Establishing Standards for the Assessment of Suicide Risk Among Callers to the National Suicide Prevention Lifeline

On January 1, 2005, the National Suicide Prevention Lifeline (1–800–273–TALK) was launched. Lifeline, supported by a federal grant from the Substance Abuse and Mental Health Services Administration (SAMHSA), consists of a network of more than 120 crisis centers located in communities across the country that are committed to suicide prevention. Persons in emotional distress or in suicidal crisis can call the toll-free number at anytime from anywhere in the nation and are routed to the nearest networked crisis center. Callers are then connected with a trained telephone worker who can provide emotional support, assessment, crisis intervention, and/or linkages to local treatment and support resources, including emergency services.

Two major goals of Lifeline are to promote efficient access to this service so it will reach more people nationwide at risk of suicide, and to ensure high quality of services to its callers so as to more effectively prevent suicide. Lifeline established a subcommittee of American and Canadian suicide prevention experts in March 2005 to consult on developing standards and recommended practices for its network of crisis centers. Lifeline's Certification and Training Subcommittee's (CTS) extensive review of research and field practices yielded recommendations that are embodied in Lifeline's Suicide Risk Assessment Standards, which were phased in for implementation beginning in January 2007, with the expectation of network-wide adherence by September 2007.

In this article, we provide the background on the need for these standards; describe the process that produced them; summarize the research and rationale supporting the standards; review how these standard assessment principles and their subcomponents can be weighted in relation to one another so as to effectively guide crisis hotline workers in their everyday assessments of callers to Lifeline; and discuss the implementation process that will be provided by Lifeline.

THE NEED FOR EVIDENCE-BASED RISK ASSESSMENT STANDARDS

Because of their accessibility, crisis hotlines are in a unique position to intervene with individuals at various points along the pathway to suicidal behavior, including the moments or hours prior to fateful decisions. This special contribution to suicide prevention is undermined if staff members are unable, unwilling, or reluctant to persistently inquire about and explore suicidal thoughts and feelings with callers.

Recently completed SAMHSA-sponsored evaluations of crisis hotlines' processes and outcomes employed monitoring of hotlines and follow-up of callers to hotlines. These studies provided overall evidence in support of crisis hotlines' role in responding to crisis and suicidal callers, while raising some concerns about the quality of suicide risk assessments. Kalafat, Gould, Munfakh, and Kleinman (this issue) studied 1,085 suicidal and 1,617 nonsuicidal crisis callers to eight crisis hotlines. The hotlines had agreed to use standardized, evidence-based suicide risk assessments and measures of crisis states, assessed near the start and at the end of their calls; and, for those who consented, at a follow-up call approximately 3 weeks after the original call. Significant reductions in crisis and suicide status occurred during the calls and continued to the follow-up. Notably, in response to an open-ended question at follow-up as to what was helpful about the call, 11.6% (n = 44) of suicidal callers said that the call prevented them from killing or harming themselves.

Follow-up assessments were conducted with 801 of the 1,617 callers who had been categorized by centers as nonsuicidal crisis callers. At follow-up, 52 (6.5%) reported having suicidal thoughts when they had originally called the centers, and 27 of these said they had told the crisis worker of these thoughts. Crisis centers had not conducted risk assessments for these callers, and these callers were more distressed than callers who did not report suicidal thoughts. This study highlighted the need to inquire about suicide on crisis calls, particularly with more distressed callers (see Gould, Kalafat, Munfakh, & Kleinman, this issue; Kalafat et al., this issue).

In another study Mishara and colleagues (this issue) silently monitored 1,431 calls to 14 centers. Overall, when emotional or cognitive changes occurred from the beginning to the end of the calls, such changes were positive. Their report concluded that the centers had helped a significant number of callers and may have saved some lives. For example, at the end of the calls, 52.3% of callers were less confused and more decided about next steps, 48.7% were less helpless and more resourceful, and 40% were more hopeful.

Of the 1,431 callers, 723 were not asked about suicidality. Of the 474 who were asked or who reported suicidal thoughts, no questions about the means were asked on 46% of the calls. In 159 instances when the helper was aware that the caller was considering suicide and had determined what means to use, the helper asked only 30 of those callers if an attempt was in progress. Questions about prior attempts were asked of only 104 callers. The study characterized these risk assessments as not following the accreditation guidelines of the American Association of Suicidology or procedures mandated by their own center directors.

