Another face to Colombia's Zika virus data...

I was checking out the interactive Zika virus data graphs that are hosted by the Pan American Health Organization-World Health Organization (PAHO-WHO). 

From http://ais.paho.org/phip/viz/ed_zika_epicurve.asp.[2]
Click on countries to see different data.
Click on image to enlarge.

Last week I wrote about the various ways that the apparently same data were being graphed. In the space of a week - the format of the PAHO-WHO version has changed again. Quite a lot. 

It now looks more like the Colombian weekly epidemiological report again.[3; graph below]

Excerpted from the latest Colombian National Institute for Health.[3]Click on image to enlarge.

It's kinda disappointing to think that these are some of the better data on Zika virus in the public domain around today.
I'm not even going to ask what's happened to the PAHO-WHO data for Brazil this week!

From http://ais.paho.org/phip/viz/ed_zika_epicurve.asp.[2] 
Click on countries to see different data.
Click on image to enlarge.

References...

1. http://virologydownunder.blogspot.com.au/2016/04/the-many-faces-of-colombian-zika-virus.html
2. http://ais.paho.org/phip/viz/ed_zika_epicurve.asp
3. http://www.ins.gov.co/boletin-epidemiologico/Boletn%20Epidemiolgico/2016%20Boletin%20epidemiologico%20semana%2016.pdf

Colombia...no recent confirmed Zika virus positives or reporting lag or...?

UPDATE #1: 24APR2016

Okay. Let me rephrase an issue I had from one of my posts yesterday-the one looking over the Colombian data. The relevant text was...

"I'm still not clear how the tally of confirmed cases is rising, but not showing up on the Colombian NIH graph (the red bars; the grey ones are clinical suspected cases). 

I've excised and posted the Colombian NIH graph to the left. 458 cases this week should show up clearly using that axis - it was the second biggest tally, just below Week No. 8 (also missing?) and 1.6X bigger than Week No. 4 which has been plotted and shows up clearly. I'm just assuming that the graph person has forgotten but would love to know if there is something else going on."

Below I've excised the relevant graphs from the most recent epidemiology report (Week No. 13 [2]) and also the one that first plotted confirmed alongside suspected Zika virus (ZIKV) cases (Week No. 11 [1]).

You can in fact see that the red bars have grown over the 3 week period - but that none of the 700 confirmed cases reported over that time have visibly made it into a bar after Week No. 5. So my question is whether that means:

1. no newly confirmed ZIKV cases have been detected/become ill during the past 8 epidemiology weeks (keeping in mind that Week No. 13's report only reports data to 02APR2016)
2. there is an 8-week lag in getting lab data into these reports?
3. Someone forgot to update the graphs
4. Colombia is no longer testing for ZIKV

 ..or perhaps something else altogether?

Week No. 11.
Data from [1]

Week No. 13.
Data from [2]

The lag seems to be the leading theory so far (see Tweet below)...but that's quite a lag. 

@MackayIM @INSColombia graph reflects adjusted reported cases per wk. not #s on table. "New" confirmed may be old. Wk12 only 1,100 on graph

— Gabby Molinolo (@gabby_molinolo) April 10, 2016

I find it strange that no-one seems to be critical of the testing lag here and yet it was the topic du jour for MERS in the Kingdom of Saudi Arabia. 

Is it simply about the perception of resources - the Americas are perceived as resource-poor and the Saudis as resource abundant? 

In the area of lab testing, I'm not at all convinced that the Saudis were at all resource rich though. I don't think they were in any better position at the start of their outbreaks. Especially if the resource we measure is understanding about testing to understand the aetiology of an emerging infectious disease.

References...

1. http://www.ins.gov.co/boletin-epidemiologico/Boletn%20Epidemiolgico/2016%20Bolet%C3%ADn%20epidemiol%C3%B3gico%20semana%2011.pdf
2. http://www.ins.gov.co/boletin-epidemiologico/Boletn%20Epidemiolgico/2016%20Bolet%C3%ADn%20epidemiol%C3%B3gico%20semana%2013.pdf

Updates...

1. Added a new 3rd option which was not carried over from the orginal post 10APR2016  - and pushed the previous 34rd to 4th

Plotting the Brazil microcephaly data...

These graphs are made using data obtained from Week 13's Brazil Ministry of Health microcephaly report, which you can also find here.

The chart above reports the second quietest week of 2016 in terms of the number of newly suspected microcephaly cases in Brazil up to 02-April-2016.

In the second graph, we can see a steeper rise in suspected microcephaly diagnoses that were discarded as being not microcephaly - the biggest since these microcephaly reports began. This rise was greater than the rise in confirmed microcephaly diagnoses - which was the 5th biggest rise since reporting began. 

Zika virus (ZIKV) detections jumped by the largest weekly amount (up by 40) since the Brazilian reports began noting laboratory confirmed detections in these reports. 

