Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 May 26; 334(7603): 1083–1085.
PMCID: PMC1877957
Sudden Infant Death

Protection for the innocent

Jonathan Gornall, freelance journalist

Concerns over the prosecutions of Angela Cannings, Sally Clark, and Trupti Patel have led to wide ranging reforms to the way unexplained infant deaths are handled. Jonathan Gornall describes how the reforms should make a difficult situation less traumatic for both parents and professionals

“They came at dawn. Two police cars, full of officers, brake in front of Hope Cottage in Wilmslow. Sally is in the kitchen in her dressing gown. There is a knock at the front door …”

This is how John Batt, a solicitor who was part of Sally Clark's legal team, recorded the moment almost 10 years ago when her already shattered world collided with what was then the harsh reality of the investigation of sudden unexpected death in infancy (SUDI) in England and Wales.1 Much has happened since Mrs Clark's arrest in 1998 for the murder of her two infant sons, Christopher and Harry: her imprisonment in November 1999, the failure of her first appeal in October 2000, her successful appeal in January 2003, and, on 16 March this year, her own death.2 The consequences of Mrs Clark's case, devastating for her family, have also been far reaching for the medical and other professionals involved and for the child protection system as a whole in England and Wales.3 4 5 6

A positive legacy is, however, emerging from the tragedy. From April next year all sudden unexpected infant deaths in England and Wales will be investigated in accordance with a new multiagency protocol, introduced as part of the reforms of the 2004 Children Act and as a direct result of the Clark case. The overriding principle, set out in the statutory guidance for the operation of the protocol, is now this: “Each unexpected death of a child is a tragedy for his or her family, and subsequent enquiries/investigations should keep an appropriate balance between forensic and medical requirements and the family's need for support.” The guidance continues: “Families should be treated with sensitivity, discretion and respect at all times, and professionals should approach their enquiries with an open mind.” 7

Mrs Clark's appeal in 2003 was followed the same year by two other high profile cases. In June, Trupti Patel was tried and acquitted of the murder of her three children and in December, Angela Cannings, jailed in 2002 for the murder of two of her children, was released after the Court of Appeal found her convictions unsafe.

As debate raged in the media about the three cases, the royal colleges of pathologists and paediatrics and child health set up a joint working group to examine how to improve the investigation of sudden unexpected death in infancy. It was chaired by Helena Kennedy QC, then president of the National Children's Bureau.

In her report, published in September 2004, Baroness Kennedy concluded there was “great variance around the country as to how a sudden infant death is handled by doctors and the police” that “creates a very complicated patchwork in which good practice can often be found, but stories of insensitivity and failure are still sadly and angrily being told.”8

One of the 11 members of the working group was Peter Fleming, professor of infant health and developmental physiology at the University of Bristol, who had been responsible for pioneering work in preventing cot death and the investigation of sudden unexpected infant death. The template chosen for the protocol in the Kennedy report was one that had been developed by Fleming and colleagues for their investigations in Avon.9 10

As the Kennedy report was being finalised, the 2004 Children Act was going through parliament. One of its measures obliged local authorities to set up local safeguarding children boards responsible for establishing panels to investigate each unexpected death of a child in their area. This function, which has been optional since April 2006, becomes compulsory from next April, and the statutory guidance for the panels was produced by a working group of professionals, many of whom also worked on the Kennedy report, including Fleming.7

The statutory process by which the Kennedy protocol will have become obligatory by April next year has been astoundingly fast. This, says Fleming, “reflects a huge concern within the professions that professionals and families were being let down by the system because nobody knew quite what was required of them. We were getting it wrong in both directions. Families were suffering and so were professionals.

“From the point of view of protecting families, always having a team of professionals involved means that an over-zealous police officer or an over-suspicious paediatrician is less likely to have a disastrous effect.”

New procedures

The protocol, which establishes a standard routine for a collaborative multiagency response to every sudden unexpected infant death, details what is expected from ambulance crews, accident and emergency staff, child protection coordinators, coroners, coroners' officers, general practitioners, health visitors, midwives, paediatricians, pathologists, police, and social workers. It emerged out of the Avon cot death studies, which began in the early '80s.10 11 Fleming and colleagues had introduced their first structured arrangements for home visits and multiagency review of sudden unexpected infant deaths in 1984, and the benefits quickly became apparent: “Every one of the major risk factors for cot death was first identified here in the '80s, and it was through that approach,” says Professor Fleming, who believes that not only will the protocol ensure more robust and reliable investigations of sudden unexpected infant deaths but that it could also lead to a halving in the number of deaths from sudden infant death syndrome. One of the responsibilities of the new boards will be to collate data on deaths in their regions and report important factors to the Confidential Enquiry into Maternal and Child Health. The inquiry will in turn aggregate the regional data and draw nationally applicable conclusions.

The protocol fundamentally changes the way sudden unexpected infant deaths are investigated. A key shift is that the investigating police officer should visit the home with a paediatrician. “If your child was ill, you wouldn't ask a policeman to tell you what was wrong with it,” says Professor Fleming.