It should be noted that failure to conduct appropriate suicide risk assessments or to pursue clients' suicidal communications is not unique to crisis hotline staff, as this has been found also among professional mental health providers (Bongar, Maris, Berman, & Litman, 1998; Coombs et al., 1992) and among primary care physicians (Williams et al., 1999). Nevertheless, this finding for organizations, many of which include suicide intervention as a primary part of their mission, prompted the CTS of the Lifeline Network to make the development of standards for evidence-based risk assessment a first priority.

Primarily due to their accessibility to callers in immediate suicidal crisis, crisis hotlines must engage in the assessment of imminent risk. As telephone services, crisis hotlines face unique challenges in conducting suicide risk assessments and intervening with suicidal persons. Crisis workers must establish and maintain rapport with callers with whom there is less control than in face-to-face situations, who may be using a phone service primarily because they wish to retain this control, and/or may be reluctant to commit to face-to-face contact or ongoing treatment. They may also be using a phone service because they are currently in an acute state of distress or suicidality.

The challenge, then, is to conduct a systematic and thorough risk assessment within the connection and flow of a telephone contact. To accomplish this, crisis staff must be thoroughly familiar with the current risk and protective factors for suicide, and be comfortable enough with the topic to weave the risk assessment into the course of the call. Most importantly, crisis staff must be assured that persistent assessment of suicidal thoughts, feelings, and plans, as well as of alternatives and inhibitors, is the most effective way to reduce callers' isolation, anxiety, and despair, and to begin the exploration of alternative ways of addressing their problems.

THE PROCESS OF DEVELOPING SUICIDE RISK ASSESSMENT STANDARDS FOR NETWORK CENTERS

Empirical Basis for the Standards

Empirical research and clinical experience suggest that suicidality is a multifaceted phenomenon. Research to date indicates that three facets—suicidal desire, suicidal capability, and suicidal intent—cover the domain of the phenomenon (and importantly, are not redundant with one another; Beck, Brown, & Steer, 1979; Joiner, Rudd, & Rajab, 1997; Joiner et al., 2003). We believe a fourth facet—buffers against suicidality—also needs to be included to provide a full framework for suicide assessment in the context of crisis center hotline work. These four facets, as well as their subcomponents, were influenced by and are compatible with the “IS PATH WARM?” warning signs mnemonic (where I = ideation, S = substance abuse, P = purpose-lessness, A = anxiety, T = trapped, H = hopeless, W = withdrawal, A = anger, R = recklessness, and M = mood changes; Rudd et al., 2006). In what follows, the four facets are described, some research on each is summarized, and the inter-relations of the facets are discussed (the four core principles and their subcomponents are summarized in Figure 1).

THE INTERRELATIONS OF THE FOUR FACETS, AND ATTENDANT IMPLICATIONS FOR CRISIS CALLS

Suicidal desire occurring independent of suicidal capability and/or suicidal intent typically presents a low-risk-of-suicide scenario. When desire combines with capability and/or intent, then suicidal risk may dramatically increase and the intervening impact of buffers may enter into the equation. Below are representations of possible combinations of these factors, but it is important to emphasize that assigning risk status to callers should not interfere with or take precedence over establishing empathic contact with callers.

Starting with the clearest—and highest risk—scenario, when suicidal desire, suicidal capability, and suicidal intent are all present, risk is high, and this is essentially true regardless of the presence of buffers. This last phrase is key, as clinical experience points to many people who have died by suicide even in the presence of buffers. When desire is paired with either intent or capability (but not with both), risk is lower but still considerable, and the determination of whether risk is particularly high rests with the safety afforded by buffers—if safety is high, risk is more moderate (though still elevated and in need of regular monitoring); if safety is low, risk is approximately as high as when desire, capability, and intent are all present. Suicidal desire is best viewed as an indicator of acute distress or a symptom of a mood disorder, and does not entail significant risk on its own. Capability and intent are more pernicious, and here again, the safety afforded by buffers is partly determinative. If safety is high, capability and/or intent do not convey the higher risk categories, but may convey moderate risk and require regular monitoring. If safety is low, capability and/or intent is a more serious concern, and requires active intervention, though probably not to the level of rigor or immediacy occasioned by the combinations of desire, capability, and intent (see Figure 2 for graphic representation).

It is important to note that formulating an individual's risk for suicide is best practiced through a highly collaborative process, whereby efforts to engage and intervene with the caller are often seamlessly interwoven throughout the worker's assessment process. For example, research has shown that an individual's self-assessment of suicide risk may outperform clinical judgments (Joiner et al., 1999), suggesting that workers can further enhance their assessment by asking the caller to rate his/her own risk of suicide. In the work by Kalafat et al. and Gould and colleagues (both this issue), the “intent to die” assessed at the end of the call was the best predictor of the caller's later suicidality, indicating that interventions during the call itself can affect the degree to which the caller is ultimately assessed to be at risk.

