Random Analytica

Random thoughts, charts, infographics & analysis. Not in that order

Month: October, 2014

Random Analytics: West African Ebola Outbreak (to 18/19 Oct 2014)

The World Health Organisation (WHO) has updated it latest advice on the Ebola Outbreak. Ebola Response Roadmap Situation Report 22 October 2014. Summary:


A total of 9936 confirmed, probably, and suspected cases of Ebola virus disease (EVD) have been reported in five affected countries (Guinea, Liberia, Sierra Leone, Spain, and the United States of America) and two previously affected countries (Nigeria and Senegal) up to the end of 19 October. A total of 4877 deaths have been reported.

The outbreaks of EVD in Senegal and Nigeria were declared over on 17 October and 19 October 2014, respectively.

EVD transmission remains persistent and widespread in Guinea, Liberia, and Sierra Leone. All but one administrative district in Liberia and all administrative districts in Sierra Leone have now reported at least one confirmed or probably case of EVD since the outbreak began. Cases of EVD transmission remain lowest in Guinea, but case numbers are still very high in absolute terms. Transmission remains intense in the capital cities of the three most affected countries. Case numbers continue to be under-reported, especially from the Liberian capital Monrovia.

Of the countries with localized transmission, both Spain and the United States continue to monitor potential contacts. On 21 October the single patient with EVD in Spain tested negative for the disease for a second time. Spain will be declared free of EVD 42 days after the date of the second negative test unless a new case arises during that period.

On 22 October 2014, WHO convened the third Emergency Committee on Ebola under the International Health Regulations (2005).

Here are three charts or infographics of the West African Outbreak with data confirmed by WHO to 18/19 October 2014.

01 - Ebola_WAfricaOutbreakTableau_141023


The West African Ebola Outbreak (2013-2014) infographic details the cases, fatalities, the provisional Case Fatality Rate (CFR) and health spend per capita in 2012 $USD.

Midway through October the biggest data find for me has been the difference between the index cases in Senegal and the United States. Lots of similarities but IMO four very important differences:

  1. Both cases entered the country without disclosing the possibility that they may have been exposed to Ebola;
  2. Both cases presented to hospital shortly after arrival in the destination country;
  3. Difference One: In the case of Thomas Eric Duncan who travelled to the USA he presented at hospital but was released back into the community. He presented again three days later (interestingly a key and important fact left out of the official CDC Overview) with more extreme symptoms. In the Senegal case the patient sought treatment in Dakar giving no indication he may have Ebola but was not released and a day later Senegalese authorities linked him to other cases. He was subsequently quarantined;
  4. Difference Two: In the USA contract tracing was slow to be implemented while PPE protocols were inadequate which resulted in the spread of the disease amongst Health Care Workers (HCW). In Senegal no HCW were infected.
  5. Difference Three: The USA index case died of the disease; potentially due to those critical days where he was without care while in Senegal the patient recovered and has returned to Guinea.
  6. Difference Four (and this is the key one which you can see in the infographic): The average health spend per capita in the USA is $8,895 (in 2012 $USD money) while Senegal makes do with just $51 per person. That’s 174 times less than the USA.

02 - Ebola_WAfricaOutbreakChart_141023

The West African Ebola Outbreak (by Nation) chart looks at the split between those recovered or still in treatment and those deceased. As per the latest update from WHO there is an element of underreporting going on in the worst impacted countries (Liberia, Sierra Leone and Guinea).

03 - WestAfrica_Cases~FatalitiesMonth_141023

The West African EVD Outbreak – All Cases/Fatalities by month details the epidemiological curve of the outbreak from March 2014 when the first cases and deaths became apparent. Each column is visualised by the flags of the seven impacted countries.

The mid-month data-find from this chart is that the epidemic curve in relation to deaths continues to increase month on month yet the case load seems to be pulling back. All the anecdotal evidence coming out of West Africa points to an increasing, rather than a decreasing case load as individuals, families and population centres avoid health care facilities. In past months, the case load mid-month would be approximately equal to the previous month. Best case scenario would be that West Africa is getting on top of the disease but the facts-on-the-ground don’t support this. I’d expect to see some revisions to the data in coming weeks and months.


I’ll update and post my charts again when the data becomes available for the entirety of October. In the meantime if you want to keep across the West African outbreak detail via some alternate sources then I would recommend Crawford Kilian and for more regular epidemiological posts Maia Majumder, MPH and Ian Mackay, PhD.


