Abstract
Individuals with a mental illness may be particularly vulnerable during police interviews. Assessing fitness for police interview is vital for ensuring procedural fairness. This article reports the findings of a retrospective review of 31 police interviews of mentally ill persons charged with murder (n = 18) or attempted murder (n = 13) who appeared before the Queensland Mental Health Court. Police interviews were conducted for all murder and 50% of attempted murder cases. Possible or overt mental illness symptoms were present in all interviews. Symptoms of mental illness were pervasive in 36.7% of interviews, intermittent in 43.3% of interviews and seldom in 20% of interviews. Support persons were present for 9.7% of interviews, and legal representation was not present for any interview. These findings highlight the need to enhance access to support persons during interviews. Intersectoral collaboration between mental health services, forensic medical officers, police, public guardians and the legal sector is needed.
Key words: criminal justice, fitness, interview, medico-legal, mental illness, police, unsoundness
Introduction
One of the purposes of a police interview is to obtain admissible evidence regarding the commission of an offence (Ventress, Rix, & Kent, 2008). Procedures for collecting information about a criminal charge through police interview need to be developed with consideration for the suspect’s rights during the criminal justice process. In Australian jurisdictions,1 people who are detained for interview in relation to a criminal charge are protected by a number of fundamental rights that are designed to support fair and judicial criminal justice processes. In Queensland, these include the right to remain silent in the face of questions, time limits around the interview process, access to a lawyer and additional protections for certain vulnerable persons (Queensland Government, 2018). Individuals should be informed of these rights. Multiple factors can impair a person’s ability to understand and exercise legal rights, including: language and cross-cultural communications barriers; and physical, intellectual or mental disability (Bartels, 2011; Medford, Gudjonsson, & Pearse, 2003). Police interviews that progress without due consideration of a person’s fitness for interview risk obtaining evidence that is inadmissible, inaccurate, unfairly disadvantageous or incriminating to an individual (Gall & Freckelton, 1999; Norfolk, 1996).
The disproportionately high prevalence of mental illness among people in contact with the criminal justice system is well established (Fazel & Seewald, 2012; Ogloff, Warren, Tye, Blaher, & Thomas, 2011). This fact alone warrants dedicated consideration of procedural fairness for people with mental illness, including during the process of a police interview (Gregory, 2004; Norfolk, 1996). Correct identification of mental illness among those involved in the criminal justice process is needed to ensure appropriate and timely intervention, including diversion from the criminal system where appropriate (Kent & Gunasekaran, 2010). From the point of arrest through to the interviewing stage, police have the opportunity to intervene if it is considered that an individual has a mental illness that requires treatment and care (Ogloff, Davis, Rivers, & Ross, 2006). Police and prosecutors have discretion not to proceed in a matter if they believe that there is evidence supporting a reasonable doubt about mental capacity or if they believe that it is not in the public interest to proceed (Office of the Director of Public Prosecutions, 2016; Queensland Police Service, 2018).
Fitness for interview
A major difficulty with determining a person’s fitness for interview by police is the absence of standardised definition and criteria for the assessment (Gall & Freckelton, 1999). This undermines objectivity and complicates decision-making processes for both police and medical practitioners. There has been some research undertaken in the United Kingdom to identify appropriate criteria for assessing fitness for interview. Three broad criteria have been identified: (i) understanding of police cautions; (ii) orientation to person, place and time; and (iii) whether there is a likelihood that answers may be provided that could be seriously misconstrued by a court (Gudjonsson, 1995). A study of police detainees in the United Kingdom, which used the above criteria, identified that paranoid beliefs, lack of understanding of police caution and simple questioning, as well as being confused and disoriented, were factors that psychiatrists and forensic medical examiners considered to be reasons for unfitness for interview (Gudjonsson, Hayes, & Rowlands, 2000).
Decisions regarding fitness for interview are time and context specific. For example, a person may be assessed as fit for an interview for a simple offence (e.g. understand their rights and the implications of being interviewed) but not for an interview relating to a more serious charge, where the potential criminal justice outcomes related to the charge may be more complex (Norfolk, 2001). Alternatively, a person’s lack of capacity to participate in a police interview process may not become apparent until the interview is underway (Gudjonsson et al., 2000).
Despite the high rates of mental illness among Australian police detainees (Heffernan, Finn, Saunders, & Byrne, 2003; Ogloff et al., 2011), limited Australian research exists on police interviews of people with mental illness. Research is needed to inform processes for detecting and providing support for people with mental illness, to ensure that criminal justice processes are fair and appropriately support vulnerable persons during the interview process.
This retrospective study is of a sample of individuals with mental illness found not criminally responsible for serious violent offences in Queensland, Australia. The study aims reported in this article are:
to estimate the prevalence of symptoms of mental illness during a police interview among those subsequently found not criminally responsible by reason of insanity (a defence called unsound of mind, USM, in Queensland); and
to describe the specific police practices that were initiated to respond to the needs of people with symptoms of mental illness during a police interview.
Method
Design
This study was a retrospective analysis of available police interview video-recordings, audio-recordings or transcripts contained in the Queensland Mental Health Court files for individuals who met the study’s inclusion criteria (see below). The Mental Health Court is a specialist Supreme Court in Queensland, Australia that makes determinations regarding a person’s criminal responsibility (if USM2 at the time of an offence or not) and fitness for trial. The Court conducts proceedings in an inquisitorial manner and is assisted by two psychiatrists who facilitate explanation of clinical evidence. The Court may make a forensic order that authorises involuntary detention, treatment and care for a person found USM or not fit for trial.
