Incomplete homicide-suicide in Hunan China from 2010 to 2019: characteristics of surviving perpetrators

Homicide followed by suicide (H-S), which is also named “homicide-suicide” and sometimes known as “murder-suicide”, is a lethal event in which one person kills another and subsequently ends his or her own life [1]. H-S is a relatively rare incident compared to homicide or suicide alone. Studies on this topic have been carried out in many regions of the world, such as Australia [2], Oklahoma [3] and England and Wales [4, 5]. But most studies were performed in the United States [1, 6], where the incidence rate was reported to be ranged from 0.05 [5, 7] to 0.3 per 100,000 persons annually [8, 9]. Although a slightly higher rates (0.89 per 100,000) was fond in the South African [10], its incidence is still quite low. However, H-S incidents may produce long-lasting consequences for the involved families and the whole society [11], including strong emotional impact on survivors, especially when children are involved. In addition to significant financial burdens related to complex investigations of the incident, H-S also brings challenges for mental health specialists in the assessment and intervention of such events.

Theoretical studies have been focused on typology and theory of H-S. Various classification frameworks have been proposed to classify the different subtypes of H-S. In 1992, Marzuk et al. [9] proposed a typology system based on victim–perpetrator relationship and motives. In 1995, Felthous et al. [12] integrated the psychopathological factors into Marzuk et al.’s classification. In 2007, Harper et al. [13] classified the H-S on thematic context. In 2014, Joiner et al. [14] proposed the “perversion of virtue” framework, which argues that H-S is the result of primary suicidal intent that lead to the perversion of four virtues. Later, Knoll et al. [15] proposed a classification scheme includes both elements of the Marzuk typology and the Felthous methods. Most recently, Jordan et al. [16] developed an empirically based typology that consist of eight H-S subtypes. Besides, some theories have been suggested to apply to H-S, such as the strain theory, stream theory, psychodynamic theory, psycho-evolutionary theory, social integration theory, attribution theory, and masculinity theory [11, 17]. These classification frameworks and theories mainly involve factors such as motivation, relationships between perpetrators and victims, and psychopathology.

Except theoretical researches, other studies investigated the prevalence of H-S, mostly focus on the complete H-S with a successful suicide, and its related factors, and a few of them compared the profiles of H-S with either homicide only or suicide only. Generally these studies found that about 75% of H-S involved a female victim, and most perpetrators were adult males; incidents with multiple victims are also common [18‐21]. 90% of the victims and perpetrators were related or known to each other, and most of them were identified as intimate partners [21]. According to a study of H-S in mainland China, an intimate relationship between the perpetrators and victims was identified in 95% cases [22]. Weapons associated with H-S incidents often included sharp instruments, blunt objects, and poisons; other methods such as strangling and drowning were also included [4]. Studies in the United States, where firearm ownership is allowed, showed that the gun was involved in the majority of deaths associated with H-S incidents [1, 3]. Through comparing H-S incidents to homicides, a study showed that when the motivation is associated with sickness, stress or possessiveness, the perpetrators are more likely to die by suicide [23].

Although we have a certain understanding of H-S, there is still little known about of why some suicide decedents killed others before death. It is generally believed that a considerable number of individuals who died by suicide after killing others are psychiatrically disturbed [24] [20]. The estimated prevalence of mental illness in different samples ranged from 20% [25] to 91% [26]. Recently, a study in England and Wales estimated that 40% of H-S perpetrators had a history of seeking medical care or consultation within a month before the incident [19]. The variation at least partly attributed to different definition of mental illness and different sources of information.

Data of H-S, especially about mental disorder, are often retrieved unsystematically due to problems in information acquisition. Most information in the existing literature come from the homicide or suicide death system, such as National Violent Death Reporting System in United States [1, 6], Office for National Statistics in England and Wales [4], and Hong Kong Police Criminal Records Bureau [22]; part of the data even came from the news [27]. Descriptive characteristics of the perpetrators have been well established by existing literature, but little is known about the mental health factors related to these incidents.

