Abstract
Although mental state defenses frequently are raised in cases of infanticide, legal criteria for these defenses vary across jurisdictions. We reviewed outcomes of such cases in states using M'Naughten or model penal code (MPC) standards for insanity, and the factors considered by the courts in reaching these decisions. LexisNexis and Westlaw searches were conducted of case law, legal precedent, and law review articles related to infanticide. Google and other Internet search engines were used to identify unpublished cases. Despite the differing legal standards for insanity among states, the outcomes of infanticide cases do not appear to be dependent solely on which standard is used. The presence of psychosis was important in the successful mental state defenses. This case series suggests that states that use the stricter M'Naughten standard have not been less likely than states with an MPC standard to adjudicate women who have committed infanticide as not guilty by reason of insanity.
Infanticide, defined as the killing of children less than one year of age, has evoked questions and controversy in both psychiatry and the law. Infanticide is etiologically distinct from neonaticide, the killing of an infant in the first 24 hours of life.1
Over the years, when a woman with postpartum mental illness has killed her child, the courts have responded in vastly different ways. In some courts, such women have been found not guilty by reason of insanity (NGRI) and ordered to undergo psychiatric treatment, while in others, they have been found guilty and sentenced to prison, sometimes for life. Some scholars have argued that differing state insanity standards result in disparate sentencing.2 In this article, we explore this hypothesis and examine cases tried in jurisdictions with either the M'Naughten or model penal code (MPC) standard. We conclude by summarizing relevant data for forensic mental health professionals serving as experts in such cases.
Postpartum Disorders: Definition and Epidemiology
The postpartum period is the time of highest risk in a woman's lifetime for the development of a mental illness.3 Although the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) defines the postpartum period as the immediate, relatively short time following delivery,4 in research studies this period has been defined as ranging from a few days to 12 months.5
Postpartum-onset psychological problems generally fall into one of three categories: so-called baby blues, postpartum depression, and postpartum psychosis. Baby blues affect up to 70 percent of mothers5 and include mild symptoms of crying and irritability which end by the second week after childbirth. Postpartum-onset depression affects from 7 to 15 percent of mothers and is characterized by sadness, crying, self-blame, loss of control, irritability, anxiety, tension, and sleep difficulties.6 Factors such as personality, negative feelings toward the infant, marital discord, poor social supports, and ambivalence over parenthood appear to predispose women to the onset of depressive symptoms in the postpartum period.7
Postpartum psychosis, the most severe of postpartum mental disorders, occurs in 1 to 2 of 1,000 births and frequently requires hospitalization to stabilize symptoms.8 The risk of postpartum psychosis rises to one in seven in mothers with a past episode of postpartum psychosis.3 Onset usually occurs within the first three months after childbirth, with a majority of cases appearing within 3 to 14 days.9 The disorder is characterized by an extended period of loss of contact with reality that may include auditory or visual hallucinations, delusions, or rapid mood swings. The auditory and visual hallucinations may focus on violence toward self or the infant.8 Studies show that women with postpartum psychosis are often victims of domestic violence or abusive childhoods and often have histories of abandonment or substance abuse.10
Risk factors for postpartum psychosis can be of both biologic and genetic origins.11 One study found that the rate of postpartum psychosis in mothers with bipolar disorder was 26 percent, with the rate in mothers with bipolar disorder and a family history of postpartum psychosis increasing to 57 percent.12
Postpartum Psychosis: How Is It Different?
Several characteristics of postpartum psychosis differentiate it from psychosis of other etiologies. First, research supports that postpartum psychosis is associated with prominent cognitive disturbances, ranging from mental confusion and indecision to intrusive and bizarre thoughts.13 Women with this disorder can appear delirious and disorganized.14 Second, symptoms can arise and dissipate suddenly, with the woman appearing lucid one moment and psychotic the next.15 Third, although DSM-IV-TR classifies it as a psychotic disorder not otherwise specified,4 research shows that the syndrome can have both affective and psychotic components.16
Legal Aspects: Postpartum Syndromes and the Courts
It has been estimated that four percent of women with postpartum psychosis commit infanticide.17 Since the 1980s, courts have permitted mothers with postpartum psychosis to assert the insanity defense. Postpartum psychosis and other postpartum mental health syndromes (e.g., postpartum depression) are also admissible in sentencing hearings as mitigation on the basis of diminished capacity.
