Sunday, 13 February 2011

Colonel Mahoney, in Porton Down, with the flawed data

"There are known knowns; there are things we know we know.
We also know there are known unknowns; that is to say we know there are some things we do not know.
But there are also unknown unknowns - the ones we don't know we don't know."

We know there may exist unknown unknowns but, in the instances of the 7 July Inquests, should unknown unknowns form the basis of an open, just and transparent inquest process?

One of the most shocking revelations from these proceedings has been the fact that no internal post-mortems to identify the actual cause of death of the victims were carried out at the Resilience Mortuary set up in the grounds of the Honourable Artillery Company.

This shock was expressed by one of the first doctors on the scene from the BMA, Dr. Awani Choudhary, when giving evidence about his attempts to save the life of Gladys Wundowa:
Q. You've described your concerns about a possible neck injury. Is that what you meant by feeling that she was in immediate danger or did you mean something more by that?
A. More, I thought -- you see, I have felt her pulse, and by feeling the pulse you can have many, many information, and I thought that she was bleeding from somewhere, I have not seen the post-mortem report, but I thought that she was bleeding from somewhere, and I suspected that she is probably bleeding from either in her chest or abdomen, because her pulse was going very fast, and it was becoming weaker and weaker and, within 10 minutes, you could not feel her pulse near the wrist, but you could feel it just about near the elbow. So if the post-mortem says that she was not bleeding from anywhere, just had a spinal injury, I will be surprised.
Q. Can I ask you this: you've just mentioned bleeding. Do I take it from what you say that you mean internal bleeding?
A. Internal bleeding, yes.
Q. Since you ask about the post-mortem, can I simply inform you that, as with all the other casualties of the day, no internal post-mortem was conducted into Gladys Wundowa, so unfortunately, much as we would like the answers to the questions that you've asked, they don't --
A. I don't want to go through the medical too much. There are two types of shock a patient can have. What we call it, neurogenic, and the other, we call it hypovolemic. In the neurogenic pulse -- you see, the pulse will be pretty well, but the blood pressure will be low, you see, and in the hypovolemic shock, because the blood had been lost, it will be other way round, and I'm absolutely sure that she had had internal injury as well as a spinal injury, and I'm absolutely surprised that a post-mortem has not been done through and through.
Q. Well, Mr Choudhary, that isn't a matter to concern you.
A. Sorry.
Q. I was simply informing you so that we didn't chase any red herrings, but we don't need to concern ourselves about that matter.

