Dr John Crippen
"That a mother who has had two of her children die of cot death should then be unjustly convicted of killing them defies belief, and yet that is what happened to solicitor, Sally Clark. She was sentenced to two life sentences for killing her children."
I didn't that anyone thought that Sally Clark's children had died of cot death.
There was some evidence of bacteria in CSF, suggesting an infectious cause of death. Reading through the post-mortems (which are freely available at her own website), both children also had evidence of traumatic injuries, which would be highly unusual for such young babies.
It was Roy Meadow's dogmatic refusal to consider any other factor than his own (deeply flawed) statistical theory -- documented over the last few years (and in several cases) by Private Eye -- that should have led to him being struck off.
His stubborn refusal to contemplate that his pet theory could be wrong (and on the kind of medical theory that even I could have exploded through elementary genetics) led to a number of innocent people being imprisoned.
And if a doctor could not be bothered to understand the science properly, how are a jury supposed to do so?
DK
Absolutely, DK. The BBC quotes Meadows as saying "This is an important decision for paediatricians and all doctors, nurses, teachers and other professionals who may have to express difficult and sometimes unpopular opinions in the course of giving evidence in court. They should be able to do so without the fear of prosecution by the GMC or other professional regulators."
Difficult and unpopular is irrelevant. His opinion was based on professional knowledge and understanding which he ought to have had, if he was holding himself out as an expert, but did not; and was not willing to learn when given the opportunity. He was struck off for ignorance, arrogance and pig-headedness. Or rather his arrogance led to the serious professional misconduct of refusing to learn or admit you might be wrong, which led to him being struck off.
At the original trial the defence should have questioned his ability to make claims about probabilities, after all, he is not a statistician and I assume didn't have a lot of epidemiology training. Whatever happens Roy Meadows will never be an expert witness again.
Easy to preach on some blog board that he made a mistake so he should be struck off. Being struck off is not the only action the GMC could have taken, they had the options of less severe possibly more appropriate penalties (fixed-term suspension, suspension until retraining, etc) but chose to strike Prof Meadows off the register completely. Why the severity? To 'maintain public confidence in medical self-regulation' (in the GMC's own words). Justice is not served when the degree of punishment is meted out by the court of public opinion.
I know of a number of doctors no longer willing to be expert witnesses and the courses for training expert witnesses are now undersubscribed. This is even for cases which have nothing to do with child protection.
This goes beyond expert witnesses as well. If you are going to hold doctors to never mistakes in everything they do, you are going to end up being dissatisfied with a very expensive healthcare system.
The question of how juries should deal with expert evidence is an interesting one. Experts deal in probabilities and juries deal in sure/not sure. It's a real square peg/round hole situation. And it is very rare that the way in which probability works is properly explained to a jury. They might be told that the chances of two people having the same DNA are, say, five million to one, which sounds astronomical, but often can translate to several people theoretically having identical DNA. Which means that actually in a case where there is no evidence other than DNA there is potentially tons of reasonable doubt - but you don't necessarily have a juror who can do the maths...
Incidentally, Collins J (High Court, not Court of Appeal, John - I expected better of you) is a bit of a softy and I suspect his ruling was motivated more by the fact that Meadows is a bit long in the tooth and he didn't want to trash him at the end of his career.
Kathy
Yes, apologies for that. A slip. Thanks for pointing it out.
John
Neurology trainee at Queen Square
Huf!
This isn't "some blog board" - the is NHS BLOG DOCTOR!!
Huf huf!
John
And you are a lawyer and everything!
Mind you, it was an appeal. I am such a pedant. I should stop.
None of the expert witnesses, or the original pathologist involved in the Clarke cases considered the deaths to be due to SIDS.
We now use the term SUDI in preference (sudden and unexpected death in infancy). SIDS is an unhelpful term, implying a single disease process or syndrome. The deaths in 'cot death' babies are likely to be due to a multiplicity of causes-infections, variability in heart rate/temperature control, cardiac problems and so on, many of which can't be identified at autopsy with any degree of certainty.
And it is inaccurate to say that SUDI or SIDS can happen to any baby anywhere-there is a much higher risk in certain groups, particularly in babies whose parents smoke, single parents with low income, premature infants and so on.