Ethical clearances for this study were granted by the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (no. HREC/15/QRBW/504) and Queensland Police Service research committee (approval date 13 June 2016). Permission to access the data was also obtained from the Mental Health Court President and the data custodian of Queensland’s Consumer Integrated Mental Health Application (CIMHA), a clinical information system that supports mental health clinicians who deliver state-delivered mental health services.
Inclusion criteria (case identification)
For the purposes of this study, mental illness does not include a person with a sole diagnosis of cognitive or intellectual disability.
Eligible cases were identified via the CIMHA database by the data custodian, and the names of individuals who were on forensic orders between 2009 and 2014 were provided to the research team. From this pool, cases were included in the present study, if: (i) the individual was found USM by the Queensland Mental Health Court between 2009 and 2014; (ii) they were charged with murder or attempted murder; and (iii) they were 18 years of age or older at the time of the index offence.
Exclusion criteria for this study were: (i) USM finding was solely attributable to intellectual or cognitive disability; or (ii) the file did not contain adequate information for data analysis.
The primary data source was the police interview contained in the Mental Health Court file, in the form of a transcript, audio-recording or video-recording. All files were reviewed at the Mental Health Court registry. A recorded interaction between police and a person found USM was considered as an interview. In the event an interview was not undertaken, police witness statements were examined to determine the reason for an interview not being conducted.
The interviews consisted of formal interviews, usually held at a police station, and ‘field’ questioning. Field interviews were usually conducted at the scene of the alleged offence, in hospital, en route to a police station or prior to a formal interview. If several interviews were held by the same officers on one day with breaks between, these interviews were treated as a single interview. However, where successive interviews of different types were conducted, or multiple interviews were conducted at different time periods, these were counted as separate interviews.
Procedure
The Mental Health Court files for all persons who met the inclusion criteria were examined by B.C. or B.G. using a standardised proforma that was developed following a literature review and consultation with police. The proforma was piloted on 10 cases and was modified following discussions between B.C. and B.G. A summary of the items collected is outlined in Table 1. Approximately 85% of the interviews accessed were reviewed by two investigators (B.C. and B.G.) to check the collected data and review any discrepancies. Any differences of opinion were resolved through discussions to reach consensus.
Table 1.
Data items extracted from Mental Health Court Files.
| Demographics | Age (years); Country of birth (Australian born, not Australian born); Aboriginal and Torres Strait Islander background (Indigenous, non-Indigenous); Gender (male, female, intersex); Marital status (single, married, divorced/separated); Employment status (employed, pension/unemployed); Education (<Grade 8, Grade 8–10, Grade 11–12, trade qualification, university qualification); Residence (family, married, with others, alone). |
| Offence/s and Mental Health Court | Date of hearing/s; Offence/s and date offence/s committed; Mental Health Court outcomes; |
| Police interview details | How did person come to the attention of police; Date and start/finish time of interview/s; Interview type;Interview format; Number of pages for transcript; Number and rank of police present; Gender of police;Other people present; Interview purpose explained by police and cautions provided; Was the presence of an independent person offered by police; Did the person accept the offer; Individual confessed and when (if an individual stated that they were not sure if they killed a victim, then this was recorded as a partial confession); Was the individual challenged about their account; How did Police explain the purpose of the interview; How did the individual express an understanding of the purpose of the interview; How was the individual asked about their account of the offence; Did the individual provide an account of the offence; Did the officers demonstrate empathy to the individual throughout the interview. |
| Mental health matters | Individual asked about any illness; Individual asked about mental illness; Individual asked about treatment and medication for mental illness; Did the individual report having current or past treatment for mental illness; If applicable, was the individual’s history of mental illness known by police. |
| Cognitive impairment/ intoxication or other communication vulnerabilities at interview | Was there evidence that the individual had difficulty answering questions; Was there evidence of any other communication vulnerabilities; If mental illness or communication difficulties were identified, was the presence of an independent person offered again; Did the individual ask to stop the interview; Was the individual asked about consumption of substances; Did the individual state that they were intoxicated; Did the individual state that they had consumed substances. |
| Symptoms at interview | Are clear symptoms present at any time in the interview; Are symptoms pervasive. |
| Mental state examination | Examples of symptoms present during interview; Appearance and behaviour (video interview only); Mood, Affect; Speech; Thought form; Thought content; Perception; Cognition; Insight. |
A psychiatrist (J.P. or E.H.) or registrar (C.C. or Z.S.) completed the mental state and symptom ratings for all files. The mental state assessment considered appearance and behaviour (video interview only); mood, affect; speech; thought form; thought content; perception; cognition; and insight. Symptom presence was rated as: unable to be assessed; nil; possible; or overt. Symptom pervasiveness was rated as: unable to be assessed; nil; seldom; intermittent; or often. General symptom ratings were also made by non-medically trained authors (B.C. and B.G.) during data collection and were compared with the ratings of a psychiatrist or registrar. Where differences of ratings occurred, these were discussed with J.P. or E.H. who provided the determining decision for symptom ratings.
Data analysis
Data were collated into a data file and analysed using STATA Version 14 (StataCorp, 2015). Analyses were primarily descriptive with percentages based on valid values (i.e. missing values were excluded from calculations). The denominator for analyses reported in Table 2 is the number of cases (people) included in this analysis, categorised in terms of whether an interview was conducted or not. The denominator for Tables 3–5 is the number of interviews included in this study, categorised in terms of whether the interview related to a murder or attempted murder charge.
Table 2.
