The National Trajectory Project (NTP): Examining Profiles \u0026 Recidivism through a Gendered Lens

$Page 1: The National Trajectory Project (NTP): Examining Profiles \u0026 Recidivism through a Gendered Lens$
Tonia Nicholls, University of British Columbia and BCMHSUS

Anne Crocker, McGill and Douglas Institute

Michael Seto, University of Ottawa Yanick Charette, Douglas Institute and Université de Montreal

Catherine Wilson, University of British Columbia and BCMHSUS

Gilles Côté, Institute Philippe Pinel de Montreal and Université Trois Rivieres

Malijai Caulet, Douglas Institute

The National Trajectory Project (NTP): Examining Profiles & Recidivism through a Gendered Lens.

APLS March, 2014

Introduction

• Research has yielded somewhat equivocal findings (e.g., de Vogel et al., 2012; Heilbrun et al., 2008; Nicholls, Cruise, Greig, & Hinz, accepted)

• Largely concentrated in correctional samples

• This may be owing to small samples of female forensic consumers

– inability to comment on subgroups (diagnostic groups); – low base rate events (sexual offences)

•Many experts assert that management and treatment of justice involved women should be ‘gender‐specific’.

• Mandate of Review Boards is to protect public safety while safeguarding the rights of the NCRMD accused.

– RB must consider (Criminal Code s. 672.54)

• Protection of the public from dangerous persons• Mental condition of the accused• Reintegration of the accused into society• Other needs of the accused

Make a disposition decision that is the “least onerous and least restrictive to the accused.”

• To the extent that dangerousness, mental condition and other needs vary by gender these issues should be taken into consideration in treatment planning.

Rationale: The Present Study

Objectives and Hypotheses

Objective 1 : Compare demographic and psychosocial characteristics and of men and women found not criminally responsible .• Women will have more greater needs / poorer functioning.

Objective 3 : Examine the characteristics of the index offence.• There will be differences in the relationship with the victim of the offence by gender but women will be no less likely to have perpetrated serious index offences against persons.

Objective 2 : Report symptoms present during the index offence and the primary diagnoses at the time of the NCRMD finding.• Men & women will have similar symptoms at index offence but unique diagnoses.

Objective 4 : Contrast criminal histories and recidivism by gender.• Women will have less extensive criminal histories and will be less likely to recidivate.

Objective 1 : Compare the demographic and psychosocial characteristics.• Women will present with unique socio‐demographic profiles typically indicative of greater marginalization, more mental problems and poorer functioning than men.

3/11/2014 5

Participants

84%

16%

Participant GenderMen (1,519)Women (280)

• N = 1799

[One of] the largest samples of forensic women to date

BC

ON

QC

Men N = 189 (85.1%)

Women N = 33 (14.9%)

Men N = 403 (83.3%)

Women N = 81 (16.7%)

Men N = 927 (84.8%)

Women N = 166 (15.2%)

DEMOGRAPHIC AND PSYCHOSICAL CHARACTERISTICS

Gender differences and similarities

3/11/2014 7

Results‐ Demographic characteristics

• Women were older at the time of the index offence

Canadian born English speaking French speaking Other Aboriginal status

Men 65.7% 61.1% 26.9% 12.1% 3.0%Women 68.1% 63.6% 27.7% 8.7% 2.9%Total 66.0% 61.4% 27.0% 11.6% 2.9%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

% of P

articipants

p‐value < .05

•Age: Men = 35.8 (12.5); Women = 40.6 (11.2) K-W (1,1989) = 53.2, p < .001

3/11/2014 3/11/2014 8

High school In relationship Live alone Live wspouse/family

Supervisedsetting Homeless Other Paid work

Men 47.3% 14.3% 30.3% 43.5% 8.6% 10.1% 7.5% 16.2%Women 60.1% 27.6% 38.4% 46.1% 6.9% 4.5% 4.1% 13.9%Total 49.2% 16.4% 31.6% 43.9% 8.3% 9.2% 7.0% 15.8%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

% of P

articipants

p‐value < .05

Results– Psychosocial characteristics

• Women: more educated; in relationships; residing alone

• Men: more likely to have been homeless

3/11/2014 9

Results

– Psychosocial Characteristics

Women were more educated; more often in relationships

Variable Male

n (Valid %) or M (SD)

Female


Statistic Total


High school completed 507 (47.3%) 115 (60.1%) χ2(1, 1266) = 10.81, p < .001 623 (49.2%)

