Random Analytics

Charts, Infographics & Analytics. No Spinning the Data. No Juking the Stats

Month: July, 2013

Random Analytics: Hendra! (to 1 Aug 2013)

Recently I wrote and completed some analytics about the Ebola Haemorrhagic Fever (EHF) which lives in the rain forests of Africa. The natural host of that disease are the fruit bats of the Pteropodidae family.

Queensland is also host to a range of fruit bats. It has also become the host of a new and highly dangerous virus. Originally coined the equine morbillivirus it was later named the Hendra Virus (HeV) after the suburb which hosted the index cluster. It is closely related to another deadly variant, the Nipah virus which emerged from Malaysia in 1999.

HeV is a rare, emerging zoonotic virus (a virus transmitted to humans from animals), that can cause respiratory and neurological disease and death in people. It can also cause severe disease and death in animals, namely horses, resulting in considerable economic losses for horse breeders.

Key Facts (last updated July 2012 via the CSIRO):

  • In 1994, CSIRO and Queensland researchers discovered a completely new virus now called Hendra virus;
  • Scientists believe fruit bats are the natural ‘host’ of Hendra virus, meaning the virus is carried by bats but has little effect on them;
  • Hendra virus is not highly contagious but if transmitted to horses and humans it can be lethal
  • CSIRO has shown that a prototype vaccine can protect horses against Hendra virus;
  • Pfizer Animal Health has made the Equivac HeV vaccine available, under permit, for accredited veterinarians to administer to horses.

Since August 1994 to the 28th July 2013 there have been 48-clusters of the disease in Australia which have resulted in four human deaths. The confirmed human cases stand at seven giving it a very high Case Fatality Rate of 57%, fortunately in in a very low sample size.

It has also caused the deaths of 90-horses. From that number I have been able to verify, with a reasonable amount of certainty, that at least 54-horses died of HeV (60%) and that 36 (40%) were euthanized. Since 2011 two dogs have also become infected; one recently and both were subsequently euthanized. The practice of euthanasia has occurred prior to death due to humane reasons or as part of the AUSVETPLAN (2013) directive which prescribes euthanasia in all animals that test positive for HeV.

Most primary source subject matter experts on the subject state that the horse CFR is approximately 75%. If you interested in the research then I’d suggest the 2011 FAO Report Investigating the role of bats in emerging zoonosis.

A Quick History of HeV

01 - Hendra_CasesbyLocation_130728

Aug 1994 – Oct 1995: Mark P, a 35-year old Mackay canefarmer and horse breeder assists his wife, a veterinarian, undertake necropsies on their property after two horses die of mysterious circumstances. Mark would become ill and be admitted to hospital with suspected aseptic meningitis. 13-months later he relapses and dies 21 October 1995 thus becoming Hendra’s only victim to die of HeV linked late-onset encephalitis. Further testing reveals the presence of HeV in his brain and samples taken from the two horses also confirm HeV making this the retrospective index case.

Sep 1994: The first known outbreak of the disease occurs in the northern Brisbane suburb of Hendra which is a hub of horse racing and the location of two race-tracks. After tending to a sick horse Vic R, a 49-year old local trainer and his 40-year old stable-hand become sick. Vic would die within the fortnight while the stable-hand slowly recovered and was released in October 1994. 20 horses would die during this index outbreak with the necropsies undertaken onsite. From Queensland Department of Primary Industry (now QDAFF) it would lead to the discovery of a new disease by the CSIRO Australian Animal Health Laboratory.

02 - HendraSite_1994

Photo: Queensland Department of Primary Industry officer’s onsite at the Hendra (1994).

Oct 1994 – May 2008: During the next 13-years and 8-months there are only 7 HeV virus clusters recorded, each involving a single horse and all but one are located within Queensland. In the Gordonvale cluster (Oct 2004) a young female veterinarian assisting on a necropsy is infected, she becomes ill but does not die. In October 2006 the first case of HeV is recorded in Mullumbimby, New South Wales.

