More testing shows more iceberg

COVID-19 case numbers are rising quickly in many parts of the northern hemisphere. Already some totals have outstripped the peaks seen in the first wave of the pandemic. Europe and The United States, in particular, look to be in for a harsh winter. As we look at testing and hospital outcomes, we can see reasons for why wave 2 looks different from wave 1 (if wave 1 ever truly finished) in March/April. In particular, we often see just the tip of the iceberg when it comes to outbreaks and epidemics. But in wave 2 more testing shows more iceberg.

Initial testing shock

One thing is pretty clear: laboratories and biotechnology companies worldwide were unprepared for the scale of a pandemic. This wasn’t a slow-moving SARS-CoV or a poorly transmitting MERS-CoV. This was a fast-moving, fully armed, well-equipped, respiratory virus. Labs couldn’t keep pace. And even though test development was super-quick, the fuel to feed and power what those tests needed, quickly became scarce.

Despite the lab challenges posed by Ebola virus epidemics in Africa just a few years ago, and Zika virus epidemics across the world, testing wasn’t ready enough for pandemic 2020. In today’s world, a pandemic was always going to mean using real-time PCR-based tools to detect the virus. These were the most sensitive tests we had. Also, we didn’t have a better tool because no replacement had hit the mainstream. Despite the recurring promise of new platforms pitched during recent outbreaks and epidemics, PCR hasn’t faced a mainstream challenger since it’s real-time iteration hit the brights lights in 2009.

After COVID-19 worldwide wave 1, sample throughput increased although turnaround times still seem to blow out whenever cases rise quickly. The improvements have meant that we no longer just see just the “tip of the iceberg“, but much more of what lies beneath the waterline.

NOTE: After the initial spread of the last (influenza) pandemic and in every annual flu epidemic, we seek to test only a portion of cases – usually those who are sick – to get a good idea of how much and which viruses are around. In this pandemic, we’re trying to test more than that. We’ve asked more from our testing than ever before, and it’s not clear that our pandemic plans ever suggested we do that, which may be why lab capacity planning may have suffered. One major reason for all this testing is so that we can contact trace and then quarantine and isolate to interrupt transmission. In some countries, however, tracing isn’t functioning because there is an overwhelming number of cases. Testing without a follow-up action is perhaps a poor use of resources. Once we can protect the vulnerable with vaccines and treat COVID-19 with useful and specific antivirals, the need for extensive testing will probably recede.

The tip of the iceberg

You’ve probably heard that phrase a million times; the “tip of the iceberg”. It’s used during outbreaks to remind us that we only ever detect and record some of the infections due to a pathogen; usually, the easiest ones to find and test. The comparison is to the iceberg; we only see that small piece of the whole which protrudes above the waves.

The graphic demonstrates this for COVID-19. Initially, we only tested those people linked to travel from China and close contacts, then those with pneumonia then we moved to test local hotspots, those in hospital, and eventually, we started testing anyone with symptoms and some without (for contact tracing or to gauge the level of asymptomatic versus presymptomatic transmission), even coming to your street to test. Now we know that its possible to find SARS-CoV-2 in those with a range of symptoms – which means we are seeing more of COVID-19 (the iceberg).

A beautiful visualization of the combination and permutations of signs and symptoms in patients infected with SARS-CoV-2.
From Australia’s COVID-19 fortnightly epidemiology report No. 25.

The infographic above also includes what we currently consider to be the COVID-19 infection fatality ratio (IFR). The IFR is the percentage of deaths due to COVID-19 among all cases, not just those that were tested through relative convenience.

The case fatality ratio (CFR) however, is more simply the number of deaths divided by the number of cases (usually presented once everyone’s infections have run their course).

The CFR doesn’t include real infections that were missed because they were mild, or asymptomatic or those people just weren’t tested. The IFR tries to capture all of these.

A blanket IFR value (“point estimate”) was recently calculated to be 0.68%.[5] But it doesn’t seem like it;s that low when you look over the numbers.

In the US, the CFR was, at the time of writing, 2.6%.[6] That is 3.8x higher than the global IFR estimate.

