The proportion of fatal cases (PFC)...

This is excerpted and altered a little, from a more influenza A(H7N9) virus slanted article to be found here. But I think it deserves its own page.

In July 2013 I coined a term on VDU to avoid the use of the term Case Fatality Ratio/Rate/Risk (CFR). 

My term was the Proportion of Fatal Cases (PFC). I use the term on VDU and have published it an article.[1] I have no expectations that anyone else will use it although I notice it made an appearance in a HeathMap story[2] by Maia Majumder.

The PFC is a percentage calculated as the currently known number of fatalities divided by the number of total lab-confirmed cases including fatalities, regardless of whether surviving cases are inpatients (hospitalized) or outpatients.

The PFC is just a number - it's not meant to imply that every case that ever happened is included - it never could. It does not account for those cases who will die later on, either directly or indirectly, as a result of their infection but who may be alive at the time of calculation. 

The PFC is a snapshot to be used before an outbreak is done and dusted. It is meant as a guide to what is happening right now using the data we can get our hands on. Sometimes that means lots of data and sometimes they are very limited or just plain behind closed doors.

The CFR makes use of the number of recovered cases in its denominator.[3] So it's important to know survivor numbers. As suggested above, this requires that all the people who will recover from their infection, have recovered (and been discharged) from their infection. 

Using the CFR early in an emerging virus/disease outbreak, when what usually brings in outbreak to our attention is death, is great for selling papers, but not helpful realistic in a bigger picture sense. 

The CFR is most useful at the end of an epidemic/pandemic, but not so much when data-in-hand is poor during the early days of many outbreak. 

Of course, some will take a PFC and multiply it by the world's population as an estimate of how many are going to die if the virus reaches pandemic levels. That's not helpful or accurate. Just accept it as that snapshot of what's happening now.

References...

1. J. P. Dudley and I. M. Mackay. Age-Specific and Sex-Specific Morbidity and Mortality from Avian Influenza A(H7N9). J. Clin. Virol. 2013. Nov;58(3):568-70. ePub Sept.
   http://www.ncbi.nlm.nih.gov/pubmed/24091087
2. http://www.healthmap.org/site/diseasedaily/article/estimating-fatality-2014-west-african-ebola-outbreak-91014
3. http://en.wikipedia.org/wiki/Case_fatality_rate

Snapdate: Charting up MERS-CoV numbers..

Another quick chart update.

Nothing too drastic to add since cases have been infrequent over recent days (see previous post for recent dailies on this [1])

Same caveats apply about the found113 (they are not included here [1]).

Detections and fatal outcomes have slowed to a plateau.

Next stop, Ramadan.

Click on chart to enlarge

References...

1. Adding in the recent MERS-CoV cases by chart...we're back to 2013
   http://virologydownunder.blogspot.com.au/2014/06/adding-in-recent-mers-cov-cases-by.html

MERS-CoV detections and deaths: is the war on MERS making headway?

For a few days now its felt as though we are seeing a more deaths in each Kingdom of Saudi Arabia (KSA) Ministry of Health (MOH) report, than we had been for a while. All very precise language there.

Currently there are about 161 fatalities among the 571 MERS-CoV detection. According to those numbers, 28.2% of people testing positive for MERS-CoV have died.

NOTE: Specific dates for some deaths are missing; date of reporting has been used instead.
The chart use 151 of approximately 161 deaths.
Click on image to enlarge.

Earlier in April, when the case numbers exploded thanks to a healthcare-associated outbreak, I wondered on Twitter if we'd see deaths "catch-up" to announcements as cases struggling for life in hospital, eventually succumbed to the damage started by their MERS-CoV infection. This is a virus with an apparently determined ability to wreak havoc in the lungs and kidneys of those who often already have an associated underlying disease of these tissues, and/or the cardiovascular system.

From the chart above, it looks like that catch-up is happening as the number of fatal outcomes (red dots) among the total of laboratory-confirmed MERS-CoV detections (green mountain) has been rising from mid-April onward. 

Most notable to me is that the proportion of fatal cases (PFC) is now also rising as the number of new detections no longer outstrips the number of deaths being reported. 

