12th Greek Australian Legal and Medical Conference
Samos, Greece 2009

INFANTICIDE, PSYCHIATRY AND THE LAW; A REVIEW OF CURRENT KNOWLEDGE AND PRACTICES

Dr Lanny Bochsler MBBS, MMed (Psych), FRANZCP

INTRODUCTION

The murder of an innocent child by a mother challenges society’s expectations of women as carers and nurturers, provoking complex reactions ranging from sympathy to anger and calls for rehabilitation to demands for retribution.

While there is much overlap between the various forms of child murder, infanticide occupies a unique position within Australian law (and other jurisdictions) in formalising psychiatric illness as a factor mitigating responsibility on the grounds that it is directly causal to the death of the infant. Though infanticide has been related to a range of socio-cultural factors and motivations, research consistently implicates mental illness as the major risk factor for infanticide in Western societies.

Accordingly, this paper:

DEFINITIONS

The terms infanticide, neonaticide and filicide are often used interchangeably but are distinguishable. The more general term of filicide refers to the murder of a child at any age. Infanticide, derived from the Latin infanticidium, describes the practice of an individual intentionally causing the death of a child (usually) under the age of one, whereas, neonaticide refers to the killing of a child within 24 hours of birth [1].

HISTORY [2,3,4,5, 6]

Recognised since paleolithic times, infanticide has taken many forms throughout history and within different cultures, including cannibalism and the sacrificing of infants to pagan gods in ancient times (most notoriously in Carthage). While such practices were considered barbaric in Ancient Greece and Rome, the exposure of unwanted newborns to the elements was common.

In certain cultures, there was formal sanctioning of such practices by the laws of the time - from patria potens in Ancient Rome, which granted the father the right to dispose of his offspring as he saw fit, to the adoption of “stubborn child laws” by the Massachusetts colony in the 1650’s, which allowed for the execution of disobedient children.

Rejected by the major religions of Judaism, Christianity and Islam, infanticide nevertheless persisted on a wide level through the ages with female gender, economic factors and congenital abnormalities as the primary motivations. This occurred despite the instigation of legal censures within many countries in an attempt to prevent infanticide. However, in response to concerns regarding the severe punishment inflicted on these often desperate and exploited mothers, countries began to recognise a distinction between infanticide and other forms of murder by assigning more lenient sentences. Russia, in 1647, became the first country to implement such an approach, and by the 19th century most European states had taken similar positions.

With the onset of the 20th century, a new perspective on the crime of infanticide was introduced, that of psychiatric illness. Within the United Kingdom this notion was embraced with the passing of the Infanticide Act of 1922 following the high-profile convictions of a number of patently mentally unwell women. The enactment of this legislation resulted in infanticide becoming a lesser crime for which probation and treatment were the usual outcome. This act was subsequently amended in 1938 [7] to increase the age limit of the victim to 12 months from the previous “new-born child”. “Lactation” was also added as grounds for mental disturbance. This allowed the medical basis for excusing infanticide to be expanded beyond the first few weeks of a child’s life:

Where a woman by any wilful act or omission causes the death of her child – aged less than a year – but at the time the balance of her mind was disturbed by reason of her not having recovered from the effect of giving birth to the child or by reason of the effect of lactation…the offence, which would have amounted to murder, is deemed to be infanticide and is dealt with and punished as of it were manslaughter [7]

This statute, which continues to be in use today, has formed the basis for most infanticide laws, with over 20 countries (including Australia, Brazil, Canada, Italy, Japan, India, Turkey, Norway, New Zealand, and Greece, but not the United States) having adopted similar provisions.

EPIDEMIOLOGY

Prevalence

While infanticide is perceived to occur infrequently in the Western world, an individual remains four times more likely to be the victim of homicide in the first year of his or her life than at any other time [5]. This risk is greatest on the first day of life, with a further 10% of filicides in the first week, and 30% within the first year of birth [8]. Unlike other forms of child homicide, within the first year of life, a child is most likely to die at hands of their mother, but after the age of one, death by a male becomes more likely (usually the father, stepfather or male companion of the mother) [9].

Currently the United States has the highest reported rate of infant homicide in the Western world of 8 per 100,000 population, which equates to an estimated death of one infant under the age of one per day.

