The Columbia-Suicide Severity Rating Scale: Validation for use as a screen for suicide risk in New Zealand prisons and probation settings

Dr Nick J. Wilson
Principal Adviser, Psychological Research, Office of the Chief Psychologist, Department of Corrections

Author biography:

Dr Nick Wilson has been with the department for over 20 years. Nick began employment with Corrections working as a clinical psychologist at psychological services in Hamilton in 1997 up until 2001 when he became a specialist applied psychological researcher. He has been involved in research into the assessment and treatment of personality disordered offenders, including high-risk offenders such as psychopaths, adult and child sex offenders, and youth offenders. Nick has also developed dynamic risk measures for the department such as RPFA-R, DRAOR and SDAC-21. Nick is currently part of the national office team lead by the Chief Psychologist.


Acknowledgements:

The author acknowledges the expertise and generosity of Dr Kelly Posner Gerstenhaber, Director, The Lighthouse Project, New York, principal developer of the C-SSRS. The research reported in this article involved the able assistance, expertise, and effort of Glen Kilgour, Principal Adviser Special Treatment Unit Development from the Office of the Chief Psychologist, New Zealand Department of Corrections, for the prison study and Jonathan Muirhead, Advisor, Corrections Services, New Zealand Department of Corrections, with the probation study. The author also thanks Simon Davies, Summer Intern, Office of the Chief Psychologist, New Zealand Department of Corrections, for his comprehensive review of the suicide literature.

This paper provides a brief summary of the findings of these studies. For a more comprehensive paper with all the detail on the statistical analysis of the psychometric properties of the application of the C-SSRS in New Zealand, please contact the author Dr Nick Wilson (nick.wilson@corrections.govt.nz).

Introduction

This article provides a summary of the results from trial administration of the Columbia-Suicide Severity Rating Scale (C-SSRS) within receiving units at two of the department’s prisons between September and December 2014 and across six probation sites in 2016-17. This research was carried out to examine the effectiveness of the C-SSRS to see if it was a tool that could be considered for use across all prisons and probation sites.

Suicide is widely accepted to be the leading cause of preventable death in correctional facilities worldwide (World Health Organisation and International Association Suicide Prevention, 2007). In comparison with the general population, offenders both in prison (Fazel, Grann, Kling and Hawton, 2011) and the community (Pratt, Appleby, Piper, Webb and Shaw, 2010) show increased rates of suicide.

Preventing suicide has thus become a major priority for correctional systems worldwide that are responsible for the offenders under their care or supervision. In terms of prevention, effective screening is a key component of a comprehensive suicide prevention programme (Hayes, 1995). This is particularly so in the correctional domain because research indicates more offenders who commit suicide do so at the beginning of their sentence (Pratt, Piper, Appleby, Webb, and Shaw, 2006; Shaw, Baker, Hunt, Moloney, and Appleby, 2004). Suicide risk can be screened for in a couple of different ways.

All offenders could be seen by a mental health professional at the beginning of their sentence (both in the community and in prison) to have their suicide risk assessed as part of a comprehensive mental health assessment. However, this option does not reflect the limited health resource available across New Zealand in general, or the large numbers needing to be assessed. In practice, therefore, identifying the most at-risk offenders relies heavily on the personal judgement of frontline correctional staff, particularly corrections officers or probation officers (Correia, 2000; Konrad et al, 2007).

A review of the international literature regarding suicide risk screening tools in correctional settings was carried out for the Department of Corrections (Davies, 2014), with a view to determining whether there were any existing ‘fit for purpose’ tools with good evidence for their validity (i.e. proven ability to accurately identify those at risk of suicide now). The issue of “currently at risk” is key, as using evidence of risk factors that correlate with lifetime risk such as depression and substance abuse or even past suicide attempts, especially if these are not recent, over identifies those potentially at risk today. Therefore, the search was for tools that focused on suicidal ideation (also known as suicide thoughts) and allowed identification in terms of risk if these thoughts were increasing in terms of action/plan.

Following this review, and subsequent consultation with a number of international corrections authorities, the Columbia-Suicide Severity Risk Screen (C-SSRS; Kelly Posner, Ph.D. New York State Psychiatric Institute, New York © 2008) was identified as having considerable research support in terms of its usefulness as a suicide risk screen across settings that had included prisons and probation (see C-SSRS items in Appendix One).