Among other things, that higher value hints that there still isn’t enough testing in the US, even though they’re doubling their target of 500,000 tests per day.[7]

NOTE: If we drill down in the US to look at CFRs in different States (or further within States), you’d see all sorts of variation.[8] Wisconsin is at 0.86%, California at 2.1%, Washington at 2.2%, New Hampshire at 4.6%, New York State at 6.7% and New Jersey at 7.1%. This is just a reminder that any given jurisdiction tells it’s own unique COVID-19 tale. Comparing different people and places is not straightforward.

More testing shows more iceberg

As testing has increased so too has the denominator for CFR calculations – that number at the bottom of the fraction below.

Here the denominator is TOTAL COVID-19 cases (lab-confirmed positives). If we can assume we’re as good at capturing the hospitalised and seriously ill COVID-19 cases now as we were during wave 1, but now we’re adding more of the milder illnesses to the denominator because more testing is available, then the CFR will decrease. Also, the time between diagnosis and death may now look more delayed (I’ll explain below) than it was earlier in the year. We may already be seeing this extra lag in our line graphs and bar charts.

A newly shaped iceberg?

Lag? What I mean is the delay between diagnosis and death in those who go on to succumb to severe disease. Right now, we’re not seeing deaths rising the way they did in wave 1.

This graph is plotted in a way that shows that deaths (bars extending below the horizontal axis) lag diagnoses (bars extending above the horizontal axis) by weeks. The axes use different values, to make the rising death curve clear.
Source: An example from Twitter by Marc Bevand

Using another nice example from Twitter by Marc Bevand we see a major reason why we’re only just starting to view an upturn in death curves among northern hemisphere countries, despite cases rising for weeks. Both the shape and size of the curve have likely changed because of all the extra testing. If you look at the blue curve for Spain below, you can see that it is predicted to start rising earlier than what was charted at the time.

The height and the width of the base of the case curve has likely changed between wave 1 and wave 2. This is an attempt to exemplify that change. We may have seen a bigger lag between rapid rise in diagnoses and deaths during wave 1, if more testing was in place back then to capture more of the iceberg earlier on.
Source: An example from Twitter by Marc Bevand

If we were conducting the degree of testing back in March/April that we are now, it might have taken longer to visualise a rise in deaths then as well. We’ve improved our capture of the 80% of COVID-19 we’d expect to find outside of a hospital. We’d become used to a two to five-week lag (predicted to be up to 11 weeks by Marc).

The message here is: don’t get too comfortable with what appears to be a changed pattern where cases are rising but deaths are not. Wait a bit longer than before.

NOTE: those mutterings that SARS-CoV-2 is less virulent? They aren’t supported by any solid evidence right now and I’m not sure they will be anytime soon. Yes, transmissibility seems enhanced by one mutation (D614G; [20]) in some SARS-CoV-2 variants, but there is as yet no evidence for enhanced or reduced disease severity or any negative impact on vaccines. Let’s nail this one to the pile of “wishful thinking” that has plagued the response to this pandemic.

Beyond the idea that more testing shows more iceberg, there have been some other changes as the first wave receded – like the use of the steroid dexamethasone to reduced the 28-day mortality rate by 17% [15,16] and prone positioning to aid oxygenation (although not survival [18) in adult pneumonia patients.[17] These changes have contributed to less severe disease and fewer deaths and should continue to do so going forward.

The COVID-19 death rate dropped a little as wave 1 progressed

Two new studies observed this. Deaths across all age groups decreased as the first wave progressed. But exactly why isn’t known.[1,2,3] Some thoughts from the authors below.

Thinking of England

In a study of national English adult COVID-19 critical care admissions, the authors saw reductions in death regardless of age (three age bands examined), sex or ethnicity (“white” or “Asian”). They also saw a drop when they looked at diabetes and kidney disease but not when looking at those with chronic respiratory disease, between March and May.[2]

The authors noted that bed saturation was highest in April which may have made individual patient management much harder. This really makes the point that if you don’t flatten the curve, you can create a situation where more death and severe disease occurs than expected.[2]

Bright lights, big city

Among an analysis of 4,689 hospitalisations in New York City between March and June, the median age and the proportion of males with underlying disease decreased.[1] But this didn’t fully explain a reduction in mortality over the study period.