The insanity defense is successful in less than 0.1 percent of all criminal trials.18 However, major population studies, including McKee and Shea,19 d'Orban,20 and Bourget and Bradford,21 found insanity pleas successful in 20, 27, and 15 percent of infanticide cases, respectively. These results suggest that, regardless of the legal or medical guidelines, the law treats women who kill their children with more leniency. Nonetheless, infanticide case outcomes vary extremely. At least two such women have been sentenced to death, whereas some have been released without a prison sentence.2
Criminal Responsibility: M'Naughten and Model Penal Code Tests
Currently, most states use either the M'Naughten test or the model penal code test (or a variation of either) to assess an insanity plea.22 The M'Naughten test is purely cognitive: the defendant must show whether she understood her actions or that her actions were wrong.23 The model penal code (MPC) test includes both a cognitive and a volitional prong: the defendant must show she understood right from wrong as well as whether she was able to conform her conduct to legal requirements.24
Some psychiatrists and legal scholars have criticized both tests for failing to allow for degrees of incapacity. The presence of postpartum psychosis may significantly impair a woman's mental stability but still not satisfy the cognitive test or even the irresistible-impulse test (i.e., the volitional prong).25 For example, after killing her child, a defendant with postpartum psychosis may recognize that the act was wrong but, because of the debilitating effects of her mental illness, may not have appreciated its wrongfulness at the time of the offense.26
To illustrate how infanticides are adjudicated in jurisdictions that use different insanity standards, we reviewed a series of such cases. Our analysis considered issues raised by the court in considering postpartum disorder-related defenses.
Methods
We identified cases by searching LexisNexis and Westlaw with the terms (or variations of these terms) postpartum psychosis, postpartum depression, postpartum defense, insanity, and infanticide. We reviewed all identified cases from 1969 to 2010 in which a woman was charged with killing her child when the child was less than one year of age. The study excluded cases in which a defendant harmed but did not kill her child.
Because LexisNexis and Westlaw record only appellate level cases or those that have been published, Google and Lexis media searches were conducted to supplement our search results with relevant cases reported in the news. Current medical and legal literature on infanticide were also reviewed. Because our search methodology would not have consistently identified cases resolved at the trial court level, it cannot be considered comprehensive. However, it is likely that this search methodology would have identified the major themes in how published and appellate-level cases of infanticide are adjudicated.
Results
The search yielded 34 cases. Twenty-six were adjudicated in M'Naughten standard states, seven in states with MPC standards, and one in a state without an insanity defense (Idaho). The defendants and states in which the cases were adjudicated are shown in Appendix A. Given the relative difficulty of meeting the M'Naughten cognitive standard, it is plausible that women tried in M'Naughten jurisdictions would have less success in pleading an NGRI defense than those tried in MPC states. This hypothesis suggests that there is a higher proportion of convictions among defendants tried in M'Naughten states than among those tried in MPC states.
Table 1 illustrates that the range of outcomes is similar in M'Naughten and MPC states. In states that used the M'Naughten standard, 46 percent (12/26) were found NGRI, 46 percent (12/26) were convicted, and 7 percent (2/26) were found guilty but mentally ill. Similarly, in states that used the MPC standard, 43 percent (3/7) were found NGRI, 43 percent (3/7) were convicted, and 14 percent (1/7) were found guilty but mentally ill. Although the sample size was insufficient for statistical analysis, the pattern does not support the hypothesis that the outcome of postpartum insanity pleas in infanticide cases is entirely dependent on the legal standard for insanity in the jurisdiction in which the cases are tried. Regardless of whether states used a M'Naughten or MPC standard, insanity defenses were successful in almost half of infanticide cases involving postpartum psychosis.