IMPLEMENTATION PROCESS OF STANDARDS FOR SUICIDE RISK ASSESSMENT

In January 2007, the suicide risk assessment standards became policy for all Lifeline network crisis centers. Extensive technical assistance will be provided by the CTS and Lifeline. Some of these methods include: network-wide conference calls; newsletter articles; e-mail communications; sample suicide risk assessment questions and instruments; and individualized assistance when requested/needed. All network centers will be required to submit their suicide risk assessment instrument to the Lifeline Certification and Training Division for review to ensure that their tool meets network standards. Centers will also be encouraged to submit examples of suicide risk assessment trainings which demonstrate how they have incorporated the standards into their routine educational and skill-building activities for crisis line workers. Once reviewed by the CTS to ensure adherence to the standards, these examples will be posted and available to all network crisis centers, with the permission of the submitting crisis centers. It is expected that all Lifeline network centers will be in adherence with the new standards by late September 2007.

Lifeline is actively promoted nationally as a resource for suicidal persons. Lifeline's policy regarding the suicide risk assessment standards will require some degree of suicide risk assessment on every Lifeline call. As a suicide prevention hotline, it is essential that every Lifeline caller be assessed for suicide risk. A common misconception is that asking about suicide might aggravate or upset callers, or, in the extreme, “plant the idea in the person's mind.” Research does not support this assumption. A study examining the impact of suicide risk questions to at risk-youth (e.g., impaired from substance abuse, depressed, or past history of suicide attempt) as well as a general youth population found that neither group was distressed nor more suicidal following the introduction of the questions (Gould et al., 2005). However, as noted earlier, research has shown that failure to routinely ask hotline callers about suicide can allow for a significant number of suicidal persons to be missed (see Kalafat et al. and Mishara et al., this issue).

Lifeline's administrators recognize that a full suicide risk assessment covering all four core principles will not be appropriate for some callers. Therefore, for every Lifeline call, Lifeline's policy will require that telephone workers ask the callers about suicidality. The CTS will be recommending that crisis center staff ask a minimum of three “prompt questions” that, if answered affirmatively, could prompt a full scale assessment (e.g., “Are you—or the person you are calling about—thinking about suicide?”). These questions will address current suicidal desire, recent (previous 2 months) suicidal desire, and past suicide attempts. Clearly, it is important to elicit current suicidal desire given that the caller is calling the Lifeline at that specific moment. What is happening in the caller's life that motivated him/her to reach out by calling? If the caller denies current suicidal ideation, inquiring about recent suicidal ideation (e.g., past 2 months) may provide insight into the caller's emotional state. In addition, a caller may feel more ready to acknowledge previous thoughts/behaviors rather than to discuss the more immediate situation. Depending on how the crisis center worker responds, discussing previous suicidal desire and/or attempts can increase rapport and trust leading to disclosure of current suicidal desire, if present. Inquiring about previous suicidal attempts also allows for the telephone worker to engage the caller in a discussion about what happened before, during, and after the attempt, which has the potential to increase awareness of the caller's coping skills, reasons for living, and awareness of available resources.

Centers can incorporate these standards and recommendations into their current risk assessments by simply adding those subcomponents of the standards that are not addressed in their assessments; or, by adopting an alternative risk assessment instrument that addresses all the subcomponents. The CTS also recognizes that telephone workers conducting risk assessments need not address each subcomponent in a rote, survey-like manner. Often, risk status can be judged based on clear statements by callers; by their answers or elaborations in response to a few questions; or by obvious indicators, such as an attempt in progress (for example, the caller reporting the ingestion of a lethal dose of pills).

Lifeline's Certification and Training Division will offer free (to Lifeline networked centers), evidence-informed trainings on how to incorporate the suicide risk assessment questions into the dialogue with a caller. These trainings will also include how to establish rapport with callers to enhance assessment and intervention practices, as well as how the assessment can be utilized in the context of collaborating with callers to better ensure their safety.

In closing, we emphasize that these guidelines are specific to crisis hotline contexts, in which factors like appearance and observed behaviors are inaccessible. Further, it should be noted that the framework described herein, though research-based, has not itself been empirically evaluated. Relatedly, we view these guidelines as current as of early-mid 2007, but we note that they may evolve as more information on them is gathered. Indeed, we are planning longitudinal, empirical work to evaluate and improve the standards, and we encourage others to do so as well.