In his most famous investigation, thirty-nine children who had recurrent episodes of apparently life-threatening events, usually outside hospital but sometimes on a hospital ward, were referred to a specialist ward where they were secretly filmed. In thirty-three cases, these ‘events’ turned out to be induced by a parent. The covert footage showed incidents not just of emotional abuse but also poisoning, strangulation and, especially, suffocation. In fact, there were no fewer than thirty attempts at suffocation in just this small group.
Thanks to the surveillance, professional intervention protected the children. But between them those children had forty-one siblings, of whom twelve had died suddenly and unexpectedly. Once the parents were exposed, four actually admitted to killing eight of these siblings. When these ‘sibling’ cases were reviewed it was found that, in eleven of the twelve cases, the cause of death given by the pathologist who had performed the post-mortem was SIDS. In the twelfth case, the cause given had been gastroenteritis, but further investigation now revealed the child had been poisoned. Fifteen more of the forty-one siblings were later discovered to have suffered abuse.
Naturally, there was widespread shock at these findings and considerable disbelief. I feel, as a result of David Southall’s work, we started to move out of an age of innocence. A lot of people, however, preferred innocence. It was hard to accept that there were children who needed protection from the very adults who should be protecting them.
Suspicion gradually became an almost routine response to the unexplained death of a baby, and that must have felt very unfair to the innocent. So unfair that David Southall faced vitriolic and vocal opposition. The ethics of his covert video surveillance of parents came in for particular criticism. I fear that, without it, he never would have been believed, so incredible did his results seem at that time.
Parents (as well as many police officers and even social workers) who had once regarded any concern from outside the family for the welfare of children as an invasion of privacy, were now forced to accept change. Adult victims of all sorts of parental abuse were talking publicly about their childhoods and a light was shining onto family secrets in a way that had been impossible when privacy was king. That light was held by professionals working with children – health visitors, doctors, nursery staff – who were now encouraged to report any suspicions of abuse.
As child protection became a subject for national debate, the discussion about unexplained baby deaths left theory behind and became concentrated in the specifics of just one case. I am talking about the trial of Sally Clark.
SIDS had once taken a cruel harvest from right across the social spectrum, but as the middle classes became informed about risk factors and minimized them, SIDS began to look like something more likely to affect the poor. Until, in 1996, Sally Clark, a well-off, middle-class solicitor and policeman’s daughter whose background and bearing reflected that of so many respectable, professional, working mothers, lost first one baby to SIDS. And then, in 1997, another.
The first baby, a boy, was eleven weeks old. Sally Clark’s husband, Stephen, also a lawyer, was out at the office party when his wife found the baby unconscious. She called an ambulance, but for some reason she was unable to open the door when it arrived. The baby the paramedics discovered had no pulse and had been cyanosed – meaning his lips and fingers were blue – for some time. But he was only officially declared dead after an hour when resuscitation attempts failed.
The pathologist who conducted the subsequent post-mortem was worried by particular injuries: a split and bruise at the top of the baby’s mouth, inside the lips, and the absence of any obvious cause of death. He had these injuries photographed and it is very unfortunate that the photographer’s camera was not working properly and the resulting pictures were of such poor quality that they were to prove absolutely unhelpful in the furore that followed. That was extreme bad luck: I’ve only come across such decisive camera failure one other time in my entire career.
When the pathologist discussed his worries with the police and the coroner’s officer, he had to agree that the split inside the baby’s lip could have been caused by the resuscitation attempt. In the case of these particular injuries, such a cause would be rare – damage to that part of the mouth is indicative of abuse – but it is certainly possible that they occurred during the frantic hour of resuscitation.
The police and the coroner’s officer chose to believe the cause was the attempted resuscitation and there was no further investigation. They were dealing with a well-off, professional family, not the kind of criminals or child abusers they normally encountered.
The baby had been X-rayed and no broken bones found. The pathologist had taken histology samples and these were pretty much normal. All except for the very small possibility – certainly not a probability – of some increase of inflammatory cells in a sample from the lungs.
He might have exonerated the parents by giving SIDS. Or, in view of the baby’s injuries, he might have aired the possibility of an unnatural death by giving ‘Unascertained’. Times were changing fast but the fact is, in 1996 he may also have been working for a conservative, one might even say Luddite, coroner: in those days any lawyer or medical practitioner with a few years under their belt could become a coroner. I do not know the coroner involved in this case but many of them, exclusively those who did not have a medical qualification, were still struggling with the concept of SIDS because it was a cause of death based on no definable evidence. And some coroners also disliked the word ‘Unascertained’, particularly when applied to babies. They wanted to say kind, comforting things to grief-stricken parents, not, ‘We don’t know why your baby’s dead.’ Limitations like this could leave a pathologist in limbo.
For whatever reason, the pathologist seized on the slides from the lung sample with their suggestion of inflammation and decided that the baby had died of natural causes: in fact, lower respiratory tract infection. The child’s death was treated as natural.