Of note, the legal issues raised in successful insanity defenses did not differ substantially by jurisdiction. In both M'Naughten and MPC states, the courts required that the defendant meet the cognitive prong of the insanity statute to be adjudicated NGRI. With one notable exception, the MPC courts did not rely on the volitional prong to exculpate women who committed infanticide.27
Analysis of Insanity Defenses in Postpartum Cases
Following are descriptions of some of the characteristics of cases of infanticide brought before the courts and the rulings handed down.
Regardless of Jurisdiction, Postpartum Psychosis Was Present in All of the Cases With Successful NGRI Adjudication of Infanticide
In both M'Naughten and MPC jurisdictions, successful NGRI defenses involved women with acute psychosis, generally labeled postpartum psychosis, but also labeled major depression, severe, with psychotic features, with postpartum onset. Below are some case examples of successful NGRI defenses.
M'Naughten Jurisdictions.
Among the cases reviewed, courts in Texas, a M'Naughten state, have found six mothers NGRI since 2003. Two of these cases are highlighted here. Andrea Yates was charged with capital murder for drowning three of her five children. After a mistrial, she was found NGRI in her second trial.28 In the first trial, the jury found Ms. Yates' ability to call the police after killing her children to be strong evidence that she understood her acts. However, during the retrial, several defense experts testified that, because of her psychotic state, she did not know right from wrong, was incapable of knowing what she did was wrong, or believed her actions were reasonable.
In the second case, in 2009 Otty Sanchez dismembered her three-and-a-half-week-old son and allegedly ate part of his brain. She stated that the devil told her to commit the act. Three psychiatrists testified that she had postpartum psychosis. She was adjudicated NGRI.29
In California, several similar cases have been adjudicated NGRI. Shontelle Cavanaugh, who had been jailed for almost five years in the suffocation death of her infant daughter, argued at trial that postpartum depression and an eating disorder led to a psychotic episode.30 The presence of psychosis was a primary contributor to her NGRI adjudication in 2010.
In New York, Ann Green, a former pediatric nurse, was found NGRI for killing two of her newborn children in 1980 and 1982 and attempting to suffocate her third child in 1985.31,32 She blamed postpartum psychosis for her actions. She testified “that she had seen hands she did not recognize holding pillows over the newborns' faces.”31 After being found NGRI, she was ordered to undergo a psychiatric evaluation in a state mental hospital as an outpatient.33
Model Penal Code Jurisdictions.
MPC states have also adjudicated women NGRI on the basis of postpartum psychosis. In Vermont, Michele Remington, killed her six-week-old son and then shot herself in the chest.34 She survived and was charged with first-degree murder. However, she was found NGRI after four psychiatrists agreed that she had postpartum psychosis. The case never went to trial because attorneys for both sides agreed that she had severe postpartum psychosis.
Regardless of Jurisdiction, Postpartum Depression, Posttraumatic Stress Disorder, Dissociative Disorders, and Personality Disorders Were Not Sufficient for Successful NGRI Adjudication
M'Naughten Jurisdictions.
In State v. Adams, the Louisiana Court of Appeals upheld a jury finding that the defendant's postpartum mental disorders did not meet the wrongfulness test of the M'Naughten standard.35 The court held that Ms. Adams, who placed her three-month old child into a clothes dryer, understood the difference between right and wrong at the time of the offense. She was convicted of second-degree murder and sentenced to life in prison. Despite the expert testimony of a psychiatrist who assigned a diagnosis of major depressive disorder, severe, with psychotic features, with postpartum onset in partial remission, the jury believed that the condition did not impair the defendant's ability to distinguish wrongfulness. Their decision was based on inconsistencies in her behavior, such as her attempt to cover up her actions, the absence of delusions, and the absence of a psychotic motive. Other psychiatrists testified that she had a major depressive episode with postpartum onset and also that she met criteria for posttraumatic stress disorder. However, that she did not experience delusions or hallucinations at the time of the killings was interpreted to mean that she did not meet the M'Naughten criterion.
