G Model
AJP-631; No. of Pages 7 Asian Journal of Psychiatry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Asian Journal of Psychiatry journal homepage: www.elsevier.com/locate/ajp
Review
Identification of suicidal ideations with the help of projective tests: A review Devvarta Kumar a,*, S. Haque Nizamie b, Priyadarshee Abhishek b, Lavanya Tumkur Prasanna a a b
Department of Clinical Psychology, National Institute of Mental Health and Neuro Sciences, Bangalore 560029, India Central Institute of Psychiatry, Ranchi 834006, India
A R T I C L E I N F O
A B S T R A C T
Article history: Received 2 May 2014 Received in revised form 24 July 2014 Accepted 27 July 2014 Available online xxx
Identification of the presence of suicidal ideations in an individual is crucial for the timely intervention. However, these ideations may remain unidentified as an individual with serious intentions of self-harm may not express them explicitly. Various projective tests can provide crucial clues to clinicians about the presence of suicidal ideations in an individual’s mind. The present review is intended to update clinicians working in the field of suicide prevention about salient findings on these tests which can serve as a ready reckoner for them. We also highlight the status of research in this domain. ß 2014 Published by Elsevier B.V.
Keywords: Projective test Suicide assessment S-Con
Contents 1. 2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rationale for the use of projective tests in the identification Rorschach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. The single-sign approach. . . . . . . . . . . . . . . . 2.1.1. 2.1.2. The constellation approach . . . . . . . . . . . . . . Suicide ideation indicators on other projective tests . 2.2. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.. of .. .. .. .. .. ..
............... suicidal ideations ............... ............... ............... ............... ............... ...............
1. Introduction Prevention of suicide is a daunting but an indispensable task for any society. According to the World Health Organization (WHO), suicide claims one life every 40 s (WHO, 1999). In India, a total of 1,35,445 people died due to suicide in 2012 alone (NCRB; National Crime Record Bureau, 2012). The data increases manifold when it comes to counting people with failed attempted suicides (WHO, 1999). A timely identification of the presence of suicidal ideations can save numerous lives. Unfortunately, suicidal ideations are
* Corresponding author. Tel.: +91 80 26995188. E-mail address: [email protected] (D. Kumar).
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
000 000 000 000 000 000 000 000
often not expressed explicitly and the subtle signs of their presence may come to notice only after an individual’s suicide attempt (e.g., realization that the individual was giving away his valuables before attempting suicide). It becomes more important in case of vulnerable individuals (e.g., persons with depression) who may have lower threshold to attempt suicide and the occurrence of suicidal ideations can lead to fatal steps. Therefore, it is imperative that clinicians are able to identify people who have suicidal ideations but may not express them explicitly. A few projective psychological tests, by virtue of the capability to detect traits and states related to suicidal behavior (e.g., impulsivity, sense of hopelessness), can be of help in identification of the presence of suicidal ideations in an individual’s mind. The present narrative review is intended to update clinicians about the
http://dx.doi.org/10.1016/j.ajp.2014.07.004 1876-2018/ß 2014 Published by Elsevier B.V.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 2
D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
indicators of suicidal ideations on some commonly used projective tests. We compile the existing data by means of a narrative review rather than a meta-analysis for two reasons. First, the studies are too less and the sample sizes of most of the studies are very small. Therefore, conducting a meta-analysis may result in more research artifacts than facts. Second, the variables studied in different studies are quite varied and, thus, it is difficult to merge data of these studies to conduct a meta-analysis. 2. Rationale for the use of projective tests in the identification of suicidal ideations The use of projective tests for the purpose of identification of the presence of suicidal ideations is based on the assumption that a preoccupation with self-harm intentions may reflect on the ambiguous materials of these tests. Also, the individual may use the testing situation to communicate suicidal intents (Sullivan and Bongar, 2009). We discuss here a few projective tests in the context of suicide assessment. 2.1. Rorschach For more than 50 years, the Rorschach test has been researched to explore its potential value as a tool to assess suicide risk (e.g., White and Schreiber, 1952; Blatt and Ritzler, 1974; Exner and Wylie, 1977; Blasczyk-Schiep et al., 2011). The Rorschach researchers have followed two approaches in suicide assessment—the single-sign approach and the constellation approach. 2.1.1. The single-sign approach This approach refers to the use of a specific Rorschach sign to predict suicide. It has the advantage of quick identification of suicidality on the basis of an easily identifiable Rorschach variable (Hansell et al., 1988). For instance, White and Schreiber (1952) proposed ‘whirling’ as one of the responses on Rorschach to be indicative of a mind preoccupied with death wishes. Subsequently, Thomas and Duzynski (1985) reviewed Rorschach responses of more than 1000 medical students and found that whirling and its synonyms are found more frequently in the protocols of those individuals who later commit suicide. The association of the word ‘whirling’ (and its synonyms) with suicidal ideations has been confirmed by another research in which individuals with prior suicidal ideations chose more number of spinning phrases than those who did not exhibit suicidal ideations on a different test (Lester et al., 1990). Likewise, research has shown that people having suicidal ideations frequently give transparency responses (Blatt and Ritzler, 1974; Rierdan et al., 1978; Kestenbaum and Lynch, 1978). Transparency responses mean a perception in which the individual sees an object in its transparent state (for example, ‘‘a jelly fish and the inner bones can be seen’’). Different explanations have been given for the presence of these responses in the protocols of individuals with suicidal ideations. The transparency responses are construed to be indicative of the merging of self into hated object or loss of the sense of self leading to feeling of loss or helplessness (Blatt and Ritzler, 1974). On the other hand, Rierdan et al. (1978) proposed that the transparency responses indicate the perception of death as a transitional phase or continuity and not an end. The color-shading responses have also been considered a sensitive indicator of suicidality. The color-shading blend is considered indicator of the failure to cope effectively with painful affect (Fowler et al., 2001a). Also it could be reflective of a mixed and confused emotional experience having both pleasure and pain components (Exner and Wylie, 1977). Appelbaum and Holzman (1962) found a relationship between the presence of the color-
shading responses and a history of attempted suicide. Later research supported the importance of the color-shading responses in the identification of suicidal ideations (Appelbaum and Colson, 1968; Colson and Hurwitz, 1973). Petot (2002) found that the color-shading responses are seen more frequently in the Rorschach protocols of children with suicidal ideations. The importance of color-shading responses in the identification of the presence of suicidal ideations can be understood by the fact that it has been included in most of the suicide constellations (which will be discussed in the next section). Apart from these, a few other signs such as the poor form level human movement responses and responses with morbid contents are also seen in the Rorschach protocols of individuals with suicidal ideations and have been included in suicide constellations (Petot, 2002; Silberg and Armstrong, 1992). However, the absence of a sign in an individual’s Rorschach protocol should not be taken as absence of suicidal ideations as some people may lack the ability to communicate the cognitively complex responses (such as transparency and cross-sectional responses) (Blatt and Ritzler, 1974). Also, it is not necessary that the presence of a sign is always related to current suicidality. For example, Hansell et al. (1988) found that the transparency responses were unrelated to current suicidal risks. 2.1.2. The constellation approach The instability and the lack of statistical power of single signs limit their use in predicting suicide (Kendra, 1979; Fowler et al., 2001b). The chances of both false-positives and false-negatives are high if an impression regarding the vulnerability of an individual to attempt suicide is formed on the basis of a single indicator. Therefore, attempts have been made to identify a set of Rorschach signs that can be used as a constellation to predict suicide. One of the earliest proposed constellations is the Martin’s checklist which consists of 17 Rorschach signs (Martin, 1951 as cited in Weiner, 1961). Limited research applying the Martin’s checklist to the Rorschach protocols of suicide completers, attempters and nonsuicidal groups have found the efficacy of this checklist in discriminating the suicidal and non-suicidal groups but not the suicide attempters and completers (Datson and Sakheim, 1960; Weiner, 1961). Kendra (1979) interpreted the Rorschach protocols by the Piotrowski method and found that six variables (FM, Fc’, C, F%, H% and W:M) could successfully classify suicide attempters, suicide effectors and psychiatric controls. In 1977, Exner and Wylie proposed a suicide constellation based on the Comprehensive System (CS) which is known as the ‘S-Con’ (Exner and Wylie, 1977). Research has highlighted some remarkable aspects of the S-Con. It consists of 12 Rorschach variables (initially 11 variables were identified and the 12th variable was added later) and a cut-off of eight is considered sensitive in detecting people with significant suicidal ideations (see Table 2). The cut-off of this constellation is based on interesting empirical findings. Exner and Wylie (1977) compared Rorschach protocols of patients who had later committed suicide with the protocols of non-suicidal individuals. When a cut-off of eight was used, the S-Con was able to identify around 75% of suicidal patients and 100% of non-patients. This finding was confirmed in a cross-validation study (Exner, 1993). Important to note that the suicide completers, identified by the S-Con, had committed suicide within 60 days of the test administration. Therefore, it is suggested that the protocol of any person having a score of eight or more on the S-Con must be taken very seriously. Moreover, it should be kept in mind that approximately 25% of people who commit suicide do not have a score of eight on the SCon and, hence, even the lower score should be taken as an indication for further careful assessment.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