To capture psychological and contextual circumstances preceding the complete H-S with a successful suicide, the psychological autopsy [24, 28, 29] method was applied. Clinical data were collected from medical records, such as record of mental health service institutes or sometimes records from supervising clinicians responsible for the patients’ care. But, due to the lack of information and detailed personal history, the data of the perpetrators’ behavior and mental health history are prone to errors. Although the interview of relatives or close friends of the perpetrators were involved in some psychological autopsy reports, the accuracy of the information is unknown. In addition, the incidence of violent behavior or mental symptom are likely to be underestimated, as the lack of such information does not necessarily mean that the behaviors or symptoms did not occur. Due to the death of the perpetrators, it is difficult to obtain information on the mental state of the perpetrators at the time of the event. Researches showed that the rate of mental disorders varied greatly from 20 to 91% [25, 26], suggesting the difficulty to determine a diagnosis of mental illness after the perpetrator’s death. As the systematic analyses of mental illnesses are helpful for a better understanding of such events [30], a direct interview with perpetrators is needed. And the incomplete H-S events, in which the perpetrators survived, provide the possibility for research in this area.

Besides, this problem is still poorly understood in China, as H-S studies were rarely conducted in the Chinese context. Only some studied in Hong Kong (one case report [31], and two descriptive studies [22, 32]), and a research in mainland China using data from Chinese news from 2000 to 2014 [27] provided some information related to H-S. Furthermore, mental and psychological conditions related H-S are often difficult to decipher, due to the death of perpetrators. With the absence of accurate motivations and assessments of mental state, it is difficult to acquire precise and complete information of factors that contribute to the occurrence of H-S. Therefore, in order to get a better understanding of the phenomenon in China and its psychological characteristics, this study includes the incomplete H-S events occurred in Hunan, China from 2010 to 2019.

The definition of H-S differs between studies, and the interval between the homicide and the suicide ranged from 24 h to 30 days [33, 34]. For most cases, the suicide occurred immediately after the homicide. In this study, the incomplete H-S referred to homicide or attempted homicide, followed by attempted suicide within 3 days. This study was launched by the Forensic Assessment Centre of the Second Xiangya Hospital, and was approved and supported by the other 7 qualified forensic institutions. The study protocol was approved by the Ethics Committees of the Second Xiangya Hospital; and all the researchers signed a written commitment to protect the privacy of subjects.

In China, perpetrators suspected of suffering from a significant mental disorder are mandated to undergo a forensic psychiatric assessment by qualified forensic institution [35]. There are 8 qualified forensic psychiatric institutions in Hunan province located in the south-central of China. The province’s population was enumerated at 65.701 million in 2010 and reached 69.18 million by the end of 2019. Detailed descriptions of forensic psychiatric assessment in China can be seen in previous literature [36, 37]. Usually, the assessment consists of an interview, electroencephalograph testing and intelligence quotient testing; no less than two experts to do the forensic psychiatric assessment, and the diagnoses according to the International Classification of Diseases 10 was made by these two experts. If the opinions of these two experts differ, both opinions should be recorded in the report. But in this study, we did not found any different opinion in the incomplete H-S cases.

Incomplete H-S cases were identified by scanning forensic archives of all criminal suspects who had received forensic psychiatric assessments in Hunan province, from 1 January 2010 to 31 December 2019. All the data were retrieved from the forensic archives, including the perpetrator’s demographic information, criminal files, medical records, as well as details and conclusions of the forensic psychiatric assessments. Three aspects of information, i.e., the subject’s demographic information (such as the gender, age, residence, marital status), clinical information (such as consultation history due to mental problems, history of psychiatric treatment, current diagnosis) and criminal information (age at the current crime, methods of homicide and/or suicide, number of victims, information of victims), were collected using a purpose designed standard data collection form that was specially designed for the present study.

The main findings are presented as numbers with proportions. In consideration of the obvious gender difference between homicide and suicide, we also compared the features of perpetrators between different gender. All the statistical analyses were computed with the SPSS (Version 22.0; IBM), with the significance level was set at 0.05.

To our knowledge, this is the first study on the characteristics of incomplete H-S, in which the perpetrator survived and received a forensic psychiatric assessment in mainland China. We focused on gender differences, psychosocial and clinical information, hoping these aspects of incomplete H-S would help in the identification and prevention of individuals at risk and to some extent, improve our understanding of H-S.

Most of the perpetrators in incomplete H-S were males, which is consist with previous studies about H-S. However, the male-female ratio is not as great as those found in previous H-S studies. For example, 95% H-S perpetrators were men in the first study of H-S in mainland China [27], while a study in Hong Kong showed that 68.7% of H-S perpetrators were male [32]. As for the gender ratio of victims in incomplete H-S, we have found slightly more males than females, which is inconsistent with previous results in H-S. In a study in mainland China, Densley reported that 50% of the victims in H-S were female, compared to 33% of males and 17% of cases where the gender of the victims was unknown [27]. A study in Hong Kong reported that 64% of the victims in H-S were female [22]. The most likely explanation for the variation is that the cases in our study are incomplete H-S incidents with surviving perpetrators, while cases in previous studies are H-S incidents include both complete and incomplete cases. The present study is based on individuals who had undergone forensic psychiatric assessment and it is possible that more male perpetrators had died, leaving them no chance to receive an assessment.