This compares with 4.5 per 100,000 for England and Wales, 4.3 per 100,000 for Scotland and 4.0 per 100,000 for Australia [10]. Emerging evidence [11] suggests that these figures may significantly underestimate the extent of maternal infanticide as a result of underreporting, inaccurate coronial rulings and some bodies having never been discovered. Controversially, some authors [12] have attributed a large proportion (up to 20%) of Sudden Infant Death syndrome deaths to homicide.

Once a prevalent public health problem, neonaticide is now uncommon in the Western world with only a handful of cases reported in each country annually [13], although precise figures remain elusive. The widespread use of contraception, the relaxation of the abortion laws and changes in society’s attitudes towards single mothers have been speculated as the reasons behind this decline [13].

Risk factors for infanticide

Research suggests maternal psychiatric illness is the primary risk factor for infanticide [3,5,11,14] and is implicated in up to 60-80% of deaths [9]. The limited research available has also highlighted the additional risk factors of a lack of prenatal care, low levels of education and being the second or subsequent child of a mother aged less than 19 year-old, which has especially been associated with infanticide beyond the first day [15].

The risk factors related to neonaticide appear to be distinct from infanticide of a child beyond their first day [1, 16]. Mental illness is not considered to be a major contributory factor [1]. Mothers who commit neonaticide tend to be younger (late teens to early twenties), unmarried, immature, and have few, if any plans, for the birth and care of their child [17]. Their pregnancies are usually concealed and denied up to the point of birth, although this might ebb and flow throughout the pregnancy. Denial may also extend to the parents and partner of the birth mother [18]. In addition, there is often a history of intermittent amnesia, a family background of emotional neglect, role confusion and a strong association with past abuse of the mothers, with reported rates of up to 50% [3]. In contrast, those who commit infanticide beyond one day of age are usually older and often married. Psychologically they tend to have higher rates of depression, psychosis, suicidal ideation, and suicide attempts [17].

Many of the typical features of neonaticide are highlighted in the case below:

Ms C: a 17 year-old high school student from a small town in rural Victoria. On returning home from work one day her mother found a dead infant wrapped in bloody clothing in the washing basket. C had given birth alone on her hands and knees, cut the umbilical cord with scissors, wrapped the baby in a blanket and punched it repeatedly to “make it go away”. The pregnancy had been concealed from her family and the father (a fellow student). She reported having not thought about the baby during the pregnancy but was now distressed and remorseful over what she had done. There was no evidence of depression or psychosis. She was subsequently convicted of infanticide and placed on probation.

Methods

In a review of 132 cases of child death in the United States [19] the predominant causes of death recorded were head trauma, throwing the child from a height, strangulation and suffocation. These methods were equally common to both female and male perpetrators.

CLASSIFICATION

“There are many forms of infanticide, each with different causes and manifestations” [13]

Many classification systems have been proposed in an attempt to identify, categorise and understand the range of reasons behind child murder. Resnick [16], following a review of 88 cases of maternal filicide, proposed a seminal typology based on the motive of the offending parent:

Subsequent classificatory systems have generally been refinements of Resnick’s approach. D’Orban [20] focused on categorising women with regards to the source of the impulse to kill the child (i.e. parent, child or situation) and included the phenomena of “battering mothers” which, under this model, was attributed to the victim’s behaviour. Bourget and Gagne [21] proposed a model accounting for intent, motive, psychiatric illness and associated circumstances in an attempt to identify each case as predictable or unpredictable.

The above systems relate to the more general crime of filicide, as such their generalisability to infanticide is uncertain. A contemporary typology specifically addressing infanticide has been proposed by Oberman [3] and consists of 5 broad categories: neonaticide, assisted / coerced, neglect-related, abuse-related and mental-illness-related types. While again similar to Resnick’s, this system emphasises neonaticide as a unique subtype. In addition, it seeks to subtype mental illness-related infanticide into acute and chronic forms.

PSYCHIATRIC ILLNESS AND INFANTICIDE

Psychiatric disorders in the postpartum have been noted since the time of Hippocrates, though the notion of a causal link between pregnancy, birth and the development of maternal mental illness was only popularised in the late 19th century by the French psychiatrists, Esquirol and Marce [22].

Subsequently, research has highlighted a number of additional predictors for the development of postpartum psychiatric disorders, including: a history of child abuse, family psychiatric history, poor parenting role models, unstable/ or absent partner, rape or sexually transmitted diseases history, drug abuse and poverty [8].