Fowler et al. (2001b) studied three groups—patients with a history of lethal suicide attempt, patients with a history of parasuicidal attempt and college students without any significant history of suicide attempts. They found that among the nine psychiatric and demographic variables, an S-Con score of seven or more was the sole predictor of near-lethal suicide attempts within 60 days. The authors concluded that a strong prediction of near lethal suicide in the absence of prediction of the para-suicidal behavior indicates the discriminant validity of the S-Con. Furthermore, the inability of other CS variables and measures of psychopathology to predict near-lethal suicide attempts suggests that the S-Con score is associated with the behavior of suicide attempt itself and not with a general tendency of impulsivity or self-destructive behaviors. Also it is important to mention here that a decrease in one score from the cut-off suggested by Exner and Wylie (1977) indicates the need for flexibility in considering the S-Con score. The S-Con (with some modifications) has been used for the assessment of suicidal ideations in adolescent population also. Arffa (1982) compared four groups of adolescents on a Rorschach signs constellation. The constellation had 10 signs from the S-Con and two additional signs. The four groups of adolescents were previous suicide attempters (attempted suicide within 10 months prior to hospitalization), unit suicide attempters (suicidal action deemed by clinical staff to be serious enough to warrant one-toone observation for longer than three hours), dual suicide attempters (presence of two or more suicide attempts) and the control group (adolescents without any history of suicide attempts). A score of four or more was found to differentiate suicidal from non-suicidal group with 92% accuracy. Interestingly, none of the signs individually discriminated the group, thus, questioning the validity of the single-sign approach. Silberg and Armstrong (1992) developed a new Rorschach Suicidal Index for Adolescents (RSIA) consisting of six variables (see Table 2). They found that the presence of four out of six RSIA variables could correctly identify 64% of suicidal depressed adolescents. On the basis of commonly observed features on the Rorschach protocols of suicidal adolescents the authors concluded that these adolescents are introspective, exhibit painful arousal, have preoccupation with things of morbid nature, are impulsive, have misperceptions of people and show distorted reasoning skills. In a recent study, the incremental validity and the effectiveness of RSIA in suicidal adolescents were investigated (Blasczyk-Schiep et al., 2011). The findings of this study were in consonance with Silberg and Armstrong’s (1992) study. The presence of, at least, four indicators could correctly identify 69% adolescents who had serious suicide attempts in past. The index showed incremental validity in predicting suicide. Also, the scores correlated with the objective measures of suicide risk such as the score on the Reasons for Living Inventory (RLI; Linehan et al., 1983) and the number of suicide attempts. Though various studies have supported the efficacy of the constellation approach in predicting suicide (Exner, 1993; Fowler et al., 2001b), there are studies that have challenged its worth. For example, a few studies indicate that the Rorschach protocols of near-lethal suicide attempters and suicide completers may not cross the cut-offs on the S-Con or related constellations (e.g., Passalacqua et al., 1996; Petot, 2002). Therefore, it warrants a cautious use of these constellations. It is better if these scores are used in conjunction with scores of a set of tests. In fact, these two sets of findings (for and against the constellation’s ability to predict suicide attempts) implicate that the constellation score should be used for the identification of cases and not for ruling out the presence of suicidal ideations. Overall, as suggested by Viglione and Hilsenroth (2001), the S-Con should not be used to rule out
3
suicide risk but to increase awareness about self-destructive behavior and suicide. 2.2. Suicide ideation indicators on other projective tests There are various other projective techniques which, by the virtue of the nature of test, can have significant role in the identification of the presence of suicidal ideations. The Thematic Apperception Test (TAT; Morgan and Murray, 1935), the Rosenzweig Picture Frustration Study (RPFS; Rosenzweig et al., 1947) and the Human Figure Drawing Test (HFDT; Machover, 1949) are such tests. It is disappointing to note that there is an acute dearth of research exploring the potentials of these tests in the identification of the presence of suicidal ideations in an individual. We will briefly discuss here some of the suicide related research on these tests. Morbid contents in an individual’s percepts can be indicative of a mind preoccupied with morbid events and the TAT stories can be a medium to reflect these preoccupations. A few important observations in the TAT protocols of individuals with a history of suicide attempt is worth mentioning. Schneidman (1961) proposed that suicidal individuals have a tendency to think dichotomously (due to their narrow thinking) in terms of either facing their hopeless situations or opting for death. Litinsky and Haslam (1998) tested this hypothesis with the help of the TAT. They administered four cards (out of 1, 2, 3, 5 and 14. Among these 1, 3 and 14 were common for all subjects whereas between 2 and 5 some subjects were administered card 2 and rest card 5. Cards 3 and 14 are considered sensitive in eliciting suicidal responses in individuals with suicidal ideations due to the depicted stimuli. Cards 1, 2 and 5 were included as neutral cards from the perspective of suicide). Findings revealed that suicidal patients had more dichotomous thinking for which the authors used the term ‘complete polarity’ (CP). It was found that these patients reflected two prominent kinds of CP—‘CP gender’ and ‘CP not.’ The ‘CP gender’ concerned polarized choices involving male and female alternatives. The ‘CP not’ indicates any writing in a TAT story which mentions a possibility and add the words ‘or not.’ For example, ‘‘tries his best whether he feels like or not.’’ This study also revealed that suicidal patients write shorter stories than non-suicidal patients. However, some other studies failed to find much difference in the content of the stories of individuals with and without a history of suicide attempt (Broida, 1954; McEvoy, 1961 as cited in Lester, 1970). With very few studies, it is difficult to comment on the research support for the utility of TAT in assessing suicide risk; however, individuals writing stories with self-harm and other morbid contents must be further evaluated for the suicide risk. The Rosenzweig Picture Frustration Study (Rosenzweig et al., 1947) which assesses an individual’s characteristic manner of reacting to a frustrating situation can also be helpful. Suicide is conceptually aggression turned towards self (Seigler and Osmond, 1966), hence, it is expected that people with suicidal ideations will exhibit more intra-aggression on this test. Lester (1970), in a review on the use of psychological tests in predicting suicide, has mentioned that there is some indication that suicidal individuals score less on extrapunitive scale of the Rosenzweig Picture Frustration Study in comparison to non-suicidal individuals; however, there is lack of unequivocal support for this finding. In fact, similar to the TAT this test is also poorly researched and in the context of suicide assessment it is difficult to mention any definitive findings. The Human Figure Drawing Test (HFDT; Machover, 1949) can be a medium for people to express their self-harm wishes. For example, slashes and loops on HFDT can be indicators of the suicidal wishes (Virshup, 1976). Zalsman et al. (2000) in their study of adolescents with a history of suicide attempt or serious
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
4
Table 1 A few important single-sign and constellation based studies on the efficacy of Rorschach in assessing and predicting suicide. Studies
Objectives
Method
Weiner (1961)
Cross-validation of Matrin’s checklist and its association to suicidal tendencies
From the clinical files Rorschach records of 24 psychiatric patients with a history of suicide attempt and 63 non-suicidal patients were interpreted
Blatt and Ritzler (1974)
Kestenbaum and Lynch (1978)
Exner and Wylie (1977)
Rierdan et al. (1978)
Kendra (1979)
Arffa (1982)
Hansell et al. (1988)
Silberg and Armstrong (1992)
Passalacqua et al. (1996)
Major findings
Two signs (C or CF appear first on VIII-X cards and P < 3 with F + % > 60) were found more (p < .05) in the protocols of suicidal group 8 Martin’s signs correctly classified 79% of the suicide attempters From the clinical files Rorschach records of 12 Transparency and cross-sectional responses To test the hypothesis that the transparency and translucency psychiatric patients who had already occur in greater frequency (P < 01) in the responses occur primarily in patients committed suicide and 12 psychiatric patients suicidal group than non-suicidal group with suicidal intent without any history of suicide attempts were interpreted To test the hypothesis that individuals From the clinical files Rorschach records of 10 No between group differences were found in who commit suicide give a greater psychiatric patients who had already transparency, cross-sectional and penetration number of transparency, crosscommitted suicide and 10 psychiatric patients responses sectional and penetration responses without any history of suicide attempts were (‘penetration’ refers to those responses interpreted that represent decay) than matched non-suicidal individuals From a Rorschach protocol pool, the records of A constellation of 11 variables was identified. To determine if any constellation of variables are seen more frequently on the following groups were included: 59 A composite of 8 was found to correctly the Rorschach records of people who individuals who had already committed suicide identify 75% of the effected group and 45% of commit suicide or attempt suicide (within 60 days of test administration), 31 the attempters. Also it was found to occur at a individuals who had attempted suicide (within very low frequency in the control groups 60 days of test administration) and 33 individuals on whom the Rorschach test was administered within five days after having attempted suicide. Apart from these, three control groups were taken: 50 depressed inpatients with no history of a suicide attempt, 50 schizophrenia in-patients without any history of suicide attempts and 50 individuals without any psychiatric history From the clinical files Rorschach records of 14 The Rorschach records of individuals who To cross-validate Blatt and Ritzler (1974) findings that individuals who individuals who committed suicide after their committed suicide had significantly more commit suicide give a greater number hospitalization in a short-term care facility and number of transparency and cross-sectional of transparency and cross-sectional 14 individuals from the same facility who had responses than the individuals from the control responses than matched non-suicidal not attempted suicide, were interpreted group. individuals Protocols were interpreted by Piotrowski’s To develop a Rorschach variables based From the clinical files Rorschach records of scale for suicide prediction three groups of 125 individuals each of suicide method and out of 44 variables, six weighted effectors, suicide attempters and psychiatric variables were able to successfully classify controls without any history of suicide were suicide effectors, suicide attempters and interpreted psychiatric controls To evaluate a multiple sign approach, From the clinical files Rorschach records of 48 A cut off score of 4 was found to be predictive consisting of 10 signs from the Exner inpatients from child and adolescent psychiatry of suicidal behavior. and Wylie (1977) list and two unit (divided into four groups of 12 each of Individual signs did not differ significantly additional signs (the ‘deterioration previous attempters, unit attempters, dual among the four groups. content’ and a specific area from card attempters and the control group) were VII) in evaluating suicidal potential in interpreted. adolescents. To test the validity of using single-signs 41 Depressed inpatients, based on their scores In general, the signs were not indicators of (color-shading, transparency and cross- on the Hamilton Rating Scale for Depression current suicidality sectional responses) on Rorschach as (HRSD) were divided into moderately to Transparency responses were unrelated to indicators of current suicidal risk. severely