The dose–effect relationship in PTSD: the South African Constitutional Court Case of AK v. Minister of Police (2022)

Abstract

The decision of the South African Constitutional Court in AK v. Minister of Police has implications for law enforcement agencies that fail the victims of crime. In this matter, the plaintiff sued the Minister and others for damages after officers had failed to rescue her from the perpetrator(s) of a protracted sexual assault and to conduct an adequate criminal investigation afterwards. The judgment deals with a noteworthy psycholegal issue, namely, whether the police are liable for any harm resulting from the plaintiff’s continued trauma exposure beyond the moment a competent search would have likely resulted in her rescue, when her trauma exposure was already prolonged. In this paper, we consider whether the interpretation of the expert evidence is consistent with the relevant research. Research suggests the psychological harm caused by police negligence may have been more than a matter of extended trauma exposure but also of supplemental trauma appraisals.

Keywords:

• PTSD
• dose–effect relationship
• personal injury law
• expert witnesses
• police
• interpersonal violence
• South Africa

Introduction

That interpersonal violence and other awful events can cause significant psychological distress has been known for centuries (Breslau et al., 1998; Ford et al., 2015). However, despite this long history, posttraumatic stress disorder (PTSD) is a relatively new clinical construct that first appeared in the Diagnostic and Statistical Manual of Mental Disorders (DSM) in 1980 (American Psychiatric Association) and was refined for each subsequent edition of the manual (Friedman et al., 2014). The diagnostic criteria for PTSD join a set of debilitating symptoms to an aetiological event, currently defined as the direct or, in some instances, indirect exposure to actual or threatened death, serious injury or sexual violence (American Psychiatric Association [APA], 2022, p. 301). This combination of cause and consequences has had predictable legal implications (Stone, 1993) and explains why PTSD is ubiquitous in civil litigation when plaintiffs seek damages for psychological injuries from those who are liable (Pitman et al., 1996; Smith, 2011).

The paradox of PTSD, according to McNally (2004), is that traumatic events must be cognitively appraised to have any psychological impact—appraisals that are inevitably subjective and shaped by a person’s historical and cultural contexts. However, the DSM committee tasked with revising the diagnostic criteria has at times been concerned that any definition of the stressor criterion that relies on the subjective meaning of traumatic events rather than an objective standard risks the overdiagnosis of PTSD, with a host of epidemiological and forensic implications, not least of which is the fear that traumatised litigants would flood the courts (McNally, 2004). Moreover, the concern that subjective psychological symptoms are easily feigned or exaggerated is why in some legal jurisdictions, compensation for psychological injury is only permissible when accompanied by physical injury (Freckelton, 2015).

In this regard, South African jurisprudence is relatively progressive. The landmark decision of Bester v. Commercial Union Versekeringsmaatskappy, decided in late 1972 and reported the following year, held that a psychiatric injury resulting from ‘nervous shock’ in the absence of bodily injury could be grounds for a delictual claim (Ahmed & Steynberg, 2015). According to the judgment (see Bester v. Commercial Union Versekeringsmaatskappy van SA Ltd, 1972), the initial claim followed a tragic incident in 1969, when a young boy was fatally struck by a car while running across the road just behind his older brother. Though physically unharmed, the older brother suffered psychological symptoms that would most likely be described as posttraumatic stress today. The then Durban and Coastal Division, guided by the legal interpretation of the time and the wording of the relevant Act, found that in the absence of physical injury, delictual damages could not be claimed for the surviving child’s psychiatric injury. However, the Appeal Court disagreed, deciding instead that because the brain and nervous system are part of the same human body, a psychiatric injury is, in fact, a bodily injury (Bester v. Commercial Union Versekeringsmaatskappy van SA Ltd, 1972). Since then, South African courts have accepted that physical and psychiatric (or psychological) injuries may independently warrant compensation. When a claim for psychiatric injury is made, the courts rely on the evidence of psychologists or psychiatrists, who typically render their expert opinions on the diagnosis, causality, prognosis and recommendations for treatment and rehabilitation (Wald & Taylor, 2009).

