Unexpected infant death: lessons from the Sally Clark case

Roger W Byard
Med J Aust 2004; 181 (1): 52-54. || doi: 10.5694/j.1326-5377.2004.tb06162.x
Published online: 5 July 2004


  • In November 1999, in the United Kingdom, a woman was convicted of the murder of her two infant sons. An appeal against the conviction was dismissed in October 2000, but the conviction was quashed by a second court of appeal in January 2003.

  • Review of the autopsy findings showed that standard procedures had not always been followed, thus limiting verification of the alleged findings. Some potentially important diagnoses and conclusions were also altered over time.

  • This case and its sequelae demonstrate the difficulties that may arise if cases are not fully investigated by pathologists with specific training or experience in paediatric forensic pathology, with all of the results being clearly summarised and discussed in autopsy reports.

  • Trying to clarify findings, diagnoses and circumstances of death at a later stage may simply not be feasible, owing to a wide variety of possibilities other than inflicted injury.

  • This type of case has unfortunately led to mistrust of the medical and legal systems and has made the investigation of such emotive and tragic cases all the harder.

On 9 November 1999, a 35-year-old lawyer, Sally Clark, was convicted of the murder of her two infant sons by a majority of 10 to 2 in the Chester (UK) Crown Court. An appeal against the conviction was dismissed on 2 October 2000. Following review by the Criminal Cases Review Commission and amid considerable media attention, the conviction was quashed by a second court of appeal1 on 28 January 2003. The Crown did not seek retrial, a decision that the court of appeal agreed with.

The Clark case raises a number of important issues for police, physicians and lawyers concerning the handling of cases of unexpected infant death. Fundamental errors were made in the assessment of autopsy findings, in the interpretation of the results and in the application of statistics to such cases. Sally Clark was wrongly jailed for 3 years and the credibility of medicolegal processes was significantly damaged. To quote Sally Clark on her release, “There are no winners here”.2

To briefly summarise the cases: on 13 December 1996, Christopher Clark, at the age of 11 weeks, was found dead by his mother. The cause of death was initially recorded as lower respiratory infection, but this was changed to suffocation following a review of the findings after his 8-week-old brother, Harry, had subsequently been found dead by his mother on 26 January 1998. Both infants had appeared well before their deaths. Both infant autopsies were performed by a pathologist with no specific training in paediatric forensic pathology.

Review of autopsy findings

Review of the pathology evidence reveals some very disturbing features. For example, it was not possible to verify whether Christopher in fact had any injuries, as the postmortem photographs were of poor quality, the bruises were not incised and rephotographed to distinguish them from simple postmortem lividity, and no histological samples were taken. Of note, the alleged bruises were not observed by medical staff or police officers at the hospital where resuscitation was attempted. It is also possible that some of the reported injuries may have been caused by resuscitative efforts.

Intra-alveolar haemorrhage, which was not noted in the initial autopsy report, was used, along with haemosiderin-containing macrophages, to support suffocation as the cause of Christopher’s death.3,4 However, both of these findings are recognised as non-specific. Intra-alveolar haemorrhage is common, and is altered by prolonged postmortem intervals, attempts at resuscitation, the areas of the lungs where samples are taken and the position of an infant’s body after death.5 Haemosiderin-containing macrophages are also found in a variety of circumstances.6,7 It is also possible that they may have resulted from Christopher’s independently verified nose-bleed 2 weeks before death. It did not appear that his nasal passages were examined at autopsy for an alternative source of the haemorrhage, such as a vascular malformation. There was no evidence of lethal lower respiratory infection, despite this being reported as the cause of death in the initial autopsy report. Based on the uncertainties in this case, it appears most appropriate to classify the death as “undetermined”. There was certainly no verifiable evidence of inflicted injury.


The initial cause of Harry’s death, based on a collection of typical and not-so-typical findings, was reported as shaken-impact syndrome. The diagnosis of retinal haemorrhage, which was quite crucial to support the possibility of shaking, was, however, incorrect. The retinal vessels were found, on independent review, to be merely congested. Retinal haemorrhages had also not been present when Harry had been examined at the hospital where resuscitation was attempted. Although not a recognised feature of shaken-impact syndrome,8 cerebral laceration was also used to support a diagnosis of shaking. However, these injuries were subsequently determined to be postmortem artefact arising at the time of brain removal. Given the unreliability of these findings and the absence of typical features of shaken-impact syndrome, such as retinal haemorrhages, subdural haemorrhages and axonal damage, the cause of death was changed to suffocation. To support this, several unusual features were proposed, including episcleral haemorrhages within the soft tissues around the eyeballs. However, this has not been reported in the literature as a marker for suffocation and is most likely an artefact caused by disruption of delicate vessels at the time of enucleation.

Other findings, such as retinal haemosiderin deposition and inflammation, acute inflammation and bruising of the paraspinal muscles, extradural haemorrhage and swelling of the spinal cord, cerebral hypoxia–ischaemia, cerebral haemorrhage, and haemosiderin staining of the meninges, were either not verifiable, were incidental findings, or were postmortem artefact.

While the alleged costochondral fracture dislocation of the right first rib could not be verified, as it had not been photographed or sampled histologically, there did appear to be evidence of a healing fracture of the right second rib on histology. This would be an unusual birth injury and may well have resulted from non-lethal compressive chest trauma in the weeks before death.

