Abstract
Background
In emergency psychiatry high levels of suicidality are encountered. However, empirical studies about the extent of suicidal ideation remain scarce, and data are often gathered retrospectively via medical records. The present observational study aims to assess suicidal ideation, planning, and actions experienced by patients presenting with psychiatric emergencies.
Methods
In a consecutive sample of 136 patients, a standard psychiatric examination including the Chronological Assessment of Suicide Events (CASE Approach) was completed. An evaluative monitoring form for the CASE Approach was designed to record and quantify the garnered information.
Results
Almost 60% of the patients reported suicidal ideation with some planning. Approximately 25% had experienced over 10 hours of suicidal ideation on their worst days in the past 2 months; 27% had procured their method of choice, and another 27% had practiced a small, moderate, or serious suicide attempt.
Conclusion
Findings suggest more intense suicidality in emergency psychiatry than has been reported in previous empirical studies.
Keywords
Introduction
A veterans health administration study delineated that many patients who die by suicide, denied suicidal ideation (SI) to clinicians. 1 In his review of the literature, Berman found that about 75% of patients—when last interviewed before their death—made such denials. 2 Several studies have also indicated that the time duration from the decision to act to the initiation of suicide is often short.3–5 Opinions vary on how such results should be interpreted.
On the one hand, one may interpret the results as indicating that suicidal behavior can be characterized as impulsive, and that, in many cases, there is no specific suicidal plan or significant SI prior to the attempt.3,6–10
Others have disagreed with this interpretation citing that it appears illogical considering the difficulty inherent in proceeding with a lethal attempt. These theorists suggest that the apparent lack of SI in such patients is better explained by the marked hesitancy of patients—particularly those in immediate danger of suicide—to share their ideation including their method of choice (MOC) for fear of consequences such as the removal of the MOC and/or involuntary commitment.2,11–15
Anestis and colleagues argue that suicidal behavior is rarely if ever impulsive—that it is too frightening and physically distressing to engage in without forethought. 11 They postulate that, with many patients who have been reported as acting impulsively, their ability to act “impulsively” occurs because of significant prior SI and planning (sometimes weeks or even months in advance). This suicidal forethought desensitizes the patient and allows them to procure and practice with their MOC making it immediately available for “impulsive action” on the day of the attempt. In their meta-analytic study, they found a minimal correlation between the magnitude of impulsivity as a personality trait and suicidal behavior.
The work of Anestis was presented through the lens of the interpersonal–psychological theory of suicidal behavior, IPTS 16 and its iterations. 17 In this model, one of the 3 elements that can function as a potential harbinger of imminent suicide is the concept of acquired capability which develops through desensitization caused by repeated exposure to painful and/or frightening experiences associated with death.
Some of this desensitization may be related to externally generated desensitizers, which are often historically distal to the suicide attempt, such as exposure to domestic violence, military combat, and medical illness. But some of this desensitization may also be related to internally generated desensitizers such as the patient's own suicidal thoughts and actions some of which may be historically distal while others may be occurring within weeks, days, or minutes of the attempt. Van Orden et al 17 argue that, “. . . . the most direct route (but not the only route) to acquiring the capability of suicide (ie, the most potent painful and provocative experience) is by engaging in suicidal behavior, through suicide attempts, aborted suicide attempts (preparing for the attempt and nearly carrying it out), or practicing/preparing for suicidal behavior (eg, tying a noose; buying a gun with intent to engage in suicidal behavior; imagining one's death by suicide)” (p. 593).
Other theories of suicide such as the 3-step theory of Klonsky and May, as well as O’Connor's Motivational-Volitional model, include acquired capability as an integral component in the development of suicidal action.18,19 In these “ideation-to-action models,” one would expect to see significant amounts of SI, planning, and behavior in the months approaching a suicide attempt.