Postpartum mental illness is common, with up to 40% of women reporting significant emotional disturbance during this period [8] with the initial three months following birth representing the peak life-time prevalence period for mental illness in women [23, 24]. This key period also corresponds with 58% of maternal infanticides [3].

Most relevant diagnostically to the issue of infant harm are the acute disorders of postpartum-onset depression and psychosis. The relapse related to the puerperium of well-established chronic mental illnesses (especially schizophrenia, life-long depression and bipolar disorder which, if left untreated, has a relapse rate up to 70% [8]) has also been strongly associated with infanticide [3]. Substance abuse and personality disorders have been noted as common co-morbidities (with reported rates of up to 67% [25]) and the role of mother-infant relationship disorders (related to the maltreatment of infants) and dissociative disorders (due to their reported association with neonaticide) have been a focus of increasing research interest. In a study of psychiatric diagnosis in maternal infanticide, Resnick [19] reported the following rates: schizophrenia 22%, psychosis (other) 24%, no diagnosis 22%, personality disorder 12%, severe depression 11%, bipolar disorder 2%, intellectually disabled 2%, neurosis 2%, delirium 2% and epilepsy 1%.

Postpartum depression

Depression in the postpartum period occurs in approximately 15% of women [8]. This disorder is distinguishable from the more prevalent “baby blues” by the persistence, range and severity of the symptoms, but is otherwise similar to depression occurring in other contexts. Common features include; disinterest, sleep problems, fatigue, feelings of guilt, hopelessness and inadequacy, thoughts of dying and suicide, and a lack of love for the new child. In severe depressive states, psychotic features can emerge and can involve delusions of possession (i.e. being controlled by others) and persecution, hallucinations, confusion and stupor [26].

Data on the predictors of depression-related infanticide are scarce, although the presence of psychotic symptoms should be a cause for alarm. Research suggests that non-psychotic depressed women are unlikely to commit infanticide but if they do, their motive is likely to be altruistic [9] as described in the following vignette:

Ms A: young, newly married Australian mother, she became severely depressed 3 weeks after giving birth to a healthy baby daughter. During this time she experienced strong feelings of believing that she was a bad mother, obsessional thoughts about harming her baby and increasing suicidal ideas. Believing that her baby could not safely survive in the world without her, she held her in her arms when she jumped in front of a train.

Postpartum (puerperal) psychosis

Occurring in 1 in 600-1000 births, this group of disorders can be due to a diverse range of causes, though research suggests a strong link with manic-depression in psychosis with an acute postpartum-onset (present in up to 90% [26]). Postpartum psychoses have been reported to differ from other psychotic episodes because of alterations in cognition and confusion. While this syndrome (also referred to as cycloid or acute polymorphic psychosis), has long been connected to childbirth, it also occurs in more than 10% of non-puerperal psychotic patients [13].

In relation to infanticide, psychosis is the most significant psychiatric diagnosis and is present in up to 70% of perpetrators. Delusions regarding the child or child’s safety, as well as command hallucinations, represent the main symptoms of concern [3]. Compared with non-psychotic women who kill their children, psychotic women tend to be older and more educated, have a history of substance abuse, psychiatric hospitalisation and treatment, and suicide attempts [25]. They are also more likely to kill multiple victims, use a weapon to kill their children and lack premeditation, with the development of delusions only days prior to the offence. This is in contrast to non-psychotic women who often reported thinking about their children’s death days or weeks beforehand [12].

Other disorders

The British peri-natal psychiatrist, Brockington [13] has proposed that severe mother-infant relationship disorders (also known as attachment disorders), leading to hatred and rejection of the infant, should be psychiatric considerations in understanding the fatal abuse and neglect forms of infanticide. This position is controversial as such disorders are not established diagnostic entities.

The role of mental illness in the perpetrators of neonaticide is likewise contentious. Depression and psychosis are rare, with fear and desperation as commonly cited explanations [14]. Other authors have argued for dissociative states related to denial of the pregnancy as a key factor. In a review of sixteen women charged with neonaticide, Spinelli [3] found high scores on dissociation scales suggesting possible dissociative states at the time of the event.

This has led to calls for “depersonalisation disorder” as a defence [3]. As is the case with other criminal acts, dissociative pathology has been put forward as a defence to reduce responsibility, at times arguably in advance of accepted medial knowledge. In particular, determining the threshold with regards to culpability for such syndromes is unclear; should people overpowered by aggressive impulses, claiming that they feel “taken over” and distanced from what is happening also qualify for a specific defence against homicide?

