IAFMHS 2013 Bjørkly Safe 10 06 13

OsloUniversity Hospital

www.forensic-psychiatry.no HELSE SØR-ØST

Clinical experiences with Version 2 and 3 of the HCR-20 in the SAFE

pilot projectStål Bjørkly, Molde University College.

Gunnar Eidhammer and Lars Erik Selmer, Clinic of Addiction and Psychiatry, Vestre Viken Trust.

Johnny Wærp, Section for Medium Security, Gaustad, Oslo University Hospital.

Alexander R. Flaata and Marthe Kjerstad, Section for Maximum Security, Dikemark, Oslo University Hospital.

Centre for Forensic Psychiatry,Oslo University Hospital,

Oslo, Norway


HELSE SØR-ØST

Main research objectives

- Risk assessment - Risk management - Prospective follow-up- Naturalistic design


HELSE SØR-ØST

Setting and sample

One maximum security unit and three medium security units+ Forensic patient+ Major mental illness+ Serious violence towards others + Discharge within 6-12 months


HELSE SØR-ØSTwww.forensic-psychiatry.no

Sample characteristics (n= 22)

Mean age: 32 Gender: 19 male Relationship with partner: No:17 Highest education:Compulsory school (8/2

failed), High school (11/10), University (3/2) Employment/Income: Social support Ethnicity:17 native Caucasian Norwegians



Sample characteristics cont.

Previous psychiatric treatment: 1 – 12 months: 7 1 – 2 years: 3 Over 2 years: 12

Primary diagnoses: Paranoid Schizophrenia 14 Hebephrenic Schizophrenia: 4 Personality disorders: 4


HELSE SØR-ØST

Repeat dynamic risk assessment

Hallucinations (PSYRATS-A) Delusions (PSYRATS-B) TCO (PERI) Hopelesness (The Hopelesness Scale) Dissociation (DES-II) Insight (SAI-E) Warning signs (SWAB, FESAI, ERM) Situational vulnerability (REFA)



Focus in this presentation

• To compare HCR-20 version 2 and 3 • Clinical risk factors and risk

management



Internal consistency and clinical utility

Gunnar Eidhammer, Lars Erik Selmer and Stål Bjørkly



Participants and setting

Setting Raters Assessment of 20 forensic mental health patients All patients were male



Sample (n = 20)Diagnoses Most severe offense N

Paranoid Schizophrenia Homicide 3Serious physical assault 3Arson and homicidal threats 1

Schizophrenia Serious physical assault 3Rape 1Knife stabbing 1

Psychosis due to substance abuse

Serious physical assault and homicidal threats

3

Homicidal threats 1

Paranoid PD Double homicide 1Serious physical assault 1

Antisocial PD Serious physical assault and threats 1

APD/Psychosis Serious physical assault and threats 1



Internal consistency ICC: Two-way mixed effects model (rater effects random and measures

effects fixed) Cronbach's alpha (95%CI)

H-items .854 (.726 - .940)C-items .586 (.256 - .809)R-items .812 (.662 - .913)All items of HCR-20 and V3 .842 (.708 - .935)C-items: p < .001; Other items: p < .000

Significant differences for H-items (t=-2.797, df=19, p<.012) and C-items (t=-4.040, df,=19, p<.001), but not for R-items (t=.218, df=19, p<.830).



Clinical utility: coding Change from 0, 1, 2, omit (V2) to y, p, n, o (V3)

makes the V3 more adapted to clinical practice. Mitigates the risk of importing an actuarial

approach into the use of the V3 Protects against misuse of numbers in decision

making processes The new structure reduces the risk of empirically

unfounded conclusions



Clinical utility: Presence / Relevance

Presence: The principal common risk factors Evidence-/knowledge-based

Relevance: The principal individual risk factors Person-specific



Clinical utility: Risk formulations

Risk formulations inform and enhance clinical practise through individualized risk management strategies

An example: Combining the Early Recognition Method (ERM) and R item analysis to develop risk management strategies



Concluding remarks The 7-step structure 1. Gather and document basic case information 2. Identify presence 3. Assess relevance 4. Integrate information into case formulation 5. Identify risk scenarios 6. Recommend risk management strategies7. Risk judgment documentation

V3 = A more systematic, comprehensive and individualized violence risk analysis

“Risk analysis” = assessment + management


HELSE SØR-ØST

Changes in repeated HCR-20 measurement

Johnny Wærp

www.forensic-psychiatry.no


HELSE SØR-ØST

Clinical Comparison HCR-20 V2-V3

Clinical items V2 C1 Lack of insight C2 Negative attitudes C3 Active symptoms of

major mental disorder C4 Impulsivity C5 Unresponsive to

treatment

Clinical items V3 C1 Insight C2 Violent ideation or

intent C3 Symptoms of major

mental disorder C4 Instability C5 Treatment

supervision response


HELSE SØR-ØST

Rihanna

Diagnosis F 60.3 Emotionally unstable personality disorder (DSM:Borderline)