Emergency psychiatry settings are important mental health organisations for crisis intervention for persons with acute psychiatric problems and/or SIs. Empirical evidence supports the notion that up to 38% of patients who present in emergency psychiatry settings are associated with some potential for suicide.20,21 In Dhossche's study, 311 medical records were reviewed and suicidal behavior was considered present if current SI or attempts—within 24 h of or during emergency evaluation—were noted in the medical record. 20 Further, De Winter et al in their study recorded 14,705 psychiatric emergency services assessments in a period of 5 years and found that approximately one third of all patients displayed indications of suicidal behavior and 9.2% were at risk for attempting suicide. 21
Despite such research providing consistent evidence of the presence of SI in patients presenting with psychiatric emergencies, some research has indicated surprisingly low amounts of SI as evidenced by Hall's findings that in patients admitted to psychiatric units following severe suicide attempts, 14% reported a specific suicidal plan prior to their attempt and 29% reported that they had persistent and serious thoughts before the suicide attempt. 9 The hesitancy of patients—particularly those with intense intent—to share their ideation, including their MOC and actions taken upon it, has been frequently cited in the literature.3,12–15 Nonetheless, empirical studies exploring the patient's self-reported extent and intensity of suicidal thought (in terms of death wish, SI, planning, MOC, extent of action with MOC, and intent) among patients presenting with psychiatric emergencies remain scarce.
In the Netherlands, strangulation, auto-intoxication, and railway suicide are the most prevalent methods of completed suicide, accounting for up to 75% of all suicides. 22 Dutch emergency psychiatric settings have a regional function, with referrals from local general practitioners, public health emergency departments, and police stations. Usually, the first assessment is accomplished by a community mental health nurse together with a doctor (or sometimes a clinical psychologist), under the indirect or direct supervision of a psychiatrist. To assess suicidality, the Chronological Assessment of Suicide Events (CASE) Approach is recommended by the national multidisciplinary guideline regarding the assessment and treatment of suicidal behavior. 23
The CASE Approach is a well-known, broadly applied interviewing strategy that functions as a screening tool for the presence of SI that can immediately be flexibly expanded to uncover a comprehensive understanding of the patient's planning, actions, and intent regarding their MOC as well as alternative methods (AMs).14,15,24,25 Being a stand-alone interviewing strategy, it can be utilized in any clinician's personalized interview format or in more specific protocols such as the Collaborative Assessment and Management of Suicidality (CAMS). 26 In the Netherlands, the CASE Approach has been recommended by the national multidisciplinary guideline in 2012 as the standard interviewing approach for uncovering SI, and subsequently implemented in a nation-wide research study on effective means of training mental health professionals at 48 psychiatric sites across the Netherlands. 27 In Queensland, Australia, the CASE Approach has been implemented and investigated as a part of the Zero Suicide Framework, a systems approach to suicide prevention within a health service. 28 Empirical evidence showed a reduction in repeated suicide attempts after an index attempt and a longer time to a subsequent attempt for those receiving multilevel care based on the Zero Suicide framework. 29
Despite a growing acceptance of the role of acquired capability in the field of suicidology, to date, empirical studies have not shown the intensification of proximal SI as such models would suggest. Indeed, research has indicated a relative lack of such proximal SI. The current study was designed to determine more accurately the amount of SI, planning, and actions experienced by patients presenting with psychiatric emergencies. Determining whether proximal SI is common in high-risk patients and/or those patients who die by suicide is, arguably, one of the more pressing gaps in our understanding of suicidal phenomenology for it has both theoretical and practical implications. On a theoretical level, if much more proximal SI exists than previous studies have indicated, an important empirically supportive “missing piece” to the ideation-to-action models will have been uncovered. On a clinical level, if present, the question arises: Is there a clinical interviewing approach that can enhance the validity of the patients’ self-reporting of their proximal SI, planning, and actions?
In the current observational study, the extent and intensity of SI reported by patients presenting in a psychiatric crisis were assessed utilizing the CASE Approach. The aim of the study is to investigate the intensity of SI, planning, and actions experienced within patients in an emergency psychiatry setting. It is hypothesized that the intensity of the SI present in patients with psychiatric emergencies might be higher than previously reported.
Methods
The study included 136 patients: 96 patients of Altrecht Mental Health Care (crisis resolution team, Utrecht, The Netherlands) and 40 patients who stayed in the acute admission wards of the psychiatric hospital of Eleos Mental Health Care, Zeist, The Netherlands. Patients were referred to the crisis resolution team or were visited by 2 team members during outreaching visits (including to a hospital emergency department) because of suicidality or because of presumed (acute) psychiatric problems. 30 Patient files were not included in the present study when the patients were more difficult to interview because of aggressive or disorganized behavior in case of severe psychosis, cognitive disability, or substance abuse. Referrals because of other reasons than suicidality were included (including psychotic states and states of anxiety) but were underrepresented. The patients underwent standard psychiatric examination, including medical history, mental status examination, and suicide assessment, by a psychiatrist, psychologist, or resident psychiatrist, as usually in a combined assessment with a community mental health nurse.