Data Sources

[1] World Bank. Health expenditure per capita (current US$). Accessed 23 October 2014.
[2] World Health Organisation. WHO: Ebola Response Roadmap Situation Report 22 October 2014. Accessed 23 October 2014.

Random Analytics: Ebola in the USA? Time for some context

Back at the end of September when the Thomas Eric Duncan Ebola case was first diagnosed in Texas and hit the headlines Ebola went from being an African issue to become an American problem. Social media exploded on the subject, Twitter mentions on Ebola tripled overnight and news agencies across the Western world struggled to keep up with the renewed interest.

Most of the analysis has been very good but some of the commentary has been downright crazy.

CNN did a story on how some US Republicans are calling for stronger border controls by linking Ebola with ISIS. As crazy as the idea of mixing the Ebola and ISIS narrative is CNN itself asked the same question prompting The Independent to write ‘The Isis of biological agents?’: CNN is asking the stupid Ebola questions.

1 - CNN_Ebola~ISIS

Stupid is as stupid does is not just reserved for the United States. In my own country of Australia a recently returned nurse was hospitalised on fears she may have contracted the disease after registering a mild fever. This incident had Bob Katter, the federal member for Capricornia suggesting a ban on people travelling to Australia from West Africa and calling for all returning Health Care Workers to be put into government isolation upon their return (the nurse in question was actually in self-imposed home isolation).

Anyway, I thought the debate on the one case in the United States against the 8,398 in Western Africa (as per the most recent WHO update) was worth putting in context. An infographic:

2 - EbolaCasesbyFlagSize_8Oct2014

The West African Ebola Outbreak – Cases by Country is a look at the six countries impacted by the disease with data to 7/8 October 2014. Each individual country is represented by their flag and the size of their flag is a percentile proportion of total cases.

As you can see Liberia has 48.5% of all cases, followed by Sierra Leone (35.1%), Guinea (16.1%) and Nigeria (0.2%). Both the USA and Spain have one case each which equates to 0.000119% of all cases and too small to be represented by a flag.

The story of Thomas Eric Duncan is tragic and the entry of a disease like Ebola into the West is scary but we need to get some context on how bad the situation is.


To the Lost


Data Sources

[1] Hooten, C. ‘The Isis of biological agents?’: CNN is asking the stupid Ebola questions. The Independent. 7 October 2014. Accessed 12 October 2014.
[2] World Health Organisation. WHO: Ebola Response Roadmap Situation Report 10 October 2014. Accessed 12 October 2014.

Random Analytics: Ebola in Liberia (to 5 Oct 2014)

The Liberian Ministry of Health and Social Welfare has released its latest numbers with their data confirmed to 5 October 2014.

1 - EbolaInLiberia_5Oct14

The Ebola in Liberia infographic charts the spread of the Ebola Virus Disease (EVD) by county. Each figure represents 10 lives. Note: The cases versus deaths in Maryland County are not a typo with the Ministry recording 8-cases and 9-deaths.

According to the latest data from Liberia a further 28-cases were recorded in the 24-hours from 4 to 5 October 2014. By county the new cases were reported in:

  • Grand Cape Mount (x1)
  • Lofa (x3)
  • Margibi (x11)
  • Montserrado (x13)

At the same time another 21-fatalities were recorded. New fatalities by county were reported in:

  • Grand Cape Mount (x1)
  • Margibi (x4)
  • Montserrado (x16)


To the Lost


Data Sources

[1] Ministry of Health and Social Welfare. Liberia Ebola Sitrep no. 142. Government of Liberia. Accessed 10 October 2014.
[2] Ministry of Health and Social Welfare. Liberia Ebola Sitrep no. 143. Government of Liberia. Accessed 10 October 2014.

Random Analytics: 100-days of MERS

Given that we are now half-way through the annual Hajj I thought I might spend some time looking at the Middle Eastern Respiratory Syndrome Coronavirus (MERS-CoV) which has been with us for some years now but seems to have fallen off the radar in favour of the maladie de jour, Ebola.

What publically sourced data is available is limited. In the past 100-days there have only been 25-notified cases (23 in Saudi Arabia, 1 in the United Arab Emirates and an exported case to Austria). The Kingdom’s updates are as brief as ever, the World Health Organisation (WHO) has clumped together a monthly update with only high-level data while the world’s attention is completely focussed on Texas and West Africa. Not only is the data limited but the Saudi’s have again reviewed their data and found a further 17 cases prior to 3 June that were missed. Ian Mackay wrote an excellent open letter to the KSA Ministry of Health in relation to that oversight (recommended reading).

