Random Analytics: H7N9 Analysis (to 21 Apr 2013) Short Version

***** Please note that this post reflects data, graphs and infographics of the Avian Influenza A(H7N9) virus and was updated with public source information between 21– 22 April 2013 *****


The avian influenza A(H7N9) virus which first struck Shanghai in mid-February when an 87-year old Shanghai resident fell ill with flu/pneumonia like symptoms, then subsequently died two-weeks later continues to cause disruption in China.

Like previous outbreaks of respiratory disease, this new strain designated by the World Health Organisation (WHO) as avian influenza A(H7N9) has the potential to cause major issues for the World’s second largest economy and like SARS previously to spread internationally.

I’m not a medical specialist but I do like data so each day (when available) I’ll try to present a range of analytics using the most up to date public and primary sourced information available so for any interested parties they can get an evidenced based snapshot of the evolving situation.

Case Fatality Rate

In the past 24-hours there have been 102-cases including 20-fatalities. This equates to a case fatality rate of 19.6%. For
context SARS was 10.9%.

9 persons (8.8%) are known to have recovered and have been discharged from hospital or treatment.

The most recent fatalities reported were on the 21st April 2013 via Xinhua and WHO.

Here is an infographic looking at those hospitalised, confirmed fatalities and patients that have had confirmed recoveries. Where I have been able to confirm the details I have broken the groups into male/female sub-groups.

1 - H7N9_Infographic_130422


The next graph is an epidemiological analysis of cases, including deaths and patients who have recovered where I have been able to verify the onset of illness. Currently I have been able to verify the illness onset dates of 91-patients which have been included in the following graph. (Note: Where a range of dates for onset of avian influenza A(H7N9) have been used I have input the earliest date until I can verify the actual onset).

Please note that WHO issue a similar report on a weekly basis.

2 - EpedemioCurve_130422

Cases by Municipality/Province

The next infographic is a look at the 102-cases by Municipality (in the case of Beijing and Shanghai) or Province.

At this stage the bulk of the official reported cases have been in the Zhejiang Province (37.3%) followed by the Shanghai Municipality (33.3%) and Jiangsu (22.6%).

At a regional level the cases are almost entirely confined within the East China Region (96.1%), represented by the blue section in the infographic.

Outside of the East China only Henan and Beijing have recorded instances of H7N9, although to date no deaths have been reported in these areas.

Analysis: Shanghai was the originating municipality and had the bulk of the provincial cases. During the period 20 – 21 April the neighbouring province of Zhejiang equaled Shanghai for case numbers (33 each) and the most recent announcements have stretched this further (38 compared to 34).

Also worth noting that currently Zhejiang has a much lower percentile of fatalities (13.2% against 32.4%). Being at the epicentre of the outbreak while the disease was an unknown could be an explanatory factor in the much higher Shanghai case fatality rate.

3 - CasesByRegion_130422

Presentations by Age Group

Here is a look at those presenting with avian influenza A(H7N9) by their age cohort.

Of the official 98-cases where age data is known the average age was 59.3-years and 45.9% of the patients were aged 65 or greater.

Currently only four cases were under the age of 24, three children all under the age of 10 and a 21-year old female from Jiangsu. This does not include the 4-year old asymptomatic case in Beijing reported in mid-April.

Analysis: A definite and steady trend that this respiratory disease is targeting older cohorts and men especially. For those aged 65 or over the male to female ratio is 2.75:1

4 - PresentationsByAge_130422

Age at Death

Of the 16-cases where death has been verified the average age was 61.3-years and 43.8% of the deaths were recorded in those over the age of 65.

The youngest person to die was a 27-year old butcher from Shanghai with a history of hepatitis B. Being amongst the first victims of the disease he did not receive antiviral treatment until the second week (the recommended window for this treatment should be two days) and succumbed to the disease after 12-days.

5 - DeathByAge_130422


With the new information made available over the past 24-hours, the item of most interest data-point is the ascendancy of Zhejiang in terms of case numbers, overtaking Shanghai for the first time.

What is also noteworthy is the lack of town/city detail of the victims of that province that has been provided to the public via official or semi-official means (i.e. Xinhua). Zhejiang has had 38-cases to date but I’ve only come across one primary sourced reference to a city (Huzhou) while neighbouring Jiangsu with 23 cases has provided eight city level locations.

I’m not a China expert so if anyone has a reasonable answer to this lack of detail around Zhejiang please drop me a line via Twitter (@gmggranger)


Response from Crawford Kilian (@Crof) via his H5N1 blog:

The answer, I suspect, lies in a tedious search through Google-translated reports from Zhejiang Chinese media…

There he details, in some detail five recent Zhejiang cases via Hangzhou Ribao before stating:

My point is that this one report gives the location of each of the recent cases, and a diligent search should turn up earlier cases. Maybe some of these people were shipped in to hospitals in Hangzhou from rural districts (or not – Chinese cities don’t always distinguish between rural and urban). But we amateur epidemiologists should be content with gaining a sense of where these cases are being treated.

Fantastic advice! I totally agree Crawford and I’ll be tracking down some of this legacy detail in coming days to flesh out my database.


Updates (22 Apr 2013)

  • Added I/O section with Zhejiang response from Crawford Kilian:
  • Corrected all instances of Mortality Rate to Case Fatality Rate. Tks to Ramon Hernandez (@HlthAnalysis) for picking this up.

Updates (23/04/2013)

  • Updated the WHO SARS case fatality rate with a later study that put the CFR at 11.9% (not 9.6%).