Tag Archive : case fatality rate

/ case fatality rate

(Part 1 of a three-part series.)

Less than four weeks ago I was looking forward to a week in Colorado for some late-season skiing and to check up on our summer home, about to be vacated by six Latin American ski-season guests. [Lacking a Real-ID, the Spaniel wasn’t going to be able to make this trip]. Within a few days, the COVID really hit the fan. On March 16, I flew from Atlanta to Vail on a Boeing 757 with three passengers on board. The service was amazing.

Suddenly, our nation was in panic, its people hunkered down, the economy in free fall. The ski resorts closed for the season, public assemblies were banned, meetings were cancelled, restaurants restricted to take-out service – all soon followed by state-wide “stay-at-home” orders. (Bizarrely, as of this writing eight US states have decreed that marijuana dispensaries are an “essential business” and have permitted them to stay open. Because, you know, this is a perfect time to be dropping our inhibitions).

Predictions are hard, especially about the future.

Cynics commonly remark that financial experts have predicted thirty of the last three recessions. One might also say thirty of the last three pandemics have been similarly predicted. There is an observable tendency to overpredict disaster that is deeply rooted in human personality and yet not exactly irrational. People are risk-averse and instinctively choose minimizing loss over maximizing gain across a variety of situations. When an unknown number of lives are at stake, it is perfectly reasonable to err on the side of caution. Medical people in particular possess a mindset oriented toward doing “whatever it takes” to save a life (within the realm of possibility). That is probably what you hope for when your number comes up.

In rapid succession we went from a mysterious new virus showing up in Wuhan in December, to a deadly epidemic in China in January, to global spread in February, to economic shutdown in March. The only data we had were the numbers from China, coupled with a concern that, if allowed to proceed unchecked, hundreds of millions might die.

The Number Games

Just how many could die from COVID? Everything depended on the numbers, particularly the mortality rate. Would COVID be just another flu-type pandemic, with an average mortality rate of 0.1% and an estimated 650,000 annual deaths worldwide? We don’t shut down the global economy for the flu. The initial numbers for COVID looked much worse. As recently as February 24, Chinese health officials and the WHO were estimating a mortality rate of 3-4%. By early March, terrifying reports from Italy described an exponential rise in cases, overwhelmed hospital systems, and overflowing morgues. This didn’t look anything like flu, which fells an estimated 8,000 Italians each year over a period of many months. It was probably the situation in Italy more than any other factor that galvanized America into action.

On March 16, the Imperial College COVID-19 Response Team predicted that without active suppression, the US should expect deaths approaching 2.2 million or more. To avoid a similar fate, we had to “flatten the curve”. It might never keep you from catching it, but at least when you did there might be room in the hospital.

Just one day later, internationally renowned researcher John Ioannidis of Stanford says “hold on, there.” Is this a once-in-a-century pandemic? Or a “once-in-a-century fiasco”? Ioannidis was concerned that – based on current data – the mortality rate might be much lower, on par with a bad flu outbreak. While death rates among the elderly were quite high, death rates among the young were very low. Ioannidis suspected the overall fatality rate might lie between 0.05 and 1.0%. At the upper range, it would still be ten times deadlier than the flu. He speculated that the death toll could even be as low as 10,000. Drastic measures might be truly necessary, but we needed better data to know for sure.

Needless to say, there’s quite a difference between 10,000 – fewer than a typical flu season – and 2.2 million. Well, that was three weeks ago. Most developed countries instituted strict social distancing and business closures. New COVID cases in Italy peaked on March 21 and have been declining since. Deaths in Italy peaked 6 days later – March 27 – and have since dropped by almost half. Spain followed a similar trajectory about one week later.

On March 29 Dr. Anthony Fauci predicted, on a somewhat more positive note, that US deaths could be limited to the range of 100,000 to 200,000 if restrictive policies were successfully implemented. At that time, the total US death count remained below 3,000. He was using projections from the Institute for Health Metrics and Evaluation (IHME) at the University of Washington.

In subsequent weeks, both deaths and total cases have fallen at the low end of projections. The total US death count was revised downward to 82,000 on April 5, then 60,000 (real-time projection on April 7). COVID remains primarily a threat to the elderly. Worldwide, 95% of deaths are in persons over 60 years of age. More than half are over 80. In New York, 83% of deaths have been in persons over 60, and 37% were over 80.

So what is the case fatality rate?

The case-fatality rate (CFR) is a simple ratio of deaths divided by the number of people who catch it. We have a pretty good grasp of the deaths. Some may go uncounted while others may be falsely counted, but we have to go with what we’ve got. The real problem is knowing the denominator – how many people have or have had the disease. It is universally acknowledged that many have been infected with the virus and have little or even no symptoms. We call those “subclinical infections”, and this is a known phenomenon for many infectious diseases. One of the most famous cases concerned the notorious “Typhoid Mary”, a cook and asymptomatic typhoid carrier who moved from household to household in the early 20th century infecting every family in her wake. There have been many reports of COVID transmission by asymptomatic carriers.

Untold others may have mild symptoms but either didn’t seek testing or couldn’t get it. The estimated proportions of undiagnosed cases have ranged from as low as 25% to as high as 94% worldwide. Meaning, for every 100 who get sick, there might be as few as 33 or as many as 1,600 who have been infected but are not reflected in the data. Anything approaching the high end of that range would mean a CFR far below 1.0%.

On March 30, a team of researchers publishing in Lancet calculated a CFR of 0.66% – near the middle of the range suggested by Ioannidis. (While the US total death count has already exceeded his lowest estimate of 10,000, that was assuming only 1% of the US population got infected, with a CFR of 0.3%. We may already be well past 1% of the population).

In any case, it is well-established that the case fatality rate is very high in the most elderly, but also extremely low in individuals under 40.

Where are we now?

Looking back three weeks at the wildly divergent projections of IHME and Ioannidis, we’ve seen a convergence of estimates and are probably at or very close to peak infection. The daily new cases in the US topped 34,000 on April 4 and have fluctuated within a narrow range since April 2. US deaths are currently just below 2,000 per day and the IHME projects we are at or near a peak.

The US health care system has not experienced the disastrous overload experienced in Italy. Many hospitals in New York City were swamped, but we were able to scale up very quickly. This revealed one pitfall behind some of the sunnier forecasts – even if the death rate were low, it could still be disastrous if the illness spread very quickly and everyone fell ill at once. And COVID certainly spreads quickly. By one recent analysis, it may be five times more infectious – and spread that much more quickly – than the common flu.

Fortunately, the worst predictions were unrealized. Much yet depends on what the next week or two bring us. No one is quite sure how or when the economy can resume. Only now has it been possible to undertake antibody testing, which may ultimately give a more accurate picture of the total scope of infection and reveal who may be immune and thus able to safely circulate in public. [The Spaniel wishes to remind you that you can’t catch COVID from your dog, but cats are bad news]. 

In the next post, we’ll look at some important lessons to be learned, and how Christians should – and should not – respond to crisis.