However, research findings cast doubt on the quality of some psycholegal reports and expert psychological testimony. For example, a survey of South African lawyers, albeit some time ago, reported that psychologists who render this assistance to the courts too often lack objectivity (Allan & Louw, 2001). Adversarial allegiance refers to the tendency of experts to emphasise findings that support the legal case of the party that instructed them in the first place (Murrie & Boccaccini, 2015). Various factors may be at play: selection effects occur as lawyers choose experts who are likely to agree with them, while experts tend to choose cases consistent with their assumptions. Additionally, unconscious cognitive biases (Meintjes-van der Walt & Olaborede, 2019), personality differences (Miller et al., 2011), and personal attitudes to criminal justice (Neal, 2016) all influence expert witnesses’ opinions. Finally, psycholegal work can be lucrative, and experts may have a perverse incentive to assist the legal team responsible for their payment. Worst of all, the ‘hired gun’ phenomenon refers to the perception that some experts unscrupulously present false findings for financial reward, bringing enormous discredit to their profession (Lerm, 2015). Given all of this, it is no surprise that the interrater reliability—a measure of consensus—of forensic evaluations in adversarial contexts appears to be lower than in ordinary clinical contexts (Guarnera & Murrie, 2017).

Additionally, even those well versed in these professional and ethical requirements face challenges. For example, the once and for all rule stipulates that a plaintiff should claim damages for past and prospective loss in a single legal claim (Potgieter et al., 2012). However, the complexity of forensic assessments, which often rely on ambiguous, incomplete and/or historical information, incurs a degree of inherent unreliability (Guarnera et al., 2017). The inevitable uncertainty involved can only increase the likelihood that experts might disagree. This is compounded by the fact that assessment procedures are mostly unstandardised, varying from one expert to the next (Neal & Grisso, 2014). The dilemma for the courts is that because they inevitably lack expert knowledge in specialist areas, they must rely on experts. However, because experts often disagree, particularly in an adversarial legal system, the courts must evaluate and adjudicate expert opinions without the benefit of that specialist knowledge (Meintjes-Van Der Walt, 2003).

Given the concerns about the quality of expert evidence and that courts must adjudicate conflicting expert evidence that it does not fully comprehend, research that identifies and documents the actual psycholegal questions that are considered by the courts, drawing on the relevant research findings, is likely to benefit experts and lawyers alike. In AK v. Minister of Safety and Security and Others (2018), the plaintiff sued the Minister of Police and certain police officers for the alleged negligence of the officers who had failed to rescue her from the perpetrator(s) of a protracted sexual assault and to conduct an adequate criminal investigation following the incident, both of which were alleged to have compounded her resulting psychological trauma. The courts considered whether the Minister, in his official capacity, could be liable for the damages resulting from the trauma exposure that occurred after the point when a reasonably competent search would have most likely resulted in her rescue. Significantly, the traumatic event was already prolonged and severe by the earliest opportunity that the police had to rescue her. Various experts were involved in the litigation, including a psychiatrist and psychologist for the plaintiff and a psychiatrist and psychologist for the defence. First heard in the Port Elizabeth High Court, the case went to the Supreme Court of Appeal and finally to the Constitutional Court. The expert witnesses broadly agreed that while the prolonged exposure was probably deleterious, the effect of this could not be quantified. The correct interpretation of this ambiguous evidence was disputed. In this paper, we isolate the primary psycholegal question evident in the written judgments of the three courts (AK v. Minister of Safety and Security and Others, 2018; Minister of Police v. K, 2020; AK v. the Minister of Police, 2022) and consider the legal decisions are consistent with the relevant research.

The litigation

According to the High Court judgment, the facts of the case are as follows (see AK v. Minister of Safety and Security and Others, 2018). The plaintiff visited Port Elizabeth (now Gqeberha) to conduct business before a scheduled flight back to Johannesburg. With some free time, in the afternoon she visited the beachfront for a walk. However, she was attacked in broad daylight, abducted, taken and held captive in a secluded area, and raped repeatedly throughout the afternoon and night. She missed her flight home, and her family alerted the police. The police found her car that evening and mounted a search with a police dog and then by police helicopter but abandoned their search without finding the plaintiff. The plaintiff escaped the following morning and was assisted by a group of joggers 15 hours after being abducted.