Jun – Aug 2008: While treating a sick horse both Ben C (34 year-old veterinarian) and a Natalie B (20-year old vet nurse) became infected after taking nasal swaps. They both had flu-type symptoms and were given a five-day intravenous drug course of Ribavirin when found to have HeV before going into intensive care. Ben C survives in intensive care before dying 20 Aug 2008 (about 37-days after admission). Natalie B was discharged the day before his death. Eight horses are infected; three die of HeV and five are euthanized. The ABC Australia Story covered this case in some detail including an emotional interview with Natalie B and The Weekend Australian Magazine did a comprehensive piece in Mar 2013.

Jul – Aug 2009: Alister R, a 55-year old Rockhampton veterinarian became the seventh person ever to contract the virus. After a long day he endoscopes an additional sick horse without first returning to his car to don his PPE (Personal Protective Equipment) such as a surgical mask or gloves. Although he is one of four people to be given an experimental anti-viral treatment he becomes ill approximately 20-days later and dies on the first day of September. Prior to his own infection he stated that the next person to die of Hendra would be another veterinarian.

2011: With 18 recorded clusters in a calendar year (eight in NSW and 10 in Queensland) HeV spikes to record levels. 2011 is also the first year a case is confirmed in a dog which is subsequently euthanized.

2012: On November 1 2012 the CSIRO announced a vaccine against Hendra Virus: Equivac® HeV, the world’s first commercially available Hendra vaccine for horses. CSIRO senior pathologist Deborah Middleton discussed the vaccine via The Conversation.

2013 – Vaccination uptake low – Cases of HeV Continue

There already have been eight clusters of HeV already recorded in 2013, all in unvaccinated horses and equal to the cluster total for 2012. While a horse vaccine has been made available uptake has been low. A recent report by Rural Weekly covered the subject.

In July the NSW DPI confirmed that a dog on the mid north coast had tested positive to HeV. Currently there is no vaccine for domestic animals (cats and dogs) or humans.

The following infographic shows cluster events since the vaccine became available in November 2012.

03 - Hendra_CasesbyUVLocation_130728

Temporal Pattern

As shown in the following graph the season of winter is the peak period for HeV. Winter outbreaks account for 27 (56.3%) of the cluster onsets.

04 - TemporalPattern(CalYr)_130728

Information Underload

One of the biggest discoveries I made during my research into HeV was that there are a large number of web-sites which purport to be knowledgeable about the subject yet contain incorrect, misleading or untruthful information.

One prominent site which has been active for many years states “There is currently no cure and no vaccine”. Not only is this untrue but after discussing Hendra with a number of specialists I found that there was also a significant amount of disinformation which was impeding the uptake of the Equivac HeV vaccine program.

Another issue I found was that some primary source outlets had horribly out-dated information regarding both HeV and the Nipah virus. The World Health Organisation is especially at fault here having last updated these pages in July 2009. As an example the WHO still incorrectly states in its HeV Key Facts that there is no vaccine for horses.

When primary source health agencies, especially those with global health responsibilities don’t update their sites with the most recent and updated information (and data) they allow the cranks and the quacks space to operate in this asymmetric digital world.

If you want to know about HeV on the internet then stick to primary source materials (i.e. BioSecurity Queenland and the New South Wales DPI are both recommended).

If you NEED to know about HeV then seek professional advice, preferably with someone face-to-face.

Final Thoughts

One of the fondest memories of my early childhood (lived in the late 1970’s) is that of my brother and myself lying on our backs on the 18th-hole of the Bellingen Golf Club looking up at the sky at sunset. The fading blues would be cut in half, as if by a black knife, as tens of thousands of fruit bats would evacuate their caves in the Dorrigo Hills and venture down to the banana plantations that surrounded Coffs Harbour.

Some three decades later I find myself living in regional Queensland, in a region that has experienced a Hendra outbreak and next to a horse paddock with agisted horses. Often we will interact, feed or pat the horses.

It’s not unusual for my family to sit outside in the early evening during winter, next to a fire and watch the horses graze in the next door paddock.

Occasionally, much to the delight of my youngest son we will see a flying fox.

Out of interest I’ll continue to track HeV, a very rare zoonotic disease.

Using the evidence I know that the risks to my family are negligible.

Note: If you found this interesting you might also like to see my analytics of Ebola or H7N9.