The well-known case of Sharon Comitz reached the Pennsylvania Superior Court in 1987.36 The court accepted Ms. Comitz's plea of guilty but mentally ill to third-degree murder in the death of her infant son. However, the trial court determined that she was not severely mentally disabled at the time of the killing and sentenced her to 8 to 20 years in prison. The only question considered was whether, at the time of the murder, she was aware that her conduct would cause serious harm. The court found that she did not lack cognition and thus failed to meet the NGRI standard. She appealed on the basis of ineffective counsel, stating that her attorney was essentially unaware of atypical dissociative reactions in postpartum psychosis. A forensic psychiatrist testified that she had “dissociated” at the time of the infant's death. The prosecution psychiatrist testified that there was no evidence that she was actually psychotic at the time of the crime. (At the time of the trial, the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III) listed postpartum psychosis as an atypical dissociative disorder.37) The sentencing court found that even though she had an atypical dissociative reaction, she was aware at that time that her conduct would cause serious harm.
In an older case in another M'Naughten state, Nevada, Pamela Clark was convicted of attempting to murder her two-week-old daughter by wrapping her in a blanket and abandoning her in the bushes at the side of a road.38 Two psychiatrists and one psychologist testified to Ms. Clark's “severe post partum depression” that rendered her legally insane at the time of the abandonment. However, her family members testified that she was acting normally on the day of the abandonment. The jury found that her calmness and lack of affect when questioned by the police and her ability to make up a story that the child was kidnapped outweighed the expert medical testimony.
Model Penal Code Jurisdictions.
In Kentucky, an MPC state with a guilty but mentally ill (GBMI) option,39 the supreme court affirmed Karen Rae Mitchell's conviction for murdering her nine-month-old daughter.40 She had received a diagnosis of severe depression and anorexia. Three months before the murder, she tried to strangle her child, after which the district court ordered that Ms. Mitchell not be left alone with her. However, she subsequently suffocated the infant. The jury rejected her insanity defense, and she was found GBMI and sentenced to 20 years in prison.
In a case decided by the Supreme Court of Idaho in 2004, Michelle Tiffany was charged with involuntary manslaughter after she suffocated her infant with her hand while trying to stop him from crying.41 Although Idaho abolished the insanity defense in 1982, the court was clear in ruling that it believed that Ms. Tiffany understood the nature and quality of her acts. The lower court held (and the supreme court affirmed) that there was no evidence that her depression made her unaware that she was cutting off her child's oxygen supply. A defense psychiatrist argued that Ms. Tiffany's postpartum depression affected her ability to form the requisite mens rea. The Idaho Supreme Court commented, however, that the psychiatrist did not testify that her depression would affect her ability to understand the consequences of her conduct. The ruling noted:
…there was no evidence in this case that because of her depression, Tiffany did not understand that she was cutting off her son's oxygen supply when she placed her hand over his nose and mouth and that doing so would cause injury to his person or health. …The psychiatrist's testimony did not address the mental element necessary for the crime of injury to a child [Ref. 41].
The Cognitive Changes Associated With Postpartum Psychosis Are Not Fully Recognized by the Courts
In the first Yates trial,28 and in the Comitz36 and Clark38 trials, the fact that the mothers had shown somewhat lucid behavior after the killings weighed heavily under the M'Naughten test to show that they knew right from wrong. In Yates, the jury found it difficult to overlook that Ms. Yates had called 911 to report the killings. In Comitz and Clark, the women created kidnapping stories. Because fluctuating mental status is a common attribute of postpartum psychosis, some have argued that a mother's efforts to contact the police after the killing should not be admissible as evidence of knowledge of right and wrong at the time of the killing.42
In none of the cases reviewed, however, did the courts recognize the rapid mental status changes that can be associated with postpartum illness. The focus remained on the cognitive prong of the insanity test: whether the defendant understood the difference between right and wrong at the time of the crime.
