Monday 27 March 2017

Happy 4th birthday...

This blog!

I'm sure somene one said that a year in social media equates to 7 cat years. Maybe it 9.

Anyhoo, I'll be hunting down a cake for us all today - or maybe a muffin at the canteen.

My heartfelt thanks to those who have made writing for this blog so much fun. 

I hope we all keep learning about viruses together. Anything is worthwhile if you learn something from it. ...and have a reference to prove it was based on data...and cite that reference...and it gets cited by others...and peer reviewed....

Happy 4th!!

Tuesday 14 March 2017

SNAPDATE: H7N9 by map...

With the latest numbers out from Hong Kong's Centre for Health Protection, Week 10 marks the second week of around a dozen human spillover cases - likley to be poultry to human infections.

Click on image to enlarge.
While the number of cases at each geographic location within China is small given the millions living in each region, this week's cases are spread across 10 provinces or municipalities. A huge area. 

Guangxi province is adjacent to Vietnam. It's always worth remembering that an outbreak anywhere can turn into a threat everywhere

From these numbers, there is no obvious escalation of human cases to suggest anything has changed in the way H7N9 infections are acquired; they remain relatively rare and from animal-to-human close contact.

While the cases are still coming, the totals (see weekly and daily bars below) shows a levelling off - at last suggesting a slowing of this season's epidemic. But stay tuned.

Numbers of human H7N9 cases.
Image taken from the VDU static H7N9 graphs and numbers page.
Click on image to enlarge.

Sunday 5 March 2017

Avian influenza A(H7N9) virus in humans: lay of the land...

I'm getting a little more of a grip on the H7N9 numbers thanks to the data from the Hong Kong Centre for Health Protection (CHP) [1] - which have been a solid source in 2017.

Keep in mind that these numbers :

  1.  are imperfect because they are reported inconsistently by those who have the data and because they only contain partial detail - death details are impossible to come by.  Please keep in mind that there is no global, running-tally of H7N9 cases presented to the public, by any public health entity. There is a great line list from the citizen-run FluTrackers list, ([2]my usual go-to) but in 2017 they got swamped by these unsatisfactory data.
  2. will only represent those cases that have been lab tested. Any people who have met the criteria for being a "Case" [pick from 3-6] - which in most instance means being sick. In some instances a cases is identified because diligent doctors have followed up those people who had contact with a known case - which is called contact tracing. Sometimes these contacts may virus positive by only mildly ill or have no illness at all.
I've graphed these wrangled data using one of my older formats - to show which province, municipality or autonomous region is contributing to the peaks as the site of origin for an H7N9 case.

Sometimes where a case has been detected may not be where they were infected. I prefer to talk about where a case "acquired" their infection. 

Click on image to enlarge.
Data for this graph can be downloaded from my static H7N9 graphs page, here.[9]
We can see - perhaps - that the major contributors to the peaks in January and February's 5th Wave are Jiangsu province (green circles) followed by Zhejiang province (orange circles) then Guangdong (brown circles), Anhui (purple circles), Hubei (red squares) and a range of smaller contributor regions.

Guangdong and Zhejiang are familiar to H7N9 watchers as being hotspots for human spillover and while Jiangsu has always had a presence in the outbreaks, it has had a very big season this time around.

It remains to be seen whether a range of market live poultry market (LPM) closures has cut the flow of virus into these markets and to their many, many visitors. These closures have been in response to cases rather than to prevent the outbreaks but no obvious nationwide coordination is apparent. It seems likely that spread of infected fowl will continue until more markets close, the source is contained or the seasons for influenza spread (winter and colder shoulders) is over.

Stay tuned to the CHP update this week - last week's tally was lower than previous weeks; a blip or a trend? 

A few things about this graph.
  • It mirrors the FluTracker's line list numbering scheme up until FT816. From entry No. 817 it uses data from the CHP reports. These are PDFs but as they helpfully told me by email this week - you can extract the data yourself using Adobe Acrobat Pro. If you don't have that - I've already done that extraction and am happy to share an Excel version of it with you. Shoot me an email, leave a message or Tweet me @MackayIM.
  • The Outbreak numbering - or waves - is based on when cases appeared or stopped. Its imperfect too. There are published schemes but they also differ from each other [e.g. 7,8]. This isn't life or death - you get the idea from the obvious peaks and troughs. FYI - this year I've updated my numbering for previous outbreaks.
  • Market closures include long term or short term shutdowns or rotating closures for one or more days for disinfection followed by restocking. Each province is a populous place. Often markets are closed here or there but not everywhere in a province and certainly not all provinces at once. 
  • Data are plotted by week of illness onset (hard data to come by) or when the case was reported. The grey peaks indicate totals from all provinces for that week