Civil/marital status

In a relationship 200 (14.3%) 71 (27.6%) χ2(1, 1656) = 28.19, p < .001 271 (16.4%)

Single 1199 (85.7%) 186 (72.4%) 1385 (83.6%)

Residential status χ2(4, 1561) = 15.81, p = .003

Living alone 399 (30.3%) 94 (38.4%) 493 (31.6%)

Living with spouse, family or friends 572 (43.5%) 113 (46.1%) χ2(1, 1562) = 5.98, p = .014 685 (43.9%)

Supervised Setting 113 (8.6%) 17 (6.9%) 130 (8.3%)

Homeless 133 (10.1%) 11 (4.5%) χ2(1, 1562) = 7.86, p = .005 144 (9.2%)

Other 99 (7.5%) 10 (4.1%) 109 (7.0%)

Income

Own paid work (or partner) 189 (16.2%) 28 (13.9%) χ2(2, 1372) = 2.67, p = .263 217 (15.8%)

Pension/welfare 828 (70.8%) 154 (76.2%) 982 (71.6%)

Other 153 (13.1%) 20 (9.9%) 173 (12.6 %)

Note. At time of index offence.

MENTAL HEALTH: HISTORY, at time of OFFENCE and at time of VERDICT


3/11/2014 3/11/2014 11

Age first psych consult Age first psych hosp # hosp prior to index

Men 26.5 28.2 3.7Women 28.3 31.3 4.3

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

% of P

articipants

p‐value < .01

Results– Mental health history

•Women were older at first psychiatric hospitalization

3/11/2014 3/11/2014 12

Any Psychosis Hallucinations Delusions Suicide ideation Suicide attempt Self‐harm Homicidalideation Substance use

Men 58.0% 20.2% 46.5% 5.8% 1.4% 1.5% 5.6% 24.2%Women 55.4% 17.8% 43.9% 8.9% 3.2% 2.5% 8.6% 17.2%Total 57.6% 19.8% 46.1% 6.3% 1.7% 1.7% 6.1% 23.1%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

% of P

articipants

p‐value < .05

Results– Mental state at the time of the index offence

• Women were more likely to have been suicidal;

• Men were more likely to be using substances

3/11/2014 3/11/2014 13

PsychoticSpectrum Mood Disorder * Substance Use

Spectrum*PersonalityDisorder*

SMI & SubstanceUse Disorder*

SMI & PersonalityDisorder Other

Men 71.8% 22.2% 32.5% 9.9% 30.5% 8.9% 6.0%Women 66.2% 28.4% 21.6% 14.4% 20.1% 12.2% 5.4%Total 70.9% 23.2% 30.8% 10.6% 28.9% 9.5% 5.9%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

% of P

articipants

p‐value < .05

Results– Psychiatric diagnosis at the time of the index verdict

•Women more likely dx with mood & personality disorders

• Men had higher rates of SUDs and SMI + SU

CHARACTERISTICS OF THEINDEX OFFENCE


3/11/2014 15

Results‐ Index Offence – Offense against the person

• Women have similar rates of offenses against the person

Women (n=280) Men (n=1,519)

64.4%

35.6%

Offense against the person

Offense not against the person

65.0%

35.0%

Offense against the person

Offense not against the person

3/11/2014 16

Results‐ Nature of Most Serious Index Offence

• Few significant differences in the nature and severity of men’s & women’s index offences

Most severe offence Malen (Valid %)

Femalen (Valid %)

Statisticχ2

Totaln (Valid %)

Causing death/attempting 96 (6.3%) 27 (9.6%) χ2(1, 1799) = 4.10, p = .043 123 (6.8%)Sex offences 39 (2.6%) 2 (0.7%) χ2(1, 1801) = 3.66, p = .056 41 (2.3%)Assaults 401 (26.4%) 78 (27.8%) χ2(1, 1801) = .23, p = .631 479 (26.6%)Deprivation of freedom 27 (1.8%) 6 (2.1%) χ2(1, 1801) = .17, p = .680 33 (1.8%)Threats/other ‐against person 425 (28.0%) 68 (24.3%) χ2(1, 1799) = 1.62, p = .203 493 (27.4%)Property offences 246 (16.2%) 58 (20.6%) χ2(1, 1800) = 3.34, p = .068 304 (16.9%)Offensive weapons 95 (6.2%) 15 (5.3%) χ2(1, 1801) = .34, p = .558 110 (6.1%)Administration of justice 73 (4.8%) 10 (3.6%) χ2(1, 1800) = .84, p = .360 83 (4.6%)Disturbing the peace 8 (0.5%) 0 (0%) χ2(1, 1800) = 1.48, p = .618 8 (0.4%)Drug possession/trafficking 2 (0.1%) 0 (0%) χ2(1, 1800) = .37, p = 1.00 2 (0.1%)Dangerous driving/vehicle 54 (3.6%) 7 (2.5%) χ2(1, 1801) = .82, p = .366 61 (3.4%)Other 54 (3.6%) 10 (3.6%) χ2(1, 1799) = .00, p = .989 64 (3.6%)