CRIMINAL LAW DEFENCES

Dating back to ancient Greece and Rome, it has been acknowledged that those suffering from certain mental disorders should not always bear the full weight of responsibility for their actions. Under contemporary criminal laws, a guilty state of mind (mens rea) is crucial in order to justify a conviction. This element is negated with the successful application of the defences of insanity and diminished responsibility (including infanticide-specific statutes).

Research suggests that many cases of infanticide are perpetrated not by mothers methodically and callously disposing of their children, but by grossly unwell women (often in the grip of psychosis). The question of their criminal responsibility is inevitably and appropriately raised. On the other hand, as for any other crime, mental illness does not de facto equate to “not responsible”. These central questions have been addressed differently depending on the jurisdiction with differing legislative approaches and implications for the mother.

The predominant legal models in Western societies are infanticide-specific laws (as in Victoria) and the insanity defence (as in the United States). The more general partial defence of diminished responsibility, which if successful, results in the reduction in charges from murder to manslaughter, is an alternative legal approach in some jurisdictions.

Infanticide laws

In the Australian state of Victoria, infanticide specific laws were first enacted in 1958 and closely followed the English Infanticide Act (1938). Under the Crimes Act 1958 (Vic) [27], infanticide could be both an offence and an alternative verdict to murder (i.e. partial defence). This dual arrangement occurs in most of the states of Australia where infanticide laws are present (i.e. Tasmania, Western Australia and N.S.W.). Much like the English Infanticide Act (1938), infanticide is defined as occurring when a woman kills her child – aged less than 12 months old - due to a disturbance of mind, which is caused by the effects of either childbirth or lactation. A verdict of infanticide in Victoria limits the maximum sentence to 5 years jail, although probation is the most common outcome [28]. This is consistent with other countries with infanticide statutes, for e.g., in the past 50 years no woman in the U.K. convicted of infanticide has been incarcerated [3].

In 2004, the infanticide statutes were the focus of a review by the Victorian Law Reform Commission. This review resulted in a number of recommendations [29], subsequently incorporated into the Crimes (Homicide) Act 2005. In particular, the connection between childbirth, lactation and mental disturbance was removed due to a lack of evidence for an association between breast-feeding and mental illness, and replaced by; “the balance of her mind disturbed because of:

In addition, the age limit was extended to two years, reflecting that while the vast majority of mothers kill their children within the first year of life, the limit potentially disadvantages deserving mothers.

Insanity defences

Unlike Victoria, the United States does not recognise any distinction between homicide and infanticide, with defendants reliant on the more stringent insanity defences, with diminished responsibility defence an alternative in some states.

This lack of infanticide-specific laws relates, in part, to the American legal systems divergence from English law prior to the enactment of infanticide provisions, but also to the lack of a nationwide approach to homicide laws in general, due to its federalist system of largely autonomous states.

The main formulations of the insanity defence used in the United States are the M’Naghten Test [29] and the Model Penal Code/American Law Institute Test [30], which both require the defendant to show that they did not know what they were doing was “wrong” at the time of the act due the effects of a medically recognised “mental disease”. As such they are both considered essentially “cognitive” tests, although the Model Penal Law is somewhat broader in its inclusion of a volitional component (i.e. the ability to control one’s actions), which, if satisfied, is also grounds for an insanity verdict.

Criticisms of infanticide laws

Despite being widely adopted and perceived by many as a more humane approach, infanticide laws are not without criticism. These include:

Infanticide laws have been criticised as being unnecessary in view of the existence of the defence of mental impairment or the more general partial defence of diminished responsibility [28, 31]. As with the infanticide defence, both can also lead to enforced non-custodial supervision and treatment. Furthermore, infanticide laws are in reality a form of diminished responsibility restricted to a very specific group (i.e. biological mothers), with others excluded if their psychiatric disturbance falls short of full mental impairment. Diminished responsibility laws, if present, would allow for reduced sentences for these other groups (i.e. fathers, adoptive parents) if they are able to show an “abnormality of the mind”.

Alternatively, forwarded as a pragmatic reason to maintain rather than merge the laws, infanticide laws generally result in greater leniency towards offenders compared with diminished responsibility verdicts related to other crimes [32]. Infanticide laws also have the advantage over diminished responsibility laws of allowing the prosecution to proceed on a charge of infanticide and not murder, thereby conceding mental disturbance and relieving the defendant of the need to prove such a state, as well as avoiding the stigma associated with being charged with murder.