Demographic characteristics of interviewed and not-interviewed persons.
| Variable | Interviewed (N = 25a) |
Not interviewed (N = 14) |
Total (N = 39) |
|||
|---|---|---|---|---|---|---|
| M (SD) | N (%) | M (SD) | N (%) | M (SD) | N (%) | |
| Type of alleged offence | ||||||
| Murder | 11 (44.0) | 0 (0) | 11 (28.2) | |||
| Attempted murder | 14 (56.0) | 14 (100) | 28 (71.8) | |||
| Age | ||||||
| Mean age in years | 37.6 (17.2) | 36.7 (8.6) | 37.3 (14.6) | |||
| Country of birth and cultural background | ||||||
| Australian born | 18 (72.0) | 11 (78.6) | 29 (74.4) | |||
| Indigenous | 1 (4.0) | 1 (7.1) | 2 (5.1) | |||
| Gender | ||||||
| Male | 21 (84.0) | 12 (85.7) | 33 (84.6) | |||
| Female | 4 (16.0) | 2 (14.3) | 6 (15.4) | |||
| Marital status | ||||||
| Single | 9b (37.5) | 9 (64.3) | 18d (47.4) | |||
| Married | 8 (33.3) | 3 (21.4) | 11 (28.9) | |||
| Divorced/separated | 7 (29.2) | 2 (14.3) | 9 (23.7) | |||
| Employment status | ||||||
| Pension/unemployed | 17 (68.0) | 11 (78.6) | 28 (71.8) | |||
| Education | ||||||
| <Grade 8 | 2b (8.3) | 0 (0.0) | 2d (5.3) | |||
| Grade 8–10 | 11 (45.8) | 5 (35.7) | 16 (42.1) | |||
| Grade 11–12 | 11 (45.8) | 9 (64.3) | 20 (52.6) | |||
| Trade qualificationc | 2 (8.3) | 4 (28.6) | 6 (15.8) | |||
| University qualificationc | 1 (4.2) | 1 (7.1) | 2 (5.3) | |||
| Residence | ||||||
| Family | 16 (41.0) | |||||
| Married | 12 (30.8) | |||||
| With others | 8 (20.5) | |||||
| Alone | 3 (7.7) | |||||
Includes data for two cases with missing interviews.
N = 1 missing, percentages for variable calculated out of N = 24.
Multiple responses allowed.
dPercentages for variables calculated out of N = 38
Table 3.
Characteristics of interview type, length and police and non-police presence.
| Variable | Murder (N = 18) |
Attempted murder (N = 13) |
Total (N = 31) |
||||||
|---|---|---|---|---|---|---|---|---|---|
| M (SD) | Mdn (range) | N (%) | M (SD) | Mdn (range) | N (%) | M (SD) | Mdn (range) | N (%) | |
| Interview type | |||||||||
| Formal | 9 (50) | 6 (46.2) | 15 (48.4) | ||||||
| Field | 9 (50) | 7 (53.8) | 16 (51.6) | ||||||
| Elapsed time between alleged offence and interview days | 0.6 (0.9) | 0 (0–3) | 23.1 (46.2) | 0 (0–134) | 10.7 (32.5) | 0 (0–134) | |||
| Records | |||||||||
| Transcript available | 17 (94.4) | 13 (100) | 30 (96.8) | ||||||
| Transcript length (pp) | 76.0 (54.7) | 54 (11–175) | 34.8 (30.2) | 25 (10–123) | 58.1 (49.6) | 34 (10–175) | |||
| Audio-file | 4 (22.2) | 7 (53.8) | 11 (35.5) | ||||||
| Video-file | 4 (22.2) | 1 (7.7) | 5 (16.1) | ||||||
| Length of recorded interviews (min) | 160 (133.1) | 105 (14–421) | 94.8 (120.5) | 43 (7–345) | 130.65 (128.67) | 90.5 (7–42) | |||
| Police presence at interview | |||||||||
| Mean police present | 3.2 (2.0) | 2 (2–9) | 2.5 (1.2) | 2 (2–6) | 2.9 (1.7) | 2 (2–9) | |||
| Police presence at interview, by rank | |||||||||
| Sergeant | 12 (66.7) | 5 (38.5) | 17 (54.8) | ||||||
| Senior Constable | 14 (77.8) | 12 (92.3) | 26 (83.9) | ||||||
| Constable | 5 (27.8) | 4 (30.8) | 9 (19.0) | ||||||
| Police officer of unknown rank | 7 (38.9) | 6 (46.2) | 13 (31.9) | ||||||
| Presence by person who is not a police officer | 4 (22.2) | 6 (46.2) | 10 (32.3) | ||||||
Note. pp = pages.
Table 4.
Mental state and symptom ratings.
| Variable | Murder (N = 18) N (%) | Attempted murder (N = 13) N (%) | Total (N = 31) N (%) |
|---|---|---|---|
| Symptoms present | |||
| Overt | 14 (77.8) | 10 (76.9) | 24 (77.4) |
| Possible | 4 (22.8) | 3 (23.1) | 7 (22.6) |
| Symptoms pervasive | |||
| Seldom | 5a (29.4) | 1 (7.7) | 6b (20.0) |
| Intermittent | 7 (41.2) | 6 (46.2) | 13 (43.3) |
| Often | 5 (29.4) | 6 (46.2) | 11 (36.7) |
| Rate of speech (elevated/decreased) | 2 (11.1) | 3 (23.1) | 5 (16.1) |
| Thought disorder | 9 (50.0) | 5 (38.5) | 14 (45.2) |
| Impaired cognition | 7 (38.9) | 11 (84.6) | 18 (58.1) |
| Mood elevated/decreased | 2 (11.1) | 10 (76.9) | 12 (38.7) |
| Delusions present | 14 (77.8) | 8 (61.5) | 22 (71.0) |
| Hallucinations | 7 (38.9) | 1 (7.7) | 8 (25.8) |
N = 1 Missing in murder interviews for pervasiveness due to inability to assess, percentages for variable calculated out of N = 17.
bPercentages for variable were calculated out of N = 30.
Table 5.