Analysis of the data reveals significant gender differences in perpetrators’ suicide history, current diagnosis and motivation. Females showed a higher rate of suicide history, they were more likely to be diagnosed as depression and often motivated by delirious altruism, the idea of saving their children from the horrible fate or future [38]. Sometimes these cases are understood as “extended suicides” as the depressive mothers have also been suggested to consider their children as their continuity, and then they naturally end their children’s lives when they suicide due to the unworthy existence [9]. It suggested that the incomplete H-S in woman might be more related to depression and suicide ideation, which implied that public health measures aimed at reducing suicide might be helpful to prevent such events.

The majority of studies have indicated that psychopathology is a predisposing factor of H-S. H-S perpetrators are more likely to suffering from mental illness, with the most common disease being depressive disorder, followed by psychosis [39]. A study in England and Wales found that 62% perpetrators had mental health problems, with the most common diagnosis of depression, but few had received any mental health services recently before the incident [19]. Furthermore, depression was found to be more common in female perpetrators who killed their children [40]. In addition to depression, individuals with psychosis, especially schizophrenia, might be at higher risk for violence, including H-S, as their current delusion was associated with aggression [41].

As described in this study, the mental illnesses leading to hallucinations, delusions or depression, may well have been a factor in some incomplete H-S incidents. However, the path from mental disorders to it might not be a straight line, and mental disorders are often a part of the complex story and can be affected by a variety of other factors, such as treatment. In this study, 110 perpetrators were diagnosed as mental disorders, but only 56 had a history of consultation due to mental health issues and only one of them continued treatment. Undetected and untreated psychosis may be more closely related to incomplete H-S. Many studies have reported that most of the aggressive behaviors of individuals with psychosis occurred during the first episode of the illness [42] or before contacting the mental health system [43]. Compared with the later stage, the relationship between aggressive and psychosis is the most prominent during the first episode [44]. Treatment compliance is also considered to be closely related to aggressive behavior. A meta-analysis on first-episode psychosis showed that involuntary treatment after the treatment regimen started was closely related to aggressive behavior [42]. Similarly, non-compliance to treatment can also lead to serious aggressive behaviors. This suggests that early diagnosis of diseases and effective, long-term treatment should be highly valued.

Although the rate of mental illness is high in the subjects included in this study, it does not mean that mental disorders will lead to or significantly increase the risk for incomplete H-S; it implies that it is necessary to assess the mental state in incomplete H-S population and to improve our understanding of the role of mental illnesses, although the degree of impact of mental illnesses on such events are needed to be determined. In most incomplete H-S cases, perpetrators are required to undergo a psychiatric assessment to determine whether they were mentally abnormal at the time of the event. However, such a process cannot be completed if the perpetrators have died of suicide.

In this study, apart from the events with mental disorder-related factors, most cases were motivated by family problems. Furthermore, situational factors, including recent relationship problems, job or financial problems and physical health problems, were also associated with the incomplete H-S. This is consistent with previous studies which tried to elucidate the circumstances under which the events occurs [6, 45]. A study showed that H-S typically occurs when noxious stimuli were present, such as financial or family problems [34]. Major changes in social environment are also influencing factors. Failure in achieving highly valued goals, such as being rejected by an intimate partner, is generally considered as a contributing factor to H-S. H-S researches showed that the perpetrator-victim relationship is typically close. Logan [21] found over half of perpetrators (53.9%) in the subjects experienced conflicts with their intimate partner.

The major limitation of this study is that we just included the incomplete H-S cases. Due to the lack of completed H-S cases, the characteristics of the study sample may have deviated from those of whole H-S sample. And the group of subjects is not representative of the general population, hence the results cannot be generalized and are only suggestive of possible associations. However, a greater understanding of the reasons for incomplete H-S may lead us to ways to reduce these catastrophic events. Additionally, the data does not represent all the incidents in Hunan, China from 2010 to 2019. Cases in which the perpetrator has not been sent for assessment are not included; however, this study has provided more accurate information of H-S, especially on psychosocial and clinical information.

In summary, this study demonstrates the characteristics of surviving perpetrators in incomplete H-S events. The findings suggest that psychopathology might be a predisposing factor underlying incomplete H-S and highlight the importance of mental state assessment for the population involved in such events. Further, a better understanding of the role of mental disorders could be useful in developing appropriate interventions.