The presence of a unique homicide offence with a lesser penalty for the killing of an infant has been argued to devalue the life of that child. Likewise, the lack of a requirement in infanticide laws to establish a causal relationship between mental disorders and the killing of a child has seemingly led to a per se defence to any killing that occurs in infancy, despite evidence suggesting that women convicted of infanticide (particularly in cases of neonaticide) sometimes do not have significant mental illness as technically required by the law, and instead suffer from “social stresses and personality problems” [20].

The availability of a defence dependent on the age of the victim would appear arbitrary. This concern is highlighted in the two following scenarios:

Criminal responsibility is a legal rather than psychiatric concept and since the 19th century courts have increasingly relied on the assistance of mental health professionals to provide clarification and guidance, with infanticide being no exception.

While infanticide laws are predicated on the belief that childbirth can cause mental disturbance, which results in some women killing their children, the nature and extent of the connection remains controversial. Formal classificatory systems, such as DSM-IV TR [33], continue to not recognise postpartum disorders as unique conditions and research findings regarding a biological basis to postpartum syndromes have been inconclusive (despite findings of various neurohormonal changes [10], including rapid fluctuations of oestrogen, progesterone and other gonadal hormone levels during pregnancy and the precipitous loss of these hormones at birth, potentially acting as triggers to central nervous system neurotransmitter changes [3]). A further complication in establishing a unique biological basis to postpartum disorders is the increasing awareness of the role of external aggravating factors (for e.g. the role of familial and spousal support [35]).

The lack of recognition by the medical model underpinning infanticide laws of other potential factors relevant to the killing of a child has also been central to concerns of gender bias. Despite similar rates and types of psychiatric disturbance, as well as the presence of many of the same psychosocial stressors in perpetrators of both genders [14], no specific infanticide laws are available to men. Instead they remain reliant on insanity and diminished responsibility defences:

Mr D: a 35 yo married Australian father of a 5 month old daughter was convicted of murder and sentenced to 9 years jail after battering her to death. The pregnancy had been unplanned and unwanted and the couple were under significant financial strain.

It can be reasonably argued that based on Victorian legal precedent, if Mr D had been a woman, incarceration would’ve been unlikely.

Criticisms of insanity defences

The restrictiveness of the criteria of the mental impairment defence has been cited as a major criticism of the law, as well as a supportive argument for broader diminished responsibility-type defences (including infanticide). People with significant mental conditions who fall short of the current M’Naghten-based mental impairment defence are convicted of murder. This is arguably unjust because it fails to recognise the reduced culpability of mentally impaired people.

This appears particularly relevant to infanticide, as many mothers recognise the moral wrongness of their acts despite being undoubtedly mentally unstable.

The use of insanity defence in the setting of infanticide, particularly in the United States, has resulted in much controversy, with a number of recent high profile cases highlighting the wide disparity and apparent arbitrariness of outcomes under such statutes. These concerns are demonstrated in the two cases compared below, one of which includes the case of Andrea Yates, probably the most notorious recent example of maternal infanticide. Both cases highlighted occurred in Texas where the insanity defence is derived from the M’Naghten Rule with the defendant having to prove failure to know the act was “wrong”. However, the definition of “wrong” is not clear (i.e. whether morally or legally) and instead left to the jury to determine:

Yates, a married nurse in her early 30s with a past history of a previous untreated depressive episode and a family history of depression and bipolar disorder. Over the course of seven years, she bore 5 children, with each pregnancy followed by psychiatric problems, including mood and psychotic episodes and two suicide attempts following her fourth child. Despite medical advice to not have any further children (ignored by her devoutly Christian husband) she became pregnant for the 5th time. Following the birth she was subsequently admitted to hospital in a near catatonic state. After being discharged, though barely speaking and with no additional childcare supports in place, her anti-psychotic medication was ceased due to unclear reasons. A few weeks later she drowned her five children.

At the time of the event, she stated that Satan had directed her to kill her children in order to save them from the “turmoil of hell” and that if they were not killed before the “age of accountability” one would become a serial killer, one killed horribly and one become gay. Psychiatrists on both sides agreed that the delusions led her to believe that killing them was in their best interests but also that she knew what she had done was wrong. The prosecution argued that she knew it was legally wrong because she called 911. The prosecution psychiatrist (who had never seen a case of puerperal psychosis) testified that she must have seen a recent Law & Order episode where the woman “got off” after she killed her child due to postnatal depression.