suicidal, mildly suicidal and doubtfully current suicidal risk, but linked to past suicides, or not suicidal. Participants were administered indicative of a more enduring trait Rorschach and HRSD at the time of admission Those patients who gave these responses and re-tested when they were ‘out-of-episode’ during the episode had the likelihood of giving them in the recovered phase also, thus, suggesting that these signs may indicate a ‘‘suicidal personality’’. To see the usefulness of Rorschach in 26 severely depressed suicidal adolescents, 26 A Rorschach suicidal index for adolescents assessing suicidal potentials among severely depressed non-suicidal adolescents was created by selecting six variables from Shospitalized adolescents and 28 non-suicidal non-depressed adolescent Con of the comprehensive system. inpatients were administered Rorschach. Four of six of the variables could identify 64% of suicidal adolescents Chose a total of 65 signs from location, From the clinical files Rorschach records of 10 An average of 22 signs were found in the determinants, contents, formulae and persons who had committed suicide, 10 protocols of combined suicide group whereas special phenomenon to see how many psychiatric patients with strong selfthe non-suicidal group protocols had only 15 of these signs are frequently seen in the destructive tendencies and 20 patients without signs. Authors suggest that the occurrence of protocols of people who have any suicidal inclinations were interpreted. >20 of the proposed signs should be considered committed suicide, who have strong to be indicative of the danger of suicide. self-destructive tendencies and who do Morbid responses were seen in almost similar not have any suicidal ideations frequencies in both suicide completers and the group with strong suicidal ideations. However, it was nearly absent in the non-suicidal group.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
5
Table 1 (Continued ) Studies
Objectives
Fowler et al. (2001a)
To examine the relation between the Rorschach Comprehensive System’s SCon and lethality of suicide attempts during the course of patients’ hospitalization
Fowler et al. (2001b)
Method
Major findings
Clinical sample:104 patients (non-suicidal = 37, parasuicidal = 37, near-lethal = 30) Non-clinical sample: 50 college students The clinical sample Rorschach protocols were taken from the psychological evaluation files. In the suicidal group only those patients were included who had done suicidal activity within 60 days of the administration of Rorschach. Non clinical sample was administered Rorschach after the inclusion in the study Rorschach protocols of 122 adults inpatients To understand the underlying unconscious processes that place an (non-suicidal = 44, parasuicidal = 41, nearindividual at risk for suicide. lethal = 37) were taken from the psychological To evaluate the effectiveness of evaluation files. Protocols were interpreted for psychoanalytic Rorschach signs and presence of morbid, color-shading, construct a psychoanalytic Rorschach transparency and cross-section responses. The suicide index (Riggs Index) that would Riggs index is the composite of the summation predict near lethal suicide attempts of each instance of identification of the above within 60 days. mentioned signs
A cut-off score of 7 was found to be indicator of near lethal suicide attempts within 60 days.
Patients who later made near lethal suicide attempts had highest number of responses in each category compared to parasuicidal and the non-suicidal groups. In the prediction model color-shading blend emerged in the first step followed by transparency, morbid and cross-section responses. Near lethal patients had three times higher scores on the Riggs index than non-suicidal and parasuicidal patients. Petot (2002) To test the validity of Silberg & Three groups of children—suicidal (N = 14), Color-shading blend score clearly Armstrong’s constellation (Rorschach non-suicidal depressives (N = 15) and anxious differentiated suicidal group from the nonSuicidal Index for Adolescents) in child (N = 14) were administered Rorschach as part of suicidal group. and adolescent sample an assessment on a set of tests. Other indices of the Suicidal Index for Adolescents though did not attain the level of significance in differentiating suicidal and nonsuicidal children, in general the suicidal group had higher score on all the indices. Individuals with suicidal ideation (n = 30), Presence of at least four out of six indicators of Blasczyk-Schiep et al. (2011) To see the effectiveness of Silberg & Armstrong constellation (Rorschach history of parasuicidal behavior (n = 30), history the Suicidal Index for Adolescents could Suicidal Index for Adolescents) in of near-lethal suicidal attempts (n = 26) and identify 69% of people who had a history of assessing suicidal risk among without any history of suicide attempt (n = 30) serious suicide attempt. adolescents and young students were administered Rorschach. The index was found to have incremental validity in predicting suicide.
suicidal ideations found that a few signs on HFDT correlated significantly with severity of suicidal behavior as measured by a scale for the assessment of suicide potentials. These signs were shaky, unstable, feathery lines; shadowy, bearded lines; inconsistencies in body outlines, slash lines in the neck; slash lines on the forearms and overall impression of suicidality. 3. Discussion Individuals having suicidal ideations express them during interviews as well as on the objective psychological tests and the suicide assessment scales (Beck et al., 1985; Hillbrand, 1996; Sepaher et al., 1999; Glassmire et al., 2001). However, both false-positives and false-negatives are possible during objective assessments. From a clinician’s point of view falsepositive in the domain of suicide assessment is not very disturbing but any false-negative case is a matter of grave concern. A false-negative case on objective assessment implicates that the individual has suicidal ideations but does not express them explicitly. In such situations, the use of projective tests becomes more important as on the ambiguous stimuli of these tests, the individual may indicate his/her suicidal ideations. For example, the Rorschach responses are extremely pathological of those children who try to conceal their ideas of suicide (Petot, 1999 as cited in Petot, 2002). Our review of literature suggests that the projective tests have potentials to predict and identify suicidal risks. However, the review also suggests that most of the findings are preliminary in nature with a dearth of studies in this area necessitating further research to reach definitive conclusions. For example, many constellations of Rorschach suicide risk signs have been proposed; however, no constellation has been sufficiently researched to be
accepted unequivocally as a tool for the identification of suicide risks. The S-Con is a glaring example which has been found to have strong potentials to predict suicide at a specific cut-off (Exner and Wylie, 1977); but to the best of our knowledge, there are only two studies (Exner, 1993; Fowler et al., 2001b) examining this claim in a systematic way. Likewise, the Rorschach Suicidal Index for Adolescents has been proposed in one study (Silberg and Armstrong, 1992) and, till date, has been examined for its validity in only one more study (Blasczyk-Schiep et al., 2011). The paucity of research holds true for the single signs of Rorschach as well. Various signs have been proposed; however, the number of studies
Table 2 The single signs and constellations commonly used for the identification of presence of suicidal ideations. Single-signs 1. 2. 3. 4.
Whirling responses Transparency responses Color-shading responses Responses with morbid contents
Comprehensive system’s S-Con
Suicidal index for adolescents
1. FV + VF + V + FD > 2 2. Color shading blend >0 3. Egocentricity index .44 4. Mor >3 5. Zd >3.5, Zd EA 7. CF + C > FC 8. X + % < .70 9. S > 3 10. P 8 11. Pur H < 2 12. R < 17
1. CF + C > FC 2. FV + VF + V >0 3. Color shading blend >1 4. Weighted special scores > 9 5. M–>1 6. Mor > 0
Cut-off I7
Cut-off I4
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
6
examining any specific sign is too few to draw any definite conclusions. Blatt and Ritzler’s (1974) hypothesis that the transparency and cross-sectional responses are seen with greater frequency in the Rorschach protocols of people who commit suicide has been tested systematically in merely few studies (e.g., Kestenbaum and Lynch, 1978; Rierdan et al., 1978) and the findings are equivocal. Yet, the literature review suggests that a few Rorschach findings are relatively robust. For example, the presence of color-shading responses in the protocols of individuals having serious suicidal ideations has been reported in many studies (see Table 1). Therefore, the presence of color-shading responses on a protocol should be taken seriously. Likewise, even if there are very few studies on the S-Con, its reported predictive ability for future suicide attempts warrants that one must cautiously look for the presence of the S-Con variables in a protocol (Table 2). The research scenario is more disappointing when it comes to other projective tests. For instance, the potentials of TAT to assess suicidal ideations has been recognized (Broida, 1954); however, barring a few studies there is a near absence of research activity in this domain. Same holds true for other promising projective techniques such as the RPFS and the HFDT. Having said that, one should also admit the practical difficulties in suicide research. For example, there cannot be a longitudinal prospective study in which psychological assessment is done and the individuals are followed passively to see how many of them with the presence of the indicators of suicidal ideations actually attempt suicide (Hansell et al., 1988). Likewise, studies on individuals with past history of suicide attempts have the difficulty that the tests may not catch those dynamic psychological states that were present when an individual attempted suicide (Lester, 1970). Perhaps this is the reason that most of the studies use the data that were collected as a routine practice (for example, routine evaluation of all the admitted patients) and later if some of those individuals attempted suicide, comparisons of their test protocols were done with those who did not attempt suicide. In general, the methodological limitations of studies related to the use of projective tests in assessing suicide risks underscore the need for a cautious approach towards acceptance of the findings. For example, studies with small samples will have low power and, therefore, the inferences based on the findings of such studies need caution. Likewise, a host of other confounding variables can affect the findings. To cite one example, Meyer (1993) found that the length of Rorschach records can affect the external validity of many of the ‘Comprehensive System’ (Exner and Wylie, 1977) constellations, including the S-Con. Also, most of the studies are from the west and, therefore, caution should be used in applying these findings on individuals from other regional and cultural backgrounds. In all likelihood, the Rorschach and other projective test variables implicated in suicide risk assessment may not really be affected by the cultural background of the individual, yet, it is desirable that in future research these factors are also investigated. In conclusion, it can be said that in view of the scant research in this area, a clinician should not take the presence of a sign (or a set of signs) on a projective test as a full proof indicator of the presence of suicidal ideations; however, it must be considered as an indicator for further explorations for suicide risk. Given the serious implications of the presence of suicidal ideations in an individual, the preventive efforts demand a wider perspective. References Appelbaum, S.A., Colson, D.B., 1968. A reexamination of the color-shading Rorschach test response and suicide attempts. J. Proj. Tech. Personal. Assess. 32, 160–164.