The plaintiff subsequently instituted legal action, arguing that had the police conducted a proper search and an adequate investigation of the crime, they would have spared her significant psychological trauma. The defendant, in response, denied the causal link between the conduct of the police and the plaintiff’s psychological injuries. From their perspective, the unknown person (or people) who attacked her was responsible for the harm inflicted on her. There can be no liability if it cannot be proven on a balance of probabilities that the defendant’s conduct caused the harm (Potgieter et al., 2012) or, at least, as in this case, a portion of the harm.

The High Court found that had the police conducted an immediate search on foot, the plaintiff would most likely have been found within an hour or so; instead, her ordeal was prolonged by almost a third of the total time of her captivity, until her escape early the following morning. The eventual search, using a police dog followed by a police helicopter, was poorly coordinated. Additionally, the criminal investigation fell short of the standards reasonably expected. In particular, the failure to identify and question the bush dwellers living in the area in the days after the attack was, according to the Court, a significant lost opportunity. To date, her rapist(s) has not been found.

After one expert witness testified for the plaintiff, stating that ‘if the police had managed to find her, she would have been spared further trauma’ (AK v. Minister of Safety and Security and Others, 2018, p. 65), the High Court Judge suggested that the remaining three expert witnesses (one for the plaintiff and two for the defence) should meet and compile a joint minute to summarise points of agreement and difference. These experts agreed that the plaintiff had sustained ‘serious psychological and psychiatric sequelae’ and that the ‘prolonged and severe trauma suffered by the plaintiff cannot be divided into sub-units that are quantifiable with any level of psychological or psychiatric validity’ (AK v. Minister of Safety and Security and Others, 2018, p. 64). Moreover, while the ‘prolonged life-threatening incident carries the predominant causative weight in the psychiatric illness’ and the ‘plaintiff’s subjective experience of the quality of the SAPS investigation contributes to, maintains and aggravates the psychiatric illness’, the ‘civil litigation (embarked upon by the plaintiff) is a contributing factor to the poor treatment outcomes to date.’ (AK v. Minister of Safety and Security and Others, 2018, p. 64).

Additionally, evidence by one of the experts for the defendants, quoted in the judgment, was that although the duration was ‘incredibly important’, ‘this only refers to the difference in days or months, not hours’ (AK v. Minister of Safety and Security and Others, 2018, p. 66) and is one of many factors to affect the clinical outcome. Furthermore, the expert is quoted as stating that the plaintiff’s decision to become ‘embroiled in protracted and expensive litigation, constitutes a psychological stressor which could impact on and aggravate her psychopathology’ (AK v. Minister of Safety and Security and Others, 2018, p. 68). Another witness, presumably her treating psychologist, reported the deleterious effects whenever the plaintiff was required to engage with the investigation and the legal case. The defendants argued that the plaintiff was partly responsible for her poor mental health.

With the benefit of the testimony of one of the experts, the joint minute provided by the other experts, and the experts’ reports, the Judge found that the ‘experts agreed that the severity of the plaintiff’s trauma is directly related to the initial injury, the length of time she endured this trauma and the events which took place subsequently’ (AK v. Minister of Safety and Security and Others, 2018, p. 65). Noting that the ‘measure of the damage suffered as a result of the negligent search and investigation does not need to be quantified scientifically’ (AK v. Minister of Safety and Security and Others, 2018, p. 68), the Judge decided that the defendant was liable for 40% of the plaintiff’s damages. The suggestion that the plaintiff contributed to her poor mental health was dismissed.

