The limitations of the data: Why modelling and statistics fail to reveal the factors affecting the severity of the COVID-19 crisis
By Jack Walsh
The limitations of the data- Why modelling and statistics fail to reveal the factors affecting the severity of the COVID-19 crisis
In the UK, the finger of blame is being pointed from all sides at Boris Johnson’s Conservative government as our death toll passes 40,000, second only to the US worldwide. With the notable exceptions of Germany and South Korea, it appears that the majority of Western leaders will be held accountable for the actions they took, or didn’t take, as the death tolls first began to increase.
Many have since made the assumption that the bleak situation can be traced directly to our delayed entry into lockdown
While closely watched statistics such as the death toll for each country can give us an indication of how severe the effects of the coronavirus pandemic is in a given area, oftentimes, the data fails to paint an accurate picture of the underlying causes of why the disease has affected the UK in particular more acutely than most. Boris Johnson’s reluctance in initiating lockdown is widely credited to be the reason it is regarded as being the hardest hit by the pandemic. While countries such as Germany responded to the tragic events in Southern Europe swiftly, the conservative government dragged its feet for several days. Many have since made the assumption that the bleak situation can be traced directly to our delayed entry into lockdown on March 23. It is tempting to make a logical leap and while this undoubtedly will have affected the death toll in the UK, it is unlikely to be the overarching explanation.
The government’s stalled response… does not begin to explain the crisis
Compiling accurate data on how many lives could have been saved by entering lockdown earlier is notoriously difficult. It is unlikely an accurate picture will be painted to give clarity on how effective an earlier lockdown would have been. There are also issues with the data themselves. Many of the statistics on the pandemic are subject to sampling bias and are unrepresentative of the population as a whole, often warping perceptions over key factors such as the number of infected people and growth over time. With any statistical model, assumptions on how a virus spreads through the community are built into the system and strong assumptions are made about the future of the virus. Reports in the US in March often extrapolated their data up until August, but provided no direction or advice heading into the autumn, suggesting that governments lacked a long-term strategy to tackle the virus. As people begin to return to society, they should do so with the awareness that little is known about the future outside of conventional assumptions about epidemiology.
As the UK’s death toll began to surpass it’s European neighbours, the cabinet scrambled to find an explanation for this tragedy, claiming that only when data for the excess deaths in each country was released, could a clear picture be painted of the situation. Unfortunately for Johnson and co., the UK ranks highest in Europe for the number of excess deaths and second internationally. This has led Johnson’s detractors to trace the UK’s macabre death rate of 575 per million, third only to Spain and Belgium, directly to our inaction in the early days of the virus. To add to the misery, an analysis by the Financial Times suggests that the UK’s excess death rate remains one of the highest in Europe both when the data considers the mortality of pensions and when pensioners are excluded. In fact, almost every statistical model is indicative of the fact that the UK has been hit harder than almost any other nation worldwide.
Every statistical model is indicative of the fact that the UK has been hit harder than almost any other nation
While the government’s stalled response to the pandemic no doubt added to the severity of the virus in the UK, this does not begin to explain the crisis. To highlight this, the spread of the virus in the UK can be contrasted to that of Germany, who have recorded as little as 8,700 deaths from around 185,000 infections. By contrast, the UK has recorded almost 40,000 deaths from a near 300,000 infections. To compound the differences between the two nations, the UK has a lower population of 68 million compared to Germany’s 83 million. Sporadic suggestions that an ageing population explain the differences in the mortality rates in a given location are also predicated on false grounds; the median age in the UK is 40, compared to 46 in Germany.
Germany’s response to the onset of the virus has been lauded worldwide; yet amongst their extensive measures to limit the spread of the virus is the implementation of a mass testing system, ensuring that health officials were accurately able to diagnose those displaying few symptoms of the virus. Even people who were unlikely to suffer greatly from the virus were mandated to self-isolate for fourteen days, unable to pass on the virus to more vulnerable demographics. Officials in the UK decided to abandon the prospect of a mass diagnostic testing system in March, it is therefore unknown how many were already infected by March 23, when the
majority of restaurants, bars and businesses closed their doors. As health officials in the UK were only able to provide a diagnosis for those experiencing a marked deterioration in health, many asymptomatic victims of coronavirus were able to accelerate the spread of the virus.
The condition of Germany’s multi payer health care system provides the strongest insight into their ability to tackle the virus, simultaneously giving an indication of how exactly the UK has succumbed to a worst case scenario situation. In January, for each 100,000 people, Germany had 34 care beds with ventilators, compared to just 7 in other harder-hit countries such as the Netherlands. It is unknown what the exact numbers are for the NHS, yet reports that the absolute number of patient beds in the UK has more than halved over the past thirty years does not provide much opportunity for optimism. Unlike the UK, there have been no austerity measures in Germany for over a decade, ensuring that their health care system was, compared to its European neighbours, equipped to deal with the coronavirus. Combined with their rigorously planned and executed testing programme, it appears likely that the ability of the German health care service to identify and provide care for the majority of residents suffering from the virus before they experienced a drastic deterioration in health allowed them to avoid the worst.
The fates of many European nations were largely determined before lockdown was initiated in any country. While the UK was given time to prepare as the world witnessed the situation in Southern Europe deteriorate, a decade of austerity, understaffing and reduction in resources, a situation which continues today as NHS workers await the arrival of essential PPE, ensured that once the UK had abandoned its mass testing programme, we were destined for a death toll surpassing that of our continental neighbours.
As the UK eases its own restrictions on lockdown, the public perception is that this has been done prematurely and risks a second explosion in cases. Whether with merit or not, it is clear that the economic concerns of the government are beginning to outweigh the need for public health. As ‘the new normal’ begins to take shape, experts are warning of the possibility of a second wave. Schools in South Korea were forced to close once again after mere days after recording more cases than at any point over the past two months. It is possible that should a second wave occur, the NHS will be more equipped and able to sustain the impact. However, if we are unable to implement any effective measures for tackling COVID-19, this sort of optimism may be grounded in fantasy.
Image: Loco Steve via Flickr