Note. †Fisher's exact test is reported if the cell frequency is <5.

Nature of most serious index offence

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

Male

Female

Total

17

p < .05

• Few significant differences in the nature and severity of men’s & women’s index offences

p < .05

Results‐ Relationship with the Victim

0.05.010.015.020.025.030.035.040.045.0

MaleFemaleTotal

Able to ID victim in 92.7% of offences against persons; p < .05

• Women - significantly less likely to offend against strangers & more likely to offend against offspring or partners

3/11/2014 19

Results‐ Relationship to victim of Index Offence

Women were significantly less likely to perpetrate offences against strangers and more likely to offend

against offspring an partners


Victim Characteristics Malen (Valid %)

Femalen (Valid %)

Statistic Totaln (Valid %)

1) Stranger 221 (24.1%) 25 (15.0%) χ2(1, 1084) = 6.51, p = .011 246 (22.7%)

1) Professional 211 (23.0%) 37 (22.3%) χ2(1, 1084) = 0.04, p = .844 248 (22.9%)Police officer 112 (12.2%) 18 (10.8%) χ2(1, 1084) = 0.25, p = .620 130 (12.0%)Mental health worker 78 (8.5%) 15 (9.0%) χ2(1, 1084) = 0.05, p = .816 93 (8.6%)Other authority figure 21 (2.3%) 5 (3.0%) χ2(1, 1084) = 0.32, p = .575 26 (2.4%)

1) Family 299 (32.6%) 66 (39.8%) χ2(1, 1084) = 3.25, p = .071 365 (33.7%)Offspring 14 (1.5%) 14 (8.4%) χ2(1, 1084) = 26.67, p < .001 28 (2.6%)Partner/spouse 99 (10.8%) 30 (18.0%) χ2(1, 1084) = 7.12, p = .008 129 (11.9%)Parent 129 (14.1%) 15 (9.0%) χ2(1, 1084) = 3.07, p = .080 144 (13.3%)Other family member 57 (6.2%) 7 (4.2%) χ2(1, 1084) = 0.97, p = .325 64 (5.9%)

1) Other known person 187 (20.4%) 38 (22.9%) χ2(1, 1084) = 0.60, p = .438 225 (20.8%)Friend/acquaintance 119 (13.0%) 24 (14.4%) χ2(1, 1084) = 0.28, p = .597 143 (13.2%)Roommate/co-resident/co-patient

38 (4.1%) 6 (3.6%) χ2(1, 1084) = 0.09, p = .763 44 (4.1%)

Other 30 (3.3%) 8 (4.8%) χ2(1, 1084) = 1.00, p = .317 38 (3.5%)TOTAL 147 (100%) 324 (100%) χ2(10, 1084) = 42.58, p < .001 1084 (100%)

CRIMINAL HISTORY & RECIDIVISM


Results‐ Criminal History

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

Any priorconviction or

NCRMD

Offenceagainstperson

Otheroffence

Any priorconviction

Any priorNCRMD

MenWomenTotal

3/11/2014 3/11/2014 21

p < .05

• Men were significantly more likely than women to have a

criminal record, including offences against persons

3/11/2014 22

Results‐ Criminal History

There were very few significant differences in the nature and severity of men’s & women’s index

offences


Criminal History Malen (Valid %) or M (SD)

Femalen (Valid %)or M (SD)

Statisticn (Valid %) or M (SD)

Totaln (Valid %) or M (SD)

Any prior conviction or NCRMD finding

797 (52.5%) 88 (31.4%) χ2(1, 1799) = 47.88, p < .001 885 (49.2%)

Offence against person 504 (33.2%) 52 (18.6%) χ2(1, 1800) = 23.57, p < .001 556 (30.9%)

Other offence 683 (45.0%) 70 (24.9%) χ2(1, 1800) = 39.19, p < .001 753 (41.8%)