However, the Supreme Court of Appeal (SCA) in Minister of Police v. K (2020) overturned the judgment, finding that the High Court judgment was not consistent with the actual evidence of the experts, as summarised in the joint minute, who had indicated that the effect of the extended period of exposure could not be quantified. If unquantifiable, then any effect is presumably indeterminable. Additionally, the SCA was concerned about a legal precedent that might result in a flood of litigation. The matter was then heard in the Constitutional Court (AK v. Minister of Police, 2022), which brought finality to the litigation by deciding that even though the attack was already prolonged by the time the police started their search and despite it not being possible to quantify the impact of the additional traumatic exposure that resulted from the failed search, the irrefutable interpretation of the expert evidence was that ‘the longer the exposure, the greater the trauma’ (AK v. Minister of Police, 2022, p. 43). The apex Court confirmed the reasoning of the High Court that causation need not be determined as ‘a matter of mathematical certainty or science, but rather, causation is established on a common sense weighing up of the evidence’ (AK v. Minister of Police, 2022, p. 45). As a result, the order of the SCA was set aside; the plaintiff is entitled to recover damages.

The psycholegal question

That she was psychologically harmed due to the horrific experience was uncontested. Similarly, there was no dispute that the actual or perceived inadequacy of the investigation had aggravated her psychological injury. Instead, the primary psycholegal question is the extent to which a traumatic event is more damaging as a result of being prolonged in duration. Expert witnesses are necessarily constrained by the requirements of the court; in particular, their evidence should not exceed the logic dictated by their scientific knowledge. Therefore, a review of the relevant research findings is necessary to consider this logical boundary.

Discussion

It has long been assumed that the dose–effect relationship in PTSD is that the stressor’s magnitude correlates with the severity of symptoms (March, 1993). Indeed, from the start, DSM-III stated that ‘the severity, duration, and proximity of an individual’s exposure to the traumatic event are the most important factors affecting the likelihood of developing this disorder’ (American Psychiatric Association, 1980, p. 426). Direct experiences of traumatic events are more pathogenic than indirect experiences (Breslau et al., 1998). Interpersonal violence tends to carry a greater risk of PTSD than natural events (Breslau et al., 1998). Several studies have shown that of various forms of interpersonal violence, rape has the strongest association with PTSD among women (Kaminer et al., 2008; Kessler et al., 1995). Moreover, being raped more than once during an assault confers an additional risk of PTSD (Tiihonen Möller et al., 2014). Despite these findings, empirical support for the purported relationship between event magnitude and outcome is inconsistent (e.g., Rosen & Lilienfeld, 2008). Certainly, studies that have attempted to describe this dose–effect relationship face methodological challenges. Unlike in experimental animal studies, the accurate measurement of the magnitude of traumatic exposure is a challenge, compounded by the fact that such measurements typically rely on unreliable retrospective accounts (McNally, 2004). The magnitude of traumatic exposure differs on various dimensions, including physical proximity to the event, social connection to those affected, temporal duration of the traumatic event, number of exposures, subjective perceptions of life threat, and actual physical injury (Wilson, 2014). Additionally, any dose–effect relationship is entangled amongst various pre-, peri-, and post-traumatic variables that impact the severity and course of a person’s psychological response to a traumatic event (Brewin et al., 2000; Ozer et al., 2008). The clinical outcome of a traumatic stressor is multifaceted, with exposure duration being possibly just one such factor.

However, the issue may well be more than a matter of lengthened duration. In addition to prolonging her traumatic exposure, one might consider the police failings as stressors in and of themselves. The judgment noted that the plaintiff had ‘struggled to come to terms with the way the case was investigated’ (AK v. Minister of Safety and Security and Others, 2018, p.68). For many who report their sexual assault, the experience of the legal and health systems is a significant form of secondary victimisation (Campbell, 2008). What happens during and after the event can shape the meaning taken from the experience. A leading contemporary model proposes that posttraumatic stress becomes persistent when individuals process the trauma in a way that produces a sense of serious, current threat (Ehlers & Clark, 2000). The sense of threat arises as a consequence of excessively negative appraisals of the traumatic event and its sequelae, and a disturbance of autobiographical memory (Ehlers & Clark, 2000). Counterproductive behavioural and cognitive coping strategies prevent reappraisal and trauma memory processing (Ehlers & Clark, 2000). Significant, therefore, is the role of the person’s idiosyncratic trauma appraisals (see also Creamer et al., 2005). Typically, trauma appraisals convert a usually time-limited event into an experience with globally negative implications for the person’s future (Ehlers & Clark, 2000). These appraisals can be about the fact that the trauma occurred, the person’s behaviour during and after the traumatic event, their initial symptoms of traumatic stress, other people’s reactions, and other consequences and sequelae of the trauma event (Ehlers & Clark, 2000). In this case, the appraisals resulting from the police’s failure to rescue her or to take her criminal complaint with the seriousness it deserved may imply something negative about her relative worth as a person. Whatever the content of her appraisals, from the perspective of the Ehlers and Clark model, the experience of being assaulted, abducted and raped and the experience of being failed by the police at a time of greatest need are each likely to have resulted in distinct trauma appraisals, which, together, may have exacerbated and complicated her resulting PTSD. Put differently, the combination of experiences—the sexual assault and the failings of the police—is more likely to ‘shatter’ fundamental, necessary assumptions about her world (Janoff-Bulman, 1992).