I agree with the comments that Prof Meadow was stepping outside his area of expertise, but the defence barrister should have asked him if he was an expert in statistics. The defence team didn't do their job thoroughly, or properly, but have they been castigated by their governing body?
HI Pathologist
do have a look at Phillip Gordon's article that covers this in detail
++++++++
but the defence barrister should have asked him if he was an expert in statistics. The defence team didn't do their job thoroughly, or properly, but have they been castigated by their governing body?
That's Rondell v Worsley as I recall - where the judges (ex barristers) said that barristers could not be sued for incompetence
John
And of course, whilst the barrister profession as a whole has managed to perpetuate its cosy little monopoly, market forces operate in a very real way on individual barristers. Crap barristers don't get briefs. Or, more probably, crap barristers get crap briefs?
And, of course, the true irony is that the statistics were being quoted from a government publication. Even more ironically, there WAS a statistician involved in that publication. And unless I am very much mistaken, the court of appeal did not find the statistical error enough to change the verdict - what concerned them was incomplete disclosure of pathology reports.
Probabilities are routinely screwed up by newspapers, people at large, and even, it would seem, expert witnesses! Take any probability you see (in the news, etc.) with a grain of salt. For example, a newspaper headline (admittedly, not in one of the better newspapers) trumpeting the "1 in 48 million!!" case it has found of a father, mother and son all with the same birthday (365 x 365 x 365 = about 48 million). Is it so hard to see the flaws here? These kinds of wildly inaccurate probabilities appear in public all the time. Public, beware!
Hi John,
over here our medical investigators are also using a term called SUIDs (sudden unexplained infant death syndrome) and these are the ones that do not meet the rigid criteria for SIDs. So sure, I think it is safe to say there are other explanations for sudden infant death. However, I think the data is pretty clear that SIDs is a fairly large swath of the infants who die suddenly in the first year of life, at least here in the states. The back to sleep campaign has demonstrated a farily impressive reduction in death rates based on this definition. It's pretty hard to argue with that. SUIDs has been rising in recent years but not at a rate that would explain the drop.
Can it happen to any baby? Probably yes, because at the heart of it all, these babies smother. It's not that they have a pillow put over their face, it's that they don't have the arousal mechanism in place to wake up and gasp. From my understanding of the data, some children are at greater risk than others (children of smoking, non-breastfeeding, non-pacifier using mothers who co-bed in plush blankets, while drinking alcohol). But it can theoretically happen to any child if enough of these behavioral risks are at play. That's why the back-to-sleep campaign is endorsed by the NICHD, and just about every other society that is actively invovled with SIDs.
Um sorry. But what is the probability of father, mother and son having the same birthday? My maths is crap. Im standing in for my PCT Chief Exec this week and I would appreciate the correct calculation for.........
Sun Reader,
The odds of the Mother, Father, and son all having the same birthday are not 1 in 48,000,000. They are (if we ignore leap year) 1 in 133,225 or 1 in (365*365).
This is because it doesn't matter what the birthday of the first person in the group is.
Similarly, if we have a couple who we know have the same birthday, then the odds that their child will have that same birthday is... 1 in 365.
Now, the odds that the Mother, the Father, and the son all have the same birthday and that it will some specific date (say June 15) is 1 in 48,627,125 or 1 in (365*365*365)
JK
Oi!
Clark's childrens' deaths were NOT cot deaths.
Prof Meadows stats were wrong. But they were irrelevant to the case.
Have a look at the post mortem reports for both children. Lots of things that are difficult to explain. Also look at Sally Clark's history at the time of the deaths. You might find something interesting!
She was actually freed on a technical point. It was never established that "infection" was the cause of death, and that was only one child anyway. Read the transcripts.
I would welcome correction on this point. It thought Roy Meadows, was in part, struck off for hearing about the case on TV and then phoning up to offer his opinion. Thus pressing his opinion(now incorret)instead of being asked for it.
As i said please correct me if this is wrong.
His decision to offer an opinion lead to serious consequences for Sally Clarke. He should be accountable for his actions. But to be struck off.....hmmmm..... By striking him off the GMC might be maintaining public confidence, but what about ours. We don't get much from them for our money.