Acknowledgements: My fellow Queensland blogger, Dr Ian Mackay has commenced a Virology Down Under page for Hendra. His H7N9 & MERS-COV material is excellent so looking forward to see how this page develops. Many thanks to Trish Roderick (Heritage Collection Coordinator, Mackay Regional Council) who was able to confirm and send through scanned details of the Mark P (35M) case from local reporting.

Update (1/08/2013)

  • Updated ‘A Quick History of HeV’ section with confirmed Mark P details.
Advertisements

Peak Jobs and HR Automation

During a recent recruitment discussion on #NZLEAD I brought up the concept that not only could most of the recruitment process be automated but there was a body of evidence that was proving this methodology was now successfully competing with traditional (human) practices.

What quickly became apparent was that the HR and recruitment crowd partaking in the conversation were very uncomfortable with the idea of any sort of replacement but especially by robots. A follow-up review of the #NZLEAD Recap Recruitment Processes ignored any discussion on automated process and concentrated on human inputs only.

It’s not just the recruitment process that is susceptible to an augmentation and automation overhaul. Many components of the Human Resources role could and can be downsized via augmentation or replaced by automation. It might even be argued that after automating most of the payroll function away in the 1980’s that HR itself has reached its next ‘peak job’ phase as its functions get outsourced or further automated.

So here is my ‘Good Read Guide’ on the subject of HR automation in recent times. Got one you think I have missed? Shoot me a comment with a link as I’d love to include more HR automation stories.

HR Automation – Good Read Guide

Laurie Ruettimann: Cold Reading: Sylvia Browne, Amanda Berry & Recruiters

I thought I would get kicked off with an article that sums up the topic without realms of detail and given that it’s written by the Cynical Girl it’s also a very punchy start to my reading guide. Laurie suggests that the methodology of recruitment is little better than an “unsophisticated psychic trick”: and “that technology can solve for bias and discrimination in the hiring process”.

Naomi Bloom: HRM Analytics – Dashboards, Cockpits And Mission Control

1 - NaomiBloom_1992

One item that keeps coming up in my ongoing conversation with HR is that I believe all things can be measured (but not all things should be as you should look for value against effort). There is always lots of discussion about this in the HR space. Naomi Bloom believes that all things HR should be measured. In an earlier 2009 piece she stated:

“If the real purpose, the only purpose, of HRM is to achieve organizational outcomes, then we’d better be able to measure the effects of specific investments in HRM on those organizational outcomes. Otherwise, why would anyone trust us with a budget?”

I reached out via Twitter to Naomi Bloom, given that she has spanned the entire modern HR journey between old and new (the picture is a copy of her 1992 opus on the subject which she kindly sent me). She suggested the above recent analytics article as a primer. It’s worth a read given that analytics is a key augmentation step and who does robots better than NASA!

The Ladders: Keeping an eye on recruiting behavior

Here is a resume service provider using eye tracking technology to highlight where recruiters spend their “four to five minutes per resume”. Don’t think resume writing or reviewing can be automated…. Think about it as a series of transactions and then ask yourself, can each of these transactions become automated?

Fiona Smith (via the Australian Financial Review): Driven by data: moneyball recruitment takes away all the guesswork

On the subject of recruitment Carol Howard suggested this piece by Fiona Smith on using data and analytics to take the guess work out of the hiring process. The case study utilised is Sears Holdings Corporation which put all of its applicant data against its employee data and found that their “best employees did not come from their previous talent pool”. If robots aren’t in the process of taking over the job of recruiters, big data is certainly going to assist in the downsizing of that role.

John Sumser: The perils of Automation

Before I leave you with links to a HR future that might not need (much less) humans in it I came across this thoughtful 2012 piece by John Sumser. Very wisely John suggests that “Automation strips the fuzzy stuff out of relationships to turn them into transactions. In that process, things get much more efficient. It’s less clear that we understand what we’re leaving behind.

David Creelman (via HRVoice.org): Unending Automation

Maybe the role of HR won’t be in looking after your current employee’s but assisting those who are technologically displaced prior to their own exit. David Creelman suggests:

“Many countries do not require organizations to protect workers from technological change. If self-driving vehicles can replace your truckers then perhaps you can just send them a note wishing them luck finding another job. However, ethically we have a responsibility to at least inform workers about their longer-term prospects and preferably find ways to help. Ways to help could include early retirement, job-sharing, or retraining. HR should explore all those options.”