Appelbaum, S.A., Holzman, P., 1962. The color-shading response and suicide. J. Proj. Tech. Personal. Assess. 26, 155–161. Arffa, S., 1982. Predicting adolescent suicidal behavior and the order of Rorschach measurement. J. Personal. Assess. 46, 563–568. Beck, A.T., Steer, R.A., Kovacks, M., Garrison, B., 1985. Hopelessness and eventual suicide: a 10 year prospective study of patients hospitalized with suicidal ideation. Am. J. Psychiatry 142, 559–563. Blasczyk-Schiep, S., Kazeen, M., Kuhl, J., Grygielski, M., 2011. Appraisal of suicidal risk among adolescents and young adults through the Rorschach test. J. Personal. Assess. 93, 518–526. Blatt, S.J., Ritzler, D.A., 1974. Suicide and the representation of transparency and cross-sections on the Rorschach. J. Consult. Clin. Psychol. 42, 280–287. Broida, D.C., 1954. An investigation of certain psychodiagnostic indications of suicidal tendencies and depression in mental hospital patients. Psychiatr. Q. 28, 453–464. Colson, D.B., Hurwitz, B.A., 1973. A new experimental approach to the relationship between color-shading and suicide attempts. J. Personal. Assess. 37, 237–241. Datson, P.G., Sakheim, G.A., 1960. Prediction of successful suicide from the Rorschach test using a sign approach. J. Proj. Tech. 24, 355–361. Exner, J.E., 1993. The Rorschach: A Comprehensive System: Vol. 1, 3rd ed. John Wiley and Sons, New York. Exner, J.E., Wylie, J., 1977. Some Rorschach data concerning suicide. J. Personal. Assess. 41, 339–348. Fowler, J.C., Hilsenroth, M.J., Piers, C., 2001. An empirical study of seriously disturbed suicidal patients. J Am Psychoanal Assoc 49, 16–186. Fowler, J.C., Piers, C., Hilsenroth, M.J., Holdwick, D.J., Padawer, J.R., 2001. The Rorschach suicide constellation: Assessing various degrees of lethality. J. Personal. Assess. 76, 333–351. Glassmire, D.M., Stolberg, R.A., Greene, R.L., Bongar, B., 2001. The utility of MMPI-2 suicide items for assessing suicidal potential: development of a Suicidal Potential Scale. Assessment 8, 281–290. Hansell, A.G., Lerner, H.D., Milden, R.S., Ludolph, P.S., 1988. Single-sign Rorschach suicide indicators: a validity study using a depressed inpatient population. J. Personal. Assess. 52, 658–669. Hillbrand, M., 1996. Validity of two measures of suicidal risk—MCMI II Selfdestructive potential noteworthy items vs. the suicidal risk scale. J. Offender Rehabil. 23, 1–10. Kendra, J.M., 1979. Predicting suicide using the Rorschach inkblot test. J. Personal. Assess. 43, 452–456. Kestenbaum, J.M., Lynch, D., 1978. Rorschach suicide predictors: a cross validational study. J. Clin. Psychol. 44, 921–928. Lester, D., 1970. Attempts to predict suicidal risk using psychological tests. Psychol. Bull. 74, 1–17. Lester, D., Decker, J., Eisenberg, R.J., Ecker, C., Guerriero, J., Mielish, C., 1990. Association between whirling responses on psychological tests and suicidal preoccupation. Percept. Mot. Skills 71, 1105. Linehan, M.M., Goodstein, J.L., Nielsen, S.L., Chiles, J.K., 1983. Reasons for staying alive when you are thinking of killing yourself: the reasons for Living Inventory. J. Consult. Clin. Psychol. 51, 276–286. Litinsky, A.M., Haslam, N., 1998. Dichotomous thinking as a sign of suicide risk on the TAT. J. Personal. Assess. 71, 368–378. Machover, K., 1949. Personality Projection in the Drawing of the Human Figure: A Method of Personality Investigation. American Lectures in Psychology. Vol. 25. Charles C Thomas Publisher, Springfield, pp. 3–32. Meyer, G.H., 1993. The impact of response frequency on the Rorschach constellation indices and on their validity with diagnostic and MMPI-2 criteria. J. Personal. Assess. 60, 153–180. Morgan, C., Murray, H.A., 1935. A method for investigating fantasies: the thematic apperception test. Arch. Neurol. Psychiatry 34, 289–306. National Crime Record Bureau, 2012. Suicides in India. Downloaded from hncrb.nic.in/CD-ADSI-2012/suicides-11.pdfi on April 27, 2014. Passalacqua, A.M., Echenique, S., Herrara, M.S., Orcoyen, D., 1996. Beyond subjects’ response: detecting suicidal inclinations in the Rorschach inkblot technique. Rorschachiana 21, 75–90. Petot, D., 2002. Rorschach characteristics of depressive children with suicidal ideation. Rorschachiana 25, 165–174. Rierdan, J., Lang, E., Eddy, S., 1978. Suicide and transparency responses on the Rorschach: a replication. J. Consult. Clin. Psychol. 46, 1162–1163. Rosenzweig, S., Fleming, E., Clarke, H.J., 1947. Revised scoring manual for the Rosenzweig Picture-Frustration Study. J. Psychol. 24, 165–205. Schneidman, E.S., 1961. Psycho-logic: a personality approach to patterns of thinking. In: Kagan, J., Laser, G. (Eds.), Contemporary Issues in Thematic Apperceptive Methods. Thomas, Springfield, Illinois, pp. 153–190. Seigler, M., Osmond, H., 1966. Models of madness. Br. J. Psychiatry 112, 1193–1203. Sepaher, I., Bongar, B., Greene, R.L., 1999. Codetype base rates for the I mean business suicide items on the MMPI-2. J. Clin. Psychol. 55, 1167–1173. Silberg, J.L., Armstrong, J.G., 1992. The Rorschach test for predicting suicide among depressed adolescent inpatients. J. Personal. Assess. 59, 290–303. Sullivan, G.R., Bongar, B., 2009. Assessing suicide risk in the adult patient. In: Kleespies, P.M. (Ed.), Behavioral Emergencies: An Evidence-Based Resource for Evaluating and Managing Risk of Suicide, Violence, and Victimization. American Psychological Association, Washington, pp. 59–78. Thomas, C.B., Duzynski, K.R., 1985. Are the words on Rorschach predictors of disease and death? Psychosom. Med. 47, 201–211.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx Viglione, D.J., Hilsenroth, M.J., 2001. The Rorschach: facts, fictions and future. Psychol. Assess. 13, 452–471. Virshup, E., 1976. On graphic suicide plans. Arts Psychother. 3, 17–22. Weiner, I.B., 1961. Cross-validation of a Rorschach checklist associated with suicidal tendencies. J. Consult. Psychol. 4, 312–315. White, M.A., Schreiber, H., 1952. Diagnosing suicidal risks on the Rorschach. Psychiatr. Q. (Suppl.) 26, 161–189.