From what I recall, Meadows got it wrong in several cases, not just one.
Our 1st son was put on his back to sleep as was recommended, until he was about 3 months, by which time he would cry a lot when going to bed. It went on for 2 or 3 days, until my wife put him on his front and he promptly fell asleep.
The next few nights were the most stressful we have ever had, and we felt so bad about it !
He on the other hand, was sleeping very soundly !
The same thing happened with our 2nd, although we did not stress as much !
Dr Sniper
I would welcome correction on this point. It thought Roy Meadows, was in part, struck off for hearing about the case on TV and then phoning up to offer his opinion. Thus pressing his opinion(now incorret)instead of being asked for it.
As i said please correct me if this is wrong.
+++++
No, that was acutally another doc, David Southall, who phoned in his opinion after watching a TV documentary!! A SERIOUS busy body. I think what he did was worse than Roy Meadow.... in the climate at that time, he was lucky to keep his job
http://thescotsman.scotsman.com/...fm?
id=904722004
John
Will people please get the poor mans name right-it's MEADOW, not Meadows.
Dr C,
I've done around 100 autopsies in the last 8 years on 'cot death' babies, and the number of babies that come from typical, middle-class/professional homes with 2 non-smoking parents around I can count on the fingers of one hand. Yes, it does happen, but it is very infrequent in that group.
I see the forensic photographs taken at the scene and some of the conditions which some of these babies are living in is heart-breaking. Melanie Newbould, a paediatric pathologist from Manchester wrote an article recently about the effects of poverty on child health and the risks of cot death: she sees the same sort of cases I do (haven't got the reference to hand).
In other cases, it is the standard of parenting skills or knowledge that is the problem-not realising that you shouldn't be feeding a new born baby with just milky tea, not knowing that you shouldn't go to bed with your baby when you've just had a skinful of beer and vodka, not knowing that a sofa isn't the safest environment to put him to sleep on as he'll get wedged down the back of it.
The Back to Sleep campaign has worked wonders-the infant death rate has fallen dramatically. The cases we are seeing now are those mostly (not all) where there are unexpected natural disease present, or where there are extraneous factors, such as inappropriate sleeping arrangements that may have played a role in the death.
Looking at the statistical comments on this list, I strongly suggest that you all read the latest bad science column. If you hadn't though through the point that Ben Goldacre makes in this post, then maybe you aren't in a position to criticise the statistical errors.
http://www.badscience.net/?p=318
I know that I shouldn't comment further on either the statistics or the medicine.
"The advice has changed from prone, to side, to back even within his short career. " Why did they advice to put child prone in the beginning Any expalnation - like a statistical one?
I see the forensic photographs taken at the scene and some of the conditions which some of these babies are living in is heart-breaking. Melanie Newbould, a paediatric pathologist from Manchester wrote an article recently about the effects of poverty on child health and the risks of cot death: she sees the same sort of cases I do (haven't got the reference to hand).
In other cases, it is the standard of parenting skills or knowledge that is the problem-not realising that you shouldn't be feeding a new born baby with just milky tea, not knowing that you shouldn't go to bed with your baby when you've just had a skinful of beer and vodka, not knowing that a sofa isn't the safest environment to put him to sleep on as he'll get wedged down the back of it.
++++++
Controversial stuff; is there data to back that up?
John
Now, the odds that the Mother, the Father, and the son all have the same birthday and that it will some specific date (say June 15) is 1 in 48,627,125 or 1 in (365*365*365)
+++++
Thank you, JK.
You explained it better than I could have. To put it another way:
Go out on the street and choose 3 people at random. Ask them their birthdays. What is the chance that *all 3* will have birthdays on Christmas Day? This is clearly very unlikely. The probability is 1 in 48,627,125 (ignoring leap day).
Peter, thanks for the link to Ben Goldacre. I read him regularly and there a number of interesting points in the comments. However, my respect for his analytical abilities has just taken an enormous nosedive, when I got to the comment where he opines that the legal profession should be nationalised.