It also calls for a change of attitude among the police, who traditionally have treated sudden infant deaths like any other unexplained deaths—with suspicion. This contrasts with the Kennedy report, which stressed that “An important starting point is the acknowledgement that in the vast majority of cases where babies suddenly die, nothing unlawful has taken place.”

Professor Fleming, having worked on the protocol for the past four years with every police force in the Southwest, is convinced that forces nationwide will embrace the concept of investigating sudden infant deaths in partnership with a paediatrician.

“Initially, many of them were suspicious and concerned but once they'd become engaged they recognised that in a sense it takes a lot of the pressure off them as well. They are no longer having to look for medical clues and there's no question that two sets of eyes and ears pick up a lot more information.”

Instead of treating the scene of death as a crime scene and arriving in the small hours of the morning to seize potential evidence, trained, plain clothed officers attend the house at a civilised hour and in partnership with a paediatrician. The parents benefit from post-traumatic debriefing in the setting in which the trauma occurred and the investigators benefit because “when you talk to people about what's happened in the place where it has happened, they remember much more of the detail and, indeed, they want to give us the information.”

Geared to the importance of the first “golden hour” of any investigation, police were concerned initially that vital evidence would be lost, but the Avon researchers found that the information gathered in this way often pointed them in the direction of particular tests that needed to be done on the baby and which might not have given valid results if the pathologist did not examine the body for two or three days.

Recruitment problems

There are, however, major practical obstacles to be overcome if the protocol is to function smoothly. For one thing, Professor Fleming is under no illusion that it will be easy to recruit paediatricians to be home visitors. A report by the Royal College of Paediatrics and Child Health earlier this year found that child protection had become something of a poisoned chalice for paediatricians, who are increasingly the targets of unfounded complaints. These had had “a profound impact on the professional and private lives of some paediatricians and had influenced their willingness to undertake future child protection work.” 12

In the Southwest region, where the protocol has been used for four years for all deaths in children under 2 years, Professor Fleming and colleagues have trained at least one paediatrician in every centre, but even here “it has become very clear that most paediatricians feel very uncomfortable in that role.” As a result, he and his colleagues are now exploring the possibility that the lead role could be taken not by paediatricians but by specially trained health visitors operating full time as a central resource for entire regions. The role of the paediatrician would be limited to the review and assessment phase of the investigation.

Pathology is another vital plank of the protocol—and another specialty with staffing problems. The protocol calls for each postmortem to be carried out to a standardised format by a paediatric pathologist. If concerns have been raised about the possibility of abuse or neglect, a forensic pathologist should take the lead, assisted by a paediatric pathologist, but such specialists are, as Baroness Kennedy pointed out, “thin on the ground and often unavailable at the crucial time.” She noted that throughout England and Wales, which have 43 police forces and 28 strategic health authorities, there were just 40 paediatric pathologists.

In 2002, a working group convened by five organisations, including the royal colleges of paediatrics and child health, pathologists, and obstetricians and gynaecologists, was set up to restore and develop specialist paediatric pathology. The report left no doubt about the vital role of the paediatric pathologist in the investigation of sudden unexpected infant deaths. An objective review of 450 sudden unexpected infant deaths carried out by the Confidential Enquiry into Stillbirths and Deaths in Infancy had “demonstrated serious deficiencies when the post-mortem examination was carried out by a non-paediatric pathologist.” Essential tests had been omitted in up to 70% of children and the diagnosis was incorrect in an estimated 20% of cases. This had led to a “failure to recognise inherited conditions and on occasions led to inappropriate suspicion of harm.”13

The protocol emphasises the importance of the role of the pathologist in great detail. For instance, traditionally permission has had to be sought from the coroner before samples can be taken from a body. Under the new protocol, however, coroners will be expected to approve a routine set of up to 11 samples that can be taken in emergency or paediatric departments.

“Certain investigations need to be done very quickly in order to allow us to be clear about the answers,” says Fleming. “If the baby dies on a Friday and the postmortem isn't done until Monday or Tuesday, many of the tests will be totally useless.”

It's a lesson direct from the Clark experience. Flawed pathology lay at the heart of the case and, Professor Fleming believes, the trial would never have taken place had the multiagency protocol been in operation in 1998. Sally Clark's second appeal, in January 2003, succeeded after it emerged that the results of microbiological tests carried out on Harry had not been put before the jury at her trial. That, ruled the court, made her convictions unsafe.14

“Sally Clark was sent to prison for life for murder on the basis of a trial that didn't hear important information, which was that a spinal fluid sample had grown a pure growth of a known pathogen,” says Professor Fleming. “The problem was that the sample was taken 24 hours after that child died and the pathologist had assumed it was contaminant. The question of whether it was or wasn't can't be proved one way or the other, but it raises reasonable doubt and therefore the case should never have come to trial. Had a spinal fluid sample been taken immediately after death, we'd know the answer.”