This is the first rise (rise >1% in PFC at least) since the beginning of March and reverses a trend of relatively stable PFC which dated back to late Aug-2013. 

This bears watching both at face value but also in light of the recent concerns about extra testing and its impact on straining the KSA healthcare system. It might be more prudent to discuss the significant strain being placed on the KSA healthcare system by filled hospital beds and intensive care units and the supportive management of severely ill people detected as result of that testing. 

The testing  should be telling the MOH (even if it no longer tells us) in which hospitals the problems lie. At this stage, 21-days after the KSA Health Minster was replaced (21-April-2014), we'd expect to see the case numbers dropping as (hopefully) changes have been implemented that (hopefully) focused a stern gaze toward (hopefully) improving healthcare worker safety and (hopefully) infection prevention and control (IPC). 

21-days into the new Health Minister's watch lies well outside the upper limit of the known MERS-CoV incubation period. If changes were sufficiently sweeping and if the war on MERS was entering a phase, detection numbers should be dropping as steeply as H7N9 cases do after a wet market closure. But they are not. Not yet anyway. 

The accumulating MERS-CoV detections (green mountain) are looking to be slowing a little, which is good news. But of course with that slowing, I expect we'll see those red and back lines rise in the weeks ahead as deaths keep occurring. Which is not, and will not be reported as, good news. About the only thing to stop that numerical marker from rising would be a new outbreak of cases. Not something we want to see. And of course, it would only be a smoke-screen for the fact that MERS-CoV positive people are still dying from MERS-CoV infection and the complications that ensue.

Last point for this post. Recent KSA MOH reports are more up-to-date than they have been for a while so we can see that MERS-CoV infected people are still being admitted to hospital, even in recent days (e.g. 7-May), and MERS-CoV infected people are dying relatively recently too (e.g. 10-May).

H7N9 deaths jump significantly....

Click on image to enlarge.

Twitter was buzzing this morning with news that several sources had announced a new total number of deaths in human cases of H7N9 infection.

It was not a total surprise that there were more deaths than we had heard about, and that is for several reasons:

• In Wave 1, Spring 2013 in South east China, there had been a greater proportion of deaths than we have seen in Wave 2. That's seemed unusual.
• After Wave 1, the proportion of fatal cases (PFC; see background here) sat up as high as 33%. Wave 2's high case numbers but few reported deaths had lowered that to 18% at one point. If the virus hadn't changed and human-to-human transmission had not changed then that was incongruous
• The media were reporting higher numbers than we had data for in early Feb and in late Jan, Xinua reported 26 deaths in Zhejiang alone for 2014 - this far outstripped any publicly data available

So now we see that the tally is 112 fatal H7N9 cases among people infected with a laboratory confirmed H7N9 virus, since the outbreak began in 2013; that tally includes both waves of human cases. That makes the PFC among the 361 confirmed human cases at 31%. 

So this one new piece of news has bumped up the PFC by 10%. From 1:5 (22% last week) to nearly 1:3 cases dying after acquiring infection. 

Thankfully, H7N9 is not spreading efficiently among humans (or chickens according to reports). But these are numbers to care about.

For comparison, my Excel sheet has 64 cases with data that I can cross-check (I believe that agrees with the FluTracker's count also). 

The last media update I looked at had a tally of 77 fatal outcomes. 

So we have between 35-48 people have died without any ability for anyone outside China to link them to:

• their age
• when they became ill
• where they were
• how they may have acquired their infection
• their sex
• time to hospitalization and diagnosis
• length of stay in hospital 
• what contacts they had and how they have fared. 

I think that this is a ball that has been not just been dropped, but buried in a hole and covered over with feathers. I'm disappointed by such a gaping data loss. And don't get me started about the absence of H7N9 sequences from 2014 cases!

Sources...