Single, 34 yrs History of:

sexual abuse psychiatric hospitalisations (last 10 years) violence towards others self harm and suicidal behaviour


HELSE SØR-ØST

Repeated Measurements HCR-20 V2

T1

Total: 24

C3: 1

T2

Total 23

C3:1

T3

Total 21

C3:0


HELSE SØR-ØST

Changes in repeated measurement

Psyrats A&B T1

Score 0

Score 0

Psyrats A&B T2

Score 0

Score 0

Psyrats A&B T3

Score 0

Score 0


HELSE SØR-ØST

Perceived threat and control override symptoms (TCO)

Monitored weekly for TCO- symptoms for 18 months. At T3 a risk management plan was formulated: -

focus on TCO- symptoms - associated with irregular menstruation.

Presence of TCO- symptoms: Risk factor and, Warning sign (SWAB) For vilolence toward self/others

Patient acknowledged this relationship Medication with birth pills.


HELSE SØR-ØST

Changes in repeated measurements of TCO

T1

TCO symptoms present

Other psychotic symptoms present

T2

TCO symptoms present in pattern

Other Psychotic symptoms present

in pattern

T3

TCO symptoms not present

Only 1 other symptom present


HELSE SØR-ØST

C3 Formulation in V2 vs. in V3

V2 FormulationC3 Active symptoms of major mental disorder

T1: 1, T2: 1, T3: 0 Psyrats A

No score (T1, T2, T3) Psyrats B

No score (T1, T2, T3 ) Peri - TCO

Scores present at T1 andT2, not present at T3. TCO weekly monitoring

still shows pattern related to irregular menstruation

V3 Formualtion C3 Active symptoms of major mental disorder Sub-item: Psychotic disorders Does not meet ICD- criteria for schizophrenia. Not present and not relevant,but Shows active TCO symptoms in relevant pattern recorded on SWAB and REFA (short term)Sub-item: Major mood disorders Meets criteria for major mood disorder in remission Not present, but relevant for long term riskSub-item: Intellect/executive/ social NP tests show some memory problems Present and moderately relevant


HELSE SØR-ØST

Conclusion

Sub-items structure judgement Presence and relevance help define

formulations Specification of time frame and priority of

case enhance structure of risk mangement procedure

Formulation becomes more specific and useful for next level of treatment


HELSE SØR-ØST

A single-case illustration from a high security ward

Alexander R. FlaataMarthe Kjerstad

www.forensic-psychiatry.no


HELSE SØR-ØST

Method

Two independent raters One patient: Justin, in his late twenties Raters completed HCR-20 (V2/V3) together

with hospital staff Outcome of violence risk assessment,

violence risk management plan, and clinical utility was discussed


HELSE SØR-ØST

Justin Alcohol abuse: 12 years old Drug abuse (cannabis, amphetamine and heroine) Age at first psychiatric admission: 21 19 hospitalisations F20.0 Paranoid Schizophrenia Transferred from medium to high security ward : 24

years old Numerous severe physical assaults against mental

health professionals


HELSE SØR-ØST

Dynamics of violence in high security ward. Paranoid delusions:

Controlled by the personnel (by a device in his brain) Own thoughts broadcasted Emotional distress: Psychotic anxiety

Somatic delusions: Body changes (penis and face) Emotional distress: Dysmorphophobic anxiety He felt to be in urgent need of medical examination Reality orientation: increased risk of violence.

Antisocial (instrumental) violence: Persistent demand of sedatives, violent movies, discharge… Verbal threats, and physical violence to get what he wanted.


HELSE SØR-ØST

Comparison of clinical utility

V2 Less support in the assessment process More broad and general risk management planning

V3 Clear structure for assessment procedure Requires more spesific inquiry into how risk factors

contribute to present risk of violence Method for identifying significant individual risk

factors


HELSE SØR-ØST

Clinical utility (II)

V3 Systematic approach to develop realistic scenarios Relevance scores enhance risk assessment accuracy

Assessment of Justin: Splitting factor: Items with low relevance This contributed to different outcomes: v2: moderat

v3: low to moderate,


HELSE SØR-ØST

Conclusion

Both versions have significant clinical utility in formulating violence risk and management

V3 provides a better structure of the assessment process

V3 is more comprehensive regarding the whole process of violence risk assessment

V3 is more time-consuming


www.forensic-psychiatry.no HELSE SØR-ØST

Contact information

Stål Bjø[email protected]

Centre for Forensic Psychiatry,Oslo University Hospital,

Oslo, Norway

mailto:[email protected]

IAFMHS 2013 Bjørkly Safe 10 06 13

Documents