Discussion - Estimating the CFR for 2019-nCoV

[kiwibird]

https://flutrackers.com/forum/forum/-2019-ncov-new-coronavirus/-2019-ncov-studies-research-academia/epidemiology/894666-clin-microbiol-infect-national-case-fatality-rates-of-the-covid-19-pandemic

Interesting article posted by Tetano - many thanks Tetano - I was surprised to see Germany had such a high cfr. Am I reading it correctly. Germany has had a great run with testing and has been very thorough with track and trace - so I would assume their numbers are very accurate.

[Ronan Kelly]

systematic review and meta-analysis of published research data on COVID-19 infection-fatality rates

Gideon Meyerowitz-Katz, Lea Merone
doi: https://doi.org/10.1101/2020.05.03.20089854
This article is a preprint and has not been certified by peer review [what does this mean?]. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.

• Abstract
• Info/History
• Metrics
• Preview PDF

Abstract


Introduction: An important unknown during the COVID-19 pandemic has been the infection-fatality rate (IFR). This differs from the case-fatality rate (CFR) as an estimate of the number of deaths as a proportion of the total number of cases, including those who are mild and asymptomatic. While the CFR is extremely valuable for experts, IFR is increasingly being called for by policy-makers and the lay public as an estimate of the overall mortality from COVID-19. Methods: Pubmed and Medrxiv were searched using a set of terms and Boolean operators on 25/04/2020. Articles were screened for inclusion by both authors. Meta-analysis was performed in Stata 15.1 using the metan command, based on IFR and confidence intervals extracted from each study. Google/Google Scholar was used to assess the grey literature relating to government reports. Results: After exclusions, there were 13 estimates of IFR included in the final meta-analysis, from a wide range of countries, published between February and April 2020. The meta-analysis demonstrated a point-estimate of IFR of 0.75% (0.49-1.01%) with significant heterogeneity (p<0.001). Conclusion: Based on a systematic review and meta-analysis of published evidence on COVID-19 until the end of April, 2020, the IFR of the disease across populations is 0.75% (0.49-1.01%). However, due to very high heterogeneity in the meta-analysis, it is difficult to know if this represents the "true" point estimate. It is likely that different places will experience different IFRs. More research looking at age-stratified IFR is urgently needed to inform policy-making on this front.
...
https://www.medrxiv.org/content/10.1....03.20089854v1

[JJackson]

Cartski those numbers would be useful particularly if they include a figure for patients who get zero assistance. The Imperial epi calculations seem to indicate a best case scenario with 8 times as many patients needing ICU as there is capacity for in wave peaks (worst case model 30x). The numbers you give for CFR in the wealthy are IMO very optimistic I do not think good patient care has been missing for the bulk of cases to date so an 80 year old billionaire probably will not differ from anyone else who got a critical care bed of similar age. I know it does not feel like at the moment but this epidemic is in its initial stages with the scope to get a great deal worse if not managed aggressively.

[cartski]

I wonder if it is more informative to determine likely CFRs based on social/economic cohorts and accessible health system level.

For example, billionaires would have private hospitals in their mansions, so CFR 0.01

Middle clause would have health insurance hospitals, so CFR 0.005

Low class would have Medicare services, so CFR 3.0

Then assume 70% of cohort gets infected regardless of class.

Has anyone seen or heard of something like that?

J.

[JJackson]

Re the Diamond Princess.
N.B. I have only read what is posted above.
Fortunately for the passengers and crew, but unfortunately for calculating the CAR, the index case was identified before the 2nd generation cases showed. This allowed testing and quarantine/social distancing which will have protected many passengers from exposure which in turn will make their 17% estimate low. It does however show that this figure is more than achievable in the world at large. Pick what ever CFR you like and apply it to 17% of your population and that is how many dead you will have. The CAR could be twice this or more. How you manage to spread those deaths over time will determine if your CFR remains a constant, if you exceed your ICU/mechanical ventilator surge limit you must expect every case there was not room for to die - ICU triage by that point is not going to give space to anyone it is unlikely to save or who had a fighting chance without it. In a badly managed wave peak your x ICU beds may have 10x or more patients who meet admission criteria.

[mscox]

That is an interesting read, however it must be read with caution. What it implies is more correctly stated as follows:

Within a given population who contract covid-19 and for whom health facilities are available to treat severe and critical cases as required. Then the natural CFR will be roughly 0.5% (or 10x the flu). The absolute key part of that statement is the health system being available and functional.

If you allow the virus to infect too many people, too quickly, then expect the CFR to rapidly (order of magnitude, so 10x again) jump up as the health system overloads. This is why OUR governments need to urgently implement social distancing measures to flatten the curve!

[Shiloh]

Published 23 mins ago
Cruise ship data helps reveal coronavirus death rate: researchers
By Maxim Lott | Fox News
the coronavirus outbreak: Should I cancel my cruise? What precautions should I take?

Plan on taking a cruse during the corona virus outbreak? Here are some precautions to take.