Fleming is confident that the protocol will protect professionals, as well as parents and children: “Many paediatricians have reservations about getting into this area because they might end up having to give evidence if a case does turn out to be a non-natural death. It's not something that any of us enjoys, but we have to do it because part of our job is protecting children. Now, at least, we will be doing it within a recognised and accepted framework, so in theory we can be criticised only if we have failed to do what we should have done, rather than merely for being there.”

Information paediatricians must collect at first interview and home visit

  • Family medical history
  • Social and family history
  • Detailed medical history of mother
  • Detailed medical and developmental history of baby who has died
  • Detailed narrative account of baby's feeding, sleeping, activity, and health over past two weeks
  • Detailed, hour by hour narrative account of events in 48 hours before infant was found dead
  • The final sleep—careful description of when, where, and how baby was placed to sleep
  • Action after baby was found

Summary of protocol for dealing with sudden unexpected death in infancy (SUDI)8

  • Baby taken to accident and emergency department and, “unless clearly inappropriate,” resuscitation initiated
  • Examination by paediatric consultant and careful history taken from parents
  • Standard set of investigative samples taken immediately on arrival and after death confirmed
  • Paediatrician breaks news to parents, explains role of police and need for postmortem examination and permanent retention of blocks and slides; offers opportunity to donate tissues or organs for research
  • Member of staff allocated to care for parents, who are allowed to hold and spend time with baby and take mementoes, such as lock of hair
  • Police, designated SUDI paediatrician, and paediatric pathologist notified as soon as death pronounced
  • On-call paediatrician or SUDI paediatrician talks urgently to social services department to review relevant child protection information
  • Initial strategy discussion held between supervising police officer and SUDI paediatrician; together, they visit parents at home within 24 hours
  • If serious concerns raised at any stage about possibility of abuse or neglect, police become lead agency
  • SUDI paediatrician compiles report for coroner and pathologist, based on clinical circumstances of death, history obtained on home visit, and review of medical and social records
  • Coroner orders postmortem examination, preferably within 48 hours, by either paediatric pathologist or, if concerns raised about abuse or neglect, forensic pathologist assisted by paediatric pathologist
  • If examination shows no identifiable cause of death, whether or not concerns raised about abuse or neglect, pathologist categorises death as “unexplained pending further investigations.” In every case coroner holds inquest
  • SUDI paediatrician discusses results of postmortem examination with parents at earliest opportunity
  • Case discussion meeting held, usually in general practitioner's surgery, as soon as results of main postmortem tests available. This includes general practitioner, health visitor, midwife (if appropriate), SUDI paediatrician, other paediatrician if involved, pathologist(s), senior investigating police officer and, if appropriate, social worker
  • Purpose is to share information, agree cause of death and plan future care for family. There must be “explicit discussion of the possibility of abuse or neglect and, if no evidence is identified to suggest maltreatment, this should be documented as part of the report of the meeting”
  • Report sent to coroner, who takes information into consideration in conduct of inquest and cause of death notified
  • If serious concerns raised about abuse or neglect, social services take lead in any further investigation
  • SUDI paediatrician writes detailed letter to parents, giving information about cause of death, and meets them to explain contents of letter, answer questions, and offer care and support


Competing interests: None declared.


1. Batt J. Stolen innocence—the story of Sally Clark London: Ebury Press, 2004
2. Dyer C. Falsely convicted Sally Clark dies suddenly. BMJ 2007;334:602-3.
3. Dyer C. GMC hears case against paediatrician in Sally Clark trial. BMJ 2005;330:1463
4. Dyer C. Sally Clark pathologist removed from Home Office list. BMJ 2005;331:1355
5. Dyer O. Southall is barred for three years from child protection work. BMJ 2004;329:366
6. Williams C. United Kingdom General Medical Council fails child protection. Pediatrics 2006;119:800-2.
7. Department for Education and Skills. Working together to safeguard children London: Stationery Office, 2006.
8. Royal College of Pathologists and Royal College of Paediatrics and Child Health. Sudden unexpected death in infancy: a multi-agency protocol for care and investigation London: RCP and RCPCH, 2004.
9. Bristol Children and Young People's Services. Sudden unexpected deaths in childhood: a joint agency protocol for investigation and management in Avon and Somerset June 2002.
10. Fleming P, Blair PS, Sidebotham PD, Hayler T. Investigating sudden unexpected deaths in infancy and childhood and caring for bereaved families: an integrated multiagency approach. BMJ 2004;328:331-4. [PMC free article] [PubMed]
11. Fleming PJ, Gilbert R, Azaz Y, Berry PJ, Rudd PT, Stewart A, et al. Interaction between bedding and sleeping position in the sudden infant death syndrome: a population based case-control study. BMJ 1990;301:85-9. [PMC free article] [PubMed]
12. Royal College of Paediatrics and Child Health. An investigation into the nature and impact of complaints made against paediatricians involved in child protection procedures London: RCPCH, 2007
13. Royal College of Paediatrics and Child Health. The future of paediatric pathology services London: RCPCH, 2002
14. R v Clark, [2003] EWCA Crim 1020.

Articles from The BMJ are provided here courtesy of BMJ Publishing Group