I guess the moral is that professionals, as Dr Crippen never tires of pointing out, should know the limits of their competence and understanding and stick within them. They should do this of their own accord WHETHER OR NOT there is someone else whose job it is to remind them (like the defence team). It is part of being a professional. That's where Roy Meadow failed.
"The advice has changed from prone, to side, to back even within his short career. " Why did they advice to put child prone in the beginning Any expalnation - like a statistical one?
Rdoc |
++++++
I cannot quote you chapter and verse; it was a long time ago, but quite definitely this was the advice we trotted out at the time.
Have a look at: "The Politics of Cot Death" :-
http://aims.org.uk/Journal/
Vol15...icsCotDeath.htm
John
CESDI SUDI studies 1993-1996 'an infant of a young mother who smokes and is on income support is 40 times more likely to die than that of a 35 year, non-smoking mother in a home where she or her partner have a regular income'
The CESDI studies show SIDS babies are more likely (not always) to have parents with lower incomes, lower levels of academic achievement, higher levels of smoking, illicit drug use and alcohol and other contributory factors such as poverty, disorganised household, incorrect feeding and generally poor standards of care (failure to attend baby clinics, GP etc). So yes, there is evidence.
Mine is purely anecdotal-many of the cases of cot death I see are of babies that haven't had a great start in life.
I accept entirely that cot death can happen in apparently entirely normal, perfectly well looked after babies in clean, warm, well kept homes with parents who don't smoke or drink, but in my experience over the years this is becoming exceedingly rare.
And the three examples I gave previously were all past cases of mine-all babies who died unexpectedly but with factors in their care that may have played a role in their death. That is why I refuse to use SIDS as the cause of death for babies such as these and use 'un-ascertained' instead.
In reply to the prone position initially being recommended, I was always told it grew out of the advancement of improved neonatal intensive care in the 1960's and early 70's. It seemed that premature babies breathed a bit easier on their stomachs because this supported the diaphragm (or something?) and premies began to be nursed like that-this kind of spread to the general population leading to an explosion in cot deaths in the 70's and 80's. I have no idea if this is true or not.
But wasn't the advice given in the 40's and 50's was to put babies on their back to sleep? At least according to my Gran.
http://www.users.zetnet.co.uk/ji...5/075(1)
065.pdf
And one last comment-70% of SUDI deaths in Northern Ireland are co-sleepers.
Pathologist
What I find most intereting about prone/supine/side is the whole business of "fashion" in medicine.
I STILL find it hard to put patients in heart failure on beta blockers - you would have been struck of for doing that on a patient when I was a medical registrar if the patient died.
I feel like Galileo when approaching HRT now.
I can still remember writing out frusemide 40 + Slow K ii - routinely
I still remeber when Atenolol was a good safe first line drug for hypertension.
Examples are legion.
What I have learnt over the years is NEVER to jump on a bandwagon when an "exciting" bit of research comes out and, similarly, never abandon a tried and trusted line of treatment for something new until the evidence is overwhelming.
John
The Back to Sleep campaign did work in reducing the number of cot deaths I think. But there are suggestions that the fall in numbers isn't as dramatic as it appears because it coincided with a time when pathologists tended to stop using SIDS for registration of the cause of death and used 'un-ascertained' instead. The other possibility is that baby autopsies were becoming much more detailed with more investigations being carried out and natural causes were being identified.
Pathologist - I recall researching all this for my sister when my nephew was tiny, and finding that there was good evidence from other countries apart from the UK. Any idea whether your comments about autopsy practice and death categorisation might apply elsewhere?
Hi Potentilla,
I'm UK based, so I really only know a bit about this country. I know the USA has a very different system to us in investigating infant deaths-over here all are reported to the coroner as unexpected deaths and all have an autopsy. In USA, different states have different patterns of organisations, some with sheriffs, some with coroners, some with medical examiners and as far as I am aware, not every baby may be autopsied (very happy to be put right on this if anyone knows better). SIDS as a term/diagnosis originated in USA in the late 1960s as a helpful label for those babies who died unexpectedly with no obvious cause of death (the definition of SIDS has expanded over the years). So I think SIDS is still used more over there than here.
