Summary:
Introduction: Psychiatry and justice, medical and legal science, are called to deal with a special group of patients, those who have committed crimes and have been found not guilty due to mental illness. The treatment of these patients by the penal system is currently regulated by Law 4509/2017, which focuses on the regulation of their medical needs and the protection of their rights. But what are the characteristics (demographic, psychiatric, psychopathological, etc.) of the patients who commit a crime, and how can the prevention of violent behaviors be achieved? These questions are at the core of the problematic issue of early diagnosis and effective treatment of psychiatric patients prior to committing crimes.
Material and methods: This research aims to study the total number of patients (N = 78) who have been hospitalized and treated in the Department of Forensic Psychiatry of the Psychiatric Hospital of Thessaloniki from January 2015 to January 2020, which exclusively treats all psychiatric patients found not guilty throughout Northern Greece. The aim of this study is furthermore to compare the different categories of patients in terms of their demographic, clinical, legal and psychometric characteristics in order to identify any statistically significant differences. The research also aims to highlight groups of patients with specific characteristics from their history or from the evaluation of the results in psychometric tests, with the ultimate goal of assessing the risk, but also the assessment of the needs of these patients for treatment. In order to achieve the above objectives, it was investigated primarily whether there are statistically significant differences between patients who commit the offense in the first psychotic episode and those who commit it later in the course of the disease. Data were collected from four cross-referenced sources: patients’ medical records and court records, interviews, information provided by ward staff and patients' relatives, and completion of questionnaires (MINI, PANSS, HDHQ, Aggression Questionnaire, ZKPQ, SF36, CAGE, ASI, GAF and a questionnaire with myths of Aesop in ppt format).
Results: The sample consists of 78 patients, 70 men and 8 women. It consists mainly of Greeks (N = 75), Orthodox Christians (N = 75), unmarried (N = 50, 64.1%), of low educational level, who commit the crime at the age of 38.7 on average (± 12). More than half of the patients were diagnosed with schizophrenia at the onset of the disease (52.6%), while at the time of the crime almost two in three (73.1%) were found to suffer from schizophrenia. 55 patients had a history of hospitalization until the crime was committed (70.5%) and 56.4% of study participants had a history of substance abuse / dependence. 1 in 3 said they had used a psychotropic substance at the time of the crime (35.9%). In addition, in 25.6% of the cases the crime was related to substance use. The majority received psychiatric medication at the time of the crime (79.5%), but only 11.3% were compliant. 42.3% (N = 33) have relatives with a history of psychiatric illness, while 1 in 3 stated that they had a family history of substance abuse / dependence. Only 6.4% (N = 5) of the sample members reported a family history of criminal / delinquent behavior and 11.5% of the total had a family history of self-destructive or hetero-destructive behavior. 33.3% (N = 26) attempted to commit suicide prior to the crime and 74.4% after that. 3 in 4 exhibited aggressive behavior (74.4%) before committing the offense, while this percentage decreases to 17.9% after the offense. 31 patients (39.7%) were victimized physically or sexually or suffered school bullying in the past, while 30 (38.5%) had a history of conduct / behavior disorder in childhood. More than half experienced stressful events in the last two years before the offense (52.6%), more often the death of a close family member (41.5%) and dismissal from work (24.4%). The offenses mainly concerned homicides (42.3%), attempted homicides (25.6%), bodily harm (12.8%) and arson (11.5%). They were committed mainly in a common private space (45.1%) with a knife (46.2%), while the victims were usually family members (71.8%). 25.6% had a previous history of convictions (N = 20), while half were found not guilty by the Council of Judges, 41% by a Court of First degree and only 9% by a Court of Appeal decision. The mean total score on the PANSS scale was 109.69 (± 19.48) at the time of the offense, while on the positive subscale the mean score was 31.06 (± 6.15). The mean score on the GAF scale was 42.05 (± 10.61). The emotions of the 38 participants, as recorded in the test with the myths of Aesop (confidence, fear, anger, disgust, sadness, surprise, joy, anticipation, none of the above), were mainly sadness (24%) followed by anger (17%).