But the following year the Clarks’ second baby died too. He had been a few weeks premature but he was fine now, at two months old. Sally Clark breast-fed him and supplemented that with bottled milk. One evening, her husband went to prepare a bottle for the night feed, leaving the mother to watch TV while the baby lay in his bouncy chair. When she saw that her son was limp, she called her husband and rang an ambulance. The paramedics found the baby dead.
The same pathologist was involved and this time discovered injuries that suggested to him this baby may have been shaken, perhaps on several occasions over several days. He believed he had found haemorrhages in the eyes and spinal cord as well as some abnormalities of the ribs which suggested that there had been previous fracture or trauma.
Sally Clark and her husband were arrested on suspicion of the murder of their second child and, while they were being interviewed about his death, the pathologist quite rightly went back to his report on the death of the first child. In this he was complying with Home Office guidelines, which state that ‘if previously held conclusions can no longer be substantiated then any change of opinion must be promptly and clearly stated irrespective of any possible embarrassment’.
The pathologist did change his opinion and was to suffer prolonged and considerable embarrassment.
Examining the microscope slides once again which had, possibly but not probably, shown a lung inflammation in the first child, he decided that the cause of death he had previously given was entirely wrong. He decided that there was no inflammation. He had discovered blood in the baby’s alveoli, but had not even mentioned it before. He later said he had assumed this was simply consistent with changes after death. But he had since found that the nature of this finding might be abnormal and that it was consistent with asphyxiation.
At this point, as many experts subsequently pointed out, he might have kept the question open by revising the cause of death from ‘Lower Respiratory Tract Infection’ to ‘Unascertained’.
One expert witness explained in t
he later court case why he personally would have done this:
Unascertained … means that the child’s death may have been natural but without explanation – perhaps what the jury knows as a cot death. Or it might be that the child died unnaturally but I can’t find out why, or it might be the child died of a natural disease that I am not clever enough to diagnose and recognize …
But the pathologist did not choose ‘Unascertained’. Instead he submitted a further statement saying that he no longer believed the first child had died of natural causes. The volte face concluded that: ‘There is evidence he died from an asphyxial mechanism such as smothering.’
Six weeks after her arrest for the murder of her second son, Sally Clark was also arrested for the murder of her first. At her trial, the jury famously heard evidence from the paediatrician Professor Sir Roy Meadow, who had popularized, if he did not actually coin, the maxim, ‘One sudden infant death in a family is a tragedy, two is suspicious and three is murder unless proved otherwise.’
A memorable statistic was unfortunately added at Sally Clark’s trial which was to become associated with Professor Meadow for ever: ‘The chance of two children dying naturally in these circumstances is very, very long odds indeed. One in 73 million …’
One in 73 million was a headline grabber and may have sealed the defendant’s fate. He went on to say, ‘… it’s the chance of backing that long-odds outsider at the Grand National … let’s say it’s an 80 to 1 chance, you backed the winner last year, then the next year there’s another horse at 80 to 1 and you back it again and it wins … you know, to get to these odds of 73 million you’ve got to back that 1 in 80 chance four years running.’
Sally Clark was found guilty by a jury majority of ten to two of both murders and sentenced to life imprisonment.
I was not directly involved in this case. But it affected all of us. Her conviction, and David Southall’s work, was indicating that the murder of babies was much more common than we had all assumed and parents who murder their children are more common than we would have thought possible. Even ‘nice’, middle-class professional parents. We pathologists were called upon to offer a medical and scientific analysis within the context of society’s current thinking and I’m sorry to tell you that the purity of scientific truth rarely cuts through contemporary social attitudes.
Personally, I could never forget how I had walked an endlessly wailing, lactose-intolerant baby around the house, night after night, thinking, in so far as one can think over the noise of penetrating screams, that I would do almost anything for some sleep. I knew that the middle classes, even without the pressures of poverty or isolation, were as vulnerable as other parents to extreme desperation.
Soon after Sally Clark’s trial, a case of mine reflected what had now become the great controversy of child protection. When I saw this six-month-old baby in the mortuary he seemed to have been healthy and well cared for, but I was quickly aware of a distinctive triad of symptoms. He had a subdural haemorrhage, that is, a bleeding on the surface of the brain. The brain itself was swollen. And his eyes had haemorrhages in the retinas. These three symptoms, particularly without external signs of injury, were now regarded as the classic triad of symptoms of what was then called ‘shaken baby syndrome’.
In the 1940s a radiologist, John Caffey, reported multiple fractures of varying ages in some children and initially thought this was a new disease. Later research showed the fractures were due to repeated trauma and in the 1960s the term ‘battered baby syndrome’ was first used. Then, in the 1970s, a neurological variant, shaken baby syndrome, was recognised as a form of whiplash by neurosurgeon Norman Guthkelch. So these syndromes, and their root cause in deliberately inflicted trauma, were medically well-known. However, they were only brought to public attention in 1997 by the famous case of the medical couple in Massachusetts who had left their baby in the care of a nineteen-year-old English au pair.