Police interviewing practices.
| Variable | Murder (N = 18) (%) |
Attempted murder (N = 13) (%) |
Total (N = 31) (%) |
|---|---|---|---|
| Interview purpose explained | |||
| Yes | 16 (88.9) | 12 (92.3) | 28 (90.3) |
| No | 1 (5.6) | 0 (0.0) | 1 (3.2) |
| Not clear | 1 (5.6) | 1 (7.7) | 2 (6.5) |
| Accepts support | |||
| Yes | 5 (27.8) | 3 (23.1) | 8 (25.8) |
| No | 7 (38.9) | 9 (69.2) | 16 (51.6) |
| Not clear | 6 (33.3) | 1 (7.7) | 7 (22.6) |
| Support person present | 1 (5.6) | 2 (15.4) | 3 (9.7) |
| Person confesses | 12 (66.7) | 8 (61.5) | 20 (64.5) |
| Significantly challenged | 3 (16.7) | 4 (30.8) | 7 (22.6) |
| Asked to stop interview | 6 (33.3) | 5 (38.5) | 11 (35.5) |
| Trouble answering questions | 8 (44.4) | 9 (69.2) | 17 (54.8) |
| Language other than English | 0 (0.0) | 1 (7.7) | 1 (3.2) |
| Hearing difficulties | 0 (0.0) | 3 (23.1) | 3 (9.7) |
| Mental health and treatment | |||
| Asked about substance use | 13 (72.2) | 10 (76.9) | 23 (74.2) |
| Reported intoxication | 2 (11.1) | 0 (0.0) | 2 (6.5) |
| Asked about illness | 10 (55.6) | 5 (38.5) | 15 (48.4) |
| Asked about mental illness | 12 (66.7) | 6 (46.2) | 18 (58.1) |
| Mental health issues known | 10 (55.6) | 5 (38.5) | 15 (48.4) |
| Asked about medication | 11 (61.1) | 8 (61.5) | 19 (61.3) |
| In treatment | 11 (61.1) | 5 (38.5) | 16 (51.6) |
| Past treatment | 0 (0.0) | 2 (15.4) | 2 (6.5) |
Results
Sample selection
Figure 1 describes the process of case and interview identification and selection. Out of 43 potentially eligible cases identified via CIMHA, one case was excluded from the analysis due to the person being under the age of 18 at the time of the index offence; one case was excluded due to a forensic order being reinstated on appeal between 2009 and 2014 for an offence that was outside the study’s date range; and two cases were excluded due to the individuals not being found USM. The final cohort comprised 39 cases. For each case, the investigators sought to locate a police interview file from the Mental Health Court. Out of the 39 cases, 25 (64%) were recorded as having some form of interview; however, interviews were not able to be located in the Mental Health Court file for two cases. For the 23 cases with at least one accessible interview, 35 interviews were accessed. Four of these interviews were excluded as being non-informative (for example, they were forensic interviews focused solely on the collection of biological evidence or photographic information), resulting in 31 eligible interviews. Fifteen out of 31 interviews were categorised as formal interviews. In total, 16 people (cases) were interviewed once, six people were interviewed twice, and one person was interviewed three times. Five of those interviewed twice and the person who was interviewed three times were interviewed in relation to murder charges.
Figure 1.
Flow chart describing process of case and interview selection. USM = unsound of mind.
Sample characteristics
Interviews were conducted for all (100%) alleged murder offences and one half (50.0%) of attempted murder cases. Of the 14 persons who were not interviewed for a charge of attempted murder, five declined, three were considered by police to be unfit for interview, two were in hospital, and four declined after consultation with a solicitor.
Table 2 reports key demographic details for this cohort. The sample were predominately male (84.6%), Australian born (74.4%) and on a pension or unemployed (71.8%). Approximately one half of the sample (47.4%) had an educational level that was Grade 10 or below (third year of high school).
Interview characteristics
Table 3 summarises the characteristics of interviews by charge, file type, length of interview and presence of police and non-police persons of different seniority.
While statistical tests of significance were not undertaken, Table 3 nevertheless indicates a substantial difference between the length of interviews (in time and transcript length) relating to murder and attempted murder charges. Interviews for charges relating to attempted murder were an average of 94.8 minutes long, and transcripts were on average 34.8 pages long. By comparison, interviews for charges relating to murder were on average 160 minutes long, and transcripts were 76.0 pages long. Table 3 shows a further substantial difference in the mean time elapsed between the alleged offence and interview, with the time elapsed being much lower for interviews relating to murder charges than attempted murder charges. However, these figures are skewed due to interviews with two individuals in the attempted murder group: one who evaded arrest and was not interviewed for 134 days, and another who was first taken to hospital and had two interviews 93 days after the alleged offence.
On average, approximately three police were present for each interview. While it was not possible to determine at least one police officer’s rank in 31.9% of interviews, available data indicate that more senior officers (Sergeant) were more likely to be present during an interview on a murder charge than on an attempted murder charge.
Prevalence of symptoms of mental illness
Table 4 summarises the findings of mental state assessments relevant to capacity to be interviewed, including general symptoms ratings. Possible or overt symptoms of mental illness were present during all (100%) interviews. Symptom pervasiveness was assessed as frequent in 36.7% of interviews, intermittent in 43.3% of interviews and seldom in 20.0% of interviews.
The most common symptoms of mental illness were delusions, identified in 71.0% of interviews. However, symptoms varied across the two charge types. Delusions and thought disorder were the most prevalent symptoms in interviews relating to murder charges (77.8% and 50%, respectively) but in the attempted murder group, impaired cognition and elevated or decreased mood were the most common symptoms of mental illness. Impaired cognition was present in 84.6% of the attempted murder interviews (compared with 38.9% of murder interviews), and elevated or decreased mood was evident in 76.9% of attempted murder interviews (compared with 11.1% in murder interviews).