Before continuing, let us be perfectly clear about the exact role of a Coroner and the purpose of Inquests; it is to establish four things:
  • The identity of the deceased
  • When they died
  • Where they died
  • How they died
Back in April 2010 when the issues of the resumption and scope of the Inquests were being heard, Mr Patterson, counsel for some of the bereaved families, made it clear that the lack of internal post-mortems to establish cause of death had only just become known to them. The following lengthy section from the transcripts makes clear that this issue, along with the huge delay in identifying the deceased, was deemed by the families to be -- not unsurprisingly -- absolutely crucial to the scope of the Inquests:
As for this need for resumption, madam, the post-mortem examinations also arise as an area of concern for the families. In the case of Mrs Mozakka, for example, no post-mortem examination was held until 13 July, some six days after the bombs, and yes, of course the loss of life that was being addressed here was considerable, was unprecedented, but the question that arises is: could these post-mortems not have been carried out perhaps more quickly than they were? Are there lessons to be learnt? In particular, as to the post-mortems, we have now discovered from the scene reports that there were no internal examinations and, again, a question that arises is: why was the decision taken that there would be no internal examinations? Although the cause of death was clearly the explosions, the precise mechanism of death was not explored, and so that makes the question of survivability all the more difficult for us now and for the families when asking whether or not their loved one might have survived.
LADY JUSTICE HALLETT: If that decision was taken by a judicial officer, how do you say an inquest is going to explore that?
MR PATTERSON: It may be that your predecessor made that decision. I simply don't know. Doubtless, there were very good reasons, and it was carefully considered, but the fact remains that in looking now at survivability, we have to, for instance, in this case, assume that there were perhaps no internal injuries and that, therefore, sadly, this might be a case where this person was dealing with loss of blood and that Mrs Mozakka, for example, if she hadn't had any internal injuries that would have been insuperable, could have survived if the response had got to her quickly enough. We've read about tourniquets being applied and the like, and how often those very seriously injured did survive.
LADY JUSTICE HALLETT: I understand that argument. My question was how -- you said that one of the matters that you might wish to explore is why the decision was taken to have no internal investigation, and my question really was, how do you explore that aspect? It's one thing to explore what we do or don't know or what might have happened, given what we do or don't know, but it is a fact that no invasive post-mortems were held. If I am right in thinking that that decision was taken by a judicial officer, I just don't see how you end up exploring it unless you're going to ask to cross-examine him.
MR PATTERSON: I don't know what reasons were recorded at the time, whether they have been recorded somewhere.
LADY JUSTICE HALLETT: As far as whether or not -- as I see it, the aftermath can be divided into two categories. One is the response of the emergency services and whether or not that has impacted on people's chances of surviving the explosion. There's another aspect which you've just mentioned, which is: could the post-mortems have been carried out more quickly, are there lessons to be learned? Do you say -- which is almost a follow-on from the immediate aftermath, which is, what happens? Were the families notified quickly enough? Were the post-mortems carried out quickly enough? Could you help me on how you say those aspects impact or can be legitimately covered within the context of an inquest into the circumstances of the death as opposed to the circumstances of the investigation thereafter and the telling the families?
MR PATTERSON: I agree that there is a limit in the scope of the inquest, and the statute makes that clear, and the Coroners Rules make that clear, and clearly you have a discretion. All I can say in relation to those aftermath issues, where there is a legitimate argument to say that a line has to be drawn, and that they shouldn't be explored -- and, madam, we are realistic about this, and we know that, for instance, issues as to the way in which the family liaison officers communicated, it may be that some of those issues will be deemed to be beyond the scope of your inquest. We recognise that. It's a discretionary matter for you, and it may be that a clear, bright line can be drawn in relation to some of these issues by focusing on certain issues but not others, so that the recovery and the identification issues can be covered, the issues of the mortuary and the post-mortems can be covered, but that thereafter there may be a line that has to be drawn, and I recognise that and we are alive to that. All we would say is that this is a very good opportunity -- if you do resume the inquest, this is a very good opportunity to deal with those issues, if you take the view that they can be concisely and easily dealt with in the scope and in the course of your inquest.
LADY JUSTICE HALLETT: Even if I have a discretion, Mr Patterson, I'm bound to exercise that discretion in accordance with the law --
LADY JUSTICE HALLETT: -- and at some stage I would welcome greater assistance on -- you say that it might be a good idea to get these matters disposed of, but this is not a public inquiry; it is an inquest, if it's resumed.
MR PATTERSON: Absolutely, and we recognise that. Certainly, in the hours and days that followed the explosions, there were concerns that the families had as to the whereabouts of their loved ones and telephone calls that were made, enquiries that were made, didn't get the answers that it was hoped they would get and, for example, many of the families that I represent have concerns that, for instance, clear identification of the loved ones was found at the scene. In one case there was medication found with the name of the injured person, in another case there were identification cards found on the body. Yet, for many days, they were anxious and worried and telephoning hospitals and visiting hospitals enquiring into whether or not their loved one might still be alive.

Despite the pleas of Patterson and the pleas of other counsels acting on behalf of the families, the issues of identification and recovery of the deceased were placed outside of the scope of the inquests. Instead, Mr Hill for the MPS offered to discuss with the bereaved how their loved ones were identified, but only outside of the official public inquest proceedings.

Why the Coroners in charge of the Resilience Mortuary, and who opened the Inquests into the 56 deaths, failed or actively chose not to carry out full autopsies to establish actual causes of death, which is after all the function of coroners, has not been revealed nor examined in any way. How can this be?

