would have been infected by the disease. Each method has its flaws.
“The CFR has the benefit of having a positive diagnosis of disease; but it can somewhat overstate the actual mortality because you may have people that die with COVID-19 but not because of it,” said Nicolas. The IFR, meanwhile, is based on varying assumptions of total infections. This overreliance on assumption is one of the reasons why the published mortality rates have varied significantly across countries with different demographics, health systems, and politics. Another complication arises due to the long incubation period of the disease. Your symptoms only become severe enough to seek help anywhere from five to 14 days after infection. And a staggering six to eight weeks might elapse from an initial infection to the final outcome of death or recovery. An issue that had a significant impact on South Africa’s response to the pandemic was the lack of test kits and laboratory capacity, which contributed to a high initial CFR.
A post-mortem of governments’ responses to the pandemic will probably centre on how they formulated policy without accurate information combined with their inability to measure the outcomes of their policy decisions. “Having [inaccurate] statistics and developing public policy on these lines may not lead to the best public policy outcome,” said Nicolas. How then can insurers make an accurate assessment of mortality and morbidity during the early stages of a complex pandemic? As the months pass and experts have more time to pore over statistics there are some trends that can be taken to the bank. Top among these are that the infection rate is distributed equally across all age groups and that the mortality rate increases exponentially with age and comorbidities.
The naked truth in excess mortality rates
As we approach nine months under pandemic, we are in a position to consider statistics that are more credible than the CFR and IFR. “Excess mortality allows you to consider the impact of pandemic after the fact,” said Nicolas. “It allows us to establish and access credible data across various countries”. He explained that this measure, although not perfect, was increasingly considered as the most appropriate statistic to track the impact of the pandemic.
The World Health Organisation (WHO) defines excess mortality as “mortality above what would be expected based on the non-crisis mortality rate in the population of interest”. They add that it can be expressed as the difference between observed and non-crisis mortality rates, or as a total number of excess deaths”. It is thus possible, using public statistics for various cities, districts, and geographic regions, to compare the average or expected deaths for a certain period of time with the actual recorded deaths. This difference, published as excess deaths, provides the basis for a more accurate estimate of COVID-19 deaths.
There is an acceptance that not all excess deaths would be due to the disease. Some could be linked to delayed healthcare interventions due to lockdown or domestic violence or suicide, for example. It is also common for a mortality rate curve to show the occasional spike. But a study of deaths across Europe shows that the excess deaths recorded in June and July 2020 are more than 50% higher than what could be expected from, for example, a seasonal flu shock. This experience is echoed by South African data. The South African Medical Research Council (SAMRC), which publishes the “estimated number of excess natural deaths of persons age one and over” had reported 39087 excess deaths between 6 May and 18 August 2020. This compares to the 12264 confirmed COVID-19 deaths published by the Department of Health.
Early days; but the end is in sight
Nicolas observed that most African countries were still in the early stages of the pandemic; but suggested that the impact of the disease could be mitigated by applying new knowledge gained in the rest of the world. “At the time of preparing my slides [just prior the 26 August 2020 webinar] it seemed as if Africa was only at the beginning [of the COVID-19 pandemic], while developed world countries such as Italy and Spain were dealing with a second wave,” he said. There was some good news. Firstly, a vaccine is expected by latest end-2021. And secondly, the mortality rates reported in many African countries appear to be lower than those observed in Europe.
The presentation closed with a sombre warning. The World Bank is of the view that while the pandemic may not lead to catastrophic outcomes insofar mortality across Africa, the compounding impact or effect of the lockdown could have a massive impact on an already poor population. And the African Development Bank put a number to this prediction, suggestion that up to 49 million Africans could be pushed toward extreme poverty as consequence of the pandemic.
Writer’s thoughts:
One of my favourite quotes, popularised by Mark Twain in the US but impossible to accurately attribute, reads “There are three kinds of lies: lies, damned lies, and statistics”. And it seems inevitable that governments’ early statistically-informed responses to pandemic will be interrogated down the line. To what extent were your clients impacted by the often exaggerated statistics that accompanied the early stage of South Africa’s COVID-19 outbreak? Please comment below, interact with us on Twitter at @fanews_online or email us your thoughts editor@fanews.co.za.