1. SCMP with higher death tallies than public data indicated
   http://www.scmp.com/news/china/article/1425289/january-worst-month-chinas-human-h7n9-outbreak
2. Xinhua lists 26 deaths in Zhejiang alone for 2014
   http://news.xinhuanet.com/english/china/2014-01/21/c_133060657.htm
3. VDU blog on missing deaths
   http://virologydownunder.blogspot.com.au/search?q=deaths+h7n9
4. Mike Coston's Aviann Flu Diary take one the new data, with other sources
   http://afludiary.blogspot.fr/2014/02/chinas-moh-h7n9-fatalities-higher-than.html?m=1&utm_source=dlvr.it&utm_medium=twitter
5. FluTracker's thread with links to eth WHO report
   http://www.flutrackers.com/forum/showthread.php?p=525996#post525996
6. China's Ministry of Agriculture report of enlarged H7N9 death tally
   http://translate.google.com/translate?u=http%3A%2F%2Fwww.moa.gov.cn%2Fgovpublic%2FSYJ%2F201402%2Ft20140220_3791429.htm&hl=en&langpair=auto|en&tbb=1&ie=UTF-8
7. The WHO report under the "vaccines" section
   http://www.who.int/influenza/vaccines/virus/recommendations/201402_recommendation.pdf?ua=1

H7N9 snapdate; the number of surviving and fatal human cases...

Click on image to enlarge.

A quick resurrection of a chart which hasn't been updated in 4-months. 

Interesting that the average/median values have not changed too much except perhaps that the deaths have slid a little further towards the older age band. You can still see the older version on my dedicated H7N9 page here for comparison. 

But please note: there are a number of fatal cases not listed here because they could not be identified. 

I currently have data for 58 fatal cases since Feb-2013. This value is roughly supported by a recent ECDC report so clearly we're all in the dark. But the list may be missing at least 19 cases. Some of those are definitely missing as the total comes from official China sources in 2013 which is not supported by public case data. But in 2014, China's media report have reported a higher fatal case load than we can see from released numbers. 31 deaths in 2014 and 46 deaths in 2013 = 77 total deaths; written 2-days ago in the South China Morning Post) 

No-one seems to be making consistent note of H7N9 deaths in 2014. 

Sporadic mentions of this total or that total flutter around but little is being confirmed officially and regularly. 

Similarly, toward the end of H7N9's Wave 1, identification of fatal cases became hit-and-miss and they became detached; cases were initially identified by sex, age and locale at illness onset, but not linked to their deaths. 

Apparently the World Health Organisation can only take a passive role here as it awaits case information to be collated form the Provinces and municipalities and passed to Beijing and then it. That;s not a criticism of WHO, but I cannot understand why fatality data are so hard to consistently extract when all the other information is available. 

Its not as though no fatality data are available; there are linked data for 19 deaths in 2014, so why not the rest? 

Why not an assertion that there are not more deaths? 

It's really just about linkage and defining the outcome of the infection. 

Cases don't have to be identifiable to be reported on. 

Cases have already been laboratory confirmed as H7N9 positive. 

And why not summarize hospital discharges while we're about it? That's key to understanding the true case fatality ratio (CFR) rather than the proportion of fatal cases (PFC defined here) which, unlike the CFR, does not require knowledge of the entire course of the clinical disease to be calculated.

It is always about data and communication thereof when we seek to understand emerging infectious diseases and outbreaks. Don't have it, can't fully understand what's happening.

Sources...

1. South China Morning Post article
   http://www.scmp.com/news/china/article/1425289/january-worst-month-chinas-human-h7n9-outbreak

Tracking virus-related deaths using publicly available data...

Click on image to enlarge.

Here's the cumulative case chart overlaid with the cumulative deaths and PFC. see the story behind the term PFC here, created by VDU to avoid issues around case fatality rate/ratio (CFR) which relies on knowing when cases have recovered.

I have two PFC values charted here. In black dots, is my curated list based on fatal cases (n=40; red dots) that have been announced publicly. 

In yellow are the numbers gleaned from media releases and the WHO - the latest number being 52 fatal H7N9 outcomes. 

Somewhere towards the end of the initial H7N9 outbreak in May, we stopped seeing reports from China that could link fatal cases with those H7N9 cases they initially announced. If anyone knows of a complete public list of fatal H7N9 outcomes that contains all 52 cases with age/sex/date of illness onset/date of death/province, I would be most grateful to be made aware of it.