Retreating to ivory towers [Reference Bilo1]. Armchair experts [Reference Vedelago and Mannix2]. Pseudo-scientists [Reference Debelle, Maguire and Watts3]. Naysayers [Reference Binenbaum, Forbes, Topjian, Twelves and Christian4]. Nonbelievers [Reference Tully5]. Biased [Reference Dyer6]. Denialists [Reference Hymel7]. Deceitful, dishonest, unethical [Reference Strouse8]. Evil [Reference Fitzsimmons9]. These terms have all been used to describe those who challenge the science behind shaken baby syndrome (SBS), those who challenge the way abuse of infants is ‘diagnosed’.

Name-calling is used to delegitimise those with opposing views and to defend the legitimacy of the orthodoxy [Reference Benet-Martínez and Waller10].Footnote ¹ In psychology, legitimacy is a ‘property of an authority, institution, or social arrangement that leads those connected to it to believe that it is appropriate, proper, and just. Because of legitimacy, people feel that they ought to defer to decisions and rules, following them voluntarily out of obligation rather than out of fear of punishment or anticipation of reward’ [Reference Tyler11].

The defence of the orthodox approach to SBS and the efforts to delegitimise and silence those with opposing views have gone beyond name-calling. Legal threats have been made to pressure editors to retract a properly peer-reviewed journal article that questioned an SBS diagnosis [Reference Brook, Lynøe, Eriksson and Balding12]. An entire edition of an academic journal, which was to publish a debate by proponents and challengers of SBS, foundered after pressure was placed on the publisher [Reference McDonald13]. Experts have been charged with perjury or subjected to complaints to medical boards after testifying for the defence [Reference Hanson14]. They have faced censorship by medical associations. Workplace harassment and threats to employment have been endorsed and encouraged through calls for wholesale sackings and de-registration of doctors who question SBS orthodoxy [Reference Strouse8].

No other contemporary medicolegal issue has seen the level of controversy of SBS. Indeed, no other field of forensic science (or forensic medicine) has seen such controversy nor departed so severely from the norms of scientific discourse. Forty years after writing his 1971 article [Reference Guthkelch15] that hypothesised shaking as a cause of subdural haemorrhage in infants, Dr Norman Guthkelch lamented that, ‘while controversy is a normal and necessary part of scientific discourse, there has arisen a level of emotion and divisiveness on shaken baby syndrome/abusive head trauma that has interfered with our commitment to pursue the truth’ [Reference Guthkelch16]. The dispute over SBS has been described as ‘tribal warfare’ [Reference McDonald13].

1.1 Delegitimising and Silencing Dissent

Perhaps the most famous battle over SBS took place in 1997 in Boston, where Louise Woodward, a 19-year-old au pair from England, was tried for the murder of an 8-month-old boy in her care. Prosecution witnesses, including Dr Eli Newberger, Dr Gerald Feigin, Dr Joseph Madsen, and Dr Patrick Barnes, testified that the infant had been violently shaken. Defence witnesses Dr Jan Leestma, Dr Michael Baden, and Dr Ronald Uscinski testified that the infant had an old, chronic subdural haemorrhage that had re-bled. The trial made daily headlines on both sides of the Atlantic. Woodward was found guilty but the judge who presided over her trial quickly exercised his discretionary authority to serve the interests of justice and reduced her conviction from second-degree murder to manslaughter and imposed a sentence of time served, leading to her immediate release from custody.

After the case the battle spilled into the literature, where supporters of the SBS hypothesis attacked the defence witnesses, calling the possibility of re-bleed a ‘courtroom diagnosis’ rather than a medical diagnosis, and declared that the ‘triad’ of brain swelling, subdural haemorrhage, and retinal haemorrhage were ‘virtually unique’ diagnostic features of SBS [Reference Chadwick, Kirschner and Reece17].

Subsequent to the trial, prosecution witness Dr Patrick Barnes changed his mind about SBS, having moved beyond the child abuse literature to look instead at the science of traumatic head injury written by specialists in those fields. This led to Barnes being included in an ‘axis of evil’ doctors who questioned the SBS orthodoxy, alongside defence witnesses from the Woodward case Leestma and Uscinski as well as others such as Dr Janice Ophoven and Dr John Plunkett [Reference Fitzsimmons9].