Steve Boese: Virtual HR, or, ‘Did you ask the HR chatbot’

2 - Ivy

My final link (and my favourite) is from Steve Boese, not only a HR technology professional but also someone with a keen interest in how technology is transforming work. In this blog Steve looks at Intel’s incorporation of Ivy, the virtual HR agent who at the time of publishing could respond in 4,331 ways to staff interactions. On the subject of HR automation Steve states:

“Most of us, (admittedly me too), say of think things like ‘My job is just too complex and ever-changing for it to even be outsourced to a less-expensive human (much less a robot).’

The criticism of this potential HR future was summed up nicely by David Gordon, a recruiter, who replied to my original tweet/link with “@gmggranger it is a good read – still not fit for purpose for recruitment (yet?!), Ivy answers factual questions, recruitment is subjective”.

I’ll agree with you David (at the moment). Yet, a journey starts with a single step…

Final Thoughts

Human Resources are being asked to assist in the transition of human workforces to augmented or automated workplaces. From automated trucks in the mines, DIY checkouts at the supermarket or robotics augmenting people on the factory floor every industry is under increasing competitive pressure.

Yet HR itself seems totally against a conversation about HR automation.

I get it. You’re a knowledge worker and the things that you do for the organisation are just too complex to be replaced by SkyNet.

But maybe its time for HR to review this thinking. As Steve Boese states:

But it also seems likely that given enough time, access to ever-improving technologies, and the right economic incentives, there are enterprising people and organizations that even if they couldn’t completely automate or robot-icize everything you do, chances are a fair amount of even what we creative types do is already routine enough that the robots could do a passable, if not better (and cheaper and will less of a bad attitude), than we do.”

The kind of hollowing out of HR, last seen when payroll was automated from the 1980’s is already starting to impact on the HR function and recruitment seems to be the current automation focal point.

Better start getting involved in the conversation people!

It’s going to happen with or without you.

 

Acknowledgements: In an effort to shine some light on this subject I started tweeting HR automation stories from various writers. My twitter-sphere colleague Michael Carty of XpertHR suggested it might be good to compile these into a single resource. Great idea Michael and I hope you like the post!

Special Note: For those who have not read any of my previous articles on peak jobs and need a little background. ‘Peak Jobs’ is the idea that technology is replacing jobs faster than it’s creating them. For those more technically inclined it can also be attributed to the finalisation of the increased growth in average output (and income) per labour unit due to technological change since the 1820’s as put forward by Robert Solow (1956) or the commencement of technological unemployment as put forward by John Maynard Keynes (in the 1930’s) without the opportunity to transition into new roles as productivity increases but global employment declines.

Random Analytics: Ebola! (2013)

***** Please note that the infographics/charts of Ebola were updated with public source information to 12 July 2013 EST *****

I continue to remain morbidly fascinated by Ebola.

The World Health Organisation (WHO) states that EHF is a severe acute viral illness often characterized by the sudden onset of fever, intense weakness, muscle pain, headache and sore throat. This is followed by vomiting, diarrhoea, rash, impaired kidney and liver function, and in some cases, both internal and external bleeding. Laboratory findings show low counts of white blood cells and platelets as well as elevated liver enzymes. People are infectious as long as their blood and secretions contain the virus. Ebola virus was isolated from seminal fluid up to the 61st day after the onset of illness in a laboratory acquired case. The incubation period (interval from infection to onset of symptoms) varies between 2 to 21 days. During EHF outbreaks, the case-fatality rate for the three fatal strains has varied from outbreak to outbreak between 24.8% and 89.5%.

Key Facts (issued August 2012 via the WHO):

  • The Ebola virus causes severe viral haemorrhagic fever (VHF) outbreaks in humans.
  • Viral haemorrhagic fever outbreaks have a case fatality rate of up to 90%.
  • Ebola haemorrhagic fever outbreaks occur primarily in remote villages in Central and West Africa, near tropical rainforests.
  • The virus is transmitted to people from wild animals and spreads in the human population through human-to-human transmission.
  • Fruit bats of the Pteropodidae family are considered to be the natural host of the Ebola virus.
  • There is no treatment or vaccine available for either people or animals.