The six item C-SSRS screen is a semi-structured rater-based interview measure designed to assess active vs passive, severity and frequency of suicidal ideation and suicide behaviour. It was designed to identify a wide range of ideation and to monitor change across contacts with predictive safety referral criteria derived from longitudinal studies. The C-SSRS is used extensively across primary care, clinical practice, surveillance, research, and institutional settings. It is currently available in over 100 languages, and forms part of national and international public health initiatives involving the assessment of suicidality, including general medical and psychiatric emergency departments, hospital systems, managed care organisations, behavioural health organisations, medical homes, community mental health agencies, primary care, hospices, schools, college campuses, US Army, National Guard, Veterans Affairs, Navy and Air Force settings, abuse treatment centres, prisons, jails, and juvenile justice systems. The measure is applied across these settings by frontline responders (police, fire department, EMTs), as well as a wide variety of others such as clergy, teachers and judges, to reduce unnecessary hospitalisations (for details see cssrs.columbia.edu/the-columbia-scale-c-ssrs).

It is important to note that the C-SSRS has been administered several million times and has exhibited excellent feasibility with no mental health training required to administer it. The C-SSRS is the only screening tool that assesses a range of evidence-based ideation and behaviour items, with criteria for next steps (e.g. referral to mental health professionals); therefore the C-SSRS can be very useful in initial screenings. The C-SSRS has been associated with a decreased burden by reducing unnecessary interventions and redirecting limited resources. As such, there is evidence of its effectiveness in reducing unnecessary interventions in hospitals, schools, and, importantly in regards to this report, in correctional settings. The California Corrections Department spent $20 million on suicide-watch in 2010, which they believe could be cut in half by introducing the C-SSRS.

The C-SSRS has been validated in a number of published studies and, while relatively new having been released in 2009, it is endorsed by the USA Federal Drug Authority as the “gold standard for suicide screening” (Posner et al, 2011). The initial validation study by Posner et al, (2011) found the C-SSRS demonstrated good convergent and divergent validity in comparison to other validated approaches. The Columbia had high sensitivity and specificity for suicidal behaviour classifications compared with scales and both the ideation and past suicide behaviour subscales were sensitive to change over time. The researchers noted that the intensity of ideation subscale demonstrated moderate to strong internal consistency, with the two highest levels of ideation severity (intent or intent with plan) at baseline having higher odds for attempting suicide during the study.

In terms of how accurate the C-SSRS is in terms of suicidal behaviour prediction, previous research with a large sample of 3,776 individuals with a range of mental health issues found it had a 73% level of accuracy. This accuracy was for the report of further suicide behaviour while this group was being treated based on their entry C-SSRS scores for passive and or active suicide ideation or thoughts (Mundt et al, 2010). Further analysis of the same sample by the authors (Mundt et al, 2013) found those with the presence of suicide ideation alone were almost six times more likely to report suicide behaviour. When both ideation and suicide behaviour presence on the C-SSRS was added together, this resulted in a group who was nine times more likely to report future suicide behaviour. These results indicate the support in particular for the C-SSRS’s ability to assess the important area of ideation, as well as how the presence of ideation and past suicide behaviour combines to indicate a higher suicide risk group.

Method

The project sought to address the review of suicide screening by trialling the C-SSRS in a representative New Zealand prison environment in 2014 and then later in 2016/17 across a number of community probation sites. Two prison pilot sites, Waikeria men’s prison and Auckland Region Women’s Corrections Facility were used in 2014 (August to November) along with six community probation sites spread across New Zealand in late 2016 to early 2017.

Both the prison and the community probation study used training materials that were sourced from the principal developer of the C-SSRS, Dr Kelly Posner, and modified for use in New Zealand. The training workshops took 40 minutes and covered the C-SSRS screening measure with case examples and practice. Some general facts about suicide assessment including addressing misconceptions (i.e. causing harm from asking about suicide) were also included. In addition to the six C-SSRS questions, additional questions were added on areas typically asked in suicide screening that were sourced from current New Zealand Corrections procedures and from the literature review by Davies (2014).