Whether some of the victims may have had injuries that with prompt hospital treatment were survivable - as perhaps in the case of Gladys Wundowa, who Choudhary pronounced dead in the courtyard of the BMA at 11.10, some 80 minutes after the explosion on the bus whilst University College Hospital was within walking distance - will likely never be known.

The setting up of a Resilience Mortuary was part of the 'Mass Fatality Plan' - drawn up after the events of 9/11 - and contracted to De Boer just the day before these events. The deaths of 56 people - 3 of whom had died in hospital - spread over 3 coroner's districts, hardly amounted to the significantly greater number of '500 or more fatalities' as envisaged and anticipated by the plan:

In drawing up the Mass Fatalities document, London Resilience needed to define clearly which organisation was responsible for which role. It was vital no gaps were left if a seamless plan to deal with 500 or more fatalities was to be developed. The author needed to source everything from body bags and transport to personnel and storage, but it was the need for large mortuary space that proved among the most problematic issues.

It soon became obvious existing mortuaries would not be able to cope with a large scale crisis. We needed blue-sky thinking. We decided temporary structures were the way forward. Quick to erect, modular, portable and highly flexible they became an obvious choice, and De Boer became an obvious supplier. Based in Northamptonshire in the UK, the company had already completed several contracts for the Metropolitan Police. De Boer had proved it could provide a relocatable building almost anywhere in London within a few hours’ notice. De Boer offered a managed solution, working with sub-contractors to create different environments required by different sections of a mortuary like temperature control, security, privacy, power, lighting, water and waste facilities. The company thoroughly understood our requirements and was able to meet exacting specifications.


The De Boer team spent months visiting permanent mortuaries and attending meetings with London Resilience to suggest a suitable structure and interior design to replicate the facilities they had seen. The unit needed to accommodate everything from post-mortem facilities to family areas and from body storage to canteen and offi ce facilities. When the bodies of British residents killed by the Asian Tsunami in 2004 were flown back to the UK, De Boer was commissioned through London Resilience to provide extra space at Fulham mortuary in south west London. It proved the plans in place could work and where they could be improved. Six months later on July 6, 2005, a document arrived at De Boer’s UK headquarters finalising what had been agreed for a future crisis response. Within 24 hours the plan was being realised and implemented with the creation of a temporary mortuary in the grounds of the Honourable Artillery Company near Moorgate Underground Station in central London.

Source: Crisis Response Journal (no longer on line)

Who took the decision to invoke the Mass Fatality Plan and construct a Resilience Mortuary at the cost of millions of pounds on 7th July? Why was the decision taken to implement a plan intended for upwards of 500 dead?

It is conceivable that the Mass Fatality Plan would omit full autopsies on the basis of time and personnel constraints in order to focus instead on identification - probably sensible when dealing with hundreds or thousands of deaths from a presumably known source (flooding, pandemic, etc), but was this necessary for 53 deaths? This important issue should certainly fall within the scope of the Inquests to examine.

Due to the absence of full post-mortems, the Inquests turned to the Ministry of Defence to construct models to show the probable fatal injuries and likely causes of death for those with no obviously fatal external injuries. This required another expensive process and a team was set up under Colonel Mahoney, defence professor of anaesthesia and critical care at the Royal Centre for Defence Medicine in Birmingham, whose previous experience included Camp Bastion in Afghanistan. Mahoney was tasked in August 2010 to lead an investigation into what happened to the 18 victims thought to have lived for some time after the explosions.