There were three outbreaks of EHF in 2012, two occurring in August and a later outbreak in November.

The first two were almost concurrent outbreaks, one of Ebola Sudan in Uganda and another of Ebola Bundibugyo in the Democratic Republic of Congo. The Uganda outbreak, in the Kibaale District would infect 24 and kill 17 including 12 members of the same family. The DRC outbreak, in the Province Orientale would eventually infect 77 and kill 36.

Barely a month after calling and end to Kibaale outbreak (and coincidentally an outbreak of Marburg) the Ugandan Ministry of Health had to deal with a second Ebola Sudan occurrence. The 2nd Ugandan outbreak, this time in the Luweero and Kampala Districts would infect seven and kill four.

As recently as late May 2013 it was thought that another outbreak had occurred in the DRC. Although there was an occurrence of some disease testing found that it was not EHF as noted by the CIDRAP News scan on the 4th June 2013.

There has also been some discussion about the possible progress in the fight against Ebola via an FDA research program in late June 2013 (a good abstract on this via H5N1) I thought it might be time to review one of my first WordPress articles, a piece on Ebola completed in October 2012.

Here is some spruced up analysis and analytics, including some updated data from the 2012 outbreaks in Uganda and the Democratic Republic of Congo.

Infographic

The following infographic looks at some of the key dates and incidents in the history of the Ebola Haemorrhagic Fever (EHF). If you also want an idea of what the virus looks like I have included a photo of the disease under an electron micrograph (top right hand corner!).

01 - Ebola_Infographic_130713

Cases by Country and EHF strain

This infographic packs in a lot of information. First of all the circle sizes represent the amount of cases per country, split according to colour representing the four known strains of the disease. The known strains found in Africa are:

  • EBOV: Ebola Zaire (or the Ebola Virus). Colour: Orange;
  • SUDV: Ebola Sudan (or the Sudan Virus). Colour Green;
  • BDBV: Ebola Bundibugyo (or the Bundibugyo Virus). Colour Violet;
  • TAFV: Ebola Ivory Coast (or the Tai Forest Virus). Colour Brown.

Each country has additional details including total number of cases, fatalities and Case Fatality Rate (CFR).

It should be noted that I have not included a number of known laboratory incidents which have occurred outside of Africa and have resulted in at least one death (Koltsovo, Russia). I also have not included any details about the Ebola Reston which was exported in Philippine crab-eating macaque monkeys to the United States and Italy from 1989 through to 1992. Ebola Reston is not known to be dangerous to humans.

Note: I have created this infographic using Tableau Public software which can be viewed here.

02 - Ebola_CasesbyCountry_130713

CFR by Year and Strain

The final infographic looks at the Case Fatality Rates.

Each line represents a known occurrence of Ebola by onset year. If the outbreak has not completed within a calendar year the data is included from the date of the first onset. Thus the Dec 2008 – Feb 2009 Kasai-Occidental Ebola Zaire outbreak in the Democratic Republic of Congo which infected 32 and killed 15 is included as the 2008 EBOV line.

03 - Ebola_CFRbyTypeOnsetYear_130713

Final Thoughts

Although there have been no known outbreaks of EHF in 2013 it remains a highly visible reminder of our inability to completely dominate our natural space. Many other diseases infect, damage and kill more people but Ebola seems to have a special place in our high-speed mythology. No doubt this is due to its awful characteristics and super high case fatality rates. Amplification by movies such as Outbreak (made in 1995 but still watchable today) or by the writings of Tom Clancy and others who consistently turn the disease into a weapon adds to its macabre allure.

We are just a month away from August, the onset timings for both the opening 2012 outbreaks.

Will 2013 prove to be another deadly year?

Acknowledgements: The new and improved analysis could not have happened without Crawford Kilian and his H5N1 blog and the writings of Tara C. Smith (including her early work via Stanford University) but more recently her articles for National Geographic. Each disease I look at brings out a couple of new, highly specialised subject matter experts and Tara’s writings are excellent.

Random Analytics: Mining Workforce Planning Scan (Jun 2013)

The Mining Workforce Planning Scan is a quasi-quantitative report card built from relevant online industry magazines and media sources. Utilising 14 category metrics the scan collates relevant stories over a calendar month period to give a picture of how the industry is positioned from a workforce planning perspective.

