Tracking the COVID-19 outbreak and signals of containment

24 Mar 2020

Any conclusion drawn from the data should be viewed with caution due to the dynamic nature of a pandemic and the adundant sources of bias associated with reporting. I periodically update here the COVID-19 situation in the US, Europe, and Asia, tracking both the outbreak and signals of containment. The intent of this blog is not to feed daily news, but to present perspectives worth considering when reading the news. The graphs in this blog are interactive and best viewed on a desktop browser.

Signals of Containment

The Interplay of Confirmed and Death Cases

When should the economy reopen? To try to answer this question, we could look at the interplay of new confirmed cases and death cases.

More cases means more healthcare resource demand, and doctors and nurses have to make tough decisions. Unfortunately, more patients who need intensive care might not get it, leading to higher fatalities. We are probably going to see a peak in daily cases, and after some time, a peak in daily fatalities. This phenomenon is visible in the graph below. Passing the first peak means measures are taking effect; passing the second means our healthcare system is now able to cope. So, where do countries stand as of now?

Of course, the decision has to also depend on other factors such as the ability of testing and tracking down close contacts of those infected.

There are actually many questions that we could ask from this graph. For example:

1. Why does Germany has much higher daily confirms than Switzerland, and yet manages a much flatter death curve?
2. Why do the two peaks for the UK seem to occur at the same time while that’s not the case for the rest?

• Countries with hopes of relaxing some of the lockdown measures: Germany and Switzerland. Both of them have passed the peaks, have low daily cases (<20), and relatively flat and low death case curve (<5).
• Countries that probably need more time: They are at the edge of passing the first peak and record about 80 daily cases. What’s more worrying, though, is the evident pressure on the healthcare system. UK sees a drop in daily death cases, but that number is still high at 11; the US’s death case curve seems not at its peak yet. They probably need more time. - April 23, 2020

Daily Case Percentage Change

Look out for the 7-day moving average of the day-on-day percentage change in confirmed cases. It is important to see both the current percentage change and its trend. To easily classify the situation, we can use the following scale¹:

• \(r > 10\%\): Rapidly increasing.
• \(r < 10\%\): Increasing.
• \(r < 5\%\): Slowly increasing.
• \(r < 1\%\): Under control.

• Japan had a turning point on 23rd March where the increase of cases started accelerating. Coincidently (or maybe not), Japan and I.O.C. officially anounced the postponement of Tokyo 2020 on the next day.
• The cases in Japan have been rising at an increasing rate, now at a 10% day-on-day growth rate. Considering the exponential growth of infections, Abe, Japanese prime minister, is declaring emergency state for seven prefectures. - April 7, 2020
• Japan sees a slowdown of daily new cases. It’s been two weeks since the first declaration of “Emergency Situation” by the prime minister. On average, a 50% reduction in the number of people going out in monitored areas are observed. Meanwhile, mask sales have skyrocketed in Japan. - April 22, 2020

Google Search Interest

This figure tells us how many people in the US are searching for keywords such as “hand sanitizer” or “symptom”. I suspect that as the community spread of the virus is being contained, we can expect to see a drop in searches for words like “symptom” and “influenza”, similar to the trends shown in Singapore.

There are drastic differences in terms of the US and Singapore google search interests during this pandemic. When signs of community infection emerged in early March, people in the US were searching for “symptom” at a record-high frequency, similarly for “influenza” and “hand sanitizer”. Searches for “mask”, however, were not so heightened. The picture in Singapore looks very different. When more infections emerged inside the border in late January and early February, the search for “mask” shoot up rapidly, and masks went out of stock everywhere in Singapore. There are probably two main reasons for this:

1. A high percentage of Chinese living in Singapore;
2. As a nation that went through SARS, it feels natural for most people to wear masks when a contagion is spreading in the community.

US Testing Numbers

As the containment takes effect, we expect to see the number of positive and negative tests stabilize, and the number of tests pending result drops. As you can see, we are not there yet.