The presence of petechial haemorrhages of the left eyelid was also of concern, as this may occur with inflicted asphyxia or strangulation.9 Skin petechiae may also occur when there is sepsis, and one of the most significant findings in Harry’s case, which had not been put before the original court or first court of appeal, was the presence of a pure growth of Staphylococcus aureus from five separate sites (throat, bronchus, trachea, stomach and cerebrospinal fluid) in association with a mild acute inflammatory infiltrate in the cerebrospinal fluid. Neither lawyers for the prosecution nor the defence were aware of these results, as they had not been referred to by the examining pathologist in his autopsy report. Despite assertions from several medical experts that the results were merely postmortem contamination or incidental, I considered that Harry’s death was caused by staphylococcal sepsis.

Misleading statistics

Although it was stated in the primary court case that the chance of a second death in the same family attributable to sudden infant death syndrome (SIDS) was 1 in 73 million, this has been convincingly rebutted in a number of publications.2,10 However, as the cases would not be attributed to SIDS using the standard definition,11 the use of so-called SIDS statistics was really not appropriate. Despite this, it has been speculated that the use of such figures may have had considerable influence on the jury’s decision.10

Implications of the Clark case

This tragic case raises a number of important issues, not the least of which being that lethal pathological processes in infants are often not clear-cut and have poorly understood underlying mechanisms. These difficulties are magnified if investigations and interpretations are inadequate. The cases described here clearly demonstrate persistent problems that beset the investigation of unexpected infant deaths and their presentation to courts. A summary I prepared for the defence detailing aspects of this was cited by the second court of appeal:1

Unfortunately there were significant and ongoing problems in the investigation of these deaths. Standard protocols were not followed and essential steps such as routine dissection and histology were omitted which prevented verification of alleged autopsy findings.12 As well, a number of potentially important diagnoses and conclusions were altered over time. For example, Christopher’s initial cause of death of lower respiratory tract infection was withdrawn and observations of no significant haemorrhage within his lungs were changed to marked haemorrhage. The finding of retinal haemorrhages in Harry which was vital to sustain the diagnosis of shaken-impact syndrome was altered to no haemorrhage, brain lacerations were found to represent postmortem artefact, swelling of the spinal cord was not present and bruising of paraspinal tissues was also not able to be substantiated. This is not a unique situation with statements in the literature in recent years that “investigations into the pathology and circumstances of sudden infant death are often scanty and inexpert” with significant omissions being documented when cases were audited.13-14 The Clark brothers demonstrate difficulties that may arise if cases are not fully investigated15 with all of the results being clearly summarised and discussed in the autopsy report. Trying to clarify findings, diagnoses and circumstances of death at a later stage may simply not be feasible due to a wide variety of possibilities other than inflicted injury.

  • Roger W Byard

  • Forensic Science Centre, Adelaide, SA.


Competing interests:

None identified.

  • 1. R v Clark (2003) EWCA Crim 1020.
  • 2. Bacon CJ. The case of Sally Clark. J R Soc Med 2003; 96: 105.
  • 3. Yukawa N, Carter N, Rutty G, Green MA. Intra-alveolar haemorrhage in sudden infant death syndrome: a cause for concern? J Clin Pathol 1999; 52: 581-587.
  • 4. Becroft DM, Lockett BK. Intra-alveolar pulmonary siderophages in sudden infant death: a marker for previous imposed suffocation. Pathology 1997; 29: 60-63.
  • 5. Hanzlick R. Pulmonary hemorrhage in deceased infants. Baseline data for further study of infant mortality. Am J Forensic Med Pathol 2001; 22: 188-192.
  • 6. Byard RW. Sudden death in infancy, childhood and adolescence. 2nd ed. Cambridge: Cambridge University Press, 2004: 77-166.
  • 7. Byard RW, Stewart WA, Telfer S, Beal SM. Assessment of pulmonary and intrathymic hemosiderin deposition in sudden infant death syndrome. Pediatr Pathol Lab Med 1997; 17: 275-282.
  • 8. Krous HF, Byard RW. Shaken infant syndrome: selected controversies. Pediatr Dev Pathol 1999; 2: 497-498.
  • 9. Byard RW, Krous HF. Petechial hemorrhage and unexpected infant deaths. Leg Med (Tokyo) 1999; 1: 193-197.
  • 10. Watkins SJ. Conviction by mathematical error? BMJ 2000; 320: 2-3.
  • 11. Willinger M, James LS, Catz C. Defining the sudden infant death syndrome (SIDS): deliberations of an expert panel convened by the National Institute of Child Health and Human Development. Pediatr Pathol 1991; 11: 677-684.
  • 12. Krous HF, Byard RW. International Standardised Autopsy Protocol for sudden unexpected infant death. Appendix 1. In: Byard RW, Krous HF, editors. Sudden infant death syndrome — problems, progress and possibilities. London: Arnold, 2000: 319-333.
  • 13. Unexplained deaths in infancy. Lancet 1999; 353: 161.
  • 14. Bacon CJ. Cot death after CESDI. Arch Dis Child 1997; 76: 171-173.
  • 15. Byard RW. Inaccurate classification of infant deaths in Australia: a pervasive and persistent problem. Med J Aust 2001; 175: 5-7. <eMJA full text>


remove_circle_outline Delete Author
add_circle_outline Add Author

Do you have any competing interests to declare? *

I/we agree to assign copyright to the Medical Journal of Australia and agree to the Conditions of publication *
I/we agree to the Terms of use of the Medical Journal of Australia *
Email me when people comment on this article

Online responses are no longer available. Please refer to our instructions for authors page for more information.