In 2001, Plunkett published a seminal article showing that short falls could cause the findings associated with SBS [Reference Plunkett18]. Later the same year, Plunkett gave testimony at the murder trial of Lisa Stickney, a day care worker accused of shaking to death a 14-month-old boy in her care. Stickney claimed that the boy had fallen from a chair. Plunkett testified that such a fall could explain the boy’s death, contradicting the opinions of doctors who had testified for the prosecution. After Stickney was acquitted, the district attorney charged Plunkett with ‘false swearing’. The other defence expert in the case, pathologist Dr John Antonius, was the subject of a complaint to the state board of medical examiners.

Both Plunkett and Antonius were eventually cleared, but at considerable personal cost in terms of time, money, and stress. The prosecutor contacted other prosecutors in SBS cases in other jurisdictions to inform them that Plunkett was being investigated for perjury. Lawyers for Dr Plunkett asserted that the charges were a ‘blatant attempt to discredit Dr Plunkett to prevent him from testifying for the defence’. The message sent to those who challenged the SBS orthodoxy in court was also clear. ‘“It’s scary,” said Chicago neuropathologist Dr Jan Leestma, who had testified for the defence in several SBS cases, “and there comes a point where you start to ask yourself, ‘Do I want to keep on doing this?’”’[Reference Hanson14].

It was also 2001 when Dr Jennian Geddes and her collaborators published two ground-breaking articles that transformed our understanding of the findings associated with SBS [Reference Geddes, Hackshaw, Vowles, Nickols and Whitwell19, Reference Geddes, Vowles and Hackshaw20]. Geddes made detailed microscopic studies of the brains of infants who had been diagnosed with non-accidental injury and found that the majority do not have torn nerve fibres, which had been the assumed mechanism of encephalopathy in SBS cases. Instead, Geddes found that they predominantly had hypoxia, a failure of oxygen supply. This was truly a paradigm shift and showed that both the brain swelling and subdural haemorrhages that had been associated with shaking or trauma could in fact have a range of causes, including non-traumatic ones (see Chapters 3 and 4).

In a 2003 article, Geddes proposed a ‘unified hypothesis’, describing a scenario in which a lack of oxygen caused a cascade of events resulting in not only brain swelling and subdural haemorrhages but also retinal haemorrhages, thus explaining all three elements of the so-called triad of SBS findings [Reference Geddes, Tasker and Hackshaw21]. As a lack of oxygen can have many causes, from cardiac issues to choking, this hypothesis had enormous implications for SBS cases.

Rather than being lauded for her breakthrough scientific achievements, Geddes’s questioning of orthodox views provoked antagonism from doctors from the UK and the United States and she became the subject of forceful attacks in expert reports, lectures, and public presentations [22]. In 2005, an appeal was held against three shaken-baby convictions. During the trial, the prosecutors made a calculated attack on Geddes’s integrity. It was ‘horrendous’, she recalled. ‘I was there for two days being cross-examined. It was an attempt to shut my theory down, to prove my research was rubbish and that I was dishonest’ [Reference Storr23]. During her testimony, Geddes admitted that, at that stage, the unified hypothesis was just that, a hypothesis, and pointed out that the orthodoxy about the mechanism of SBS was also just a hypothesis.

Proponents of SBS claimed that Geddes recanted the unified hypothesis during the 2005 trial, a claim that Geddes has emphatically denied, publishing her denial in the literature in 2009 [Reference Geddes24]. Yet claims that Geddes recanted her hypothesis persist to this day. Proponents also cite the 2005 appeals court judgment as showing that Geddes’s hypothesis should be discounted, as though a legal judgment should be used to inform the opinion of experts. Geddes took early retirement from neuropathology largely because of the toll taken by these attacks.

