The Short-Term Assessment of Risk and Treatability (START) is a succinct structured professional judgment guide for the assessment of dynamic vulnerabilities and strengths associated with risk of violence, self-harm, suicide, unauthorized leave, substance abuse, self-neglect, and being victimized by others. Working with individuals in criminal justice, psychiatric, and community mental health contexts calls for practitioners to identify the risks that individuals might pose to themselves or others (e.g., violence, suicide, non-suicidal self-injury) and to effect management plans to prevent undesirable outcomes (e.g., mental illness relapse, rehospitalization, violence, recidivism, substance use relapse, suicide, self-harm). Given the pervasive risks to self and others in these populations and the substantial personal and social burdens, risk assessment is a process with inherently high stakes. It is fraught with challenges, including the potential for liability risk and ethical issues for the assessor, reduced civil liberties for the individual, stigma for the population, and economic implications and safety concerns for the public. Considerable investment has been made in exacting how best to conduct risk assessments and to translate those findings into effective risk management plans. The START is one such endeavor.
The START has been recognized by several agencies (e.g., Accreditation Canada, UK Department of Health) and experts as a promising, leading, or best practice in the assessment and management of violence and related risks. It is being used across the globe in correctional, forensic, and mental health institutions and community programs. As of 2018, there were nine language translations (Danish, Dutch, Finnish, French, German, Norwegian, Spanish, Swedish, and Thai; with Italian, Japanese, and Russian translations in preparation) and more than 50 published articles on the START. There also is a version for use with youth aged 13–18 years, the START: Adolescent Version.
This article begins by discussing the development, key characteristics, and use of the START; then the entry focuses on the research evidence for the START.
Development and Key Characteristics
The START was developed in the early 2000s in direct response to limitations of existing risk assessment measures, as identified in the research literature and by direct care providers and clinicians. The START is unique compared to other risk assessment measures because of several key characteristics, including its focus on dynamic factors, strengths and vulnerabilities, and multiple short-term adverse outcomes.
Dynamic Factors
START assessments focus on an individual’s current functioning on dynamic factors, within the context of their history. Historical, static factors (i.e., unchangeable variables, such as a history of child abuse, criminal record) and stable factors (i.e., variables that can change but that often are difficult or slow to change, such as antisocial attitudes) provide essential insights into future behavior and effective risk management endeavors. However, research also demonstrates the value of dynamic factors (i.e., changeable variables, such as social support), particularly in imminent, acute, and short-term assessments. Moreover, good clinical and criminal justice practice demands that assessors are attentive not only to factors that increase risk but also those that can be changed through intervention.
Strengths and Vulnerabilities
The START acknowledges that clients have strengths and vulnerabilities that can coexist and that interact with each other. The START thus guides assessors to evaluate an individual’s vulnerabilities (i.e., risk factors) as well as their strengths (i.e., protective factors). Prior to the development of the START, several well-respected scientist– practitioners in clinical risk assessment lamented that the field had failed to integrate balanced risk assessments. This is akin to a financial planner only asking about a client’s debts and not taking into account the client’s savings and investments. There are many potential advantages to considering an individual’s capacities, positive assets, and skills in addition to their challenges when assessing risk. Specifically, talking to an individual about his or her strengths can help bolster the individual’s sense of self and foster motivation and engagement, contributing to improvements in clinical outcomes and, ultimately, reductions in risk. Strengths also can be embedded in the risk management and treatment plans, providing opportunities to maintain and enhance current strengths and identify what might have fostered prior periods of stability and success (e.g., residing with a supportive family member or partner). Identification and use of strengths is also an inherent part of trauma-informed care and a recovery-oriented approach. Finally, research supports the unique contributions of protective factors to the prediction of adverse outcomes, above and beyond risk factors.
Comprehensive Care Planning: Multiple Adverse Outcomes
The START guides assessors to estimate risk of diverse health and safety concerns that are found across the populations served by social, criminal justice, and mental health agencies, with the goal of informing comprehensive care planning. Marginalized and high-risk populations experience elevated rates of violence, suicide, self-harm, substance abuse, self-neglect, unauthorized absence, and being victimized by others. Frequently, these individuals present with several of these challenges over the course of their lifetime or at any one time. Research demonstrates, for instance, that individuals with mental illnesses are at greater risk of being violently victimized than they are of committing violence. As such, a comprehensive intervention strategy may be most effective in reducing risk across multiple domains.
Integrated Assessment, Management, and Treatment
Built on a combination of historical information and dynamic factors, the START integrates risk assessment with risk management and intervention. Practitioners (e.g., probation officers, case managers, nurses, psychologists, psychiatrists) can remain attentive to an individual’s persistent risk state, while tuning up management plans and interventions to best match the individual’s current needs. To demonstrate, as a baseline, an individual with a history of suicide will be considered to be at greater risk of future suicide than an individual without a history of suicide. However, if the suicide attempt was 15 years ago when the individual was a teen, the attempt was more of a gesture (e.g., low lethality, while his parents were home, following a breakup with his first girlfriend), the individual has not had suicidal ideation since, denies experiencing relevant thoughts, plans or affect, and is presently happily married, then the suicide risk would be considered low.