The amount of SI, planning, and actions of all the patients was assessed using the Chronological Assessment of Suicide Events (CASE Approach).15,24,25 Clinicians utilizing the CASE Approach repeatedly use and sequence 7 operationally defined interviewing techniques—Behavioral Incident, 31 Gentle Assumption, 32 Normalization, Shame Attenuation, Denial of the Specific, the Catch-all Question, and Symptom Amplification 15 —that were designed to improve the validity of patient reporting while enhancing engagement. Clinicians screen and explore the patient's suicidal thoughts and actions—via the validity techniques—in 4 chronological sections in order to decrease errors of omission by the interviewer: presenting suicide events, recent suicide events (previous 2 months), past suicide events, and immediate suicide events (suicidal thought arising during the interview itself) (see Figure 1, 33 p. 152).
Chronological Assessment of Suicidal Events (CASE Approach 33 ).
In this way, interviewers are able to sensitively structure the interview while gathering data with an emphasis upon creating naturally flowing conversational interviews as opposed to interviews that employ a structured checklist and that may induce an uneasy experience in both the patient and the clinical interviewer. The strategy is usually completed within 2 to 8 min depending upon the complexity of the presentation.
The CASE Approach was initially developed in the Diagnostic and Evaluation Center: a combined emergency room, full-intake assessment center, and telephone triage center at Western Psychiatric Institute and Clinic at the University of Pittsburgh, Pennsylvania, from 1982 to 1988. As well as screening for SI, it was specifically designed to elicit SI with patients who may be at immediate (next 24 h) or imminent risk (next week) and are apt to withhold information.
When implemented with fidelity to the model, clinicians do not use any form of cue sheet nor do they record any rough notes or formal notes in the medical record while eliciting SI. This distinguishes the CASE Approach from semistructured formats. Consequently, clinicians trained in the CASE Approach can devote 100% of their attention toward engaging the patient while directly observing the patient for subtle nonverbal indicators that the patient is distorting or withholding information. The importance of creating such a sense of safety and trust has been shown to be critical for valid self-disclosure of SI. 34
Immediately after the interview, the data elicited by the CASE Approach were documented by the clinician using a specially designed evaluative monitoring form. Each clinician interviewed about 10 patients.
As part of their ongoing mental health care training, all clinicians had previously been trained in the CASE Approach in the year before performing their interviews. No experimental interventions or assessment tools were used with the patients. The following additional information was collected: gender, age, and main diagnostic classification in terms of the DSM-5. The study was approved by the Scientific Committee of Altrecht Mental Health Care, which assesses all research proposals on ethical standards, in concordance with the WMA declaration of Helsinki.
Monitoring Form, Quantification, and Documentation
The evaluative monitoring form was designed to enhance reliability and fidelity for research purposes as well as providing a format for documenting and quantifying information garnered by the CASE Approach. The following aspects of the suicidal state were quantified (see Table 2A-C). Items are: (a) Patient's “Stated Intent” on a scale of 1 to 10; (b) “Reflected Intent—Progress”, including whether a MOC has been chosen and action taken upon it; (c) “Reflected Intent—MOC/Alternative Methods”, including the form of the MOC/AMs; (d) “Reflected Intent—time spent thinking about suicide on the most desperate days in the last 2 months”; (e) Number and nature of past suicide attempts; (f) Quality of the contact with the patient as scored by the clinician (options were: “good contact,” “inconsistent contact,” “the patient is difficult to reach,” and “the client appears not to be cooperating”).
The evaluative monitoring form was developed in English and feedback via e-mail from the creator of the CASE Approach was integrated. The English version was translated to Dutch by the authors and approved by the original author using the forward–backward translation method. The Dutch version can be obtained from the authors.
Statistical Analysis
Descriptive statistics of the characteristics of the sample and CASE-interview outcomes were calculated. Intercorrelations between the items of the CASE Approach and other patient characteristics (gender and age) were computed using Spearman correlation.