Efforts to silence Dr Waney Squier, who also questioned the science behind SBS, are further expository. Squier has been described as ‘one of the world’s foremost experts in babies’ brains’ [Reference Rumbelow25] and ‘probably the person [who] has done the best science in pathology [over] the last two decades’ [Reference Aspelin26]. After giving testimony for the defence in which she disagreed with conventional assumptions and questioned the scientific foundation of the SBS theory, police instigated a complaint to the General Medical Council (GMC). The GMC brought Squier before the Medical Practitioners Tribunal, where a panel consisting of a retired military commander, a former policeman, and a retired geriatric psychiatrist heard testimony from a band of doctors who support the SBS orthodoxy. The GMC argued that Squier had ‘failed in her overriding duty to the court to remain objective and to assist the court’. The tribunal sided with the establishment, finding that Squier had provided ‘deliberately misleading’ and ‘dishonest’ evidence in court. The GMC struck her from the medical register.

Considering that the Tribunal’s panel lacked the appropriate educational background to differentiate the claims regarding science, it was clear that the decision rested on the number, authority, and credentials of the witnesses, illustrating a key problem at the interface between law and science: a court’s reliance on authority.

Squier, who had formed her opinions by critically assessing what could and could not be learned from the published studies [Reference Greenhalgh27], appealed to the High Court. In reinstating Squier’s registration, High Court Justice Mitting found that she had not misstated facts from research papers and had not acted dishonestly. Nevertheless, Justice Mitting found that Squier had ‘cherry-picked’ from articles to support her arguments, stating that she ‘did not misstate the facts which she sought to extract from the four research papers; but her citation was so selective as to misrepresent each of [the articles]’ [28].

As an example, Squier had challenged the prosecution’s assertion that retinal haemorrhages were diagnostic of abuse. To support her position, she referred to limitations of that assertion identified in a 2005 study by Matthieu Vinchon. Justice Mitting found that, although Squier had correctly cited from the study, she had omitted ‘the authors’ clear conclusion about the coincidence of abusive injury and retinal haemorrhage in small babies’. The omission referred to was Vinchon’s conclusion that ‘the specificity of … severe retinal haemorrhage for the diagnosis of child abuse was 100%’ [Reference Vinchon, Defoort-Dhellemmes, Desurmont and Dhellemmes29].

Squier was not wrong to challenge the premises underlying these conclusions. No properly trained scientist would accept a result that claims 100% specificity, a big red flag for methodological flaws and circular reasoning. Vinchon himself has admitted that ‘we were a bit disturbed to find a 100% positive predictive value … because this figure does not look like a scientific result; however, from a legal perspective, we think that this is precisely what a judge hopes for’ [Reference Vinchon30]. Indeed, Vinchon had excluded findings related to retinal haemorrhages from his diagnosis model ‘because of the circularity bias’.

Justice Mitting’s reliance on a ‘conclusion’ of the Vinchon study, without comprehending its methodology or the significance of the limitations, illustrates a second key problem at the interface between science and law: judges are not scientists, are not trained to critically appraise scientific literature or clinical publications, and do not understand the fields over which they are adjudicating.

The High Court banned Squier from providing evidence in court for 3 years. The police, prosecutors, and medical establishment had successfully removed a major roadblock to gaining convictions.Footnote ² The ramifications were broader still. Courts in the UK have acknowledged that, subsequent to the Squier case, it became very difficult to find doctors willing to testify for the defence [Reference Baker31].

1.2 The Interface of Science, Law, and Medicine

It is not hard to understand how the court’s problem of lacking scientific expertise leads to the problem of reliance on credentials and authority. Nor is it hard to understand the potential for reliance on authority to become entrenched within fields of forensic sciences. It is therefore not entirely surprising that the science–law interface has long been so problematic.

However, this is not the only clash of cultures at play. Shaken baby syndrome sits not only at the interface between science and law but also at the interface between science and medicine. Squier was opining that, while the GMC and many doctors may consider SBS medicine, it is not science.

But surely medicine is scientific? A history of evidence-based medicine is instructive. In medicine, ‘clinical practice was historically viewed as the “art of medicine”. Expert opinion, experience, and authoritarian judgment were the foundation for decision making. The use of scientific methodology, as in biomedical research, and statistical analysis, as in epidemiology, were rare in the world of medicine’ [Reference Sur and Dahm32]. The birth of evidence-based medicine has been attributed to ‘an increasing awareness of the weaknesses of [these] standard clinical practices’ [Reference Sur and Dahm32]. Dr Gordon Guyatt first called this new movement scientific medicine, which was not well received due to the implication that mainstream medicine was anything short of scientific. Guyatt subsequently referred to the movement as evidence-based medicine.