Q. The essence of your task, though, lies in recognising that it's possible to compute by reference to proximity to a given device the likely injuries that will result, all other things being equal?
A. It is, all other things being equal. Clearly, a real environment can be very complex, as we will be drawing out in a moment, but you can compute in terms of proximity to an explosion by understanding the size of the explosion, the characteristics of the explosive and the pressures and temperatures that that person was subject to, you can predict the likely injuries. Ideally, as we had with the military casualties, we would have an internal post-mortem to give us the degree of detail that we need to confirm those views, but even in the absence of that, you can make an informed judgment.
Q. That is why your view is to be properly expressed as a judgment, because you didn't have internal post-mortem material here which would have allowed you actually to see the condition of the lungs in each of the deceased?
A. That's absolutely correct. There are some people unfortunately, where you can look at the pictures and the environment and say, "I can feel very, very confident this person would not have survived and, in fact, would have likely died instantly", and there are other circumstances where we had to spend a lot more time reviewing the factors and trying to come to a balanced conclusion, and there there are no absolutes.
Q. It's important we understand, therefore, that you weren't asked to address the question of what was the cause of death in each case, because you were doing something different, which was to assess what was likely to have been the cause of death, because, of course, you had to use modelling and the other devices open to you because of the absence of internal post-mortem material?
A. That's right. Where we felt we had a clear cause of death, a particular injury that was demonstrated in the clinical documents, the external post-mortem report and that tied clearly with our views from the scene and the witness reports, we've highlighted that. In other circumstances, we've just had to come to a conclusion as to what was the likely cause of death, given the blast environment that we have calculated and assessed, but accepting the error bars that exist within that environment.

The many caveats included in the final report of this modelling of the likely causes of death of the victims from a blast, included the crucially important, if slightly bewildering, caveat that there was no idea how the explosives allegedly used actually worked. We will ignore for now the not insignificant logistical issues regarding how it might be possible to allege that explosives were used whose characteristics were unknown:
MR KEITH: Colonel, may I now turn to the methodology that you adopted in this case by particular reference to the appendices and the structure of the individual reports? There are a number of methods, or there were a number of methods, open to you to assist you in reaching your overall view as to the likelihood of survivability, which was a likelihood based on the balance of probabilities. Is that right?
A. That's correct, yes.
Q. Firstly, there is, as we've discussed, past research on the effects of particular explosive devices and their construction?
A. Yes.
Q. Secondly, research on the likely effects on the body and its tissues and organs of the detonation of an explosive device?
A. Yes.
Q. In that regard, is it possible to compute through the use of highly advanced computers and modelling systems something known as blast loading, that is to say the effect on the body and its tissues of the detonation of a device?
A. Yes, you can. It has to be with caveats because we do -- as we did -- have to make assumptions, we had to make assumptions on the explosive output of the device.

The modelling of the blast zone and subsequent blast injuries used what was thought to be the equivalent quantity* of highly-explosive TNT to construct the models. TNT was chosen because the explosive output of the alleged home-made devices -- apparently based on hydrogen peroxide and an organic substance defined as 'piperine' -- was a totally unknown parameter and therefore could not be modelled.

So, assumptions are piled upon presumptions and houses of cards are built on shifting sands. These are the openly-stated unknown unknowns from which the bereaved families are meant to learn how they lost their loved ones. (* How the equivalent quantity of an unknown explosive could be reproduced with TNT when the very thing to which it is meant to be equivalent is declared to be unknown, is anyone's guess.)

The entirely unknown nature of the alleged explosives is not the only caveat contained in the final report produced by Colonel Mahoney's team. Others include:

Q. But your approach must, overall, be read subject to a number of caveats?
A. Yes.
Q. Firstly, as you mentioned, there was no invasive post-mortem in any case.
A. Yes.
Q. Secondly, the X-ray examination was limited, as you've just said, to fluoroscopy?
A. Yes.
Q. Thirdly, although you have photographic evidence, in some cases the photographs were difficult to interpret, for reasons I won't explore with you?
A. Yes.
Q. Fourthly, although there was some objective evidence as to the location of some of the deceased, there is a distinct lack of clarity as to where and how they were moved by the emergency responders?
A. Yes, and certainly in a crowded environment such as King's Cross there are -- some of the deceased, you can't be certain where they were within that dense crowd. [ibid p4:5 -25]
This point was driven home by Mr Patterson in questioning of the Colonel on the second day of his testimony:
I think it's clear from what you said yesterday, Colonel, that all these hundreds and thousands of hours that you and your team spent analysing these various issues, the work that you had to put in would have been a lot easier if you'd had the benefit of internal examinations. Is that right?
A. Yes, if we look at our military casualties, they have internal post-mortems and the majority of them have a post-mortem CT scan. So they have a complete, whole-body CT scan which looks for fragments, makes sure that there's no retained ordnance, but also means that you have a clear record of internal injuries.
Q. So, for example, if an expert like yourself or a family want to know about the internal injuries and whether, for instance, there was this leathering effect in the lungs that you spoke of yesterday, an internal examination might answer that, X-rays might answer that. Is that right?
A. We rely on a combination of both.
Q. We are denied both in this case, is that the position? We don't have either?
A. Well, the only information that we have evidence of is an external examination and the fluoroscopic examination which comments on fragments but does not comment on internal injury.