In the interests of ensuring all questions were administered, staff in both the prison and the community study were directed to ask all six C-SSRS questions rather than stopping if the respondent said no to Q1 or Q2.

The four additional questions for the prison study were:

  • First incarceration: “Have you been in prison before?” (YES/NO)
  • Relationship stress: “Are you currently having any relationship stress or problems?” (YES/NO)
  • Substance abuse: “Have you been drunk or intoxicated with drugs anytime in the last month?” (YES/NO)
  • Psychiatric/psychology involvement: “Have you ever seen a GP, counsellor, psychologist or psychiatrist for an emotional, mood, or social problem?” (YES/NO).

In the community probation study the following five additional questions were asked in addition to the C-SSRS questions:

  • Relationship stress: “Are you currently having any relationship stress or problems?” (YES/NO)
  • Substance abuse: “Have you been drunk or intoxicated with drugs anytime in the last month?” (YES/NO)
  • Mental health: “Are you currently receiving treatment for any mental health difficulties where delusions, hallucinations, or mood disturbances are part of the problem?” (YES/NO)
  • Psychiatric/psychology involvement: “Have you ever seen a GP, counsellor, psychologist or psychiatrist for an emotional, mood, or social problem?” (YES/NO)
  • Risks for self and/or others: “Are you or your dependents at risk of abuse (physical/psychological/sexual)?” (YES/NO).

Results

Prison study

The C-SSRS was administered with 721 prisoners in the two prison settings at time of reception into prison during the study period (Wilson and Kilgour, 2015). This sample was predominately male (n=586) with a smaller but still representative sample of female prisoners (n=135). The results of the C-SSRS administration along with the four additional questions were analysed and the following key results were found:

  • There was a small but significant group of assessed prisoners with suicide ideation (16.2%) and past suicide behaviour, that typically happened more than 12 months ago (14.8%).
  • Of those with suicidal ideation, a smaller group (8%), around 57 prisoners, indicated active suicidal ideation that involved either method, intent or a plan.
  • A number of prisoners did not say yes to passive ideation questions, but as all six questions were asked in this study, they said yes to some of the active ideation questions. So it was worth persevering in asking all six C-SSRS questions.
  • No relationship was found with the C-SSRS questions and the additional question on whether this was there first time in prison. However, the other three additional questions were related to yes responses to the C-SSRS questions. The strongest relationship was found between the question on lifetime contact with mental health professionals over past emotional and social issues and the six C-SSRS questions.
  • Gender comparison. Comparison of the six item scores for the C-SSRS found no statistically significant differences between the two pilot prison sites (male and female prisons).
  • Ethnicity, age and offending risk comparison. Prisoners receiving an at risk status upon reception was not found to be related to age, ethnicity, and remand/sentence disposition or to RoC*RoI (Bakker, Riley, and O’Malley, 1999). The only significant factor for being assessed as currently at risk was whether the prisoner had been assessed historically as at risk during former imprisonment.

Feedback on the C-SSRS from corrections officers: During the study, completed forms were collected from available receiving staff. Staff said they had no issues completing the C-SSRS and saw it as simple to use. They also reported that they had not faced any resistance from prisoners being asked the questions in the C-SSRS. Finally, they hoped that if the C-SSRS was to be used that it was incorporated into the computerised Integrated Offender Management System (IOMS) as they found it gave information they had not considered in the current prison at risk assessment.

Community probation study

A total of 337 C-SSRS screening assessments were completed by community corrections staff at six representative probation sites (Muirhead and Wilson, 2017). The screening assessments were administered during the period between November 2016 and March 2017. The sample was, as expected, predominately male at 77.4% (n=261) with 19.9% female (n=67), and 2.7% unknown (n=9). The ethnicities of those who completed the C-SSRS forms were 51.9% Māori (n= 175), 28.5% NZ European (n=96), 5.6% Pasifika (n=19), 11.6% unknown (n=39), and 2.4% other (n=8). The average age at the time of assessment for those whose ages could be calculated (n=320) was 32 years old, ranging from 17 to 71 years of age.

