As deaths from the coronavirus pandemic climb relentlessly, it is already becoming clear that the official toll is an under-estimate and that significant numbers of deaths caused directly and indirectly by the virus are not being recorded as such.
These discrepancies in the death tolls are most obvious in those parts of the world where the pandemic has hit the hardest.
In Wuhan in China, where the virus first emerged, very few cases of COVID-19 are now being reported and life is slowly returning to normal as lockdown restrictions that have been in place since late January are lifted. For some time, there have been reports that China has under-stated the full extent of the coronavirus pandemic and now – as Wuhan officials release the ashes of the dead to relatives – information is emerging to give credence to those reports.
The city’s official coronavirus death toll is just over 2,500; this apparently does not include people who were not tested for coronavirus before they died or those who were infected but may have died from pre-existing conditions. But based on the numbers of burial urns being delivered to funeral homes, some residents estimate that the coronavirus death toll could be as high as 26,000.
No-one is surprised that China might seek to be less than transparent about pandemic numbers, but early analyses, using the “excess mortality” approach, of deaths from badly affected European countries show that it isn’t just deaths from coronavirus infection that are missed.
In Spain there is evidence that two to three times as many people are dying of COVID-19 as are officially being registered. For example, in Madrid in the period 10-16 March, 794 deaths were expected (based on data from previous years) but 1318 deaths were recorded. Only 192 of the excess deaths were attributed to COVID-19, leaving 524 deaths unexplained.
In Italy there are similar findings, as shown in the graph below (from Professor Biondi Zoccai at Sapienza University). An analysis of reported deaths for towns across Northern Italy indicates that the number of unexpected deaths is more than four times the number officially attributed to COVID-19.
Some of this discrepancy is due not to lack of transparency but to how coronavirus deaths are tracked and required to be reported. And then there is the added problem of determining the true number of coronavirus infections when testing is not widespread.
In Spain, for instance, deaths of untested people in private homes and aged care residences are not being included, say regional officials. In France, until very recently, only COVID-19 deaths in hospitals were recorded and again, people who died in nursing homes or at home were omitted from the statistics. French health officials have admitted that the death toll from coronavirus is much higher than the daily government tally.
While there is a growing body of evidence that the numbers of deaths directly due to coronavirus are significantly undercounted, it is much harder to assess the impact that coronavirus is having of deaths from other causes.
Even during a pandemic people continue to fall ill, have heart attacks, strokes and complications from diabetes, need ongoing cancer treatments, experience accidents and trauma, and, of course, there are babies to be delivered. Sadly, in the United States, gun violence continues (so far this year it has killed as many Americans as coronavirus). Some of these patients will need intensive care and ventilators.
Hospitals must figure out how to juggle these patients who require ordinary urgent care with those who are sick from coronavirus. The concern is that coronavirus patients will crowd out other patients whose needs are also urgent.
In New York City however, where medical services are already stretched beyond capacity, there is evidence that some patients are just not making it to hospital. The 911 service is daily overwhelmed with calls and paramedics and emergency medical workers are making life-or-death decisions about who goes to a hospital, and who is left behind.
When older adults call with a medical issue, paramedics fear taking them to the emergency room, where they could be exposed to the virus. They are often simply told to see their doctor. Because of the high level of calls, there are huge delays in responding and if it’s a call from someone with chest pain and difficulty breathing, this requires donning gowns and personal protection equipment on the assumption it’s a coronavirus case. For a cardiac arrest, those time delays are usually the difference between life and death.
Hospitals are unsafe environments for chronically ill and immuno-compromised people. Victoria’s first two coronavirus deaths were cancer patients at The Alfred hospital, and a further five cases of COVID-19 have been confirmed among patients and staff there.
Because people infected with coronavirus can be asymptomatic, simply segregating infected and uninfected patients is impossible. As a consequence, deaths in the time of pandemic will be directly due to coronavirus (in patients whose testing status may or may not be known) or from another condition but perhaps with an accompanying coronavirus infection. The only way to determine the true impact is through widespread testing, including post mortem testing and post-recovery antibody testing.
It’s impossible to know how long this pandemic will last. Longer term there are reasons to fear that there will be consequences from people going without or delaying needed community-based care. In Australia there is anecdotal evidence from comments on twitter that doctors, especially specialists like cardiologists, are seeing fewer patients in their clinics. Breast screening programs are suspended, dentists are closed or providing only emergency services, rural pharmacies are experiencing “massive” shortages of essential medicines.
Only once the world has recovered from the pandemic will we be able to get a full picture of how many people had the disease, how many people died from it, and how many people died because the healthcare system was overloaded with COVID-19 patients.
As an addendum, there are some interesting anomalies in the data that are worth pointing out.
The Centre for Evidence Based Medicine at the University of Oxford finds that all-cause mortality in England and Wales for the period 31 January (when coronavirus was first confirmed in the United Kingdom) to 20 March is 3 percent less for 2020 than the average for 2015-19. It is postulated that this reduction in deaths is due to public health measures impacting on all respiratory deaths.
Reduced social activities may also mean decreased trauma-related deaths due to decreased traffic accidents. There is some additional evidence for this from San Francisco, which has seen a recent reduction in moderate trauma cases as a result of accidents.
Dr Lesley Russell is an Adjunct Associate Professor at the Menzies Centre for Health Policy and a Non-Resident Expert at the United States Studies Centre at the University of Sydney.