Reliable data on coronavirus's death rate has been scarce, with early numbers coming from China.

But results from the cruise ship Diamond Princess could help provide a reliable estimate, medical researchers believe.

An examination of ship data concludes that widely-reported death rates are simplistic and too high, because people with mild symptoms often don't bother to come in for testing and don't get counted.

But on the Diamond Princess cruise ship, mild cases weren't missed, because every passenger and crew member was tested a month ago.

"It's really valuable data that we just wouldn't have otherwise, because all people were tested," Dr. Timothy Russell, a mathematical biologist at the London School of Hygiene and Tropical Medicine, part of the University of London system, told Fox News.

"The 3.4% death rate the World Health Organization has reported... there is criticism from nearly all scientists that's too high, because of the under-reporting of cases."

Read more: https://www.foxnews.com/us/cruise-sh...te-researchers

[JJackson]

Excellent article thanks Shiloh. I am glad to see a balanced piece explaining why CFR calculations are a nightmare and impossible to pin down at the very beginning of an epidemic, which is where we are now. The current crude CAR is 0.00% (to two decimal places) if containment fails 30% may be a fair guesstimate. How our hospital systems cope with that will be a major determination of an areas CFR, not even Wuhan has had any experience with that scenario and they had 40,000 HCWs come in from outside - when there was an outside.

[Shiloh]

Covid-19 Is Nothing Like the Spanish Flu
A widely-cited stat about death rates seems to argue otherwise, but it's surely incorrect. So how'd it end up in the research literature?

Coverage of the novel coronavirus pandemic teems with monstrous and sometimes contradictory statistics. Among the most vexing figures flitting across our screens, and spreading via text and Tweet, is the case fatality rate (CFR)—the proportion of known infections that result in death. Early in the Covid-19 pandemic, World Health Organization officials announced an average CFR of 2 percent. Later on, they revised it up to 3.4 percent. In contrast, numerous epidemiologists have argued that the global case fatality rate is closer to one percent. These might seem like small differences, but when multiplied across large populations they translate to significant discrepancies in overall deaths.

Some experts have emphasized the difficulty of calculating the fatality rate of an emerging pandemic, explaining that current estimates are biased by a deficit of testing and by the lag time between onset of illness and death. Despite this counsel, news coverage and social media discourse has obsessed over CFRs and how they compare across pandemics throughout history. A popular refrain is that the new coronavirus has a frighteningly high fatality rate of at least two percent, which is supposedly comparable to that of the 1918 influenza pandemic, also known as the Spanish flu—one of the deadliest viral outbreaks in history. The truth is that this comparison is severely flawed and that the numbers it relies on are almost certainly wrong...

Read more: https://www.wired.com/story/covid-19...e-spanish-flu/

[Shiloh]

Why Korea's CFR is low: https://asiatimes.com/2020/03/why-ar...-rates-so-low/

[Ronan Kelly]

Adjusted age-specific case fatality ratio during the COVID-19 epidemic in Hubei, China, January and February 2020
Julien Riou, Anthony Hauser, Michel J Counotte, Christian L Althaus
doi: https://doi.org/10.1101/2020.03.04.20031104
This article is a preprint and has not been peer-reviewed [what does this mean?]. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.
AbstractInfo/HistoryMetrics Preview PDF
Abstract
The coronavirus disease 2019 (COVID-19) epidemic that originated in Wuhan, China has spread to more than 60 countries. We estimated the age-specific case fatality ratio (CFR) by fitting a transmission model to data from China, accounting for underreporting of cases and the time delay to death. Overall CFR among all infections was 1.6% (1.4-1.8%) and increased considerably for the elderly, highlighting the expected burden for populations with further expansion of the COVID-19 epidemic around the globe.
...

https://www.medrxiv.org/content/10.1101/2020.03.04.20031104v1

[gsgs]

are there any estimates about the CFR, when hospitals are no longer available ?
At the peak of a wave, in countries where they didn't contain it and which get a full wave.
It had been said that 20% of cases need hospitalisation and the lower end of estimates
of infected people was ~30% within 1 year . Assume 1/3 of these die because they are
not hospitalised that would give ~7%CFR and 6.5M deaths in USA within one year,
not counting the additional deaths from other diseases/causes due to interruptions
in health care and other areas.

is 1/3 realistic ? I'm usually counting that 1/3 of the listed severe COVID cases will die

[Ronan Kelly]

A 7th Diamond Princess passenger from the 696 confirmed cases that stayed in Japan has died bringing the cfr in that cluster to 1%.