7
World Health Organization, 1999. Figures and facts about suicide. World Health Organization, Geneva. Zalsman, G., Netael, R., Fischel, T., Freudenstein, O., Landau, E., Orbach, I., Weizman, A., Pfeffer, C.R., Apter, A., 2000. Human figure drawing in the evaluation of severs adolescent suicidal behaviour. J. Am. Acad. Child Adolesc. Psychiatry 39, 1024–1031.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
AJP-631; No. of Pages 7 Asian Journal of Psychiatry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Asian Journal of Psychiatry journal homepage: www.elsevier.com/locate/ajp
Review
Identification of suicidal ideations with the help of projective tests: A review Devvarta Kumar a,*, S. Haque Nizamie b, Priyadarshee Abhishek b, Lavanya Tumkur Prasanna a a b
Department of Clinical Psychology, National Institute of Mental Health and Neuro Sciences, Bangalore 560029, India Central Institute of Psychiatry, Ranchi 834006, India
A R T I C L E I N F O
A B S T R A C T
Article history: Received 2 May 2014 Received in revised form 24 July 2014 Accepted 27 July 2014 Available online xxx
Identification of the presence of suicidal ideations in an individual is crucial for the timely intervention. However, these ideations may remain unidentified as an individual with serious intentions of self-harm may not express them explicitly. Various projective tests can provide crucial clues to clinicians about the presence of suicidal ideations in an individual’s mind. The present review is intended to update clinicians working in the field of suicide prevention about salient findings on these tests which can serve as a ready reckoner for them. We also highlight the status of research in this domain. ß 2014 Published by Elsevier B.V.
Keywords: Projective test Suicide assessment S-Con
Contents 1. 2.
3.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rationale for the use of projective tests in the identification Rorschach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. The single-sign approach. . . . . . . . . . . . . . . . 2.1.1. 2.1.2. The constellation approach . . . . . . . . . . . . . . Suicide ideation indicators on other projective tests . 2.2. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.. of .. .. .. .. .. ..
............... suicidal ideations ............... ............... ............... ............... ............... ...............
1. Introduction Prevention of suicide is a daunting but an indispensable task for any society. According to the World Health Organization (WHO), suicide claims one life every 40 s (WHO, 1999). In India, a total of 1,35,445 people died due to suicide in 2012 alone (NCRB; National Crime Record Bureau, 2012). The data increases manifold when it comes to counting people with failed attempted suicides (WHO, 1999). A timely identification of the presence of suicidal ideations can save numerous lives. Unfortunately, suicidal ideations are
* Corresponding author. Tel.: +91 80 26995188. E-mail address: [email protected] (D. Kumar).
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
000 000 000 000 000 000 000 000
often not expressed explicitly and the subtle signs of their presence may come to notice only after an individual’s suicide attempt (e.g., realization that the individual was giving away his valuables before attempting suicide). It becomes more important in case of vulnerable individuals (e.g., persons with depression) who may have lower threshold to attempt suicide and the occurrence of suicidal ideations can lead to fatal steps. Therefore, it is imperative that clinicians are able to identify people who have suicidal ideations but may not express them explicitly. A few projective psychological tests, by virtue of the capability to detect traits and states related to suicidal behavior (e.g., impulsivity, sense of hopelessness), can be of help in identification of the presence of suicidal ideations in an individual’s mind. The present narrative review is intended to update clinicians about the
http://dx.doi.org/10.1016/j.ajp.2014.07.004 1876-2018/ß 2014 Published by Elsevier B.V.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 2
D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
indicators of suicidal ideations on some commonly used projective tests. We compile the existing data by means of a narrative review rather than a meta-analysis for two reasons. First, the studies are too less and the sample sizes of most of the studies are very small. Therefore, conducting a meta-analysis may result in more research artifacts than facts. Second, the variables studied in different studies are quite varied and, thus, it is difficult to merge data of these studies to conduct a meta-analysis. 2. Rationale for the use of projective tests in the identification of suicidal ideations The use of projective tests for the purpose of identification of the presence of suicidal ideations is based on the assumption that a preoccupation with self-harm intentions may reflect on the ambiguous materials of these tests. Also, the individual may use the testing situation to communicate suicidal intents (Sullivan and Bongar, 2009). We discuss here a few projective tests in the context of suicide assessment. 2.1. Rorschach For more than 50 years, the Rorschach test has been researched to explore its potential value as a tool to assess suicide risk (e.g., White and Schreiber, 1952; Blatt and Ritzler, 1974; Exner and Wylie, 1977; Blasczyk-Schiep et al., 2011). The Rorschach researchers have followed two approaches in suicide assessment—the single-sign approach and the constellation approach. 2.1.1. The single-sign approach This approach refers to the use of a specific Rorschach sign to predict suicide. It has the advantage of quick identification of suicidality on the basis of an easily identifiable Rorschach variable (Hansell et al., 1988). For instance, White and Schreiber (1952) proposed ‘whirling’ as one of the responses on Rorschach to be indicative of a mind preoccupied with death wishes. Subsequently, Thomas and Duzynski (1985) reviewed Rorschach responses of more than 1000 medical students and found that whirling and its synonyms are found more frequently in the protocols of those individuals who later commit suicide. The association of the word ‘whirling’ (and its synonyms) with suicidal ideations has been confirmed by another research in which individuals with prior suicidal ideations chose more number of spinning phrases than those who did not exhibit suicidal ideations on a different test (Lester et al., 1990). Likewise, research has shown that people having suicidal ideations frequently give transparency responses (Blatt and Ritzler, 1974; Rierdan et al., 1978; Kestenbaum and Lynch, 1978). Transparency responses mean a perception in which the individual sees an object in its transparent state (for example, ‘‘a jelly fish and the inner bones can be seen’’). Different explanations have been given for the presence of these responses in the protocols of individuals with suicidal ideations. The transparency responses are construed to be indicative of the merging of self into hated object or loss of the sense of self leading to feeling of loss or helplessness (Blatt and Ritzler, 1974). On the other hand, Rierdan et al. (1978) proposed that the transparency responses indicate the perception of death as a transitional phase or continuity and not an end. The color-shading responses have also been considered a sensitive indicator of suicidality. The color-shading blend is considered indicator of the failure to cope effectively with painful affect (Fowler et al., 2001a). Also it could be reflective of a mixed and confused emotional experience having both pleasure and pain components (Exner and Wylie, 1977). Appelbaum and Holzman (1962) found a relationship between the presence of the color-
shading responses and a history of attempted suicide. Later research supported the importance of the color-shading responses in the identification of suicidal ideations (Appelbaum and Colson, 1968; Colson and Hurwitz, 1973). Petot (2002) found that the color-shading responses are seen more frequently in the Rorschach protocols of children with suicidal ideations. The importance of color-shading responses in the identification of the presence of suicidal ideations can be understood by the fact that it has been included in most of the suicide constellations (which will be discussed in the next section). Apart from these, a few other signs such as the poor form level human movement responses and responses with morbid contents are also seen in the Rorschach protocols of individuals with suicidal ideations and have been included in suicide constellations (Petot, 2002; Silberg and Armstrong, 1992). However, the absence of a sign in an individual’s Rorschach protocol should not be taken as absence of suicidal ideations as some people may lack the ability to communicate the cognitively complex responses (such as transparency and cross-sectional responses) (Blatt and Ritzler, 1974). Also, it is not necessary that the presence of a sign is always related to current suicidality. For example, Hansell et al. (1988) found that the transparency responses were unrelated to current suicidal risks. 2.1.2. The constellation approach The instability and the lack of statistical power of single signs limit their use in predicting suicide (Kendra, 1979; Fowler et al., 2001b). The chances of both false-positives and false-negatives are high if an impression regarding the vulnerability of an individual to attempt suicide is formed on the basis of a single indicator. Therefore, attempts have been made to identify a set of Rorschach signs that can be used as a constellation to predict suicide. One of the earliest proposed constellations is the Martin’s checklist which consists of 17 Rorschach signs (Martin, 1951 as cited in Weiner, 1961). Limited research applying the Martin’s checklist to the Rorschach protocols of suicide completers, attempters and nonsuicidal groups have found the efficacy of this checklist in discriminating the suicidal and non-suicidal groups but not the suicide attempters and completers (Datson and Sakheim, 1960; Weiner, 1961). Kendra (1979) interpreted the Rorschach protocols by the Piotrowski method and found that six variables (FM, Fc’, C, F%, H% and W:M) could successfully classify suicide attempters, suicide effectors and psychiatric controls. In 1977, Exner and Wylie proposed a suicide constellation based on the Comprehensive System (CS) which is known as the ‘S-Con’ (Exner and Wylie, 1977). Research has highlighted some remarkable aspects of the S-Con. It consists of 12 Rorschach variables (initially 11 variables were identified and the 12th variable was added later) and a cut-off of eight is considered sensitive in detecting people with significant suicidal ideations (see Table 2). The cut-off of this constellation is based on interesting empirical findings. Exner and Wylie (1977) compared Rorschach protocols of patients who had later committed suicide with the protocols of non-suicidal individuals. When a cut-off of eight was used, the S-Con was able to identify around 75% of suicidal patients and 100% of non-patients. This finding was confirmed in a cross-validation study (Exner, 1993). Important to note that the suicide completers, identified by the S-Con, had committed suicide within 60 days of the test administration. Therefore, it is suggested that the protocol of any person having a score of eight or more on the S-Con must be taken very seriously. Moreover, it should be kept in mind that approximately 25% of people who commit suicide do not have a score of eight on the SCon and, hence, even the lower score should be taken as an indication for further careful assessment.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