When the baby suddenly became ill and was rushed to hospital, he showed that classic triad of symptoms and, in a televised court case that obsessed America, young Louise Woodward was found guilty of murder. Many Americans were outraged when this charge was changed afterwards by the judge to manslaughter because he felt that there really was not sufficient evidence beyond reasonable doubt for a murder conviction. And that was because medical experts were so divided about shaken baby syndrome.
This was not the end of the story because by now shaken baby syndrome itself had become the story. Most people had never heard of it before Louise Woodward’s trial and suddenly it was in the headlines and every pathologist was on full alert for the now-famous triad of symptoms.
In fact, as a cause of death it was then and is now highly controversial and the subject of much scientific and medical argument. Shaken baby syndrome, now also called abusive head injury or non-accidental head injury, has created its own angry groups of protesters and deniers. There is an ongoing search for natural causes to explain it.
Long after Sally Clark’s imprisonment, in 2009, the Royal College of Pathologists attempted to bring together the various sides of the debate and this disparate group was able to issue a statement on what they called traumatic head injury (which was yet another name for just the same thing) reminding pathologists that, even if all elements of the triad are present, each may have other, non-traumatic causes. The statement said clearly that the triad alone is not enough to say a parent harmed a baby ‘beyond reasonable doubt’: more evidence would be needed for that. And special care should be taken interpreting the injuries when the baby is under three months old, because they might, possibly, have been caused during birth.
It looked like a consensus. However, if anything, the debate became more heated. Forty years after first describing the features of this particular type of head injury, Norman Guthkelch in 2012 reviewed its history and expressed concern:
While society is rightly shocked by an assault on its weakest members and demands retribution, there seem to have been instances in which medical science and the law has gone too far in hypothesizing and criminalizing alleged acts of violence in which the only evidence has been the presence of the classic triad or even just one or two of its elements.
At the end of the 1990s, shaken baby syndrome was very firmly on the pathologist’s radar and the six-month-old baby I saw seemed to show every relevant symptom. According to his mother, however, he had launched himself from his car seat, which she had placed on a work surface, with no prior warning that he was physically capable of doing such a thing. As a result, he had fallen out of the seat and about one metre down onto a hard kitchen floor. Even a few years earlier, I would have been reluctant to believe her. But, post-Woodward, I had very grave concerns indeed.
The mother came from an impoverished and war-torn nation and had been brought by her husband to live in London with his mother and various other family members. She spoke no English. Their accommodation was overcrowded. She still had a relationship with her husband but, because there was so little room in the family flat, had recently begun to live separately in housing that had, it seemed, been provided by the council. Her flat was clean and airy but apart from a bed and a television, she had no furniture and evidently just sat on the floor.
She seemed to spend all day there alone. Outside the family, whom she seldom saw, she only knew one person in London and that friend lived far away. I thought of my desperation with the wailing baby Chris all those years ago and pitied this woman for the utter isolation in which she lived, far from home, with this small baby.
She had no car, of course, but carried her son around the flat in his car seat. When she made meals, she took him to the kitchen and perched his seat on the work surface. He was not strapped in. One evening she was busy preparing the food, heard a loud thump (‘a sickening thud’, are words which often seem to crop up in court), and turned around to find the baby face-down on the floor. He cried immediately but his eyes rapidly became staring and his breathing erratic: it was clear something was
very wrong.
She tried to phone 999 but was unable to make herself understood. She tried to phone her husband, who spoke English, but the emergency services were still blocking the line. She ran down to the street to get someone else to phone 999 but by that time the police had already responded to her distressed although incomprehensible call.
They found a baby with blood coming from his nose and mouth, shaking slowly and losing consciousness. The police would not let the mother get into the ambulance and the baby was rushed to hospital in an apparently deteriorating condition. CT scans revealed severe internal injuries, but initially it seemed he would probably survive. However, his cardiovascular system became unstable and, in spite of all attempts at resuscitation, he died twelve hours after his mother had called the emergency services.
So far as we know, natural causes for the three individual symptoms which are together suggestive of shaken baby syndrome (like blood clotting difficulties) are very rare. While one should always be alert for these rare natural causes, I felt that a bleeding, swollen brain and bleeding retinas indicated that the baby had received a major trauma. Sometimes that trauma may be an accident, like a car crash. And sometimes, particularly if there are no external marks, it is not an accident.
In this case, I was further convinced that the baby had been shaken because, despite the internal haemorrhages, there was no fracture of the skull, no bruising of the head, no trauma at all on the outside to show that he had plummeted from a work surface onto a hard floor. He had the triad, although no trauma to the spinal cord at the neck, which is another associated symptom.
The mother was tried for manslaughter. The defence pathologist felt that the baby’s injuries meant he had been subjected to either an acceleration-deceleration trauma (shaking) or a head impact. He did agree that, in a six-month-old baby, shaking is the most common cause of such internal injuries, and added that it is sometimes unclear whether shaking or impact is the more important factor, since babies who are shaken are then sometimes thrown forcefully down. However, his overall opinion was that the brain swelling alone had killed this baby, and that the injuries therefore confirmed the mother’s story.
Unnatural Causes Page 26