Police interview practices
The interviews were reviewed in relation to a number of factors pertaining to police procedural requirements as well as questions about an interviewee’s mental state. These factors are reported in Table 5.
The purpose of the interview was explained to the interviewee in 90.3% of interviews. There were three field interviews where it appeared that the purpose had not been explained to the interviewee. However, all persons had the purpose of an interview explained to them at least once.
Although support was accepted in approximately one quarter (25.8%) of all interviews, a support person was rarely present (three cases in total or 9.7% of interviews). Family members acted as support persons for two attempted murder interviews, and in one murder interview a Justice of the Peace was requested by police to be present. Reasons given in interviews for support persons rarely being present were: (i) that support was not available due to a contact person being unavailable; (ii) distance precluding a person’s attendance; or (iii) an inappropriate person (e.g. the arresting officer or a witness to the offence) being requested. In 35.5% of interviews, the person being interviewed expressed the wish to end an interview; in one interview a support person asked to stop the interview. The majority of the interviews elicited a ‘confession’ (64.5%). In some instances, confessions had been made prior to the interview.
Table 5 also includes information regarding whether police asked about factors relevant to mental health (including questions about illness or mental illness, medication, treatment and substance use). Police were more likely to ask about substance use (74.2%) than about mental illness (58.1%) in the context of a police interview.
Discussion
Prevalence and identification of mental health symptoms
The study found evidence in all the interviews of possible or overt symptoms of mental illness; however, symptoms were considered pervasive in only 36.7% of interviews. Persons being interviewed were asked about whether they had a mental illness in approximately 60% of interviews. In contrast to questions about mental illness, police were more likely to ask about substance use. Reviews of some transcripts indicated that symptoms of mental illness may have posed a barrier to police obtaining information about mental illness. However, in some interviews, police did not ask about a person’s mental health history; it is possible that this is because they already knew the person’s history. Nevertheless, best practice would be for this information to be checked with the interviewee. Variability in this process, that this review identified, indicates a need for training in the identification of symptoms of mental illness and standardisation in how relevant history about mental health treatment is obtained.
If police do not ask or are unaware of a mental illness history, non-pervasive symptoms could be readily overlooked or insufficient weight given to them. One mechanism for improving the identification of mentally ill individuals early is the use of health screening tools for detainees in police custody. However, these tools can have low detection rates for mental illness, and, where tools are administered by police or criminal justice staff, they can have limited utility in terms of determining an appropriate course of action if mental illness is identified (Baksheev, Ogloff, & Thomas, 2012; McKinnon & Grubin, 2013). An evaluation of the standardised health screening tool used by police in the United Kingdom reported that the presence of mental illness (of any type) was only correctly identified in 52% (n = 50) of cases reviewed (N = 96; McKinnon & Grubin, 2013). Similarly, a study from two Melbourne metropolitan police stations, of 150 adults, that compared police screening and health screening, suggested that police screening processes produced high rates of false negatives, with only 47% of individuals correctly identified as having a diagnosed mental illness (Baksheev et al., 2012). Additional complexities can arise even when health staff are involved in the screening processes because these may take place after police intervention or interview has already occurred, limiting the opportunities to intervene if there are concerns about fitness for interview.
Another challenge with the use of health screening tools for the purpose of identifying those that may be unfit for interview is that they are designed to identify treatment needs, rather than questions of capacity (Nicholls, Roesch, Olley, Ogloff, & Hemphill, 2005). In contrast, assessment of fitness to be interviewed centres on capacity rather than the presence or absence of a mental illness (Ventress et al., 2008). In addition, understanding fitness for interview requires consideration of the context of the interview – for example, the nature of the task and the relevant supports that are available (Kent & Gunasekaran, 2010).
Police interview practices
A surprising finding was that only half of those charged with attempted murder were interviewed compared to all those who were charged with murder. It is unclear why individuals charged with attempted murder were less likely to be interviewed. One possible explanation is that police more rigorously investigate murders (interviews tended to be longer for murder cases). Another possibility is that individuals charged with murder were more disturbed and unable to exercise their rights. However, the estimated rate of overt symptom presence was broadly equivalent for both charge types (77.8% and 76.9% in murder and attempted murder interviews, respectively). Further research is required to explore factors influencing police decision making regarding the decision to interview or not, as well as the impact of offence seriousness on this decision making.
The purpose of the interview was explained in 90.3% of interviews. All persons interviewed had the purpose of the interview explained to them at least once, although the detail and quality of the explanation – for example, quality of the information regarding the right to contact a person – varied greatly. In some instances, police checked that the person being interviewed understood and could explain what the advice given to them meant.
A range of research-informed strategies have been suggested to enhance interviewing and communication with vulnerable subjects (Bull, 2010). For example, Bull (2010) identified the following four steps as a means of obtaining the best evidence when dealing with vulnerable witnesses:
establish good rapport, including establishing the ground rules and advising the interviewee that it is acceptable to say if they do not understand or know the answer;
obtain as much free narrative as possible, encouraging the interviewee with prompts and open-ended questions such as ‘tell me more about that’ and ‘what happened next?’;
ask questions of the right type in the right order. For example, open questions should precede specific questions and then closed questions. Leading questions should only be used as a last resort; and
have meaningful closure, including a summary of the interviewee’s evidence and providing them with an opportunity to correct any errors; and evaluate the interview, in terms of both the information obtained and the interviewer’s performance.
Presence of a support person
There were low levels of support persons present at interview. This is particularly relevant to the issue of how police dealt with a request to end the interview. In two cases the interview was ended when requested; however, in other instances a variety of strategies were employed to continue the interview. It is difficult from transcripts to determine the level of stress or emotion associated with a request to end an interview. The practice of continuing an interview after a request to end the interview has to be considered in terms of what is standard police practice in such situations and the specific context of the request.