After a brief deliberation by the jury she was convicted of murder and sentenced to life in jail. Following a public and professional outcry regarding the outcome and the conduct of the prosecution psychiatrist the case was eventually retried which resulted in a verdict of Not Guilty by Reason of Insanity (NGRI); “we understand that she knew it was legally wrong…but in her delusional mind, we believe that she thought what she did was right”

The second case concerns that of Deanna Laney:

A 38 year-old married mother of 3 with no history of mental illness, she reportedly battered her two older children to death with a stone and left her baby permanently brain damaged because “God told her to as a test of faith”. Like Yates she had also called 911 after the act. Throughout the trial, pictures of the dead children were constantly on display and in summing up the prosecutor took a 14lb stone and hit it on the table repeatedly. All 5 psychiatrists involved supported her being found NGRI.

After initially voting 8 to 4 to find her guilty, the jury returned a verdict of NGRI as “everyone knew that God would never say that”.

Although there was much similarity in the clinical circumstances of each case, with both women psychotic and both having called 911, the outcome differed greatly with the main determinant appearing to be whether God or Satan had told the mothers to kill their children.

These two cases highlight some of the frequently cited reasons for the disparate outcomes seen in the US:

Such divergent outcomes are not uncommon in the United States, with sentences varying considerably between jurisdictions; from probation to lengthy jail time (and even capital punishment) for seemingly equivalent crimes. Compared with overseas, jail time is far more frequent in the United States; in a study of the outcomes for perpetrators following infanticide [19], 68% of mothers had been hospitalised, while 27% were in prison or on probation and 5% had suicided.

The final major criticism of the insanity defence, but also the United States’ legal approach in general, relates to the question as to the purpose of any legal response; what do we seek to gain as a society? The primary emphasis placed on punishment within the United States does not seem to be justified in the case of infanticide.

MacFarlane [35] notes that the three basic justifications for punishment are deterrence (both general and specific), retribution and rehabilitation. General deterrence refers to the notion that punishing a given defendant will serve to dissuade others who might contemplate committing infanticide. In view of the desperation that is often apparent and the sudden and impulsive nature of the act there is little to believe that this applies. Specific deterrence appears relevant only for certain categories, in particular where there has been prolonged abuse. In this circumstance, jailing the mother might influence her treatment of other children in the future.

The second major justification is retribution. This rationale is longstanding and based on the idea that society can punish one who unreasonably harms another. However, the presence of severe illness challenges the notion of the mother’s actions as undeniably unjustified, particularly when the mother believes that they may be doing good. Furthermore, this argument requires clear lines of blame to be present which is not always the case, with families and clinicians at times seemingly bearing some responsibility.

Lastly there is the justification for punishment of rehabilitation. Taking into account the deplorable state of much of the prison system in the United States where mental health treatment is scarce, this argument is difficult to support.

PREVENTION AND MANAGEMENT

In view of our awareness of a trigger (i.e. pregnancy and childbirth) and the well-established high prevalence rates of the major risk factor (severe psychiatric illness), the potential for prevention of infanticide appears high. Likewise, neonaticide, once a widespread societal problem has diminished to an intermittent, although tragic, event, providing further cause for optimism.

Despite these encouraging aspects, the prevention of infanticide remains problematic. While some specific psychiatric features, as detailed above, have been established as conferring a greater risk they rarely provide a full explanation, with the role of contextual and social factors outside of historical reasons only starting to be clarified, for example, increased rates of infanticide have also been associated with economic stress in Western nations [36]. In addition, many of the recognised risk factors, such as maternal depression and social advantage, are common among non-filicidal mothers. Finally, there remains a dearth of research addressing interventions and strategies appropriate for the continuum of antenatal and postnatal services. Nevertheless, a number of recommendations would appear reasonable in view of the available knowledge.