Haven’t we now moved on from authoritative medicine? Is there still such a divide between medicine and science? While there has been a dramatic increase in scientific methods being applied to medicine, the nature of much (not all) medicine remains different from science. Scientific research is highly specialised, while most medical doctors need to know and synthesise a broad range of topics in order to make diagnoses and provide treatments. By necessity, the training in and learning of these types of knowledge is different.

In order to synthesise and disseminate such volumes of knowledge, a system of medical guidelines exists that aims to guide decisions and provide criteria regarding diagnosis, management, and treatment in specific areas of healthcare. Modern guidelines are authored by a group of experts in each particular field. Although these guidelines should be and often are based on evidence and science, they represent a divide between the science and the doctors who rely upon the guidelines.

1.3 Scientific Reference Standards

So why does Squier believe that, even though SBS is widely accepted by many medical practitioners and authorities, ‘there is no scientific evidence to support it and there never has been’ [Reference Sweeney34]? It boils down to this: studies that claim to support the SBS hypothesis do not have robust reference standards.

In studies of diagnostic accuracy, reference standards are used to categorise participants as having or not having the relevant medical condition. This categorisation must be independent of the diagnostic index test and it must be accurate in making the classifications [Reference Cohen, Korevaar and Altman35]. For example, a biopsy can provide accurate evidence for skin cancer and thus can be used as a reference standard. If one were to use a set of findings such as the evolving appearance, asymmetry, and blurred edges of a mole as a method to diagnose skin cancer, one can determine the accuracy of that diagnostic method by comparing it to the results of the reference standard – that is, the biopsy.

To assess the accuracy of using the findings of encephalopathy and subdural and retinal haemorrhages as an index test for diagnosing abuse, an accurate reference standard is required.Footnote ³ Even when a differential diagnosis is included in the diagnostic process, attempting to exclude all other possible causes, that process needs to be assessed for accuracy, and this requires a reference standard. Controlled experiments cannot be done by shaking infants to categorise infants as abused, a common ethical problem in medicine: one cannot give someone a disease to see how it progresses. Other reliable reference standards are therefore required.

Independently witnessed shaking events represent a possible reference standard, but in the 50 years since the SBS hypothesis was proposed, there are virtually no documented, independently witnessed shaking events that have resulted in the findings associated with SBS. Witnessed shaking events do occur; they just do not result in SBS-associated findings [Reference Thiblin, Andersson and Wester36].

The vast majority of research on SBS instead uses reference standards that invoke the very findings that are used in the diagnostic index test, either directly by using criteria such as ‘no history accounting for patient’s serious head injury’ or indirectly by using conclusions of medical or multidisciplinary teams or results of legal judgments, which have themselves used elements of the diagnostic index test in their determination [Reference Recce and Sege37].Footnote ⁴ Using diagnostic test results as a reference standard for testing the accuracy of a diagnostic test leads to unreliable results based on circular reasoning [Reference Eva Lai Wah, Tz Sung and Nelson38] (see Chapter 10).

Attempts to circumvent these problems have used confessed cases as the reference standard [Reference Adamsbaum, Grabar, Mejean and Rey-Salmon39]. These attempts have been hopelessly inadequate as they ignore the wealth of research into false confessions and interrogation and interview methods. These issues are examined in detail in Chapters 11 and 12.

Again, differences between science and medicine can explain, in part, the different perspectives taken on this methodology. In a medical setting, when there is no sound reference standard, a panel of experts may be invited to assign a reference diagnosis for use in research [Reference Reitsma, Rutjes, Khan, Coomarasamy and Bossuyt40]. The accuracy of such diagnostic methods can then be assessed by following their implications for the treatment and recovery of the patients [Reference Handels, Wolfs and Aalten41]. If the diagnosis was correct, the patient will respond to the treatment, so there is a feedback loop providing information about the accuracy of the diagnosis. In SBS, however, no such feedback loop exists, so there is no way of determining the accuracy of the classification of abused versus non-abused, as made by a multidisciplinary or medical panel. Without such a feedback loop providing a measure of accuracy, the risk of bias and circularity is high.