Just how much more flawed data was input into this modelling process? After all, the modelling only became necessary because of the decision taken not to conduct full autopsies. One other extremely important factor to consider: the positions of the survivors on the trains -- no modelling was undertaken for bus victims -- which would be vital information to compute how injuries were dispersed within any blast zones. This lack of evidence required that the final report placed Phil Beer closer to the seat of the explosion than he was actually positioned on the Piccadilly Line train:
Now, to answer that categorically, what we would really need to see was the injury patterns in all the surrounding people and relate the injury patterns to that individual to those other injury patterns, which has not been part of our original instruction. [ibid p25:12]

A. If we accept our reasoning that Mr Beer was close to the seat of the explosion, based on the injuries that we've described and our initial reasoning that we think he had a high blast lung -- a high likelihood of blast lung because of the overpressure, that would push you more towards injury being -- survival being less likely. If you think, or there's evidence, that Mr Beer had a different chest injury, that would make survival potentially more likely. But on the evidence that we've got our -- my interpretation and the interpretation of my team would place him closer to the seat of the explosion than was indicated on the map. But to answer it -- to answer that appropriately, we'd really need to see an injury map of all the injuries around him or where the position you believe he may have been, and only then, by mapping all the injuries and the outcomes of people, can you say, yes or no, that's how the explosive products propagated. I can't be more precise than that. [ibid p26:2]
Ms Gallagher explores further in a testy session with the Colonel:
Q. Well, Colonel, it sounds as if we agree, because what I was going to put to you next simply was this: that despite the incredible and very wide expertise on the part of your team, the task that you were engaged in is necessarily an imprecise science, particularly given the imprecision and uncertainties in relation to the raw data with which you were working. So you could clearly, with precision, identify certain matters, but generally, you're working with a very wide number of variables, very complex, closed environments and where, as you've said yesterday, and again today, if you shift one of those variables, you can have a very different outcome?
A. No question at all.

She then points to the caveats included in this report from Mr Hepper, the blast engineer on the team, on the extreme limitations of the modelling methodology absent of device and detonation details:
Q. So really what we're engaged in is a very nuanced discussion around your conclusions, which are based on this material and these variables, Colonel. In addition to the raw data received and relied upon, could I just also have on screen [INQ10552-17]?
This is appendix A, so it's Mr Hepper, your colleague, Mr Hepper.
A. Yes.
Q. It's paragraph A4.3.6. This relates to the use of computational modelling to simulate the blast environment, and all I was going to say, Colonel, was that in itself carries its own caveats, which Mr Hepper refers to here. So here he refers to the work of Stuhmiller, and he says he "highlights that the possibility to validate these models is limited", and over the page, on page 18 [INQ10552-18]: "... although the technology has progressed ... the confidence and fidelity of these models is still limited."
A. No question about it. [ibid p44:1]

None of the reports on blast lung explain the high incidence of electrocution experienced by so many passengers in the carriages, the subject of a future blog article. Nor do the reports cover how some survivors who were very close to the centre of the 'blast overpressure' survived without this injury (see passengers 56 to 60 in this bus diagram and Greg Shannon & William Walsh on the Aldgate train).

An expensive modelling process based on so many unknowns. Unknown explosives, unknown internal injuries, unknown survivor injuries, unknown X-rays. An expensive mortuary facility that failed to provide the basic information that the bereaved required, along with an expensive Inquest process that is failing to provide real answers to the key questions that it is the core function of an inquest to answer.
"If you want a more definitive answer on the injury patterns of closed-space bombings on a London train, then your task would be: look at all the deceased, look at all the living and look at all the injury patterns, and from that you'd have a more -- you'd have a greater fidelity of your model. But that wasn't the task."