Absence of legal representation is of particular concern given the severity of the offences in question. In many jurisdictions, approaches to addressing negative outcomes for potentially vulnerable individuals throughout police interview processes have primarily focused on support person models. These models aim to provide legal representation or assisted decision-making support for police interview processes involving a vulnerable person, including persons with a mental illness (Brewin & Bailey, 2005; Bull, 2010; Gudjonsson et al., 2000; Nemitz & Bean, 2001). However, there are challenges with implementing a model that mandates the presence of a support person because of the need to have clear guidelines about their role and the practicalities of ensuring that adequately trained and experienced persons are readily available (Brewin & Bailey, 2005). It has been identified that providing training in the health, police and legal sectors about the support person role can increase awareness of the importance of identifying vulnerabilities, including mental illness, in addition to highlighting the benefits of support persons in facilitating communication during police questioning (Brewin & Bailey, 2005).
The extent to which support persons are utilised in Australia is not known. A review of the legislation, policies and procedures in Australia for supports and protections provided to vulnerable persons, including persons with a mental illness, during police interviews has identified some consistency across jurisdictions (Bartels, 2011). For example, Queensland, New South Wales and Victoria all consider the role of support persons in interviews, where capacity may be in question, in legislation. South Australia, Tasmania, the Australian Capital Territory and the Commonwealth also provide limited legislated requirements for interviewing vulnerable persons. However, these requirements relate primarily to communication; specifically, they focus on procedures for interviewing people who are non-English speaking or who are unable to read or write, and thus require that police provide alternative means for obtaining statements (e.g. via interpreters, or video- or audio-recordings; Bartels, 2011).
Confessions
The proportion of persons who confessed to the offence (an admission regarding committing the offence; 69.6%) is higher than rates identified in the literature of between 36% and 58% (Pearse, Gudjonsson, Clare, & Rutter, 1998). It was notable that many of the confessions occurred at an early stage of the interview, including before an interview had begun. Based on the available police material in the Mental Health Court files, it was difficult in some cases to determine exactly what communication had occurred prior to the interview. Field interviews sometimes documented confusing scenes, with multiple police and other persons present.
Mental illness has been identified as a factor that increases the risk of a person making a false confession (Norfolk, 1997). The comparatively high rates of confession in this study, and the implications for individuals if a confession is falsely provided, lend further support to the case for improving the identification of fitness issues and support needs for people with mental illness who are being interviewed by police. A confounding factor in these data is the comparatively low levels of educational attainment when compared against Australian benchmarks. Australian Bureau of Statistics data indicate that 66% of adults aged 20–64 years have at least one non-school qualification (Australian Bureau of Statistics, 2017). By comparison, only 21% of persons in our sample had a non-school qualification. The possible impacts of educational attainment is compounded by the prevalence of impaired cognition in the sample.
Implications for policy and practice
As a first step, a more detailed definition of, and criteria for, fitness for police interview are needed. There is a pressing need for policy and procedure development regarding fitness for police interview and mental illness. Policies and procedures could include:
development of criteria for assessing fitness;
referral pathways to health staff for fitness assessments;
requirement to have police assessment reviewed by health staff prior to interview;
pathways for engaging mental health practitioners in police interviews, particularly for serious charges;
further development of publicly available guidelines on the role, function and purpose of support persons; and
training in the role requirements for a cohort of persons or service able to perform the function of support persons (in addition to family members and informal support persons).
Policies and procedures relating to fitness for interview should be further developed, acknowledging that undertaking a police interview at the time of the offence can be of significant value to the investigation. It is important that any operational processes are well defined and consistently implemented. Early identification of those with mental illness not only helps protect the rights of individuals, but can avoid the unnecessary expenses of progressing criminal charges that are ultimately dismissed or a trial that is stopped due to problems with the evidence obtained from a vulnerable suspect.
Police need to be able to access information about a person’s mental health status readily prior to interview so they can take this into consideration in determining whether a person is fit for interview. Timely access to accurate information about a person’s mental health status is paramount to ensuring fair and judicious interview processes for both interviewees and the broader community.
The primary function of the police interview is investigative – that is, to collect information from persons (e.g. the suspect, witnesses, victims and persons of interest) about the commission of a crime. However, while interviews are time sensitive (evidence may disappear) and time limited (questioning is limited to 8 hours), the police must also respect the rights of persons being interviewed and demonstrate ‘best practice’ policing of vulnerable persons. Notably, from the perspective of police, allowing the person charged to participate in an interview affords them procedural fairness, in terms of being able to put forward their views about the offence. In this context, the timely identification of support needs for interview is an important element in ensuring a person’s rights are protected throughout this process.
There is a clear and pressing need to address the overall absence of support persons during police interviews. To assist in this, facilitating police access to information from relevant appointers of substitute decision makers, such as Civil and Administrative Tribunals, in order to help identify and liaise with an appointed guardian, should be considered. Further, models for trained support persons, such as that utilised in the United Kingdom, could be considered (McKinnon & Grubin, 2013). In such models, support persons are available for all persons with a mental illness subject to a police interview (McKinnon & Grubin, 2013). The primary role of the support person is to offer assistance to explain questioning or legal processes, and observe and advise where required during the police interview (Brewin & Bailey, 2005; Gudjonsson, 1995). Although this model has assisted to increase efforts by police to identify mental health issues and facilitate support, there remains difficulties in finding suitably trained support persons who are available at short notice to support interviewees (Brewin & Bailey, 2005).
In the longer term, the development of a structured professional judgement tool or screening tool for assessing fitness for interview may be useful.