For lots of reasons I haven’t updated my rudimentary MERS-CoV Db in a couple of months and what I found during my data-entry this morning I thought was intriguing enough to do an infographic with MERS notifications going back just 100-days.

1 - MERSbyCity_141005

The 100-days of MERS infographic details the 25-cases that have been notified between the 29th June to the 6th October 2014. The Riyadh count includes the young lady who travelled from Afif to Austria and one case where the KSA Ministry of Health provided no details (thus the figure is represented as a man).

Just to cover off the basic points in the infographic, there have been 25-cases since 29 June and two notified deaths (assuming that FluTrackers case number #863 is the 76-year old male from Najran who died on 25 September, thus a provisional Case Fatality Rate (CFR) of 8%, which is extremely low compared to the current 42.4% during the outbreak in the KSA. Of the 24-cases with details, four were female, the ages ranged from 27 to 76 and the average age was 54.1

Now to the really interesting data-points, some queries and a counter-factual.

  • A quick look at my Db tells me that during the period July – September 2013 there were approximately 56-cases of MERS (not including any that formed part of the 113 that were belatedly added without details). My first question: Is MERS on the decline given that the epidemiological curve seems to have declined by half since last year?
  • Even though the cases are very low the spread of the disease is extremely widespread. Over the past 100-days MERS has cropped up in Abu Dhabi (882km west from Riyadh via Route 10), Najran (974km south via Route 10), Taif (994km south-east via Route 40) and Arar (1,157km north-east via Route 65). My next question. Can someone explain why the cases are so low but seem to be so widespread?
  • There have been seven confirmed cases in Riyadh which has a population of 4-million and six cases in Taif, population approximately 500,000. Is there any reason why Taif is currently overburdened with the limited amount of cases?
  • The provisional CFR over the past 100-days seems very low at just 8%. Is that due to better care, less cases, better surge capacity, declining potency or another reason?
  • My last data point is really a counter-factual on the data that has been presented over the past three-months. The release of a second tranche of non-notified cases (this time 17 as compared to the previous 113) has to be questioned more deeply. You can always allow for a mistake but two is either a conspiracy or a cock-up. If it is a conspiracy are the Saudi’s ‘juking the stats’ in order to protect travellers from the Hajj? Are the Saudi’s using the current Ebola outbreak to limit the amount of information they are sharing? If it’s a cock-up why was it allowed to happen a second time in the lead up to the Hajj.

In Summary

Looking at the previous 100-days of data has me asking a number of questions. Is MERS on the decline? Why is the CFR so low? Why are the cases so widespread? Why has one small city got as many cases as the capital?

There are two incontestable facts. One: During the past 100-days MERS-CoV has been widespread across Saudi Arabia and the United Arab Emirates. Two: I also know where it hasn’t officially been.


Make of that what you will…

Random Analytics: Ebola across Africa (to 1 Oct 2014)

We are pleased to say we have controlled the spread of the epidemic.” Francois Fall, Foreign Minister, Guinea (14 April 2014).

First and foremost, I want the American people to know that our experts, here at the CDC and across our government, agree that the chances of an Ebola outbreak here in the United States are extremely low.” Barack Obama, President, United States of America (16 September 2014)

As the United States grapples with its first case of the Ebola Virus Disease (EVD) it is worth noting that the disease is ravaging West Africa causing thousands of deaths, destroying capabilities, causing starvation and reeking further economic devastation across one of the poorest regions in the world.

According to the World Health Organisation (WHO) which issued new data on the 1st October 2014 the West African EVD outbreak, which is ongoing now has 7,178 clinical cases with 2,800 in past 21 days (39.1%); and 3,338 deaths (a provisional Case Fatality Rate or CFR of 46.5%). Health Care Workers who are the front-line workers in the fight against EVD are also overly represented in the casualty statistics. Currently there have been 377 clinical cases (5.3%) and 216 deaths (6.5%). The latest update does not include the new case which presented in the United States.

What started as a post-graduate project to build a five-minute Ebola Virus Disease (EVD) lesson utilising just five graphs and six-dot points has now turned into more than 20 charts and infographics.