The difference and independence tests based on the time of the offense (21 patients in the first psychotic episode and 57 patients in the course of the disease) showed that the patients who committed the offense in the first psychotic episode were younger (33.2 ± 11.9 years vs. 40.7 ± 11.5 years, p = 0.014), most often employed (71% vs. 43.9%, p = 0.031) and had experienced stressful events in the last 24 months before the offense (81% vs. 42.1%, p = 0.002). They committed more frequent violent crimes against life, mainly homicide and attempted homicide (90.5% vs. 59.6%, p = 0.010) and attempted suicide after the offense (52.4% vs. 10.5%, p <0.001). Their victims were more frequent family members (90.5% vs. 64%, p = 0.024). After the offense, these patients received antidepressant therapy more often (47.6% vs. 19.3%, p = 0.012). The two groups differed statistically in the Hostility subscale of the Aggression questionnaire (19.95 ± 7.49 vs. 24.00 ± 6.85, p = 0.042), and in the subscales Criticism of others (5.29 ± 2.5 vs. 6.88 ± 2.91, p = 0.028) and Paranoid hostility (3.67 ± 2.46 vs. 4.95. 2.51, p = 0.048) of the HDHQ questionnaire, where patients in the first psychotic episode had a lower score. On the other hand, the mean score in items P1 (Delusions), P4 (Excitement), P6 (Suspiciousness / Persecution) and P7 (Hostility) of the PANSS positive subscale was higher in patients in the first psychotic episode (p values 0.004, 0.038, 0.015, 0.040 respectively). No differences were found regarding the negative subscale. In the General Psychopathology subscale, patients in the first psychotic episode had a higher score in the items G2 (Anxiety), G3 (Guilt feelings), G4 (Tension), G12 (Lack of judgment and insight), G14 (Poor impulse control), G15 (Preoccupation) and overall, but lower scores in the items G1 (Somatic concern) and G5 (Mannerisms and posturing). No statistically significant differences were found in their scores in the CAGE questionnaire or the GAF scale. Patients who committed the offense in the course of the disease had a younger age of first alcohol use (15.62 vs. 14.04, p = 0.023) and more years of alcohol use (13.4 vs. 21.26, p = 0.02). Finally, patients who committed the crime during the course of the disease appear to be at a greater risk of developing violent behavior than those who committed the crime during the first psychotic episode, as shown in the HCR-20 scale at the time of the offense (22.9 vs. 26.79, p = 0.003) and at present (20.48 vs. 25.74, p = 0.001). At the time of the study the patients who committed the offense in the course of the disease had a higher score in items P2 (Conceptual Disorganization) and in the whole positive PANSS subscale (15.48 vs. 19.4, p = 0.035), but also in the items G1 (Somatic concern), G4 (Tension), G5 (Mannerism), G14 (Poor impulse control) and in the whole General Psychopathology subscale.
From the control of significant differences in the overall scores in PANSS, GAF and HCR20 recorded by the patients during the two phases of the study, it emerged that the patients' clinical features in terms of function, positive and general psychopathology scale, and HCR-20 subscales is significantly improved at the time of the research compared to the time of the offense. In contrast, no significant change is found in the negative subscale.
Then cluster analysis was applied, from which three types (clusters) of patients emerged. The first group includes patients who exhibited aggressive rather than suicidal behavior after the offense, and who had a history of conduct / behavior disorder in childhood. The vast majority of patients in this group assaulted a family member. The patients of this group are the most aggressive, impulsive, hostile, socially non-functional patients. Finally, in this group, the patients have higher scores regarding emotional withdrawal, lack of cooperation, lack of judgment and insight, impulsive control disorder and active social avoidance.
In the second group, the vast majority of patients suffered from schizophrenia, and they usually did not have a history of self-destructive or aggressive behavior after the offense, nor a history of conduct / behavior disorder in childhood. They committed more frequent property crimes compared to the other groups. They have moderate scores in terms of hostility, impulsivity, introverted and extroverted hostility and social function. They are characterized by stereotyped thinking both prior to the crime and afterwards, by early adjustment difficulties, lack of insight, lack of response to treatment, and higher score in the clinical and risk management scale of HCR-20.
