Abstract
In Portugal, offenders found not guilty by reason of insanity (NGRI), may be given a restricted order to a special hospital as an alternative to prison. In European countries there is a recognized need for data concerning this special population. The aim of the present study was to examine the characteristics of all the NGRI subjects (n = 274) detained in the country in a descriptive and retrospective survey conducted in January 2009. Offence committed, demographic factors, diagnosis at admission, background of substance abuse and diagnostic stability were recorded. Schizophrenia was the commonest diagnosis (51.5%). Mean population age was 42.6 years, with only 6.2% women. Homicide was the most common offense (41.2%). A background of substance abuse was found in 42.3% of subjects. There were significant differences in the schizophrenia and mental retardation patient groups when compared individually with the other diagnoses concerning homicide and arson as the offence. Mean duration of inpatient stay did not differ significantly between diagnoses. The findings also point to poor follow-up of the NGRI patients after admission.
Introduction
In Portugal, a suspect assumed to be mentally ill may be found not guilty by reason of insanity (NGRI) under article 20 of the Criminal Code [Criminal Code article 20: (1) An offender may be deemed not guilty by reason of insanity if due to a psychic anomaly, at the time of committing the act he is incapable of knowing the nature and quality of the act or to know that it was wrong (…)]. In these cases forensic assessment is a legally defined prerequisite for trial. The Court requests the National Institute of Forensic Medicine to designate a psychiatrist as an expert, after which the subject's mental status is evaluated through clinical examination and psychological interviews. The Court can then give a mentally disordered offender considered a dangerous NGRI a restricted order to a special hospital as an alternative to prison, complying with article 91 of the Criminal Code [Criminal Code article 91: (1) An offender who has practised an illicit act and is deemed not guilty by reason of insanity is ordered to be admitted in a recovery, treatment or security establishment when due to its psychic anomaly and the severity of the illicit act there is a founded concern of recurrence of acts of similar nature (…)]. A total of three such special hospitals exist in the country, comprising around 300 secure beds: Centro Hospitalar Psiquiátrico Coimbra, Centro Hospitalar Psiquiátrico Lisboa and Santa Cruz Bispo (this last one is located inside a prison establishment).
Although the vast majority of the patients detained in these facilities fall into the NGRI category, there is a small number of general convicted prison inmates who are currently suffering from psychiatric disease and were admitted for treatment, and a residual number of presumptive NGRI individuals preventively detained who are awaiting a Court decision. However, in contrast to other countries such as England and Wales, 1 these facilities are not used to treat difficult to manage non-offender patients, which stand clearly apart from the forensic system in Portugal.
There is a reported general growth in the number of forensic patients and overcrowding of facilities throughout Europe, 2 which calls for the need to pull together the facts and figures concerning the NGRI patients in order to inform debate. As far as we know, no studies have examined the characteristics of the whole population of NGRI patients in the country. This is consonant with the recognized paucity of basic information and evidence concerning the cases admitted to specialized forensic facilities across Europe. 2 Bearing these needs in mind, this study reports on offending behaviour, psychiatric diagnoses (including substance abuse), demographic factors and duration of inpatient treatment/security measure of the total population of NGRI inpatients in the country.
Method
This is a retrospective and descriptive cross-sectional study. All 274 individuals deemed NGRI detained in the specialized Psychiatric Forensic units mentioned above in January 2009 were included in the study. Data were gathered primarily via a retrospective case-note review and recorded by a clinically trained research psychiatrist/psychologist and included demography, diagnosis (both admission and current), criminal behaviour and conviction. Conditions such as previous history of drug dependence and alcoholism were also recorded. Diagnoses were coded under the ICD-10 classification. Diagnosis of mental retardation in a NGRI subject implied IQ testing besides clinical examination. In cases of schizophrenia and concomitant mental impairment, the former was considered the main psychiatric diagnosis if the patient was psychotic at the time of committing the offence and fulfilled the diagnostic criteria for schizophrenia.
Chi-square test was used to compare categorical variables while analysis of variance (ANOVA) and Kruskal–Wallis test were used for continuous variables. File integrity was generally high and a fairly complete data-set could be obtained for all variables, except for mean duration of inpatient stay (valid n = 202). Absent values were confirmed to be data missing completely at random (MCAR) for the former cases and a listwise deletion approach was used when performing the analysis. Missing values for duration of inpatient stay were coded using multiple imputation. Statistical analysis was performed using SPSS version 17 and P values ≤0.05 were considered significant.
Results
Demographic characteristics
The demographic characteristics of the population are summarized in Table 1. Mean age was 42.6 (SD = 13.2) years, range 18–86. Mean age according to diagnostic subgroup was: schizophrenia 43.7 (SD = 12.5) years, mental retardation 36.8 (SD = 9.9) years, personality disorder 42.8 (SD = 15.7) years, other 47.8 (SD = 16.0) years. There was a significant difference in age according to diagnosis (Kruskal-Wallis test, df = 3, P < 0.01). Almost all (93.8%) patients were men. No differences concerning diagnosis were significant between sexes. Only 22.3% patients were known to have been married or were married at the time of study.
Demographic characteristics
Criminal and psychiatric features
Full details of admission diagnosis, type of crime and substance abuse are shown in Table 2. Type of crime, substance abuse and mean duration of inpatient stay according to diagnosis at admission are shown in Table 3. In this study 51.5% patients were diagnosed at admission with schizophrenia, 20% mental retardation, 6% personality disorder and 22% had other diagnoses. This last group included patients diagnosed with bipolar disorder, psychotic depression, mental disorders due to psychoactive substance, schizoaffective disorder and delusional disorder.