The relationship between trauma appraisals, duration of exposure, and the risk of PTSD has been explored in a study of victims of sexual assault. Kaysen et al. (2010) found that while there was a small but significant relationship between the duration of the stressor and a measure of PTSD symptoms in the aftermath of the sexual assault, the intensity of trauma appraisals predicted PTSD symptoms three months later. While the study was interested in the participants’ perceptions of threat during the sexual assaults rather than their appraisals resulting from the trauma sequela, the findings lend further support to the suggestion that subjective trauma appraisals mediate the severity of PTSD symptomology. Taken together, the research findings reported are, for the most part, consistent with the plaintiff’s ultimately successful claim that the police failures exacerbated or otherwise complicated her psychological injury, though a key mechanism in the relationship between the conduct of the police and the plaintiff’s psychiatric injury may have been missed.

Conclusion

In AK v. Minister of Police (2022), the South African Constitutional Court established a legal precedent by recognising a dose–effect relationship in PTSD. This important decision holds significant implications for the South African police, often found wanting (Lamb, 2021) in a country sometimes described as the rape capital of the world (Jewkes & Abrahams, 2002). Though a search was conducted and an investigation followed, and even though the plaintiff had already experienced severe psychological trauma before the police could respond, their efforts were deficient, and they are liable for their contribution to her psychological injury.

The psychological research findings reported here are, for the most part, consistent with the decision of the Constitutional Court that the negligence of the police impacted the severity and/or complexity of the plaintiff’s psychological trauma. Significantly, the impact may extend beyond merely prolonged suffering, encompassing supplemental trauma appraisals.

Though research suggests that the conduct of the police is likely to have had an impact on the plaintiff’s psychopathology, factual causation is just one of several prerequisites of legal causation. Liability in personal injury law is limited, among other things, by the proximity of the causal connection between the defendant’s actions and the resulting harm (Potgieter et al., 2012). The decision arguably extends this boundary and broadens the circumstances in which plaintiffs may seek damages. The Minister of Police contended that such an outcome would have a chilling effect on the ability of the police service to conduct its work (AK v. Minister of Police, 2022).

However, the judgment sets parameters for the circumstances in which organs of state can be held liable for compounded traumatisation. Significantly, the High Court characterised the police errors as both serious and significant (AK v. Minister of Safety and Security and Others, 2018). Their conduct is likely to have undermined public trust in law enforcement, another important consideration (AK v. Minister of Safety and Security and Others, 2018). The implication for the police and other essential organs of the state is that even when the factual causal connection is remote, serious and significant failures that undermine the public’s trust, leading to further traumatisation, may be actionable. In a context marked by pervasive gender-based violence, the police are obliged to employ all reasonable and appropriate measures to safeguard the constitutional rights of women (AK v. Minister of Police, 2022).

In the absence of the arrest and conviction of the person (or people) who attacked her, the civil court provided the only legal redress available to the plaintiff. The outcome, after more than a decade of litigation, is a vindication of her dignity and rights. The contribution of psychology and psychiatry to personal injury law remains an important aspect of the administration of justice in South Africa and elsewhere.

Ethical standards

Declaration of conflicts of interest

Charles Young has declared no conflicts of interest.

Mohammed Nagdee has declared no conflicts of interest.

Amanda Pieterse has declared no conflicts of interest.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.