Finally, in the third group, almost all patients suffered from schizophrenia or psychosis, many had exhibited self-destructive behavior after committing the crime, but no one exhibited aggressive behavior. Also, the percentage of people who had conduct/behavioral disorders during childhood in this group is lower compared to the other two groups. None of them committed a crime against property, while the majority committed crimes against life. Also, most of them assaulted a member of their family. These patients have the lowest scores in terms of hostility, introverted and extroverted hostility and impulsivity, and better social function. However, this group has the most severe psychopathology regarding delusions, suspicion and persecution, anxiety, guilt and previous violent behavior before the offense, but the lower risk for violent behavior at present.
A logistic regression analysis was performed using the “Forward LR” method. Besides history of hospitalizations before the perpetration of the crime (x2 (1) = 21.592, p <0.001), only the total Hostility score (x2 (1) = 5.820, p <0.016) recorded in the Aggression questionnaire was important as a factor for the separation of our sample in the two groups studied. In particular, for perpetrators with a higher score in Hostility, the probability of committing the crime in the course of the illness is higher, than that of those who show lower levels of hostility (OR=1.15), meaning that patients who score high in the item Hostility of the Aggression Questionnaire remain at risk for committing a crime in the course of the illness.
Using the "Enter" method, the only variables that significantly affected the time of the offense are age (x2 (1) = 4.903, p = 0.027), the administration of psychiatric medication at the time of the offense (x2 (1) = 8.176, p = 0.004), the experience of stressful events (x2 (1) = 4.805, p = 0.028) and the relationship between the perpetrator and the victim (x2 (1) = 4.811, p = 0.028). In particular, the probability of the crime being committed in FEP in relation to the crime committed during the course of illness is higher for those who are younger. Also, those who have experienced stressful events in the last two years before the crime are 7.69 (1 / 0.13 = 7.69) times more likely to commit the crime in FEP than later, compared to those who did not experience such events. Finally, when the victim is a member of the family, it is more likely that the offender is in the FEP group (1 / 0.11 = 9.09).
Our findings showed distinct differences between psychotic patients who commit crimes during their first psychotic episode and patients who commit crimes in the course of their illness and are consistent with the international literature. Our study answers the questions posed as following:
• 26.9% of Greek NGMIs in Northern Greece are in their first psychotic episode. They are usually men in their 30s, who commit homicide or attempted homicide. Half of them attempted suicide after the crime. They seem to be driven by delusions and ideas of persecution to assault mostly members of their family
• Greek NGMIs who commit the crime in the course of the illness are usually men in their 40s who were often violent before the crime, had previous convictions, had been hospitalized but were not adherent to the treatment prescribed
• Patients in their FEP are younger. Their psychopathology concerning delusions and ideas of persecution is more severe, but their scores in Hostility are lower
• 56.4% of the whole sample was abusing or was addicted to either alcohol or other substances. In 25,6% of the cases the perpetration of the crime was related to abuse/use of alcohol or other substances.
Findings from logistic regression analysis point out that patients who score high in the item Hostility of the Aggression Questionnaire remain at risk for committing a crime in the course of the illness. Moreover, the probability of the crime being committed in FEP in relation to the crime committed during the course of illness is higher for those who are younger. Also, those who have experienced stressful events in the last two years before the crime are 7.69 times more likely to commit the crime in FEP than later, compared to those who did not experience such events. Finally, those who assault a family member are about 9 times more likely to act during the first psychotic episode than during the course of the disease, compared to those who assault people outside the family environment.
It is imperative to design and develop mental health services in order to identify and treat patients involved in antisocial and criminal actions in a timely and effective manner in order to prevent these behaviors and reduce the stigma of mental illness. Early intervention to control symptoms and initiate rehabilitation along with interventions aimed at medical conditions that coexist more frequently with psychiatric disorders could also reduce the risk of lifetime disability. Facilities that will treat these patients after the offense should be specialized and aim at reintegrating these double stigmatized psychiatric patients.
