Sunday, March 29, 2020

South Korea's Slowly Creeping Case Fatality Rate



#15,152


Barely a month ago, South Korea appeared to be on the same trajectory as Italy and Iran, with hundreds of new COVID-19 cases being reported each day.  After a shaky start, however, South Korea declared `war' on the virus, and successfully turned the tide.
They aggressively tested, and isolated cases. Potentially exposed individuals were quarantined, and social distancing along with hand and respiratory hygiene was heavily promoted. 
As a result, Italy now has 10 times as many confirmed cases South Korea, and 60 times as many deaths. And even though Iran's numbers are highly suspect, South Korea has a fraction of their case count or deaths as well.

Perhaps even more impressive, South Korea's Case Fatality Rate (CFR) has run far lower than any other similarly affected nation, although it continues to creep higher with time.
Three weeks ago, South Korea surpassed 7,000 cases, but had reported only 50 deaths.  A preliminary CFR of .7%, or about 1 death in every 140 cases. While still 7 times higher than seasonal flu (.1%), this was touted as extremely encouraging news. -(Typo fixed - MPC)
Over the past 21 days, the number of COVID-19 cases in South Korea has increased by 2,500 (36%), but the number of deaths has more than tripled to 152 (see chart below).


Deaths are lagging indicators in a pandemic, as aggressive medical care can stave off patient outcomes for days or even weeks.  While South Korea's apparent CFR remains low compared to many other countries, it is now no longer below the 1% mark, and it is still slowly climbing.
Instead of 1 death in every 140 cases, South Korea's CFR has risen to 1 in 63, or 1.58%
When we get serological testing on a large scale, we may find that the number of uncounted mild or asymptomatic cases drives the CFR back below 1%.  At least that's the hope.
But the reality is, no one actually counts flu cases, or keeps good track of flu-related deaths (see The Lancet: Estimates Of Global Seasonal Flu Respiratory Mortality). Every number we have for influenza incidence or its CFR is a guesstimate.  
Trying to compare estimated flu statistics with hand-counted COVID-19 cases and deaths (both of which are assumed to be badly undercounted), may not actually tell us very much.  We can get a number, but its probative value is unknown.
It may take years of research before we can get a good feel for the true burden of COVID-19. And that impact is likely to vary widely around the world, depending upon the quality and quantity of healthcare provided. 
While trying to work out the CFR based on incomplete information is an interesting academic exercise, for now our best gauge of the severity of COVID-19 is probably how well modern healthcare delivery systems are able to cope with COVID-19 cases compared to a regular flu season.

And if you look at hospitals in Italy, Iran, Spain - and increasingly in New York City - there is simply no comparison.