In terms of the distribution of scores for the C-SSRS and the additional five risk questions, the community results reflected similar frequencies of suicidal ideation to the prison study. Key results of the community study were:

  • A small group (15.1%) indicated that they had experienced passive suicidal ideation in the past month (Q1 or Q2) compared to 16.2% in the prison study.
  • Nine percent of the sample had shown past suicidal behaviour, with half of that group engaging in the behaviours more than a year ago. Only 6.5% (22 people) were experiencing some form of active suicidal ideation, answering yes to at least one of Questions 3, 4, or 5.
  • Similar to the prison study, a small number of respondents in the community study indicated active ideation (Q3-5) even if they said no to some of the passive ideation questions (Q1-2).
  • While all the additional questions had a relationship with the yes responses to the C-SSRS, the strongest relationship with suicidal ideation came from the questions relating to current mental health issues and past psychiatric/psychologist involvement.
  • C-SSRS and key demographic variables. The C-SSRS item scores were not found to have a significant relationship static risk as measured by RoC*RoI score. However, a relationship was found between younger age and suicidal ideation with increased ideation scores for those in their mid-twenties. No statistically significant relationships were found between gender and suicidal ideation, and ethnicity and suicidal ideation.

Feedback on the C-SSRS from probation staff: More than half the staff were positive about the measure and its benefits. They highlighted the ease of use and how comprehensive it was to ask about all aspects of suicidal ideation, rather than treating suicide as a singular construct. Those expressing some negative views found the C-SSRS questions too blunt and uncomfortable to ask or repetitive. The concern about the direct language likely reflected some misunderstanding of the need for plain and specific language to prompt relevant, accurate responses to this important safety issue. This discomfort suggested training materials for staff on using the tool should provide clear evidence that asking plain, somewhat repetitive questions does not increase suicide risk. In fact, rather the opposite is true.

Discussion

The C-SSRS represents an improvement in the theory and practice of suicide risk assessment. In particular, the focus on suicidal ideation and behaviour increases the likelihood that staff will pick up on current (and historical) risk, and moves beyond a reliance on current stressors. While such stressors are often salient, they can over-predict suicidal behaviour given the high presence of stressors for both prisoners and community-based offenders. The C-SSRS validation studies undertaken at two prisons as well as six probation sites for representative samples of New Zealand offenders found that the measure was able to discriminate on the basis of suicidal ideation and behaviour.

Both studies across these two different settings found similar percentages of offenders who had the presence of either passive or active suicidal ideation (16.2% in prison and 15.1%). This percentage is similar to that found in international studies using the C-SSRS (Mundt et al, 2013). In terms of active ideation – in other words, current thoughts of how the person could kill themselves, wanting to do this now, or having a plan/intention to kill themselves – there was convergence between the two Corrections studies. In the prison study 8% had active ideation and in the probation study a smaller group of 6.5% had similar active thoughts involving either method, intent or a plan. This small difference between the community and prison samples likely reflected the higher risk when offenders are placed in prison in terms of a greater degree of experience of psychosocial stressors.

In terms of the C-SSRS question relating to past suicide behaviour, the prison sample with 14.8% compared well with the 9.8% community sample. The Mundt et al, (2013) sample of patients with serious mental health or health issues had a higher 27% lifetime report of past suicide attempts. Therefore, across the two Corrections samples, the C-SSRS was able to differentiate those with current passive or active suicide ideation, assisting the ability of staff to correctly identify – based on this simple screen – which offenders should be the focus of further assessment or intervention to manage current risk of suicide behaviour. This ability to identify at risk using the C-SSRS was found across both male and female offenders, and across age and ethnicity and risk of re-offending.

The additional risk questions in the main provided only a small amount of information that was directly related to current risk. Across both the prison and the community studies a key additional question area related to either past or current contact with mental health professionals over mental health issues. Another result that was not expected was finding a small number of offenders who did not respond with a yes response to the passive ideation questions (Q1-2) but who did then indicate a yes response to one of the active ideation questions (Q3-5). One can speculate that this may be due to a range of factors such as poor comprehension by the offender or that the persistence in questioning overcame barriers such as embarrassment. The key safety issue is that asking all of the C-RRS questions, rather than stopping with ‘no’ responses to the passive ideation questions, was supported as a good practice. The writer consulted with Dr Posner (personal communication, 24 July 2017) who endorsed that asking all questions as a matter of policy did no harm, improved accuracy, and added at best another minute to an assessment.