Identifying those at risk

As mental illness represents the major amenable risk factor, the early recognition and assessment of psychiatric illness, both antenatally and postnatally, should be a priority. This is especially the case with postpartum psychosis as up to 4% of untreated mothers commit infanticide [37]. All assessments should include an exploration for altruistic ideation or acute psychotic symptoms which may be present in mothers who are psychotic, depressed, manic or delirious. Additionally, asking about childrearing practices, parenting problems and feelings of being overwhelmed may elucidate crucial information in those reluctant to share their delusional ideas. However, with the reduction in the average hospital length of stay after delivery, many cases may go undetected. This can be potentially countered by the adequate provision and training of postnatal support services, the use of screening tools (for e.g. the Edinburgh Postnatal Depression Scale, which is easy to administer, reliable and valid), and education of the community. The case below highlights many of these important aspects related to identifying those at risk of infanticide:

Mrs G: 32 year-old Filipino mother of a two year old, living in Australia with her husband. She delivered her second child uneventfully but was noted by Maternal Child Health Nurse to be anxious about breastfeeding and recommended that she see her General Practitioner (GP). Her husband was also reportedly worried about her and took her to the GP but didn’t give much history. The GP felt that she might have been depressed and prescribed antidepressant medication. Two days later she drowned her two children and then hung herself.

Later it emerged that the husband had told the GP that she had been “behaving strangely” and that she had in fact made three bizarre and out of character suicide attempts; strange behaviour in the first month postpartum should be considered postpartum psychosis until proven otherwise and represents a psychiatric emergency [4]. Furthermore, her past history and family history had not been recorded; this woman had had a previous psychiatric admission and her father had suicided, both significant risk factors for postpartum mental illness.

Early identification of warning signs would also appear to be a potential means of preventing deaths from fatal maltreatment, particularly due to violence. Bennie and Sclare [38] noted up to 50% of parents had seen a doctor immediately preceding the crime, in some cases to complain about the child’s behaviour and to seek help in management.

Once those at risk due to mental illness have been identified, ensuring appropriate intervention becomes a necessity. While a detailed discussion of the treatment of perinatal disorders is beyond the scope of this article, ensuring adequate monitoring and support, the potential use of prophylactic pharmacotherapy in the at risk, and a lesser threshold for hospitalisation should be considered for mentally ill mothers of young children. Other important interventions include ensuring adequate treatment of substance abuse and the provision accessible psychiatric services for at-risk populations prior to pregnancy.

The role of the law in prevention remains unclear with little evidence of deterrence. The United States with arguably the most punitive approach nevertheless has the highest rates in the Western world with little sign of decline. Likewise Scotland, unlike England and Wales, has no provisions for mandatory treatment and probation, yet differs little in terms of rates and features of the events suggesting that the former approach is no more effective in preventing or deterring infanticide, although undoubtedly more expensive and less efficient [3].

The aftermath

The treatment of a woman who has killed her child presents considerable challenges for any involved clinician who, due to its rarity, is unlikely to have developed a broad experience. These difficulties are further complicated by the lack of information available regarding the recovery process. Some suggested guiding principles [39] include ensuring the identification and treatment psychiatric illness when present (including addressing the substantial risk of suicide by the perpetrator following the crime), assisting the woman to acknowledge the role of illness in the offence and supporting her to regain some aspect of the mothering role (if applicable).

For the clinician, countertransference issues form a core component of management; recognising and addressing any feelings of distaste and anger is essential in order to ensure the provision of appropriate care.

CONCLUSION

Society’s approach to infanticide has been an evolving process with improvements in our knowledge of the causative factors, particularly the centrality of mental illness to this crime, and the enactment of legislation reflecting these changes.

However, infanticide remains a tragic and difficult event with much to be done in the areas of prevention and management. As professions at the frontline, psychiatry and the law face a number of key challenges which require the balancing of complex (and often competing) considerations. These include:

References

[1] Resnick P. Murder of the newborn; a psychiatric review of neonaticide. Am J Psych 1970;126:58-64

[2] Williamson, Laila (1978), “infanticide: an anthropological analysis”, in Kohl, Marvin, Infanticide and the value of life, NY: Prometheus Books, pp. 61-75

[3] Spinelli M (ed). Infanticide: Psychosocial and Legal Perspectives on Mothers Who Kill. Washington DC, American Psychiatric Publishing, 2003

[4] Spinelli M. Maternal infanticide Associated With Mental Illness: Prevention and the Promise of Saved Lives. Am J Psychiatry 2004; 161:1548-1557

[5] Marks M, Kumar R. Infanticide in England and Wales. Med Sci Law 1993; 33: 329-39)

[6] Craig M. Perinatal risk factors for neonaticide and infant homicide: can we identify those at risk? Journal of Royal Society of Medicine. London, Feb 04, vol 97;2:57-62