Fowler et al. (2001b) studied three groups—patients with a history of lethal suicide attempt, patients with a history of parasuicidal attempt and college students without any significant history of suicide attempts. They found that among the nine psychiatric and demographic variables, an S-Con score of seven or more was the sole predictor of near-lethal suicide attempts within 60 days. The authors concluded that a strong prediction of near lethal suicide in the absence of prediction of the para-suicidal behavior indicates the discriminant validity of the S-Con. Furthermore, the inability of other CS variables and measures of psychopathology to predict near-lethal suicide attempts suggests that the S-Con score is associated with the behavior of suicide attempt itself and not with a general tendency of impulsivity or self-destructive behaviors. Also it is important to mention here that a decrease in one score from the cut-off suggested by Exner and Wylie (1977) indicates the need for flexibility in considering the S-Con score. The S-Con (with some modifications) has been used for the assessment of suicidal ideations in adolescent population also. Arffa (1982) compared four groups of adolescents on a Rorschach signs constellation. The constellation had 10 signs from the S-Con and two additional signs. The four groups of adolescents were previous suicide attempters (attempted suicide within 10 months prior to hospitalization), unit suicide attempters (suicidal action deemed by clinical staff to be serious enough to warrant one-toone observation for longer than three hours), dual suicide attempters (presence of two or more suicide attempts) and the control group (adolescents without any history of suicide attempts). A score of four or more was found to differentiate suicidal from non-suicidal group with 92% accuracy. Interestingly, none of the signs individually discriminated the group, thus, questioning the validity of the single-sign approach. Silberg and Armstrong (1992) developed a new Rorschach Suicidal Index for Adolescents (RSIA) consisting of six variables (see Table 2). They found that the presence of four out of six RSIA variables could correctly identify 64% of suicidal depressed adolescents. On the basis of commonly observed features on the Rorschach protocols of suicidal adolescents the authors concluded that these adolescents are introspective, exhibit painful arousal, have preoccupation with things of morbid nature, are impulsive, have misperceptions of people and show distorted reasoning skills. In a recent study, the incremental validity and the effectiveness of RSIA in suicidal adolescents were investigated (Blasczyk-Schiep et al., 2011). The findings of this study were in consonance with Silberg and Armstrong’s (1992) study. The presence of, at least, four indicators could correctly identify 69% adolescents who had serious suicide attempts in past. The index showed incremental validity in predicting suicide. Also, the scores correlated with the objective measures of suicide risk such as the score on the Reasons for Living Inventory (RLI; Linehan et al., 1983) and the number of suicide attempts. Though various studies have supported the efficacy of the constellation approach in predicting suicide (Exner, 1993; Fowler et al., 2001b), there are studies that have challenged its worth. For example, a few studies indicate that the Rorschach protocols of near-lethal suicide attempters and suicide completers may not cross the cut-offs on the S-Con or related constellations (e.g., Passalacqua et al., 1996; Petot, 2002). Therefore, it warrants a cautious use of these constellations. It is better if these scores are used in conjunction with scores of a set of tests. In fact, these two sets of findings (for and against the constellation’s ability to predict suicide attempts) implicate that the constellation score should be used for the identification of cases and not for ruling out the presence of suicidal ideations. Overall, as suggested by Viglione and Hilsenroth (2001), the S-Con should not be used to rule out
3
suicide risk but to increase awareness about self-destructive behavior and suicide. 2.2. Suicide ideation indicators on other projective tests There are various other projective techniques which, by the virtue of the nature of test, can have significant role in the identification of the presence of suicidal ideations. The Thematic Apperception Test (TAT; Morgan and Murray, 1935), the Rosenzweig Picture Frustration Study (RPFS; Rosenzweig et al., 1947) and the Human Figure Drawing Test (HFDT; Machover, 1949) are such tests. It is disappointing to note that there is an acute dearth of research exploring the potentials of these tests in the identification of the presence of suicidal ideations in an individual. We will briefly discuss here some of the suicide related research on these tests. Morbid contents in an individual’s percepts can be indicative of a mind preoccupied with morbid events and the TAT stories can be a medium to reflect these preoccupations. A few important observations in the TAT protocols of individuals with a history of suicide attempt is worth mentioning. Schneidman (1961) proposed that suicidal individuals have a tendency to think dichotomously (due to their narrow thinking) in terms of either facing their hopeless situations or opting for death. Litinsky and Haslam (1998) tested this hypothesis with the help of the TAT. They administered four cards (out of 1, 2, 3, 5 and 14. Among these 1, 3 and 14 were common for all subjects whereas between 2 and 5 some subjects were administered card 2 and rest card 5. Cards 3 and 14 are considered sensitive in eliciting suicidal responses in individuals with suicidal ideations due to the depicted stimuli. Cards 1, 2 and 5 were included as neutral cards from the perspective of suicide). Findings revealed that suicidal patients had more dichotomous thinking for which the authors used the term ‘complete polarity’ (CP). It was found that these patients reflected two prominent kinds of CP—‘CP gender’ and ‘CP not.’ The ‘CP gender’ concerned polarized choices involving male and female alternatives. The ‘CP not’ indicates any writing in a TAT story which mentions a possibility and add the words ‘or not.’ For example, ‘‘tries his best whether he feels like or not.’’ This study also revealed that suicidal patients write shorter stories than non-suicidal patients. However, some other studies failed to find much difference in the content of the stories of individuals with and without a history of suicide attempt (Broida, 1954; McEvoy, 1961 as cited in Lester, 1970). With very few studies, it is difficult to comment on the research support for the utility of TAT in assessing suicide risk; however, individuals writing stories with self-harm and other morbid contents must be further evaluated for the suicide risk. The Rosenzweig Picture Frustration Study (Rosenzweig et al., 1947) which assesses an individual’s characteristic manner of reacting to a frustrating situation can also be helpful. Suicide is conceptually aggression turned towards self (Seigler and Osmond, 1966), hence, it is expected that people with suicidal ideations will exhibit more intra-aggression on this test. Lester (1970), in a review on the use of psychological tests in predicting suicide, has mentioned that there is some indication that suicidal individuals score less on extrapunitive scale of the Rosenzweig Picture Frustration Study in comparison to non-suicidal individuals; however, there is lack of unequivocal support for this finding. In fact, similar to the TAT this test is also poorly researched and in the context of suicide assessment it is difficult to mention any definitive findings. The Human Figure Drawing Test (HFDT; Machover, 1949) can be a medium for people to express their self-harm wishes. For example, slashes and loops on HFDT can be indicators of the suicidal wishes (Virshup, 1976). Zalsman et al. (2000) in their study of adolescents with a history of suicide attempt or serious
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
4
Table 1 A few important single-sign and constellation based studies on the efficacy of Rorschach in assessing and predicting suicide. Studies
Objectives
Method
Weiner (1961)
Cross-validation of Matrin’s checklist and its association to suicidal tendencies
From the clinical files Rorschach records of 24 psychiatric patients with a history of suicide attempt and 63 non-suicidal patients were interpreted
Blatt and Ritzler (1974)
Kestenbaum and Lynch (1978)
Exner and Wylie (1977)
Rierdan et al. (1978)
Kendra (1979)
Arffa (1982)
Hansell et al. (1988)
Silberg and Armstrong (1992)
Passalacqua et al. (1996)
Major findings
Two signs (C or CF appear first on VIII-X cards and P < 3 with F + % > 60) were found more (p < .05) in the protocols of suicidal group 8 Martin’s signs correctly classified 79% of the suicide attempters From the clinical files Rorschach records of 12 Transparency and cross-sectional responses To test the hypothesis that the transparency and translucency psychiatric patients who had already occur in greater frequency (P < 01) in the responses occur primarily in patients committed suicide and 12 psychiatric patients suicidal group than non-suicidal group with suicidal intent without any history of suicide attempts were interpreted To test the hypothesis that individuals From the clinical files Rorschach records of 10 No between group differences were found in who commit suicide give a greater psychiatric patients who had already transparency, cross-sectional and penetration number of transparency, crosscommitted suicide and 10 psychiatric patients responses sectional and penetration responses without any history of suicide attempts were (‘penetration’ refers to those responses interpreted that represent decay) than matched non-suicidal individuals From a Rorschach protocol pool, the records of A constellation of 11 variables was identified. To determine if any constellation of variables are seen more frequently on the following groups were included: 59 A composite of 8 was found to correctly the Rorschach records of people who individuals who had already committed suicide identify 75% of the effected group and 45% of commit suicide or attempt suicide (within 60 days of test administration), 31 the attempters. Also it was found to occur at a individuals who had attempted suicide (within very low frequency in the control groups 60 days of test administration) and 33 individuals on whom the Rorschach test was administered within five days after having attempted suicide. Apart from these, three control groups were taken: 50 depressed inpatients with no history of a suicide attempt, 50 schizophrenia in-patients without any history of suicide attempts and 50 individuals without any psychiatric history From the clinical files Rorschach records of 14 The Rorschach records of individuals who To cross-validate Blatt and Ritzler (1974) findings that individuals who individuals who committed suicide after their committed suicide had significantly more commit suicide give a greater number hospitalization in a short-term care facility and number of transparency and cross-sectional of transparency and cross-sectional 14 individuals from the same facility who had responses than the individuals from the control responses than matched non-suicidal not attempted suicide, were interpreted group. individuals Protocols were interpreted by Piotrowski’s To develop a Rorschach variables based From the clinical files Rorschach records of scale for suicide prediction three groups of 125 individuals each of suicide method and out of 44 variables, six weighted effectors, suicide attempters and psychiatric variables were able to successfully classify controls without any history of suicide were suicide effectors, suicide attempters and interpreted psychiatric controls To evaluate a multiple sign approach, From the clinical files Rorschach records of 48 A cut off score of 4 was found to be predictive consisting of 10 signs from the Exner inpatients from child and adolescent psychiatry of suicidal behavior. and Wylie (1977) list and two unit (divided into four groups of 12 each of Individual signs did not differ significantly additional signs (the ‘deterioration previous attempters, unit attempters, dual among the four groups. content’ and a specific area from card attempters and the control group) were VII) in evaluating suicidal potential in interpreted. adolescents. To test the validity of using single-signs 41 Depressed inpatients, based on their scores In general, the signs were not indicators of (color-shading, transparency and cross- on the Hamilton Rating Scale for Depression current suicidality sectional responses) on Rorschach