This problem of inadequate reference standards has been known for decades [Reference Geddes, Hackshaw, Vowles, Nickols and Whitwell19], yet researchers of SBS continue to use the same flawed methods [Reference Sidpra, Jeelani and Ong42].Footnote ⁵ A key to science is the insight that ‘human knowledge is constantly threatened by error. Error may arise as the result of mistakes, false assumptions, entrenched traditions, belief in authorities, superstition, wishful thinking, prejudice, bias, and even fraud’ [Reference Hoyningen-Huene43]. Science has a strong focus on detecting and eliminating these sources of error [Reference Hoyningen-Huene43]. The SBS community, by contrast, has allowed inadequate research standards and the resulting bias to remain the cornerstone of their methodology for decades.

1.4 A Clash of Reviews

In 2016, the Swedish Agency for Health Technology Assessment and Assessment of Social Services (SBU), an independent national body that evaluates methods used by medical and social services, reported that there was only ‘very low quality’ evidence that the ‘triad’ of findings associated with SBS is diagnostic of shaking [Reference Elinder, Eriksson and Hallberg44]. The systematic review had excluded research that did not have scientifically rigorous reference standards. A 2003 systematic review by an Australian physician reached similar conclusions after giving similarly low weight to research with poor reference standards [Reference Donohoe45]. Considering that guidelines for systematic reviews mandate risk of bias assessment for study inclusion, it is not scientifically controversial to assess studies with poor reference standards as having a high risk of bias [Reference Guyatt, Oxman and Vist46]. Yet both the 2003 and 2016 reviews became very controversial in the medical community, which railed against both reports and called for retraction [Reference Debelle, Maguire and Watts3]. This is explored in detail in Chapter 10.

Seemingly in response to the findings of the SBU systematic review, and with an eye on the Daubert standard for admission of forensic evidence in US courts, a consensus statement was released in 2018 in support of the SBS orthodoxy [Reference Choudhary, Servaes and Slovis47]. In contrast to the Swedish report, however, the authors of the consensus statement made no attempt to follow methodological guidelines [Reference Belton, MacDonald, Wright and Hamlin48]. No system was outlined or followed for the creation of the panel of authors. There was no author with expertise in methodology, nor even a methodology section. No quality assessment was made or criteria outlined for the inclusion of studies [Reference Liberati, Altman and Tetzlaff49]. These methodological omissions and flaws mean that the consensus statement is not a document of scientific significance. However, the statement was endorsed by multiple medical societies, which provided it with an air of authority. Indeed, the consensus statement has proven to be a document with medical and judicial significance.

1.5 The Biggest Controversy Surrounding Shaken Baby Syndrome

In courtrooms, believers in the SBS orthodoxy continue to portray a degree of certainty in their ‘diagnosis’ of abuse that is regularly interpreted by judges and juries as being ‘beyond reasonable doubt’. And therein lies the real controversy that surrounds SBS: not the name-calling, nor the retraction of articles, not the intimidation, complaints, and legal proceedings brought against those who question the SBS orthodoxy. The real controversy is that thousands of parents and caregivers throughout the world have been accused of abusing children, been convicted and given long jail sentences or even been sentenced to death, while thousands more have had their children removed from their care based on the testimony of highly credentialed medical expert witnesses who believe in the SBS orthodoxy, but who lack a scientific foundation for those beliefs.

Dissenting opinions are crucial to science. According to Nobel Prize–winning physicist Richard Feynman,

Those challenging the orthodoxy argue that there remains far too much uncertainty in the accuracy of diagnostic methods in shaken baby cases to conclude that abuse has occurred.

This book is a testament to the difficulty of silencing dissenting voices in science. In the book, those who question the SBS orthodoxy provide detailed, compelling reasons to severely doubt the orthodoxy and provide directions for future research to make progress in our understanding of the causes of the syndrome that should be known as retino-dural haemorrhage of infancy.