Any prior conviction 756 (49.7%) 82 (29.2%) χ2(1, 1801) = 40.28, p < .001 838 (46.5%)

Any prior NCRMD finding 133 (8.8%) 14 (5.0%) χ2(1, 11799) = 4.44, p = .035 147 (8.2%)

3/11/20143/11/2014 23

Results‐ Duration of Time Under Review Board and Recidivism by Gender

Results

• Women spent less time under Review Board than men but recidivated at significantly lower rates

• Women were detained in custody for a shorter period of time compared to men

‐(Mantel‐Cox Chi² = 6.05, p = .014)

‐ Review Boards unconditionally discharged women faster than men ‐(Mantel‐Cox Chi² = 19.72, p < .001)

YET…

• Women had a lower rate of recidivism after their verdict then men ‐ (Mantel‐Cox Chi² = 8.98, p = .003)

3/11/20143/11/2014 24

OR (95% CI)Aboriginal status 1.62 (0.72 - 3.63)Age at the index offence 0.99 (0.95 - 1.03)Diagnosis (non-exclusive)

Psychosis 1.52 (0.76 - 3.06)Mood 1.93 (0.93 - 4.00)Substance 0.71 (0.50 - 1.01)Personality 2.23 (1.44 - 3.45)***

Presence of psychiatric history 1.10 (0.80 - 1.51)Age at first offence against persons 1.05 (1.01 - 1.09)*Presence of criminal history

NCRMD 0.66 (0.34 - 1.28)Criminal 0.46 (0.29 - 0.71)**Against Persons 1.25 (0.70 - 2.24)

Index Most severe offence (Others as reference)Homicides/attempted 1.37 (0.73 - 2.57)Assault and sexual assaults 1.10 (0.67 - 1.80)Other crimes against persons 0.98 (0.60 - 1.62)Property crimes 1.22 (0.66 - 2.26)

-2LL = 1207.72; χ2(17) = 96.85; p < 0.001; Nagelkerke pseudo-R² = 10.6%; * p < 0.05; ** p < 0.01; *** p < 0.001

• Personality disorder dx, age 1st offence against person, and prior criminal conviction predict gender of accused

Results

Conclusions

Hypothesis 1 : Women will present with unique socio‐demographic profiles = even greater mental health problems; poorer functioning than men.• Heterogeneous population; more homeless men + women more educated• Not consistently the case that women have greater needs/lower functioning

Hypothesis 3 : There will be differences in relationship with the victim by gender but women will commit serious index offences against persons• Women perpetrated more • Women more likely to offend against partners & offspring• Men more likely to offend against strangers

Hypothesis 2 : Men & women will have similar mental state at index offence. • Several significant differences = Women more mood + suicidal + PDs; Men more SUD

Hypothesis 4 : Women will have less extensive criminal histories but will be no less likely to recidivate.• Men had substantially greater criminal justice involvement (forensic & corrections)• Women were significantly less likely to reoffend

• Strengths– Cross‐Canada sample – One of largest samples of forensic patients (women)

• Increased generalizability• Capacity to speak to low base rate events; subgroups

– Triangulated data collection (files + RCMP + health)• Limitations / Future directions

– Secondary data (may over‐/under‐estimate)• Missing data• Difficult to address diagnoses, protective factors, trauma• Provincial differences in record keeping

– We did not build in gender‐specific variables (FAM)

Discussion

Discussion

• Policy: Sensible service delivery to women (or men) in forensic psychiatric services requires a sophisticated consideration of the individual’s needs– But is this reflecting the need for a major conceptual shift or not?

• Research: further evidence required to clarify: – Is this gender specific or simply client centered?

• can be integrated in manner that complements and enhances traditional psychiatric and psychological models, rather than replacing them (e.g., RNR)

•Despite gaps in evidence, there is a good case to be made that female patients present different clinical challenges than men

Discussion

Risk

•Match level of service to risk (dose / don’t over‐intervene)• Women in custody often at higher security than necessary

Need

• Target criminogenic needs• Women have similar though fewer criminogenic needs

Responsiviity

• Tailor interventions to strengths of the individual• Women ‐more educated, socially integrated (?); less criminally entrenched

Practice Impact: Overall, a complex multi‐problem population. One size does not fit all (Latessa, APLS 2014)

Andrews, 2012

Special thanksAcknowledgments

https://ntp‐ptn.org/

The National Trajectory Project (NTP): Examining Profiles \u0026 Recidivism through a Gendered Lens

Documents