Corrections staff involved in the two validation studies for the C-SSRS reported the measure was simple to use and added information to their consideration of suicide risk. Although, the community study did identify some issues in relation to the need for more training to cover the need for the direct language used in the C-SSRS and also, to a degree, why repetition of questions is important when the risks of missing suicide risk is so great.

The administration of the C-SSRS increased the ability of receiving staff to effectively assess the risk of suicide through accessing both passive and active suicidal ideation. This increased the amount of quality information on where the person was in terms of suicide continuum. Use of the C-SSRS by Corrections covers the key assessment areas identified across the relevant literature of suicidal ideation, degree of intent to commit suicide, planning of the act, and previous suicide attempts into the decision of at risk status (Liotta, Mento, and Settineri, 2015).

Implementation of the C-SSRS in New Zealand Corrections

Subsequent to the New Zealand prison study of the validity of the C-SSRS, in 2015 the screening tool was implemented into the standard prison reception assessment completed in the computerised Integrated Offender Management System. No significant issues have been found following the rollout of the measure across all New Zealand prisons with staff reported to have quickly adapted to the inclusion of the C-SSRS questions. The successful trial of the C-SSRS in the six probation community sites in early 2017 has also resulted in a recent decision to roll out the C-SSRS across all probation offices in their management of offenders. Training has been completed with probation staff in support of the implementation which began in August 2017. The adopted policy in administering the C-SSRS in both prison and probation community settings has been to ask all six questions in the measure to ensure risk responses are not missed.

Final comment

The implementation of the C-SSRS into the frontline assessment of current suicide assessment in New Zealand Corrections has provided staff with a simple but effective screening tool that is grounded in theory and best practice. The validation of the C-SSRS across the prison and community setting provides confidence in the ability of the measure to add valuable information in the assessment of offenders who are at risk of suicide.

One of the advantages of the C-SSRS is its broad clinical validation and applicability across different settings. There are opportunities in the future to discuss Corrections’ adoption of the measure with other key stakeholders, in particular Secure Future, NZ Police, and health agencies including mental health and forensic services. This has the potential to improve information-sharing about ‘clients’ who are at risk of suicide and streamline referrals for urgent or follow-up interventions.


References

Bakker, L., Riley, D., & O'Malley, J. (1999). RoC Risk of Reconviction: Statistical models predicting four types of re-offending. Christchurch, NZ: Department of Corrections, Psychological Service.

Correia, K. M. (2000). Suicide assessment in a prison environment: a proposed protocol. Criminal Justice and Behavior, 27, 581-599. doi:10.1177/0093854800027005003

Davis, S. (Feb 2014). Suicide Assessment: Suicide Screening Tools in Correctional Settings, Literature Review. Summer Intern Report prepared for the Office of the Chief Psychologist, Service Development, Department of Corrections, Wellington.

Fazel, S., Grann, M., Kling, B., & Hawton, K. (2011). Prison suicide in 12 countries: an ecological study of 861 suicides during 2003-2007. Social Psychiatry and Psychiatric Epidemiology, 46, 191-195. doi:10.1007/s00127-010-0184-4.

Hayes, L. M. (1995). Prison suicide: an overview and a guide to prevention. The Prison Journal, 75, 431-456. doi:10.1177/0032855595075004003

Konrad, N., Daigle, M. S., Daniel, A. E., Dear, G. E., Frottier, P., Hayes, L. M., & Sarchiapone, M. (2007). Preventing suicide in prisons, part I: recommendations from the International Association for Suicide Prevention Task Force on Suicide in Prisons. Crisis, 28, 113-121. doi:10.1027/0227-5910.28.3.113

Liotta, M., Mento, C., & Settineri, S. (2015). Seriousness and lethality of attempted suicide: A systematic review. Aggression and Violent Behavior, 21, 97-109.

Mundt, J.C., Greist, J.H., Gelenberg, A.J., Katzelnick, D.J., Jefferson, J.W. & Model, J.G. (2010). "Feasibility and Validation of a Computer-Automated Columbia-Suicide severity Rating Scale Using Interactive Voice Response Technology." Journal of Psychiatric Research, doi:10.1016/j.jpsychires. 2010.04.025.