Cumulative Case Progression

---

• Japan has a relatively flat curve. However, there are legitimate concerns that Japan has been under-testing its population to know what is really going on. Assuming the true CFR is 1.2%², Japan’s current fatality number, 77, indicates that at least 6,417 people have been infected. However, only 3,139 cases are officially confirmed as of now. Also, Japan has conducted 486 tests per one million population. In Singapore, that number is 11,110. - April 5, 2020.
• For the first time, Singapore is going into a national “Shelter in Place” mode. The timeing is not surprising as some degree of wide-spread community infection is going on. The number of unlinked cases, those yet to find the source of infection, spiked over the last few days; Singapore also recorded 12 new clusters of infection just over the past five days (One of them is right across the river from my house). - April 5, 2020.
• Singapore sees a steady increase in confirmed cases, mainly in foreign worker dormitory clusters. However, if we look at the progression of confirmed cases in Singapore, it’s an almost perfect example of what “flatten the curve” looks like. For the most part, the cases double every ten days, whereas cases in some of the worst-hit countries double every one to three days. - April 8, 2020

Death Cases

---

• Germany and Switzerland fare well in this regard and manage to record comparatively low CFRs. Austria, too, has managed one of the lowerest CFRs among European nations. Austria, Germany, and a large part of Switzerland are German-speaking.🤔
• While the CFRs in Switzerland and Germany have been comparatively low, they are steadily climbing. Switzerland is probably the first country in Europe to flatten the curve, which conducted one of the highest number of tests per one million population. - April 10, 2020

• How does selection bias affect CFR?

  [In Italy], a change in strategy on Feb 25 limited testing to patients who had severe signs and symptoms also resulted in a 19% positive rate (21,157 of 109,170 tested as of Mar 14) and an apparent increase in the death rate—from 3.1% on Feb 24 to 7.2% on Mar 17—patients with milder illness were no longer tested. In the UK, only patients deemed ill enough to require at least one night in hospital met the criteria for a Covid-19 test.

  CFR rates are subject to selection bias as more severe cases are tested, generally those in the hospital settings or those with more severe symptoms. The number of currently infected asymptomatics is uncertain: estimates put it at least a half are asymptomatic; the proportion not coming forward for testing is also highly doubtful (i.e. you are symptomatic, but you do not present for testing). Therefore we can assume the IFR is significantly lower than the CFR.

• When is CFR accurate?

  Iceland’s higher rates of testing, the smaller population, and their ability to ascertain all those with Sars-CoV-2 means they can obtain. an accurate estimate of the CFR and the infection fatality rate (IFR) during the pandemic (most countries will only be able to do this after the pandemic). Current data from Iceland suggests their IFR is somewhere between 0.01% and 0.19%.

The bottom line is, CFR is probably inflated in many countries and IFR is much lower than CFR.

Resources

Websites

1. Bloomberg: Mapping the Coronavirus Outbreak Across the World
2. Johns Hopkins University: Coronavirus COVID-19 Global Cases
3. Global MediXchange: Handbook of COVID-19 Prevention and Treatment

Articles

1. Before Virus Outbreak, a Cascade of Warnings Went Unheeded, March 19, 2020
2. To Fight a Fast-Moving Pandemic, Get a Faster Hospital, March 26, 2020
3. Spain, Europe’s worst-hit country after Italy, says coronavirus tests it bought from China are failing to detect positive cases, March 26, 2020
4. Why Is Germany’s Coronavirus Death Rate So Low?, March 30, 2020
5. This 3-D Simulation Shows Why Social Distancing Is So Important, April 14, 2020

Data sources

• Japan: NHK
• Singapore: Ministry of Health
• Other countries: JHU Coronavirus COVID-19 Global Cases
• US testing numbers: The COIVD Tracking Project
• Search interests: Google Trends

Footnotes