There is also a difference of opinion between pathologists-a general adult pathologist undertaking a baby post-mortem is more likely to use the term SIDS than a paediatric pathologist. There was a 'fashion' few years ago to call these deaths 'interstitial pneumonitis' (the histological feature seen in a viral chest infection) which would mean that this wouldn;t have been coded as a cot death, but as a death due to natural causes. In practice, many of my babies have some degree of IP-all babies get snuffles and colds so I think it's unlikely that this has killed them, but you never know. In some cases, it may the interaction between an adverse sleeping environment and an otherwise relatively innocuous chestiness.
I also know that FSID (Foundation for Study of Infant Deaths), the charity, prefers pathologists to use SIDS as a diagnosis rather than un-ascertained, but the vast majority of paediatric pathologists don't like it-SIDS is not a diagnosis in itself, it is the lack of a definitive diagnosis, and as such, isn't accurate.
I fully believe that there are babies dying of natural disease that we don't yet have the ability to recognise, or babies who have a heritable predisposition to infant death on a molecular or genetic basis. However, in my own personal experience, shared by many of my paediatric pathology colleagues, the majority of my babies have home circumstances that are less than ideal.
Badly wrong John, and no correction. I'm disappointed in you actually, you are usually better.
SIDS was no part of the case.
Why did the meme for putting babies on their side or front to sleep come from? My view is that it leaked out of anaesthesia, which started 150 years or so ago, and produced a worry about aspiration.
Badly wrong John, and no correction. I'm disappointed in you actually, you are usually better.
SIDS was no part of the case.
++++++
I don't think it was badly wrong Adrian, because the PUBLIC perception is that there WERE cot deaths, rather than "unexpected" deaths.
Either way, deeply distressing for an innocent mother to have this happen and then be charged with murder.
Note this quote from the media:
The paediatrician left the hearing without commenting on the GMC's decision.
'Grand National' odds
During the trial, Sir Roy said the probability of two natural unexplained cot deaths in the family was 73 million to one.
The figure was later disputed by the Royal Statistical Society and other experts said that once genetic and environmental factors were taken into consideration, the odds of a second cot death in the same family were closer to 200 to one.
ALL THE TIME THE MEDIA REPORTED THESE DEATHS AS TURNING OUT BE BE "COT DEATHS". This was a HUGE red herring.
But "During the trial, Sir Roy said the probability of two natural unexplained cot deaths in the family was 73 million to one."
I have to be very careful indeed about what I say here, but if you read Sally Clarks own website, it is quite clear from the pathologist's reports that whatever happened here THEY WERE NOT COT DEATHS
But everyone thinks they were. Including the media.
"This trend was to reach its apogee in 1999 when solicitor Sally Clark was tried for allegedly murdering her two babies. Her elder son Christopher had died at the age of 11 weeks, and her younger son Harry at 8 weeks. Medical opinion was divided on the cause of death, and several leading paediatricians testified that the deaths were probably natural. Experts acting for the prosecution initially diagnosed that the babies had been shaken to death, but three days before the trial began several of them changed their collective opinion to smothering. Amongst the prosecution team was Meadow, whose evidence included a soundbite which was to provoke much argument: He testified that the odds against two cot deaths occurring in the same family was 73,000,000:1, a figure which he obtained by squaring the observed ratio of live-births to cot-deaths in affluent non-smoking families (approximately 8,500:1). The jury returned a 10/2 majority verdict of "guilty"."
++++
And Sally Clarks site says:
After four weeks' delay, during which Sally and Steve had made several enquiries as to progress of the autopsy, both were arrested. Dr Williams reported retina and brain damage attributed to "baby shaking". The case was later referred to Professors Green and Meadow, a team well known in the profession and media for their campaign that up to 40% of cot deaths are in fact abuse, who thought it "likely" but were cautious enough to then recommend the cause of death as "unascertained."
"cot death" is now the generally accepted explanation as to what really happened even though strictly medically this is not the diagnosis.
It is referred to everywhere.
But you are right, it was technically inaccurate - and of course made the trial an even bigger balls up by Meadow : what was he going on about?
So I have stuck a few words in.
John
Commenting by HaloScan