Mundt, J. C., Greist, J. H., Jefferson, J. W., Federico, M., Mann, J. J., & Posner, K. (2013). Prediction of suicidal behavior in clinical research by lifetime suicidal ideation and behavior ascertained by the electronic Columbia-Suicide Severity Rating Scale. The Journal of Clinical Psychiatry, 74(9), 887-893.

Muirhead, J., & Wilson, N. J. (2017). Evaluation of suicide screening tools used in community probation. A report on the trial use of the Columbia-Suicide Severity Rating Scale (C-SSRS) in six NZ community corrections sites. Internal report. Department of Corrections, Office of the Chief Probation Office, Wellington, NZ.

Posner, K., Brown, G.K., Stanley, B., Brent, D.A., Yershova, K.V., Oquendo, M.A., Currier, G.W., Melvin, G., Greenhill, L., Shen, S., & Mann, J.J. (2011). The Columbia-Suicide Severity Rating Scale: Initial Validity and Internal Consistency Findings From Three Multisite Studies With Adolescents and Adults American. Journal of Psychiatry; 168:1266-1277.

Pratt, D., Piper, M., Appleby, L., Webb, R., & Shaw, J. (2006). Suicide in recently released prisoners: a population-based cohort study. Lancet, 368, 119-123. doi:10.1016/S0140-6736(06)69002-8

Pratt, D., Appleby, L., Piper, M., Webb, R., & Shaw, J. (2010). Suicide in recently released prisoners: a case-control study. Psychological Medicine, 40, 827-835. doi:10.1017/S0033291709991048

Shaw, J., Baker, D., Hunt, I. M., Moloney, A., & Appleby, L. (2004). Suicide by prisoners. National clinical survey. British Journal of Psychiatry, 184, 263-267.

World Health Organisation and International Association for Suicide Prevention (2007). Preventing suicide in jails and prisons. Department of Mental Health and Substance Abuse WHO. ISBN 978 92 4 159550 6.

Wilson, N. J., & Kilgour, G. (2015). Evaluation of suicide screening tools used in prisons. A report on the trial use of the Columbia-Suicide Severity Rating Scale (C-SSRS) in two NZ prison sites. Internal report, Department of Corrections, Office of the Chief Psychologist. Wellington NZ.


Appendix One

C-SSRS items (asked with reference to the last month or since last contact)


Item 1. Wish to be Dead

Person endorses thoughts about a wish to be dead or not alive anymore, or wish to fall asleep and not wake up.

  • Have you wished you were dead or wished you could go to sleep and not wake up?

Item 2. Non-Specific Active Suicidal Thoughts

General non-specific thoughts of wanting to end one’s life/commit suicide, “I’ve thought about killing myself” without general thoughts of ways to kill oneself/associated methods, intent, or plan.

  • Have you actually had any thoughts of killing yourself?

Item 3. Active Suicidal Ideation with Any Methods (Not Plan) without Intent to Act

Person endorses thoughts of suicide and has thought of a least one method during the assessment period. This is different than a specific plan with time, place or method details worked out. “I thought about taking an overdose but I never made a specific plan as to when where or how I would actually do it … and I would never go through with it.”

  • Have you been thinking about how you might kill yourself?

Item 4. Active Suicidal Ideation with Some Intent to Act, without Specific Plan

Active suicidal thoughts of killing oneself and patient reports having some intent to act on such thoughts, as opposed to “I have the thoughts but I definitely will not do anything about them.”

  • Have you had these thoughts and had some intention of acting on them?

Item 5. Active Suicidal Ideation with Specific Plan and Intent

Thoughts of killing oneself with details of plan fully or partially worked out and person has some intent to carry it out.

  • Have you started to work out or worked out the details of how to kill yourself? Do you intend to carry out this plan?

Item 6. Suicide Behaviour Question:

  • Have you ever done anything, started to do anything, or prepared to do anything to end your life?
    Examples: Collected pills, obtained a gun, gave away valuables, wrote a will or suicide note, took out pills but didn’t swallow any, held a gun but changed your mind or it was grabbed from your hand, went to the roof but didn’t jump; or actually took pills, tried to shoot yourself, cut yourself, tried to hang yourself, etc.
  • If YES, ask: How long ago did you do any of these?
    • Over a year ago?
    • Between three months and a year ago?
    • Within the last three months?