[Ronan Kelly]

Estimating the infection and case fatality ratio for COVID-19 using age-adjusted data from the outbreak on the Diamond Princess cruise ship

Status: in-progress | First online: 05-03-2020 | Last update: 05-03-2020

Authors: Timothy W Russell*, Joel Hellewell¹, Christopher I Jarvis¹, Kevin van Zandvoort¹, Sam Abbott¹, Ruwan Ratnayake, CMMID nCov working group, Stefan Flasche, Rosalind Eggo, W John Edmunds & Adam J Kucharski.* corresponding author1 contributed equally

This study has not yet been peer reviewed. Aim

To estimate the infection and case fatality ratio of COVID-19, using data from passengers of the Diamond Princess cruise ship while correcting for delays between confirmation-and-death, and age-structure of the population. Abstract

Adjusting for delay from confirmation-to-death, we estimated case and infection fatality ratios (CFR, IFR) for COVID-19 on the Diamond Princess ship as 2.3% (0.75%–5.3%) and 1.2% (0.38–2.7%). Comparing deaths onboard with expected deaths based on naive CFR estimates using China data, we estimate IFR and CFR in China to be 0.5% (95% CI: 0.2–1.2%) and 1.1% (95% CI: 0.3–2.4%) respectively. Main text

In real-time, estimates of the case fatality ratio (CFR) and infection fatality ratio (IFR) can be biased upwards by under-reporting of cases and downwards by failure to account for the delay from confirmation-to-death. Collecting detailed epidemiological information from a closed population such as the quarantined Diamond Princess can produce a more comprehensive description of asymptomatic and symptomatic cases and their subsequent outcomes. Using data from the Diamond Princess, and adjusting for delay from confirmation-to-outcome and age-structure of the ship’s occupants, we estimated the IFR and CFR for the outbreak in China. As of 3rd March 2020, there have been 92,809 confirmed cases of coronavirus disease 2019 (COVID-19), with 3,164 deaths [1]. On 1st February 2020, a patient tested positive for COVID-19 in Hong Kong; they disembarked from the Diamond Princess cruise ship on the 25th January [2,3]. This patient had onset of symptoms on the 19th January, one day before boarding the ship [2]. Upon returning to Yokohama, Japan, on February 3rd, the ship was held in quarantine, during which testing was performed in order to measure COVID-19 infections among the 3,711 passengers and crew members onboard. Passengers were initially to be held in quarantine for 14 days. However, those that had intense exposure to the confirmed case-patient, such as sharing a cabin, were held in quarantine beyond the initial 14-day window [3]. By 20th February, there were 634 confirmed cases onboard (17%), with 328 of these asymptomatic (asymptomatic cases were either self-assessed or tested positive before symptom onset) [3]. Overall 3,063 PCR tests were performed among passengers and crew members. Testing started among the elderly passengers, descending by age [3]. For details on the testing procedure, see [2] and [3].
...

https://cmmid.github.io/topics/covid19/severity/diamond_cruise_cfr_estimates.html

[kiwibird]

https://www.taiwannews.com.tw/en/news/3884830

TAIPEI (Taiwan News) — With much more data from outside of China now available, the World Health Organization (WHO) on Tuesday (March 3) raised its estimate of the average death rate for the Wuhan coronavirus (COVID-19) from 2 percent to 3.4 percent.

Before the disease spread significantly to the outside world, the WHO through most of February estimated the mortality rate to be around 2 percent. This was largely based on highly questionable statistics from China, which showed an average daily death rate that mysteriously never fluctuated far above or below 2.1 percent for weeks.

Many scientists have questioned China's death rate statistics, as they present a near-perfect prediction model rather than sporadic spikes and troughs. Melody Goodman, associate professor of biostatistics at New York University’s School of Global Public Health, told Barron's that a regression run on China's data on fatality rates yielded a near-perfect r-squared (measurement of variance) score of 99.99 percent, which she said is never seen with real data.

"I have never in my years seen an r-squared of 0.99. As a statistician it makes me question the data," said Goodman. She said that when it comes to public health data, a high score would be a 0.7.

Goodman added that "Anything like 0.99 would make me think that someone is simulating data. It would mean you already know what is going to happen."

By February 24, the WHO Director-General Tedros Adhanom estimated that the fatality rate in Wuhan ranged between two and four percent, while the death rate in the rest of China was 0.7 percent. He claimed that his team estimated that measures taken in the communist country had "averted a significant number of cases," without providing any specifics.

However, in the week since, the number of confirmed cases and reported deaths have skyrocketed in Iran, South Korea, and Italy, providing much more data from outside China. With this data in hand, Tedros on Tuesday (March 3) patted China on the back for lowering its number of cases and announced that 80 percent of the new cases outside of China had been reported in Iran, South Korea, and Italy.

Adhanom then dramatically increased the estimated average mortality rate, saying, "Globally, about 3.4 percent of reported COVID-19 cases have died," He then pointed out that this percentage is far higher than the one percent death rate for those who contract the flu.

According to The New York Times, in the U.S., the death rate for the seasonal flu is typically around 0.1 percent, not one percent as Adhanom claimed.

The WHO head then proudly proclaimed that "evidence from China" indicated that only one percent of cases reported were asymptomatic. Yet in Japan, as of Feb. 26, 19 out of 186 infected patients, or 10.2 percent, had not shown any symptoms, excluding passengers from the Diamond Princess cruise ship, reported Japan's Ministry of Health, Labour and Welfare.