Repeated Assessments
Reflecting the focus on dynamic factors, the START is intended to be completed every 2–3 months to document change in psychosocial functioning and safety and to inform revisions to the management and treatment plans, if any. When there is evidence or anticipation of acute risk, the START should be repeated more frequently. In addition, deterioration or substantial improvement in mental state or situational circumstances could prompt more frequent reevaluation. In the previous example, if during a subsequent discussion, the client reported that his wife was divorcing him and he was feeling depressed, had passing thoughts of suicide, but had no current intention or plans, his risk of suicide would increase to moderate, at a minimum. In addition, the practitioner may consider a more in-depth suicide assessment and follow-up.
The objective of hospitalization, probation, and treatment is to prevent undesirable outcomes (e.g., rehospitalization, recidivism, substance abuse relapse, loss to treatment) and to bring about positive improvements in the individual’s mental health and well-being. Most hospitalized patients and inmates eventually return to the community, and consistent with community-based care being the preferred mode of mental health service, it is generally expected that individuals will transition through the continuum of care. As such, a measure like the START that demonstrates therapeutic progress or deteriorations is essential. In this way, the START is also highly useful for treatment evaluation research.
Description and Use
The START comprises 20 items: Social Skills, Relationships, Occupational, Recreational, Self-Care, Mental State, Emotional State, Substance Use, Impulse Control, External Triggers, Social Support, Material Resources, Attitudes, Medication Adherence, Rule Adherence, Conduct, Insight, Plans, Coping, and Treatability), each coded for both Strength and Vulnerability and for its historical relevance to safety and well-being. Specifically, Key Items are used to identify strengths that have been known to buffer risk or may act as a therapeutic lever, whereas Critical Items are used to identify vulnerabilities that have previously triggered a cycle of decompensation and increased risk. To demonstrate, an individual may have prosocial family support and may also socialize with antisocial peers; thus, the START assessor would code the presence of both strengths and vulnerabilities on the Social Support item. Assessors also are prompted to identify a specific set of symptoms and signs that are unique to the individual and may be seemingly unrelated but reliably are associated with increased risk of harm to self or others, called Signature Risk Signs.
The START was developed in the structured professional judgment tradition, meaning that assessors do not use numeric algorithms or cut scores to inform their final risk judgments. Instead, assessors estimate risk over the next 3 months for the seven outcomes (i.e., violence, suicide, self-harm, substance abuse, self-neglect, unauthorized absence, and victimization) as low, moderate, or high through consideration of Strengths and Vulnerabilities ratings, Key and Critical Items, Signature Risk Signs, and history of each of the outcomes. Once areas of elevated risk are identified, the START, in the section on Risk Formulation, guides assessors to describe the specific factors that predict or explain how, when, and against whom these risks become elevated. The START assessment is then used as the blueprint for a comprehensive risk management and integrated treatment plan.
Although the focus of the START is on recent and current functioning, assessors integrate past and present information to inform their risk assessments. Specifically, the Strength and Vulnerability ratings on each of the 20 dynamic factors are based on the individual’s functioning in the past 2–3 months or since the last START assessment. Past and present information is used to code Key and Critical Items and Signature Risk Signs and to identify history of the risk domains.
START is intended to support professionals working with diverse populations (e.g., homeless individuals, civil and forensic psychiatric patients, probation clients). User qualifications do not require that assessors hold a PhD or MD; however, several items require expertise and experience in mental health. Completion of START assessments within the context of a treatment or case management team is highly encouraged, though completion of START assessments by individual practitioners is common.
Research Evidence
Research on the START has demonstrated good to excellent interrater reliability, moderate concurrent validity, and effect sizes for predictive validity, generally in the moderate to large range. Specific Risk Estimates have demonstrated incremental validity over Vulnerabilities and Strengths items for aggression, self-harm, suicidality, and victimization, while research on user acceptance reports high rates for clinical as well as risk management utility.
Interrater Reliability
Several studies have evaluated the interrater reliability of START assessments. Generally, mean intraclass correlation coefficients are in the good to excellent range for the vulnerability, strength, and risk estimates. Researchers and clinicians also have good concordance, as do mental health professionals from diverse disciplines (e.g., nursing, psychiatry, social work). Agreement on Key Items and Critical Items is less consistent; research to date also suggests that it may vary by profession.
Concurrent Validity
Researchers have examined the association between the START and the Psychopathy Checklist–Revised, the Historical-Clinical-Risk Managment-20, the Suicide Risk Assessment and Management Manual, and the Level of Service Inventory–Revised and found START assessments to correlate with assessments completed using other measures in the expected manner. For instance, moderate positive correlations of the START Vulnerability total scores and moderate negative correlations of the START Strength total scores with Psychopathy Checklist–Revised and Historical-Clinical-Risk Managment-20 total scores have been observed.