Diagnosis at admission, type of crime and substance abuse
Type of crime, substance abuse and mean duration of inpatient stay by diagnosis at admission
*P < 0.01, **P < 0.001 (chi-square test), compared with the other diagnosis
There were significant differences in the schizophrenia (2χ2 = 9.96, df = 1, P < 0.01) and mental retardation (χ2 = 12.81, df = 1, P < 0.001) groups when compared individually with the other diagnoses concerning homicide as an offence. Differences were also found to be significant concerning arson for the schizophrenia (χ2 = 8.88, df = 1, P < 0.01) and mental retardation (χ2 = 8.57, df = 1 P < 0.01) groups. A background of substance abuse (alcohol and drugs) and mean duration of inpatient stay did not differ significantly between diagnoses. No differences concerning type of offence reached statistical significance between the sexes.
Diagnosis at admission was compared with actual diagnosis for subjects admitted more than six months before the time of study. Diagnostic stability of the overall population was 88% with stable diagnosis versus 10.6% modified and 1.4% unknown. In the schizophrenia subgroup only 4.6% had their initial diagnosis changed.
Discussion
Methodological considerations
Attempts were made to limit the effects of bias introduced by the retrospective nature of data collection by having comprehensive records available for the majority of referrals and by collecting information in a standardized manner. Missing data for most variables were considered to be data MCAR and due to the small number of absent values the list wise deletion approach did not diminish significantly the power of the analysis. ANOVA of mean duration of inpatient stay (Table 3) using multiple imputation for missing values did not reach statistical significance. Since this finding was consistent for the four performed imputations it can be considered quite robust.
International comparisons have to be interpreted within the frame that forensic populations are usually defined legally and that legislation differs between countries. Also, the cross-sectional study design dictates that care should be exercised when extrapolating results.
Findings
The average age of patients was 42.6 years, which concurs with the mean age of 42 of the only similar study in our country, which involved a subset of NGRI schizophrenic patients. 3 Compared with other European countries, this result is also similar to the mean population age of 37 years at a German Forensic Psychiatric Hospital in 20004 and to the mean age of 34.3 years of NGRI defendants in England from 1997 to 2001. 5 Patients aged 60 and above comprised 8% of the total of NGRI patients, differing from the 2% of total found in a study of elderly patients admitted in England and Wales. 6 The overwhelming majority of patients were men and only 6.2% were women, a value lower than the average proportion of female offenders in Europe. 7 Nevertheless, criminality among mentally disordered offenders consistently appears to be an overwhelming masculine behaviour throughout several countries. 5,8–11 Although no differences reached statistical significance between sexes it is noteworthy that three out of the 17 women subjects were convicted of arson and there was not a single female sex offender. As for marital status, 65% of the subjects were single, higher than the country's general population at this age. One plausible explanation would be that the subjects had been severely mentally disturbed for a long time prior to committing the crime, with the consequent difficulty in forming relationships leading to marriage.
Homicide was perpetrated by 41.2% of patients, a value far higher than in other countries, including England. 4,5,11,12 One possible explanation could be that psychiatric examination is more readily requested by prosecutors in our country in cases of homicide. The low proportion of murder charges in England can be attributed to the fact that an automatic restriction order is implied in a NGRI murder verdict, which makes some defendants seek a manslaughter verdict instead on the basis of diminished responsibility. 5
Only 6% of patients had personality disorder, a low rate compared with England and Wales 11 but comparable to New Zealand 12 and Tokyo. 8 One explanation for this finding could be that in contrast to England, where legislation clearly states that psychopathic disorder amenable to treatment should be included in the NGRI category, 13 in Portugal a large percentage of people with personality disorders, particularly psychopathic, are tried as common criminals and sent to jail, as they are considered criminally responsible according to the psychiatric report.
A background of substance abuse (alcohol and drugs) was found in 42.3% of subjects. The presence of substance abuse has been found to be associated with increased violence in mental patients in general, 14 which would explain not finding differences between diagnoses.
Duration of inpatient stay did not differ significantly between diagnoses. Therefore one can infer that the Court gives out the sentence primarily according to the offence committed. Lengthy periods of inpatient care seem to be common in NGRI patients. These inpatient stays for many years are considered to be unnecessary and can even be counterproductive. 12 However, this situation is likely to continue until forensic treatment programmes can demonstrate effective outcomes.
Concerning the subjects' diagnostic stability, it is noteworthy that in the patients diagnosed with schizophrenia only six (4.2%) and seven (4.9%) subjects had their initial diagnosis changed, respectively, at six months and two years after admission. This contrasts with the general retrospective studies concerning the diagnostic stability of schizophrenic patients, where results differ from 73.1% at six months and 55% at two years of follow-up in an Anglo-Saxon population 15 and 45.9% in an Iberian population. 16 These findings point to poor follow-up of NGRI patients after admission, despite the legal requirement in Portugal of patient reassessment every 1–2 years. 17
In short, these results raise doubts as to whether the criteria for mental examination of offenders are set in a consistent and coherent manner, as a greater proportion of homicidal NGRI patients were found compared to other European countries. Also, it remains open to discussion if forensic services have the resources to perform consistent follow-up of detained patients. Questions are also raised whether detention should be linked to the seriousness of offence or until treatment is complete.
As a closing remark, this study draws attention to the importance of further research involving follow-up and outcome of NGRI patients. It is hoped that future political agendas will address these important issues.