Here is a look at four charts or infographics of EVD across Africa with WHO data confirmed to 28 September and where applicable will include the US case.

Ebola across Africa

01 - Ebola_AcrossAfrica_141002

The Ebola across Africa infographic details the country specific outbreaks of Ebola since it was first discovered in 1976 (with a 1972 retrospective case from Zaire included). When I first completed this infographic in April the bulk of the outbreaks of Ebola Zaire had occurred in Central Africa with the Democratic Republic of Congo being the country most impacted in terms of outbreaks, cases and deaths.

Three interesting data-points have occurred over the past half-year. The first data-point is the fact that the weight of Ebola Zaire cases has shifted from Central Africa to West Africa (and now includes Nigeria and Senegal as countries where the disease has exported itself to). The second point is while EVD ravages West Africa another unconnected outbreak has occurred in the DRC.

The last point I think is really interesting and not fully reflected in the infographic is the differential between health expenditure of those countries impacted by Ebola in Africa versus that spent in the United States. In Africa the impacted countries health expenditure per capita ranges from just $15 in the DRC to $96 in Sierra Leone, while in the United States the amount is $8,895.

Notes: The 1976 – 2004 outbreaks of Ebola Sudan occurred in the bottom half of Sudan (now South Sudan). Zaire was renamed the Democratic Republic of Congo in 1997. Health expenditure per capita is expressed in $USD and represents 2012 data.

Ebola (Top 10 Outbreaks by Case Numbers)

02 - Ebola_Top10OutbreaksByCaseNos_141002

The first chart displays the top 10 outbreaks in order of case numbers. Each horizontal bar is filled with the flag(s) of the country where the outbreak occurred. In the case of the West African outbreak there are now six countries involved and the flags are in order sorted by number of cases.

With clinical cases now above 7,100 the West African outbreak has now become largest Ebola outbreak in history based on both case numbers (7,179) and fatalities (3,338). The second largest outbreak was of Ebola Sudan which occurred in Uganda (2000) when 425 became infected and 224 died.

Another couple of very interesting data-points. The first is the fact that the West African outbreak is the now the first to cross national land borders and the first to travel internationally out of Africa. The only other recording of an EVD that jumped borders prior to this outbreak was the Gabon/RSA (1996) outbreak. In that instance a doctor caught the disease in Gabon and subsequently took an international flight to South Africa where he became ill and infected other Health Care Workers (HCWs).

The second interesting point is the inclusion of the latest DRC outbreak. With 70-cases and 42-deaths recorded to date it has now moved into the 9th largest outbreak by case numbers.

Ebola (Cases by Classification and Year)

03 - Ebola_CasesbyClassYear_141002

The next chart shows cases by classification (in order they are Ebola Zaire, Sudan, Bundibugyo, Reston and Ivory Coast) by year, split into those recovered or those deceased (which follows in a red variant) and ordered from most impacted to least impacted year. Due to the severity of the latest outbreak I have removed 2014 as a column and included it as a separate pie chart.

The combined West African and DRC outbreaks has become the most significant in terms of case numbers by year, eclipsing the 1976 dual outbreaks (Zaire and Sudan) which saw 603 cases and 431 deaths (a combined Case Fatality Rate of 71.5%).

Notes: Several years had just one case. They are 1972 (a retrospective fatality of Ebola Zaire in Zaire), 1977 (a single case of Ebola Zaire in Zaire), 1988 (an accidental infection of Ebola Zaire in Porton Down, UK) and 2011 (a single fatality of Ebola Sudan in Uganda).

The West African Outbreak – Cases & Fatalities by Month

04 - WestAfrica_Cases~FatalitiesMonth_141002

The final chart shows both case numbers and fatalities by month. Each column is split into the five impacted countries (the USA will most likely find its place with the October data). The data is visualised by the varying national flags.

The obvious data-point from the infographic is that the epidemic curve continues to increase month on month with data still outstanding for the 29th and 30th of September.


Over the past 24-hours a lot has been made of the first case reported from the United States. As much as that is bad news the disease is destroying several countries as we speak so I am hopeful that the attention the US case creates might mean more support for the countries that are currently doing it very tough.


Data Sources

[1] World Bank. Health expenditure per capita (current US$). Accessed 2 October 2014.
[2] World Health Organisation. WHO: Ebola Response Roadmap Situation Report 1 October 2014. Accessed 2 October 2014.