The case of Sally Clark

Abstract

‘There are no winners here’ said Sally Clark, with remarkable equanimity, after the quashing of her conviction for the murder of her two baby sons. A prominent loser in this sad business, which culminated at the end of January, was the credibility of expert medical witnesses. One presented erroneous statistical evidence that damaged the defence, while another withheld pathological evidence that might have helped it. These errors formed the grounds for successive appeals, the second of which was successful—but only after Mrs Clark had been in prison for three years. The case raises issues about the assessment of sudden unexpected deaths in infancy (SUDI) as well as about the conduct of expert witnesses. Since the fall in incidence of SUDI, police and coroners have become more aware of the small proportion that result from covert homicide, and some coroners now prefer to instruct a forensic rather than a paediatric pathologist for any SUDI post-mortem. This may give a different emphasis to the examination, forensic pathologists focusing more on detection of foul-play, paediatric pathologists on identification of a natural cause of death. Certainly the latter tend to do a wider range of ancillary tests1. It is possible that crucial reports on the cerebrospinal fluid were withheld in the Sally Clark case because their significance was not initially appreciated. Since the large majority of SUDI, probably about 90%2, arise from natural causes, a paediatric pathologist is the more logical initial choice. Early identification of a natural cause will spare the family from unjustified suspicion and the police from unnecessary work. If suspicious features are encountered, a forensic colleague can be invited to participate. Ideally a SUDI pathologist should have both paediatric and forensic qualifications, but very few have such dual training. Currently there is a serious national shortage of paediatric pathologists. Coroners may therefore have to refer SUDI cases to centres at some distance, accepting the delay in the interests of more expert assessment. If, however a non-specialist is selected, at the least it should be someone with an interest and experience in paediatric pathology. In many SUDI even the most expert post-mortem will not identify a specific cause of death. Those that remain unexplained but are considered natural will normally be registered as sudden infant death syndrome (SIDS)—a classification of exclusion that has no positive diagnostic features. Distinction from covert homicide, in which the history is fabricated and signs are often non-existent, can be extraordinarily difficult. The most accurate picture will be obtained by completion of a jigsaw from all the relevant pieces of information, which must be collected by scrutiny of the circumstances and antecedents of the death as well as from the post-mortem. The Foundation for the Study of Infant Deaths is therefore recommending a comprehensive investigation in every case, including an early home visit by a paediatrician to talk with the parents and see where the baby died, careful review of all relevant records, a paediatric post-mortem conforming to the recommended protocol1, and a case discussion among all the professionals involved. There must be liaison with the coroner and police throughout. Everything possible must be done to identify a cause of death: families need to know why their baby died; other children must be protected both from familial disease and from inflicted harm; and parents must not be subjected to unjustified suspicion. Reports of the incidence of recurrent SIDS vary widely, depending largely on the thoroughness with which other possible causes of death have been excluded. The odds for recurrence quoted in this trial, 1 in 73 million, were derived by squaring the observed odds for a single case of SIDS in low-risk families such as the Clarks3. Such a calculation cannot be validly applied to individual families, each of which has its unique blend of genes and environmental influences that modifies the overall risk4. In the face of all these uncertainties, when giving evidence to the police or to the court doctors would be wise to acknowledge the limitations in their understanding. They should present all relevant facts in a balanced manner, offer opinions only within their sphere of expertise and take care not to overstate their case. Wrong conclusions in either direction may be disastrous: failure to detect maltreatment can result in the death of another child, while unjustified prosecution can wreck a life and a family. The likelihood of such disasters will be reduced if all SUDI are more appropriately investigated.