Limitations
Several limitations to this study should be noted. First, the study is retrospective based on interviews of those who were USM at the time of the offence. Consequently, it is not possible to identify how the case and interview characteristics found here compare with police interviews conducted with those, with or without a mental illness, who were not USM. Second, determination of symptoms of mental illness, including presence and pervasiveness, were based on review by health professionals of interviews carried out by police. While identification of symptom presence was probably reduced due to the interviews not being carried out by interviewers trained in mental health assessment, the aim of the research was to examine police interviewing. Third, as the interviews analysed were mostly in the form of transcripts, valuable information regarding the dynamics of the interview, non-verbal behaviour and speech was not available. Fourth, the absence of transcript review by police personnel means that some nuances relating to the application of formal police procedures for interviewing may have been overlooked. Fifth, the quality of the interviews available on the Mental Health Court file varied. At times, key responses by interviewees were unclear, interview start or end times were not recorded, or persons present were not clearly identified. These factors reflect the limitations of research based on naturally occurring data.
Sixth, murder offences have a low base rate. Selecting a 5-year time frame produced a relatively small sample, which was reduced due to the number not interviewed. Consequently, the sample was deemed not to warrant undertaking statistical tests of association. Seventh, this was a study conducted in Queensland, and practices vary across Australia. While there is similarity of police interviewing practices and laws governing criminal procedure and evidence, generalisability of this study to other jurisdictions needs to be tested.
Future research directions
Several areas of future research are warranted. First, future research could compare findings with an assessment of the characteristics of persons charged with other offences. That is, while the burden of proof is usually the same for all criminal matters, murder and attempted murder can involve more complex criminal justice processes. Consideration of the interview practices for less serious offences is also warranted to ensure that access to support persons and appropriate consideration of fitness for interview occur at all levels of the criminal justice system. Second, comparison with a sample of individuals interviewed by police on murder or attempted murder charges, but not found USM, would allow for more nuanced inferences to be made. Third, as the interviews were reviewed by researchers with a health background and not police, the overall strategy of the interview or the use of appropriate or inappropriate question types was not examined. Instead, particular police interview practices which could be identified by health practitioners, and which relate to determining symptom presence or the person’s understanding of their rights, were examined. It is noted, however, that the interview strategy and questioning type can have a significant impact on an interview outcome, especially for vulnerable and potentially highly suggestible interviewees. How mentally ill persons interpret the cautions and information given to them regarding their rights in a police interview is another area for potential research.
Conclusion
The primary purpose of the police interview is to gather reliable information about a criminal charge and not to look for symptoms of mental illness. In law there is a presumption of sanity, which may also guide police conduct in investigations. The most pressing need that the findings presented in this article highlight is the need to enhance access to support persons during interviews. Collaboration between mental health services, forensic medical officers, police, public guardians and the legal sector in this area is necessary to achieving this goal.
Notes
Crimes Act 1914 (Cth), Law Enforcement (Powers and Responsibilities Act 2002 (NSW), Police Powers and Responsibilities) Act 2000 (Qld), Criminal Code Act 1899 (Qld), Summary Offences) Act 1953 (SA), Criminal Law Detention and Interrogation Act 1995 (Tas), Crimes Act 1958 (Vic), Criminal Investigation Act 2006 (WA).
A person may be found of unsound mind under the repealed Queensland Mental Health Act 2000. The legal test for insanity is defined under s. 27 of the Criminal Code Act 1899: ‘A person is not criminally responsible for an act or omission if at the time of doing the act or making the omission the person is in such a state of mental disease or natural mental infirmity as to deprive the person of capacity to understand what the person is doing, or of capacity to control the person’s actions, or of capacity to know the person ought not do the act or make the omission’.
Ethical standards
Declaration of conflicts of interest
Bobbie Clugston has declared no conflicts of interest
Bob Green has declared no conflicts of interest
Jane Phillips has declared no conflicts of interest
Zara Samaraweera has declared no conflicts of interest
Carolina Ceron has declared no conflicts of interest
Cameron Gardner has declared no conflicts of interest
Carla Meurk has declared no conflicts of interest
Ed Heffernan has declared no conflicts of interest
Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (no. HREC/15/QRBW/504) and Queensland Police Service research committee (approval date 13 June 2016), and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent
A waiver of consent was approved under Queensland’s Public Health Act, 2005. (Public Health Act Approval number RD006043).
Acknowledgements
The authors gratefully acknowledge Professor Simon Bronitt for his helpful comments on the draft manuscript. The authors acknowledge the assistance of the Queensland Police Service (QPS) in the conduct of the research presented. The views expressed in this article are not necessarily those of QPS, and any errors of omission or commission are the responsibility of the author/s.
References
- Australian Bureau of Statistics. (2017). 6227.0 – education and work, Australia, May 2017 Canberra, Australia: AVS. Retrieved from http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/6227.0May%202017?OpenDocument.