Results
The demographic and clinical descriptive statistics of the 136 patients are presented in Table 1. Ninety-four women and 42 men were included: mean age was 37 years (SD = 16.0). Approximately half of the patients had a depressive disorder (with or without other comorbid disorder) as their primary diagnosis.
Characteristics of the Sample (N = 136).
The results show that 10 patients (7%) did not report any SI. Forty-six patients (34%) reported thoughts of death and/or vague thoughts of a suicide plan while denying any intention of proceeding with a plan (Table 2A, number 2-4 of A. Stated Intent). Another 22 (16%) reported having chosen a specific plan but denied any intention of acting upon it (Table 2A, number 5 of A. Stated Intent); 58 patients (42%) reported—not only having a well-defined plan but also stated they intended to proceed with their plan at some point.
The Results of the CASE Approach According to an Evaluative Monitoring Form Among 136 Patients: Stated Intent.
The item A scores are applied an ordinal scale variable.
The Results of the CASE Approach According to an Evaluative Monitoring Form Among 136 Patients: Reflected Intent (Item B) and Methods of Choice (Item C).
Abbreviation: MOC, method of choice.
The item B scores are applied an ordinal scale variable.
The Results of the CASE Approach According to an Evaluative Monitoring Form Among 136 Patients: Hours Spent (Item d) and Number of Suicide Attempts in the Past (Item E).
The item D scores are applied an ordinal scale variable.
The item E scores are dichotomized in further analyses (0 “no attempts” and 1 “any attempt in the past”).
The item F scores are dichotomized in further analyses (good contact 0 “no” and 1 “yes”).
Secondly, the majority of the patients (104; 77%) reported that they had chosen their MOC (number 1-7 of B. Reflected Intent). More specifically, 37 patients (27%) had their MOC available, and another 36 patients (27%) had undertaken a small, moderate, or serious attempt with their MOC. A minority of the patients reported an AM (39; 29%); an AM was not asked for by the interviewer in 43 patients (32%).
The most reported MOC was overdosing with medication and/or alcohol and/or drugs (39 patients; 29%); 27% reported—that on their worst days in the last 2 months—they spent more than 10 h thinking about suicide with another 9% of the patients reporting having spent 5 to 10 h. Seventy-nine patients (58%) had no history of a suicide attempt. Contact was rated as “good” by the clinicians in 83 of the 136 patients (61%). The clinicians qualified the contact with the remaining patients as inconsistent, difficult, or uncooperative.
Table 3 presents the associations between the items of the quantification of the CASE Approach. The suicidal intent scores show modest to moderate, significant correlations with more serious experience with the MOC, more suicidal thoughts on the worst days, and number of lifetime suicide attempts. Secondly, higher scores on item A (Stated Intent) and item D (hours spent) showed a modest correlation with worse contact about the SI as reported by the clinician.
Associations Between Gender, Age, and the Items A, B, D, and F, using Spearman Correlation Among 136 Patients.
Significant associations are printed in bold.
Abbreviation: TS, tentamen suicidii (suicide attempt).
N = 119.
N = 114.
N = 133.
Discussion and Conclusion
The results of this study suggest that a majority of the patients presenting with psychiatric emergencies may have Suicidal Ideation (SI), with SI defined as any suicidal thoughts or plans and/or a stated definite intention to act on such a plan. Further, the data suggest that about one third of the patients reported having had suicidal thoughts and planning for more than 5 hours per day on their worst days. Moreover, approximately 25% of the patients in the sample reported over 10 hours of suicidal thoughts and planning during their worst days in the past two months. The study reveals more extensive planning, more actions taken with their Method of Choice (MOC), and markedly more self-reported time spent experiencing suicidal thought and rumination when compared to previous empirical studies.20,21
Simpson et al 35 found in their study among 92,643 patients that the Columbia-Suicide Severity Rating Scale (C-SSRS) Screener was insensitive to suicide risk after emergency department discharge. They concluded that most patients deny suicidality prior to their fatal suicide attempt. Importantly, the level of suicidal intent may be the most powerful predictor of eventual suicide after attempted suicide. 36 Further Pompili et al 37 found in an emergency department that patients after a suicide attempt report significant more SI than nonattempters. They underline the importance of asking in a direct way about possible SI because patients may not self-report their SI. Further, they suggested that a suicide attempt in patients who need a psychiatric admission is followed by persistence of the wish to die. Olfson et al 38 described in an emergency department setting that patients with a history of suicide attempts or self-harm had a high suicide risk the year following discharge.