Also the biggest change, and not plotted here, was a drop of 245 cases under investigation, down from 4291 last week, to 4046 this week.  The second consecutive week seeing a negative adjustment.

There is still no sign of ZIKV testing in a more general sense from Brazil - just these numbers in the context of microcephaly. We rely in Colombia to get a sense of how many overall suspect cases are confirmed as ZIKV positive.

Microcephaly: diagnoses and discards and those with Zika virus...Report #15

The latest Brazil MOH report on microcephaly was released 03-MAR-2016 – Report #15. 

The following is a summary of info after translation via Google.

• There have been 5,909 (+269 since last week) reported suspected diagnoses of microcephaly sometime unspecified in 2015
• 4,222 (+115) remain under investigation
• 1,687 (+154) cases have been investigated and classified
• 641 (+58; 38% of those investigated, 11% of all reported) have been confirmed 
• 82 (+15; 13% of confirmed microcephaly cases; 1% of suspected cases) have been laboratory confirmed as Zika virus infections (no specifics on method or whether mum/foetus/infant/samples etc)
• 1,046 (+96; 62% of those investigated, 18% of all reported) have been discarded 

The MOH notes that if a diagnosis of microcephaly was confirmed, then the mother was most likley infected by ZIKV and that the figure of 82 is an inadequate representation of the number of cases related to the virus.[2]

References…

1. http://portalsaude.saude.gov.br/images/pdf/2016/marco/01/coes-microcefalia-informe-epid15-se08-2016-01mar2016.pdf
2. http://portalsaude.saude.gov.br/index.php/cidadao/principal/agencia-saude/22396-saude-investiga-4-222-casos-suspeitos-de-microcefalia-no-pais

Ebolaradication around the corner..?

The downward trend for new, confirmed, Ebola virus disease (EVD) cases continues as we can see in the graph below.

This graph plots the number (each blue data point or dot) of newly confirmed cases in each World Health Organization situation report of summary. The joining lines don't mean anything (the WHO doesn't provide any numbers to place between dots)- they just make it cleared how think go up and down. You can see a little about Monday'itis and its role in the bumpy road to zero cases here.

Below is part of one of the charts from my Ebola virus disease numbers page -  you can get to it from any page on this blog by clicking on that tab up there^.

The line indicating the linear trend of fewer cases over time has a p-value of  0.023 and was
calculated within Tableau Public Edition v9.0
Click on image to enlarge.

I've zoomed in the graph to highlight distinct data points (thanks to Ramon i.e. @HlthAnalysis for helping me learn to make the separately coloured dots I missed so much from my Excel graphs). Without the zooming we can see from the full dataset that it's getting a bit hard to see each data point as time goes by, and case numbers shrink (yay!).

Case number changes between Ebola virus disease reports...

This is one of my favourite charts for following the Ebola virus disease outbreak in West Africa because it shows how things are changing from report to report. 

It plots the total number of suspected, probable and laboratory-confirmed cases between reports - which is a measure of change over time that is not cumulative.

That's not to say that understanding this chart is easy for everyone...as with everything, what you take away from it may be heavily influenced by your own perspective and your background in reading graphs. I have written something about how to read some of the graphs on my blog here, which may be helpful too.

Uses World Health Organization data up to and including the Situation Report from the 5th-Sept, 2014.
Click on chart to enlarge.

I've marked up the last three periods between reports to highlight that the time changes differently. You can see this for yourself if you look carefully at the horizontal or "x" axis (the one that has the dates) and look at where each dot lines up with its date. Some are further apart than others. 

You can also mouse over the dots on the interactive version of the graph here. That will tell you the dates. THe subtraction is up to you though!

The lines joining the dots here suggest what is happening between the WHO Reports, but the line do not actually use any real collected values...because we don't have them to plot. 

Technically, a bar graph would be more accurate, but I find a line graph easier to read at a glance. So do remember - we don't know what is happening between those dots. We're just presuming it.

H7N9 cases in China to be reported weekly instead of daily.

Well that will let the guys at FluTrackers get some more sleep. Mike at Avian Flu Diary(who could probably do with some sleep too) notes it might not have a real impact on provincial reports but time will tell. The info is nestled at the bottom of this report.

This is a real shame for the public who have been getting access to some great real-time data break-downs, assemblies and interpretations from the flublogians. 

Realistically, its not like we're entitled to these detailed data from China. It all takes work (and workers) to compile, release the number and organize and hold the official press conferences etc...but in the age of "always on" and instant internet gratification...I think we feel that is how it should be.

We would feel entitled to the information if it were a story about a terrorist act or the latest comings and goings of a public figure or personality, or sports scores. Its been nice to see something as important as the emergence of a new human pathogen receiving the attention it has...at least so far.