- Baksheev G., Ogloff J., & Thomas S (2012). Identification of mental illness in police cells: A comparison of police processes, the Brief Jail Mental Health Screen and the Jail Screening Assessment Tool. Psychology, Crime & Law, 18(6), 529–542. doi: 10.1080/1068316X.2010.510118 [DOI] [Google Scholar]
- Bartels L. (2011). Police interviews with vulnerable adult suspects. Canberra. Retrieved from http://www.aic.gov.au/media_library/publications/rip/rip21/rip21.pdf
- Brewin S., & Bailey A (2005). Appropriate adults provided for vulnerable people questioned under caution: Implications for policy of the Criminal Justice Review in Northern Ireland. The Journal of Adult Protection, 7(1), 4–14. doi: 10.1108/14668203200500002 [DOI] [Google Scholar]
- Bull R. (2010). The investigative interviewing of children and other vulnerable witnesses: Psychological research and working/professional practice. Legal and Criminological Psychology, 15(1), 5–23. doi: 10.1348/014466509X440160 [DOI] [Google Scholar]
- Fazel S., & Seewald K (2012). Severe mental illness in 33 588 prisoners worldwide: Systematic review and meta-regression analysis. British Journal of Psychiatry, 200(5), 364–373. doi: 10.1192/bjp.bp.111.096370 [DOI] [PubMed] [Google Scholar]
- Gall J., & Freckelton I (1999). Fitness for interview: Current trends, views and an approach to the assessment procedure. Journal of Clinical Forensic Medicine, 6(4), 213–223. doi: 10.1016/S1353-1131(99)90000-7 [DOI] [PubMed] [Google Scholar]
- Gregory M. (2004). Fitness to be interviewed. Journal of Clinical Forensic Medicine, 11(5), 260–263. doi: 10.1016/j.jcfm.2004.04.011 [DOI] [PubMed] [Google Scholar]
- Gudjonsson G. (1995). Fitness for interview; during police detention: A conceptual framework for forensic assessment. Journal of Forensic Psychiatry, 6(1), 185–197. doi: 10.1080/09585189508409884 [DOI] [Google Scholar]
- Gudjonsson G., Hayes G., & Rowlands P (2000). Fitness to be interviewed and psychological vulnerability: The views of doctors, lawyers and police officers. Journal of Forensic Psychiatry, 11(1), 74–92. doi: 10.1080/095851800362373 [DOI] [Google Scholar]
- Heffernan E., Finn J., Saunders J., & Byrne G (2003). Substance-use disorders and psychological distress among police arrestees. Medical Journal of Australia, 179(8), 408. [DOI] [PubMed] [Google Scholar]
- Kent J., & Gunasekaran S (2010). Mentally disordered detainees in the police station: The role of the psychiatrist. Advances in Psychiatric Treatment, 16(2), 115–123. doi: 10.1192/apt.bp.107.004507 [DOI] [Google Scholar]
- McKinnon I., & Grubin D (2013). Health screening of people in police custody—evaluation of current police screening procedures in London, UK. The European Journal of Public Health, 23(3), 399–405. doi: 10.1093/eurpub/cks027 [DOI] [PubMed] [Google Scholar]
- Medford S., Gudjonsson G., & Pearse J (2003). The efficacy of the appropriate adult safeguard during police interviewing. Legal and Criminological Psychology, 8(2), 253–266. doi: 10.1348/135532503322363022 [DOI] [Google Scholar]
- Nemitz T., & Bean P (2001). Protecting the rights of the mentally disordered in police stations: The use of the appropriate adult in England and Wales. International Journal of Law and Psychiatry, 24(6), 595–605. doi: 10.1016/S0160-2527(01)00087-5 [DOI] [PubMed] [Google Scholar]
- Nicholls T., Roesch R., Olley M., Ogloff J., & Hemphill J (2005). Jail Screening Assessment Tool (JSAT): Guidelines for mental health screening in jails. Burnaby, BC: Mental Health, Law, and Policy Institute, Simon Fraser University. [Google Scholar]
- Norfolk G. (1996). Fitness to be interviewed and the appropriate adult scheme: A survey of police surgeons’ attitudes. Journal of Clinical Forensic Medicine, 3(1), 9–13. doi: 10.1016/S1353-1131(96)90039-5 [DOI] [PubMed] [Google Scholar]
- Norfolk G. (1997). Fitness to be interviewed – a proposed definition and scheme of examination. Medicine, Science and the Law, 37(3), 228. doi: 10.1177/002580249703700307 [DOI] [PubMed] [Google Scholar]
- Norfolk G. (2001). Fit to be interviewed by the police — an aid to assessment. Medicine, Science and the Law, 41(1), 5–12. doi: 10.1177/002580240104100102 [DOI] [PubMed] [Google Scholar]
- Office of the Director of Public Prosecutions. (2016). Director’s guidelines: As at 30 June 2016. Queensland, Australia: Department of Justice and Attorney-General. [Google Scholar]
- Ogloff J., Warren L., Tye C., Blaher F., & Thomas S (2011). Psychiatric symptoms and histories among people detained in police cells. Social Psychiatry and Psychiatric Epidemiology, 46(9), 871–880. doi: 10.1007/s00127-010-0256-5 [DOI] [PubMed] [Google Scholar]
- Ogloff J., Davis M., Rivers G., & Ross S (2006). The identification of mental disorders in the criminal justice system Retrieved from Canberra: http://www.criminologyresearchcouncil.gov.au/reports/2006-ogloff.pdf
- Pearse J., Gudjonsson G., Clare I., & Rutter S (1998). Police interviewing and psychological vulnerabilities: Predicting the likelihood of a confession. Journal of Community & Applied Social Psychology, 8(1), 1–21. doi: 10.0.5.68/135532510X500064 [DOI] [Google Scholar]
- Queensland Government. (2018). Being questioned Retrieved from https://www.qld.gov.au/law/crime-and-police/being-arrested-and-police-custody/being-questioned
- Queensland Police Service. (2018). Operational procedures manual - Chapter 3 Queensland, Australia: Queensland Police Service Retrieved from https://www.police.qld.gov.au/corporatedocs/OperationalPolicies/Documents/OPM/Chapter3.pdf.
- StataCorp. (2015). Stata Statistical Software: Release 14. College Station, TX: StataCorp LP. [Google Scholar]
- Ventress M., Rix K., & Kent J (2008). Keeping PACE: Fitness to be interviewed by the police. Advances in Psychiatric Treatment, 14(5), 369–381. doi: 10.1192/apt.bp.107.004093 [DOI] [Google Scholar]