The results of the current study are not in line with Hall's research where 84% of 100 patients with serious suicide attempts reported no specific suicidal plan prior to their attempt and 69% claimed they had no suicidal rumination prior to their attempt. Busch et al 39 had reported that 78% of the patients denied SI prior to their suicidal act.
These discrepancies between previous studies and the current one are consistent with both Berman's contention that the hesitancy of patients to share their SI (including planning) may hamper suicide assessment in clinical practice 2 and De Winter et al 21 who mentioned the possible dangers related to an underestimation of suicide potential caused by missed or incomplete assessments. An advantage of the current study is its observational design, for previous studies utilized a retrospective review of medical records. 2
In addition, when compared to previous studies, the magnitude and amount of actions taken by the patients on their MOC suggests that the use of a flexible, yet well-operationalized, interviewing strategy such as the CASE Approach might enable clinicians to more accurately detect the amount and intensity of SI.14,15,24,25 The apparent increase in validity seen with the use of the CASE Approach may be related to several factors including: its utilization of operationalized interviewing techniques from the field of clinical interviewing, its intensive exploration for SI and hidden MOC during the previous 2 months, its ability to conversationally flow from screening questions into a detailed reporting of extent of action taken, the maintenance of naturalistic eye-contact, and its emphasis upon patient engagement (no cue sheets and no note-taking during the elicitation of SI).
Strangulation was reported as a MOC or Alternative Method (AM) by a small minority of the patients, despite strangulation being by far the most frequent method of suicide in the Netherlands. 40 Perhaps the population of patients presenting to psychiatric emergency services is only a selection of those in the general population who proceed to die by suicide. Strangulation is somewhat less common in women in the Netherlands (37% vs 51% in man 40 ), and the male to female ratio of suicide rate is about 2:1. This sample has a relative overrepresentation of women; in general, female gender is associated with a higher prevalence of nonlethal suicidal behavior. 41 As soon patients are identified with any, possible suicidal intention, gender prevalence differences may be less relevant in how the presence of suicidal intent can be elucidated. Another possible explanation may be that patients did not report strangulation because some clinicians did not demonstrate full fidelity to the CASE Approach, representing one of the limitations to the study. When employed with full-fidelity, clinicians always ask directly about AMs of suicide, often specifically inquiring about the most common methods within a given culture. 15 Suicidal ideation and hours spent showed a correlation with a less (perceived) cooperation with the patient. This finding is in line with a recent study by Van Veen and colleagues, 42 who described in a similar population a positive association between intensity of suicidality and perceived “difficulty” of the patient by mental health care nurses. This poor cooperation reflects the intensity of the actual suicidal intent, elicited by the assessment and moving, in terms of the CASE method, from “withheld intent” to “reflected intent.”
The results of the current study show that a quarter of the patients had done a small, moderate, or serious suicide attempt, while another 11% of the patients has been put the MOC in hand while thinking about killing oneself. These results are in line with the concept of the factor acquired capability of the Interpersonal Theory of Suicidal.16,17 The severity of the SI, planning, and suicidal actions may function as exposure to the frightening experience associated with death.
A limitation of the study is that the results are based on a single assessment without control group and could be prone to a recall-bias; it may be that the nature of the interview could distort the memory of a patient with respect to the number of hours thinking about suicide. The use of intensive longitudinal methods, like the Experience Sampling Method can solve such bias (eg, Hallensleben et al 43 ).
In future studies, robust one-to-one comparisons of the effectiveness of the CASE Approach could be made with other screening tools and methods for eliciting SI such as the Ask Suicide-Screening Questions 44 Part B of the First Session Suicide Status Form from the CAMS, 26 or the C-SSRS. 45
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publi cation of this article: Matthias Jongkind and Bart van den Brink provide trainings on the CASE Approach. Shawn Shea is the Director of the Training Institute for Suicide Assessment & Clinical Interviewing (TISA) for which he provides their online courses and live trainings on the CASE Approach. He receives book royalties from Elsevier and Mental Health Presses.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