There Is Some Precedent for Courts to Consider Mental State as a Basis for Reducing or Overturning Decisions About Women Who Have Been Convicted of Infanticide
On several occasions, the court overturned jury verdicts that did not recognize mental state as a mitigating factor in cases of infanticide. In California in 1990, Sheryl Massip was charged with murdering her six-week-old son by intentionally running over him with her car and placing him in a trashcan, after which she reported him kidnapped.43 Psychiatrists testified that, at the time of the crime, she had been experiencing suicidal ideation and auditory hallucinations that her son was the devil. She was found guilty of second-degree murder and incarcerated. Two months later, however, a judge set aside the jury's guilty verdict, and acquitted her on insanity grounds. She was sentenced to one year of outpatient treatment. The decision was affirmed on appeal.
In Connecticut, an MPC state, Dawn March drowned her five-month-old daughter in 1989.44 She testified that demons told her to kill her child. She was charged with murder and was prepared to plead no contest to the lesser charge of manslaughter. However, the judge inquired into postpartum psychosis and learned that Ms. March had received a diagnosis of postpartum psychosis with command auditory hallucinations. The judge threw out her plea and ordered the case to trial. She was found NGRI on the basis of postpartum psychosis.45
In Indiana in 1996, on the basis of mental illness, the Indiana Supreme Court reduced the sentence of a defendant convicted of drowning her child.46 Before the killing, she had religious visions and accused her friends of being devils. After she killed her son, she believed that she was with Jesus and approached a stranger while completely naked. Four experts testified that she had severe postpartum psychosis. The jury convicted her and she was sentenced to 60 years. On appeal, the Indiana Supreme Court reduced her sentence to 40 years because of her mental illness.
Courts May Reconsider Verdicts That Disregard Evidence of a Woman's Mental Health
Courts have reconsidered decisions involving postpartum mental illness. In Iowa, Heidi Anfinson drowned her infant son in 1998.47 She had been hospitalized for depression, suicidality, and panic attacks. In 2000, she was convicted of second-degree murder and sentenced to 50 years in prison. The Iowa Court of Appeals upheld the verdict, ruling there was no authority in Iowa law for using postpartum depression as a basis for an insanity defense. However, in 2008, the Iowa Supreme Court ruled that her lawyer should have submitted evidence of depression and odd behavior that followed the birth of her son and remanded the case to the trial court for ineffective assistance of counsel. This ruling not only cleared the way for introducing mental health evidence in cases in which no insanity-related defense had been raised, but also effectively mandated introducing such evidence where appropriate.
There Are Some Cases in Which the State's Standard for Insanity Does Influence the Outcome of the Case
Of the 34 cases studied, 1 case outcome appeared to depend on the jurisdiction of the trial. In State v. White,27 the jury found the defendant NGRI after she killed her infant. The Supreme Court of Idaho, using the MPC standard, affirmed the acquittal. Unlike the other cases reviewed in this analysis, this case was decided on the volitional prong of the MPC. The court acquitted Janet White on the basis of her inability to conform her conduct to the requirements of the law. It is noteworthy that in this case, had M'Naughten been applied, she would have been found to be sane and therefore responsible.
Limitations
This analysis has several limitations. Our primary search methods, LexisNexis and Westlaw, do not capture cases that were resolved at the trial court level and therefore were not appealed. Although we supplemented these searches with Google and the LexisNexis media search engine, these methods would not have identified all relevant cases. Our case series should be considered illustrative of themes in how appellate-level and published cases of infanticide have been adjudicated when postpartum psychosis has been raised as a defense. An additional limitation is that size of the study group was insufficient for statistical analysis.