- Colonel Peter Mahoney


  1. Mahoney was instructed in August 2010 and the report produced by October 2010.

    Seems like a bit of a rush-job given that the Inquests-proper also started in October 2010.

    As no autopsies were performed on any of the deceased, and that it was over five years before Mahoney produced his report, it would be interesting to know what was written on the death certificates.

    And, without the virtually valueless apparent afterthought that was Mahoney's last-minute report on unknown unknowns, and with no autopsy data available, what did the State anticipate presenting to the Inquests?

  2. And, without the virtually valueless apparent afterthought that was Mahoney's last-minute report on unknown unknowns, and with no autopsy data available, what did the State anticipate presenting to the Inquests?

    Six key clues to attacks
    The Observer, Sunday 10 July 2005


    London has 6,000 CCTV cameras. Studying the footage will provide crucial clues for intelligence experts. Already initial images of the four bombings have been fed to experts at Qinetiq, the government's former research agency, which will create computer simulations showing where the bombs were planted and direction of the blasts.

  3. ...continued

    What can computer models tell us about the terror attacks on London?
    Alok Jha
    The Guardian, Thursday 14 July 2005 12.21 BST

    They will be able to pinpoint exactly where the bombs were placed and how much explosive was used in each attack.

    The government's Defence Science and Technology Laboratory (DSTL) has been given the task of gathering information on the bombs, which killed more than 50 people and injured hundreds. The DSTL's forensic explosives lab at Fort Halstead will head the investigation to identify and explain what happened during the explosions.

    Understandably, the investigators are keeping quiet on the exact details of their work, but explosives experts have suggested that 3D models of the blasts could be used to better understand the dynamics of the explosion.

    "We could tell them where the device was placed, we'd be able to tell them how much explosive was used," says Neil Fisher, director of security solutions at Qinetiq, the former government defence research lab. "The forensic chemists will be able to tell which type of explosive was used. We understand the science of those explosives, the chemistry, so we can analyse how much explosive was used to cause that much damage."

    The computer models are created by examining the blast scene and recording the position of every piece of debris. "When you're capturing the real world, you're doing it with a laser and a camera, so the equivalent we use takes 6m data points on each take," says Fisher.

    "You get a very accurate facsimile of the real world which you can then stitch together to make your simulation."

    Once the simulation has been created, specific programmes are then run to watch the effects of different types and amounts of explosive.

    "We would try to match those simulated effects with what's happened in the real world," says Fisher. "Sometimes you'll get blast but not a lot of heat. Others, like Semtex - you can make a lot of heat out of them as well as a blast wave."

    The modelling is done in two parts. After the simulated blast is compared with the actual event, the information is validated by recreating the explosion for real. "Qinetiq has a large number of ranges and test sites where we are actually licensed to go and do that," says Fisher.

    Computer models of explosions have a good pedigree in solving terrorist bombing mysteries. The explosion on board Pan Am flight 103, in 1988, was modelled by precursors to the program used by Qinetiq. Investigators showed that it would take only a small amount of plastic explosive - less than half a kilogram in the case of flight 103 - to bring down a Boeing 747.

  4. Six key clues to attacks
    The Observer, Sunday 10 July 2005

    ... Already initial images of the four bombings have been fed to experts at Qinetiq, the government's former research agency, which will create computer simulations showing where the bombs were planted and direction of the blasts.

    Why the need for Mahoney's efforts five years later?

    Where are Qinetiq's findings from the data they collected/were fed in the immediate aftermath of the incidents?

  5. A quick glance shows that TATP has an explosive velocity of 5300 m/s whereas TNT has an explosive velocity of 6900 m/s. The range for high explosives is generally between 4000 and 10000 metres per second, so a difference of 1600 m/s would presumably affect the results considerably. Castles made of sand fall in the sea, eventually.