Interestingly the amount of viral RNA being detected in each case also decreased. The authors had quite a list of factors which they suggest each may have added to the improved outcomes:

  • decreased bed saturation
  • increased use of corticosteroids, Remdesevir and anti-cytokine drugs
  • earlier intervention
  • community awareness
  • mask-wearing resulting in exposure to a lower viral dose

A grim winter may lie ahead for the north

Despite this good looking news, don’t celebrate yet. So far we’ve learned that more testing has probably changed the pattern we got used to during the first surge of COVID-19; more testing shows more iceberg. But if hospitals get overwhelmed, deaths will happen in greater than predicted numbers. This is a trajectory that some countries in the northern hemisphere are already on unless they take more action. It may already be too late for tinkering around the edges and lockdowns may be the only way forward-and have begun.[19]

Rapidly rising case numbers: deaths will follow

Tragically, as if we are living in a time loop, COVID-19 cases are once again steeply rising across Europe and in the United States, which the hasn’t seen case numbers drop below 13,000 per day (😯) since late March.

For reasons such as those we’ve discussed above, deaths may not yet be doing what we saw in the first half of 2020. But hospital bed saturation is again rising and thus bed capacity is falling.[4] More testing shows more iceberg, but even if there is a greater lag, death follows more COVID-19 hospitalizations which are increasing in multiple countries across the northern hemisphere.[9,10,11,12,13,14,15]

It’s imperative more is done to stop the spread of COVID-19 but it has to be done quickly. More cases today is a window into what happened up to two weeks ago. More cases today also means more deaths in the ensuing months. That’s a given. How many deaths and how quickly numbers rise, remains to be seen.

Things we can do about this

We can each take personal responsibility for our health and that of those around us. We can abide by rules and mandates. We can listen to experience, learning from and acting on what’s worked in other parts of the world; avoiding what hasn’t worked. We can squash false hopes and put aside wishful thinking. We need to seek out reality – harsh though it may be – and we should strangle misinformation and promote facts.

We can each do these things. We can do them together too.

Meanwhile, our governments can show the leadership we expect by organising and providing what’s required to address the shared responsibilities section in the infographic below. Including encouragement and financial and mental health support for the community to stay home. And on that point – if your government isn’t leading you to a healthier safer future then vote, and choose to replace it with a better government when next you get that chance. There are things we can do.


  1. Trends in Covid-19 risk-adjusted mortality rates in a single health system
  2. Improving COVID-19 critical care mortality over time in England: A national cohort study, March to June 2020
  3. Studies Point To Big Drop In COVID-19 Death Rates
  4. ‘It is terrifying’: WHO sounds alarm as daily infections in Europe more than double in 10 days
  5. A systematic review and meta-analysis of published research data on COVID-19 infection-fatality rates
  6. COVID-19 Dashboard by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University
  7. COVID Exit Strategy website
  8. Covid in the U.S.: Latest Map and Case Count
  9. US hospitals are preparing for the worst-case scenario as Covid-19 surges again
  10. U.S. sees highest number of new COVID-19 cases in past two days
  11. France sees highest number of Covid-19 patients going into hospital since April
  12. Coronavirus cases in Spain top one million as pandemic accelerates
  13. Dutch hospital airlifts patients to Germany amid virus surge
  14. Switzerland faces lack of hospital beds as coronavirus infections soar
  15. Dexamethasone in Hospitalized Patients with Covid-19 — Preliminary Report
  16. Covid-19: Low dose steroid cuts death in ventilated patients by one third, trial finds
  17. Feasibility and physiological effects of prone positioning in non-intubated patients with acute respiratory failure due to COVID-19 (PRON-COVID): a prospective cohort study
  18. Effect of Prone Positioning on the Survival of Patients with Acute Respiratory Failure
  19. Germany, France impose national lockdowns due to surge in COVID-19 infections
  20. Spike mutation D614G alters SARS-CoV-2 fitness

Visits: 1881