Predictive Validity
To date, the majority of published studies on START assessments have focused on the prediction of violence risk. The majority of studies demonstrate a significant association between inpatient aggression of various forms (e.g., any, verbal, physical) and both Vulnerability and Strength total scores; however, the association for strengths is less consistent in the research. These effect sizes are generally in the moderate to large range and are highly consistent with meta-analyses in the field. Predictive validity for the other six risk domains is promising, but variable, and studied less frequently. There is also emerging evidence of validity in predicting general offending.
Incremental Validity
Specific Risk Estimates have demonstrated incremental validity over Vulnerabilities and Strengths items for outcomes such as aggression, self-harm, suicidality, and victimization; that is, a low, moderate, or high rating on the seven risk estimates is a better predictor of adverse outcomes than simply summing up the item scores. In addition, Strength total scores have demonstrated incremental validity over the Vulnerability total scores in the prediction of aggression and general offending, although this finding is less consistent and requires further study.
Change and Distribution of Scores
Research demonstrates that when the START is administered repeatedly over time, changes are evident in Strength and Vulnerability total scores, as well as the Specific Risk Estimates, that correspond to changes in security needs and that can be monitored and evaluated in association with targeted interventions. This means that on average, patients’ Strength scores increase and their Vulnerability scores decrease as they move through the continuum of care—from locked/secure units to medium-security units and eventually to minimum security units. Moreover, even within high-risk populations, START assessments have excellent dispersion. For example, in forensic psychiatric inpatients and community-based offenders, a reasonably normal curve on both Strength and Vulnerability total scores and a good distribution across risk estimate ratings (i.e., low, moderate, or high risk) have been demonstrated. Despite sharing many demographic (e.g., male, age) and static risk factors (e.g., a history criminal offending, violence, lengthy psychiatric histories), a small proportion of individuals have very low or very high Vulnerabilities and a small proportion of individuals have very high Strengths and very low Strengths.
Perceived Utility and User Feedback
Feedback from assessors provides strong support for the perceived utility of the START in practice. Several studies have shown that assessors find the START easy to use and informative for conceptualizing short-term risks and needs. Research also demonstrates that among those using multiple risk assessment measures, START is often ranked most highly.
References:
- Cartwright, J. K., Desmarais, S. L., Hazel, J., Griffith, T., & Azizian, A. (2018). Predictive validity of HCR-20, START, and Static-99R assessments in predicting institutional aggression among sexual offenders. Law and Human Behavior, 42, 13–25. Retrieved from http://doi.org/10.1037/lhb0000263
- Desmarais, S. L., Van Dorn, R. A., Telford, R. P., Petrila, J., & Coffey, T. (2012). Characteristics of START assessments completed in mental health jail diversion programs. Behavioral Sciences & the Law, 30, 448–469. doi:10.1002/bsl.2022
- Dickens, G. L., & O’Shea, L. E. (2015). How short should short-term risk assessment be? Determining the optimum interval for START reassessment in a secure mental health service. Journal of Psychiatric and Mental Health Nursing. doi:10.1111/jpm.12232
- Lowder, E. M., Desmarais, S. L., Rade, C. B., Johnson, K. L., & Van Dorn, R. A. (2017, April 15). Reliability and validity of START and LSI-R assessments in mental health diversion clients. Assessment. http://doi.org/10.1177/1073191117704
- O’Shea, L. E., & Dickens, G. L. (2014). Short-Term Assessment of Risk and Treatability (START): Systematic review and meta-analysis. Psychological Assessment, 26(3), 990–1002. doi:10.1037/a0036794
- O’Shea, L. E., Picchioni, M. M., & Dickens, G. L. (2015). The predictive validity of the Short-Term Assessment of Risk and Treatability (START) for multiple adverse outcomes in a secure psychiatric inpatient setting. doi:10.1177/1073191115573301
- Van den Brink, R. H. S., Troqueste, N. A. C., Beintema, H., Mulder, T., van Os, T.W.D.P., Schoevers, R. A., & Wiersma, D. (2015). Risk assessment by client and case manager for shared decision making in outpatient forensic psychiatry. BMC Psychiatry, 15, 120. doi:10.1186/s12888-015-0500-3
- Viljoen, J. L., Cruise, K. R., Nicholls, T. L., Desmarais, S. L., & Webster, C. D. (2012). Taking stock and taking steps: The case for an adolescent version of the Short-Term Assessment of Risk and Treatability. International Journal of Forensic Mental Health, 11, 135–149.
- Webster, C. D., Martin, M.-L., Brink, J., Nicholls, T. L., & Desmarais, S. (2009). Manual for the Short-Term Assessment of Risk and Treatability (START) (Version 1.1). Coquitlam, Canada: Forensic Psychiatric Services Commission and St. Joseph’s Healthcare Hamilton.
- Webster, C. D., Nicholls, T. L., Martin, M.-L., Desmarais, S. L., & Brink, J. (2006). Short-Term Assessment of Risk and Treatability (START): The case for a new structured professional judgment scheme. Behavioural Sciences and the Law, 24, 747–766. doi:10.1002/ bsl.737