as (HRSD) were divided into moderately to Transparency responses were unrelated to indicators of current suicidal risk. severely suicidal, mildly suicidal and doubtfully current suicidal risk, but linked to past suicides, or not suicidal. Participants were administered indicative of a more enduring trait Rorschach and HRSD at the time of admission Those patients who gave these responses and re-tested when they were ‘out-of-episode’ during the episode had the likelihood of giving them in the recovered phase also, thus, suggesting that these signs may indicate a ‘‘suicidal personality’’. To see the usefulness of Rorschach in 26 severely depressed suicidal adolescents, 26 A Rorschach suicidal index for adolescents assessing suicidal potentials among severely depressed non-suicidal adolescents was created by selecting six variables from Shospitalized adolescents and 28 non-suicidal non-depressed adolescent Con of the comprehensive system. inpatients were administered Rorschach. Four of six of the variables could identify 64% of suicidal adolescents Chose a total of 65 signs from location, From the clinical files Rorschach records of 10 An average of 22 signs were found in the determinants, contents, formulae and persons who had committed suicide, 10 protocols of combined suicide group whereas special phenomenon to see how many psychiatric patients with strong selfthe non-suicidal group protocols had only 15 of these signs are frequently seen in the destructive tendencies and 20 patients without signs. Authors suggest that the occurrence of protocols of people who have any suicidal inclinations were interpreted. >20 of the proposed signs should be considered committed suicide, who have strong to be indicative of the danger of suicide. self-destructive tendencies and who do Morbid responses were seen in almost similar not have any suicidal ideations frequencies in both suicide completers and the group with strong suicidal ideations. However, it was nearly absent in the non-suicidal group.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
5
Table 1 (Continued ) Studies
Objectives
Fowler et al. (2001a)
To examine the relation between the Rorschach Comprehensive System’s SCon and lethality of suicide attempts during the course of patients’ hospitalization
Fowler et al. (2001b)
Method
Major findings
Clinical sample:104 patients (non-suicidal = 37, parasuicidal = 37, near-lethal = 30) Non-clinical sample: 50 college students The clinical sample Rorschach protocols were taken from the psychological evaluation files. In the suicidal group only those patients were included who had done suicidal activity within 60 days of the administration of Rorschach. Non clinical sample was administered Rorschach after the inclusion in the study Rorschach protocols of 122 adults inpatients To understand the underlying unconscious processes that place an (non-suicidal = 44, parasuicidal = 41, nearindividual at risk for suicide. lethal = 37) were taken from the psychological To evaluate the effectiveness of evaluation files. Protocols were interpreted for psychoanalytic Rorschach signs and presence of morbid, color-shading, construct a psychoanalytic Rorschach transparency and cross-section responses. The suicide index (Riggs Index) that would Riggs index is the composite of the summation predict near lethal suicide attempts of each instance of identification of the above within 60 days. mentioned signs
A cut-off score of 7 was found to be indicator of near lethal suicide attempts within 60 days.
Patients who later made near lethal suicide attempts had highest number of responses in each category compared to parasuicidal and the non-suicidal groups. In the prediction model color-shading blend emerged in the first step followed by transparency, morbid and cross-section responses. Near lethal patients had three times higher scores on the Riggs index than non-suicidal and parasuicidal patients. Petot (2002) To test the validity of Silberg & Three groups of children—suicidal (N = 14), Color-shading blend score clearly Armstrong’s constellation (Rorschach non-suicidal depressives (N = 15) and anxious differentiated suicidal group from the nonSuicidal Index for Adolescents) in child (N = 14) were administered Rorschach as part of suicidal group. and adolescent sample an assessment on a set of tests. Other indices of the Suicidal Index for Adolescents though did not attain the level of significance in differentiating suicidal and nonsuicidal children, in general the suicidal group had higher score on all the indices. Individuals with suicidal ideation (n = 30), Presence of at least four out of six indicators of Blasczyk-Schiep et al. (2011) To see the effectiveness of Silberg & Armstrong constellation (Rorschach history of parasuicidal behavior (n = 30), history the Suicidal Index for Adolescents could Suicidal Index for Adolescents) in of near-lethal suicidal attempts (n = 26) and identify 69% of people who had a history of assessing suicidal risk among without any history of suicide attempt (n = 30) serious suicide attempt. adolescents and young students were administered Rorschach. The index was found to have incremental validity in predicting suicide.
suicidal ideations found that a few signs on HFDT correlated significantly with severity of suicidal behavior as measured by a scale for the assessment of suicide potentials. These signs were shaky, unstable, feathery lines; shadowy, bearded lines; inconsistencies in body outlines, slash lines in the neck; slash lines on the forearms and overall impression of suicidality. 3. Discussion Individuals having suicidal ideations express them during interviews as well as on the objective psychological tests and the suicide assessment scales (Beck et al., 1985; Hillbrand, 1996; Sepaher et al., 1999; Glassmire et al., 2001). However, both false-positives and false-negatives are possible during objective assessments. From a clinician’s point of view falsepositive in the domain of suicide assessment is not very disturbing but any false-negative case is a matter of grave concern. A false-negative case on objective assessment implicates that the individual has suicidal ideations but does not express them explicitly. In such situations, the use of projective tests becomes more important as on the ambiguous stimuli of these tests, the individual may indicate his/her suicidal ideations. For example, the Rorschach responses are extremely pathological of those children who try to conceal their ideas of suicide (Petot, 1999 as cited in Petot, 2002). Our review of literature suggests that the projective tests have potentials to predict and identify suicidal risks. However, the review also suggests that most of the findings are preliminary in nature with a dearth of studies in this area necessitating further research to reach definitive conclusions. For example, many constellations of Rorschach suicide risk signs have been proposed; however, no constellation has been sufficiently researched to be
accepted unequivocally as a tool for the identification of suicide risks. The S-Con is a glaring example which has been found to have strong potentials to predict suicide at a specific cut-off (Exner and Wylie, 1977); but to the best of our knowledge, there are only two studies (Exner, 1993; Fowler et al., 2001b) examining this claim in a systematic way. Likewise, the Rorschach Suicidal Index for Adolescents has been proposed in one study (Silberg and Armstrong, 1992) and, till date, has been examined for its validity in only one more study (Blasczyk-Schiep et al., 2011). The paucity of research holds true for the single signs of Rorschach as well. Various signs have been proposed; however, the number of studies
Table 2 The single signs and constellations commonly used for the identification of presence of suicidal ideations. Single-signs 1. 2. 3. 4.
Whirling responses Transparency responses Color-shading responses Responses with morbid contents
Comprehensive system’s S-Con
Suicidal index for adolescents
1. FV + VF + V + FD > 2 2. Color shading blend >0 3. Egocentricity index .44 4. Mor >3 5. Zd >3.5, Zd EA 7. CF + C > FC 8. X + % < .70 9. S > 3 10. P 8 11. Pur H < 2 12. R < 17
1. CF + C > FC 2. FV + VF + V >0 3. Color shading blend >1 4. Weighted special scores > 9 5. M–>1 6. Mor > 0
Cut-off I7
Cut-off I4
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx
6
examining any specific sign is too few to draw any definite conclusions. Blatt and Ritzler’s (1974) hypothesis that the transparency and cross-sectional responses are seen with greater frequency in the Rorschach protocols of people who commit suicide has been tested systematically in merely few studies (e.g., Kestenbaum and Lynch, 1978; Rierdan et al., 1978) and the findings are equivocal. Yet, the literature review suggests that a few Rorschach findings are relatively robust. For example, the presence of color-shading responses in the protocols of individuals having serious suicidal ideations has been reported in many studies (see Table 1). Therefore, the presence of color-shading responses on a protocol should be taken seriously. Likewise, even if there are very few studies on the S-Con, its reported predictive ability for future suicide attempts warrants that one must cautiously look for the presence of the S-Con variables in a protocol (Table 2). The research scenario is more disappointing when it comes to other projective tests. For instance, the potentials of TAT to assess suicidal ideations has been recognized (Broida, 1954); however, barring a few studies there is a near absence of research activity in this domain. Same holds true for other promising projective techniques such as the RPFS and the HFDT. Having said that, one should also admit the practical difficulties in suicide research. For example, there cannot be a longitudinal prospective study in which psychological assessment is done and the individuals are followed passively to see how many of them with the presence of the indicators of suicidal ideations actually attempt suicide (Hansell et al., 1988). Likewise, studies on individuals with past history of suicide attempts have the difficulty that the tests may not catch those dynamic psychological states that were present when an individual attempted suicide (Lester, 1970). Perhaps this is the reason that most of the studies use the data that were collected as a routine practice (for example, routine evaluation of all the admitted patients) and later if some of those individuals attempted suicide, comparisons of their test protocols were done with those who did not attempt suicide. In general, the methodological limitations of studies related to the use of projective tests in assessing suicide risks underscore the need for a cautious approach towards acceptance of the findings. For example, studies with small samples will have low power and, therefore, the inferences based on the findings of such studies need caution. Likewise, a host of other confounding variables can affect the findings. To cite one example, Meyer (1993) found that the length of Rorschach records can affect the external validity of many of the ‘Comprehensive System’ (Exner and Wylie, 1977) constellations, including the S-Con. Also, most of the studies are from the west and, therefore, caution should be used in applying these findings on individuals from other regional and cultural backgrounds. In all likelihood, the Rorschach and other projective test variables implicated in suicide risk assessment may not really be affected by the cultural background of the individual, yet, it is desirable that in future research these factors are also investigated. In conclusion, it can be said that in view of the scant research in this area, a clinician should not take the presence of a sign (or a set of signs) on a projective test as a full proof indicator of the presence of suicidal ideations; however, it must be considered as an indicator for further explorations for suicide risk. Given the serious implications of the presence of suicidal ideations in an individual, the preventive efforts demand a wider perspective. References Appelbaum, S.A., Colson, D.B., 1968. A reexamination of the color-shading Rorschach test response and suicide attempts. J. Proj. Tech. Personal. Assess. 32, 160–164.