[7] Infanticide Act, 2 Geo 6, Ch 36 (Eng 1938)

[8] Buist A. Seminar on “Infanticide” 2008

[9] Bourget D, Gagne P. A review of maternal and paternal filicide. J Am Acad Psych Law 35:74-82, 2007

[10] Friedman SH, Horowitz SM, Resnick P. Child murder by mothers; a critical analysis of the current state of knowledge and a research agenda. Am J Psychiatry 2005;162:1578-87

[11] Overpeck M. Epidemiology of infanticide, in Infanticide: Psychosocial perspectives on mothers who kill. Ed by Spinelli, Washington DC. American Psychiatric Publishing, 2003, pp 19-31

[12] Stanton J, Simpson A. Filicide: A review. International Journal of Law and Psychiatry, 25, 1-14, 2002

[13] Brockington I. Review of Infanticide: Psychosocial and Legal Perspectives on Mothers Who Kill. Am J Psych161:9, 2004

[14] Friedman SH, Resnick P. Child murder by mothers; patterns and prevention. World Psychiatry 2007;6(3):137-141

[15] Overpeck M, Brenner R, Trumble A, Trifilletti L. Berendes H. Risk factors for infant homicide in the United States. N Eng J Med 1998; 339:1211-1216

[16] Resnick P. Child murder by parents: a psychiatric review of filicide. Am J Psych 126: 325-34 1969

[17] Pitt S, Bale E. Neonaticide, infanticide and filicide: A review of the literature. Bulletin of the American Academy of Psychiatry and the Law 1995, 23, 375-86

[18] Vallone D, Hoffman L 2003 Preventing the tragedy of neonaticide. Holistic Nursing Practice, 17, 223-28

[19] Resnick P. Seminar on “filicide”, American Psychiatric Association 2007

[20] D’Orban P. Women who kill their children. Br J Psychiatry 134;560-71, 1979

[21] Bourget D, Gagne P. Maternal filicide in Quebec. J Am Acad Psychiatry Law 30:345-51, 2002

[22] Marce L. Traite de la folie des femmes enceintes, des nouvelles accouches et des nourrices. Paris. JB Balliere et Fils, 1858

[23] Kendell R, Chalmers J, Platz C. Epidemiology of puerperal psychoses. Br J Psych 1987; 150:662-73

[24] Kendell R, Rennie D, Clarke J, Dean C. The social and obstetric correlates of psychiatric admission in the puerperium. Psychol Med 1981; 11:341-50

[25] Lewis and Bunce. Filicidal mothers and the impact of psychosis on maternal filicide 2003. J Am Acad Psych Law 31:459

[26] Brockington I. Postpartum psychiatric disorders. Lancet 2004; 363:303-10

[27] Crimes Act 1958 (Vic) s6

[28] Victorian Law Reform Commission. Defences to Homicide: Final Report

[29] M’Naghten’s Case, 10 Clark and Finnelly 200 (1843)

[30] Model Penal Code 4.01(1). Philadelphia, American Law Institute, 1962

[31] Submission of the Victorian Bar on Infanticide and Diminished Responsibility in connection with the Crimes (Homicide) Bill 2005

[32] Law Reform Commission (2001). Partial defences to murder: provocation and infanticide. Law Reform Commission Pub

[33] American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision. Washington DC, American Psychiatric Association, 2000

[34] Wilczynski A. Child homicide. Greenwich Medical Media 1997

[35] MacFarlane J. Criminal defences in the cases of infanticide and neonaticide, in Infanticide: Psychosocial and Legal Perspectives on Mothers Who Kill. Ed Spinelli M, Washington DC, American Psychiatric Publishing, 2003, pp133-166

[36] Gauthier DK, Chaudoir NK, Forsyth CJ. A sociological analysis of maternal infanticide in the United States 1984-1996. Deviant Behaviour. 2003; 24:393-405

[37] Altshuler LL. Hendrick V. Cohen LS. Course of mood and anxiety disorders during pregnancy and the postpartum period. J Clinical Psychiatry. 1998; 59(suppl. 2):29-33

[38] Bennie E, Sclare S. The battered child syndrome. Am J Psychiatry 1969; 125:975-9

[39] Stanton J. Simpson A. The aftermath: aspects of recovery described by perpetrators of maternal filicide committed in the context of severe mental illness. Behavioural Sciences and the Law. 249(1): 103-12, 2006