Conclusions
In this case series, NGRI acquittal in cases of postpartum mental illness generally involved defendants who were psychotic at the time of the offense. Regardless of jurisdiction, postpartum depression, posttraumatic stress disorder, dissociative disorders, and personality disorders were not sufficient for successful NGRI adjudication. Although postpartum psychosis was important for acquittal, the cognitive changes associated with postpartum psychosis have not been fully recognized by the courts. In multiple cases, courts have not acquitted mothers who were psychotic but displayed seemingly lucid behavior around the time of the crimes.
Furthermore, despite differences in standards for insanity across the country, in cases of postpartum mental illness, courts have ruled in a mostly consistent manner. Differing legal standards for insanity for the most part were not associated with different patterns of outcomes of insanity defenses across jurisdictions. A more consistent factor in the final ruling was the diagnosis assigned to the woman who committed the infanticide.
Although there is some precedent for courts to reduce or overturn sentences of women who have been convicted of infanticide, the focus of these reviews has remained on psychotic behavior. In the cases described herein, if the woman did not have postpartum psychosis, her mental state at the time of the offense was usually insufficient to meet insanity standards, regardless of the jurisdiction. Specifically, although the MPC contains a volitional prong, most of the cases in MPC states were not adjudicated NGRI on the basis of that prong alone. The factor that merited an NGRI adjudication was usually the cognitive prong.
Potential Future Directions
In recent years, there have been some proposals to change infanticide laws in the United States. For example, a bill introduced in Texas in March 2009,48 proposed an amendment to the penal code to create the offense of infanticide, punishable as a state felony, but it did not pass.
Some have expressed concern that a firmer scientific foundation about postpartum diagnoses would be needed to support legal claims relying on them. Two federal courts (U.S. District Court for the Southern District of Florida, and U.S. Court of Appeals for the Eleventh Circuit) expressed interest in possibly recognizing postpartum illness as a physical illness as it relates to health insurance policies, but would “require proof…such as a test result showing an imbalance in serotonin or norepinephrine …”.49
Considerations for the Forensic Expert in Postpartum Cases
As evidenced by the cases reviewed, the role of the forensic expert can be paramount in postpartum cases. Forensic evaluators must strive for objectivity and impartiality. They should remain aware of potential biases and the emotions, felt by both themselves and others, evoked by such horrific crimes. Forensic evaluators must understand the myriad mental health problems that can affect women during the postpartum period as well as the degree of impairment expected in a woman with such conditions.
Experts function as educators and must be aware that the jury and triers-of-fact frequently have limited understanding of the complexities of postpartum mental health changes. Nonetheless, experts must remain focused on the questions posed (e.g., state of mind at the time of the offense) and leave the legal question of sanity to the trier-of-fact.
Finally, forensic experts testifying in postpartum cases should be aware that the legal and diagnostic standards for infanticide are evolving and subject to change. For example, substantial evidence supports that fathers and adoptive mothers can experience significant psychological symptoms following the birth of a child.50 Participants in the legal system may not be abreast of new research; therefore, forensic experts should be aware of current clinical and scientific developments.
Appendix A
The defendants in the 34 cases chosen and the state in which they were adjudicated are as follows: Adams (LA), Anderson (CA), Anfinson (IO), Cavanaugh (CA), Clark (NV), Comitz (PA), Currie (MI), Dean (OH), Diaz (TX), Dupre (PA), Ferguson (CA), Fuelling (CA), Gambill (IN), Gindorf (IL), Green (NY), Householder (WV), Laney (TX), March (CT), Massip (CA), Maxon (TX), Mitchell (KY), Molina (CA), Pixley (DC), Reilly (PA), Remington (VT), Sanchez (TX), Schlosser (TX), Thompson, A. (CA), Thompson (OH), Tiffany (ID), White (ID), Wilhelm (NY), Yates (TX), Young (OH).
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2012 American Academy of Psychiatry and the Law