Body Count

After last month’s poor Employment numbers which dipped below -10 negative sentiment for the first time since the commodities crash of August through to October last year I stated that if this route were to continue then I would be very concerned, especially if the numbers being cut were more than just peripheral cost cutting.

The numbers were bad all month but the last week of Financial Year 2012/2013 was especially brutal for coal which reported a loss of more than 1000-jobs in a single week. Gold was also very negative with more jobs and reviews being shed globally.

I’m also not convinced about the big operational employment numbers forecast to come out of future mega-mines such as Kevin’s Corner (QLD) or refining capacity like the Curtis Island facility.

Looking at a recent example of refining from Portugal José Manuel Fernandes stated:

Last April, GALP [a Portuguese energy conglomerate] inaugurated a renovated refinery in Sines. At €1.4bn, it is the biggest industrial investment in the history of Portugal. It will have a huge impact on our balance of payments, because we will export diesel fuel. All of this is excellent, except when it comes to the impact in terms of jobs. Only a hundred people will benefit. That is next to nothing, and it is an example reveals the dilemma of modern economies. Huge investments, including investments in heavy industry, are capable of having a significant impact on competitivity and on the trade balance, but they create very few jobs. Sometimes they even reduce the number of employees. What is true of GALP is also true for most of the industrial sectors in Portugal, as well as for the rest of Europe.

It’s been a tough month for mining and I’m not convinced we have seen an end to the bad news.

Categories

Employment was the leading category with 38-stories (41.3%), the fifth month in a row and a record high over 18-months of data analysis. The next highest Employment tally was September 2012 with 36-stories (39.6%).

Like April, WH&S (Work Health & Safety) was the second leading category with 14-stories (15.2%) and IR (Industrial Relations) and FIFO/DIDO were equal third with 9-stories (9.8%).

If you want to get a feel for where mining is going there have been no stories recorded for SkillsShort (Skills Shortages) in June and only one article, thus far, in 2013. For context during the period Jan-Jun 2012 there were six SkillsShort stories.

What this is telling me is that mining is cutting employee’s quickly enough that new ventures have enough candidates to fill most of their hard-to-fill and critical roles and the operational critical roles (generally only around 5% of a workforce) are holding onto positions rather than risk a move.

2 - Mining_Categories_Jun2013

Positive/Negative Index

From a +4 in March Employment has been on a steady reversal in terms of sentiment. Returning a -4 in April (-8 points) it has continued its rapid decline with -11 recorded in May (-7 points) and most recently a -16 (-5 points). Only September 2012 was more negative when a -20 was recorded.

On the positive side both Engagement & FIFO/DIDO recorded the monthly high of +3. Engagement usually tracks pretty well but FIFO/DIDO as the best indicator for the month comes as a surprise. Looking at the detail there was only one negative story which looked at FIFO mental health and four positives. Two of the positives looked at changing the FIFO/DIDO workforce to better suit local conditions, one was a response to the negative press FIFO has received and another was a FIFO support consultancy which is owned and operated by a miner’s spouse.

3 - Mining_PosNegIndex_Jun2013

Mining Employment Gains & Losses

May was the second month that saw 2013 new employment numbers fewer than 1,000, although there were some employment projections as far out as 2020.

4 - Mining_Employment_Jun2013

Here’s a look at the June data.

5 - Mining_Data_Jun2013

Story of the Month

FIFO is a tough business, especially for families so it was nice to see the story on Anna Rushton (I’ve linked the original story via The West Australian) who has started her own little consultancy FIFO Success. As a mum and wife to a FIFO miner she obviously could see a business opportunity!

Final Thoughts

If we have another set of -10 or worse numbers for Employment sentiment in July we are then in a period similar to the commodities crash from August last year.

Given that production numbers are starting to ramp up (especially as coal and iron ore capacity comes on line), there is softening demand (especially from China) and declining commodity prices I wouldn’t be surprised if we continue to see more bad news for mining and miners.

 

Note: My previous Mining Workforce Planning Scan can be found at Random Analytics: Mining Workforce Planning Scan (May 2013).

Updates (10/07/2013)