Appelbaum, S.A., Holzman, P., 1962. The color-shading response and suicide. J. Proj. Tech. Personal. Assess. 26, 155–161. Arffa, S., 1982. Predicting adolescent suicidal behavior and the order of Rorschach measurement. J. Personal. Assess. 46, 563–568. Beck, A.T., Steer, R.A., Kovacks, M., Garrison, B., 1985. Hopelessness and eventual suicide: a 10 year prospective study of patients hospitalized with suicidal ideation. Am. J. Psychiatry 142, 559–563. Blasczyk-Schiep, S., Kazeen, M., Kuhl, J., Grygielski, M., 2011. Appraisal of suicidal risk among adolescents and young adults through the Rorschach test. J. Personal. Assess. 93, 518–526. Blatt, S.J., Ritzler, D.A., 1974. Suicide and the representation of transparency and cross-sections on the Rorschach. J. Consult. Clin. Psychol. 42, 280–287. Broida, D.C., 1954. An investigation of certain psychodiagnostic indications of suicidal tendencies and depression in mental hospital patients. Psychiatr. Q. 28, 453–464. Colson, D.B., Hurwitz, B.A., 1973. A new experimental approach to the relationship between color-shading and suicide attempts. J. Personal. Assess. 37, 237–241. Datson, P.G., Sakheim, G.A., 1960. Prediction of successful suicide from the Rorschach test using a sign approach. J. Proj. Tech. 24, 355–361. Exner, J.E., 1993. The Rorschach: A Comprehensive System: Vol. 1, 3rd ed. John Wiley and Sons, New York. Exner, J.E., Wylie, J., 1977. Some Rorschach data concerning suicide. J. Personal. Assess. 41, 339–348. Fowler, J.C., Hilsenroth, M.J., Piers, C., 2001. An empirical study of seriously disturbed suicidal patients. J Am Psychoanal Assoc 49, 16–186. Fowler, J.C., Piers, C., Hilsenroth, M.J., Holdwick, D.J., Padawer, J.R., 2001. The Rorschach suicide constellation: Assessing various degrees of lethality. J. Personal. Assess. 76, 333–351. Glassmire, D.M., Stolberg, R.A., Greene, R.L., Bongar, B., 2001. The utility of MMPI-2 suicide items for assessing suicidal potential: development of a Suicidal Potential Scale. Assessment 8, 281–290. Hansell, A.G., Lerner, H.D., Milden, R.S., Ludolph, P.S., 1988. Single-sign Rorschach suicide indicators: a validity study using a depressed inpatient population. J. Personal. Assess. 52, 658–669. Hillbrand, M., 1996. Validity of two measures of suicidal risk—MCMI II Selfdestructive potential noteworthy items vs. the suicidal risk scale. J. Offender Rehabil. 23, 1–10. Kendra, J.M., 1979. Predicting suicide using the Rorschach inkblot test. J. Personal. Assess. 43, 452–456. Kestenbaum, J.M., Lynch, D., 1978. Rorschach suicide predictors: a cross validational study. J. Clin. Psychol. 44, 921–928. Lester, D., 1970. Attempts to predict suicidal risk using psychological tests. Psychol. Bull. 74, 1–17. Lester, D., Decker, J., Eisenberg, R.J., Ecker, C., Guerriero, J., Mielish, C., 1990. Association between whirling responses on psychological tests and suicidal preoccupation. Percept. Mot. Skills 71, 1105. Linehan, M.M., Goodstein, J.L., Nielsen, S.L., Chiles, J.K., 1983. Reasons for staying alive when you are thinking of killing yourself: the reasons for Living Inventory. J. Consult. Clin. Psychol. 51, 276–286. Litinsky, A.M., Haslam, N., 1998. Dichotomous thinking as a sign of suicide risk on the TAT. J. Personal. Assess. 71, 368–378. Machover, K., 1949. Personality Projection in the Drawing of the Human Figure: A Method of Personality Investigation. American Lectures in Psychology. Vol. 25. Charles C Thomas Publisher, Springfield, pp. 3–32. Meyer, G.H., 1993. The impact of response frequency on the Rorschach constellation indices and on their validity with diagnostic and MMPI-2 criteria. J. Personal. Assess. 60, 153–180. Morgan, C., Murray, H.A., 1935. A method for investigating fantasies: the thematic apperception test. Arch. Neurol. Psychiatry 34, 289–306. National Crime Record Bureau, 2012. Suicides in India. Downloaded from hncrb.nic.in/CD-ADSI-2012/suicides-11.pdfi on April 27, 2014. Passalacqua, A.M., Echenique, S., Herrara, M.S., Orcoyen, D., 1996. Beyond subjects’ response: detecting suicidal inclinations in the Rorschach inkblot technique. Rorschachiana 21, 75–90. Petot, D., 2002. Rorschach characteristics of depressive children with suicidal ideation. Rorschachiana 25, 165–174. Rierdan, J., Lang, E., Eddy, S., 1978. Suicide and transparency responses on the Rorschach: a replication. J. Consult. Clin. Psychol. 46, 1162–1163. Rosenzweig, S., Fleming, E., Clarke, H.J., 1947. Revised scoring manual for the Rosenzweig Picture-Frustration Study. J. Psychol. 24, 165–205. Schneidman, E.S., 1961. Psycho-logic: a personality approach to patterns of thinking. In: Kagan, J., Laser, G. (Eds.), Contemporary Issues in Thematic Apperceptive Methods. Thomas, Springfield, Illinois, pp. 153–190. Seigler, M., Osmond, H., 1966. Models of madness. Br. J. Psychiatry 112, 1193–1203. Sepaher, I., Bongar, B., Greene, R.L., 1999. Codetype base rates for the I mean business suicide items on the MMPI-2. J. Clin. Psychol. 55, 1167–1173. Silberg, J.L., Armstrong, J.G., 1992. The Rorschach test for predicting suicide among depressed adolescent inpatients. J. Personal. Assess. 59, 290–303. Sullivan, G.R., Bongar, B., 2009. Assessing suicide risk in the adult patient. In: Kleespies, P.M. (Ed.), Behavioral Emergencies: An Evidence-Based Resource for Evaluating and Managing Risk of Suicide, Violence, and Victimization. American Psychological Association, Washington, pp. 59–78. Thomas, C.B., Duzynski, K.R., 1985. Are the words on Rorschach predictors of disease and death? Psychosom. Med. 47, 201–211.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
G Model
AJP-631; No. of Pages 7 D. Kumar et al. / Asian Journal of Psychiatry xxx (2014) xxx–xxx Viglione, D.J., Hilsenroth, M.J., 2001. The Rorschach: facts, fictions and future. Psychol. Assess. 13, 452–471. Virshup, E., 1976. On graphic suicide plans. Arts Psychother. 3, 17–22. Weiner, I.B., 1961. Cross-validation of a Rorschach checklist associated with suicidal tendencies. J. Consult. Psychol. 4, 312–315. White, M.A., Schreiber, H., 1952. Diagnosing suicidal risks on the Rorschach. Psychiatr. Q. (Suppl.) 26, 161–189.
7
World Health Organization, 1999. Figures and facts about suicide. World Health Organization, Geneva. Zalsman, G., Netael, R., Fischel, T., Freudenstein, O., Landau, E., Orbach, I., Weizman, A., Pfeffer, C.R., Apter, A., 2000. Human figure drawing in the evaluation of severs adolescent suicidal behaviour. J. Am. Acad. Child Adolesc. Psychiatry 39, 1024–1031.
Please cite this article in press as: Kumar, D., et al., Identification of suicidal ideations with the help of projective tests: A ?review. Asian J. Psychiatry (2014), http://dx.doi.org/10.1016/j.ajp.2014.07.004
