Underwriting Considerations for Dissociative Disorders Polly M

JOURNAL OF INSURANCE MEDICINE Copyright ᮊ 2000 Journal of Insurance Medicine J Insur Med 2000;32:71–78

ORIGINAL ARTICLE

Underwriting Considerations for Dissociative Disorders Polly M. Galbraith, MD; Patricia J. Neubauer, PhD

Objective.—Dissociative identity disorder (DID) has been diag- Address: Fortis Benefits Insurance nosed more frequently and is under greater scrutiny. Because of the Company, 2323 Grand Blvd, 7th number of comorbid conditions, the underwriting risks must be floor, Kansas City, MO 64108-2670. evaluated to determine morbidity and mortality implications. Correspondent: Polly M. Galbraith, Background.—The number of diagnosed cases of DID has in- MD, Medical Director; e-mail creased in recent years. The diagnosis often coexists with other di- [email protected]. agnoses such as bipolar disorder, major depression, post-traumatic stress disorder, anxiety disorder, somatization, personality disor- Key words: Underwriting, dissocia- ders, and psychotic disorder. A high incidence of substance abuse tive disorder, multiple personality and eating disorders is found in the population diagnosed with DID. disorder, psychiatric disorders, mor- Methods and Results.—A query of disability claim experience bidity, mortality, disability. with DID indicated that these claims tended to reach the maximum Received: December 1, 1999. duration for mental/nervous benefits despite case management and return to work activities. Accepted: January 16, 2000. Conclusions.—The DID psychiatric population is a complex group with mental disorders that place them in a group likely to use maximum disability benefits and who would pose increased life underwriting risk. In addition, the literature indicates a high excess risk for early mortality and excess health care expenses compared to the normal population.

psychiatric diagnosis that has been un- Most recent estimates suggest that more than Ader serious scrutiny for more than 10 6000 diagnosed cases of DID or dissociative years is dissociative identity disorder (DID), disorder, not otherwise specified (DDNOS, a formerly known as multiple personality dis- dissociative disorder with many similar fea- order (MPD) until the publication of the Di- tures but lacking in key elements that consti- agnostic and Statistical Manual of Mental Dis- tute a diagnosis of DID) have been identified orders, Fourth Edition (DSM-IV) in 1994.1 The since 1980.3 name was changed to address both the con- Clinical studies have identified that 12–30% cerns about the misconceptions regarding the of psychiatric inpatients have a dissociative nature of the disorder and the clinical tran- disorder and approximately 3–5% would be sition from a rare to a more commonplace diagnosed with DID.2,4–7 Using semistruc- diagnosis. Dissociation has been described in tured interviews of randomly selected psy- psychiatric literature as early as 1815, but chiatric inpatients, Rifkin et al8 found an in- only 200 cases were estimated prior to 1980.2 cidence rate of 1% for DID. In groups of non-

71 JOURNAL OF INSURANCE MEDICINE clinical subjects, prevalence rates of 3.5–11% ality disorder, 44% anxiety disorder, 40% were found for dissociative disorders and schizophrenia, and 31% substance abuse. 0.5–1.3% were diagnosed with DID.7,9 Given All studies of DID have found evidence of the increased assignment of this diagnosis, comorbidity with other mental and physical there is a real need to determine associated disorders. It was found that the average num- morbidity and mortality risks. ber of lifetime comorbid Axis I disorders was The essential features of the dissociative 7.3 (SD ϭ 2.5) while the average number of disorders are a disruption in the usually in- Axis II diagnoses was 3.6 (SD ϭ 2.5). This did tegrated functions of consciousness, memory, not include DID, post-traumatic stress disor- identity, or perception of the environment. der (PTSD), sleep disorders, or psychosexual The disturbance may be sudden, gradual, disorders. It was concluded the average DID transient, or chronic. DID is characterized by patient meets the lifetime criteria for about 15 the presence of two or more distinct identi- different DSM-IV disorders.11 Associated fea- ties, or personality states, that recurrently tures include post-traumatic symptoms such take control over an individual’s behavior ac- as nightmares, flashbacks, and startle reac- companied by an inability to recall important tions; thus, an additional diagnosis of post- personal information that is too extensive to traumatic stress disorder is warranted.1 Her- be explained by ordinary forgetfulness. This man and van der Kolk et al12,13 suggest that disturbance cannot be attributed to the direct DID should be diagnosed as complicated physiological effects of a substance or general PTSD, as the diagnosis is clearly related to medical condition. Amnesia or memory gaps early trauma. The Sidran Foundation14 (a DID are frequent in the personal history. Transi- advocacy group) estimates that 80–100% of tions between identities are often triggered people with DID also have secondary PTSD. by psychosocial stress.1 Saxe et al2 found that all of the psychiatric Controversy exists concerning the differ- inpatients with DID had comorbidity with ential diagnosis, which may include many another psychiatric diagnosis. Most met the other mental disorders such as bipolar dis- criteria for PTSD and major depression, either order with rapid cycling, anxiety disorders, currently or during their lifetime. Approxi- somatization disorders, personality disorders, mately two thirds met the criteria for border- and psychotic disorders.1 Self mutilation, ag- line personality disorder and substance gressiveness, and suicidal and conversion abuse. symptoms such as pseudoseizures may occur. Ross interpreted these findings to suggest The condition appears to have a fluctuating that DID is part of a normal human response clinical course that is chronic and recurring. to severe chronic childhood trauma. He fur- People with DID may also meet the psychi- ther opined that trauma is a major etiologic atric criteria for mood, substance, sexual, eat- factor in all psychiatric disorders.11 Pribor and ing, and sleep disorders.1 Some meet the cri- Dinwiddie found that abused subjects show teria for borderline personality disorder as more PTSD, panic disorder, social and simple well. Thus, there is a strong association of phobia, agoraphobia, major depression, and multiple comorbid psychiatric disorders with substance abuse when compared to general DID. The diagnosis may be difficult to make population and nonabused psychiatric sub- and confused with a number of other better- jects.15 A review of the literature found that known psychiatric disorders. Estimates sug- DID was notable for auditory hallucinations gest that persons will be treated in the mental (positive Schneiderian symptoms), depres- health system for up to 7 years before an ap- sion and suicidality, phobic anxiety, somati- propriate diagnosis is made. Ross et al10 zation, substance abuse, and borderline fea- found that prior to the MPD (or DID) diag- tures.16 In a study of 2-year outcomes for in- nosis, patients averaged 2.74 other diagnoses, patients, Ellason and Ross17 found that people including 64% affective disorder, 57% person- with DID endorse a mean of 7.3 Axis I di-

72 GALBRAITH ET AL—UNDERWRITING CONSIDERATIONS FOR DISSOCIATIVE DISORDERS

agnoses at the onset of diagnosis. Ross and In a study of 51 sexual abuse survivors, Norton18 suggest that DID might be a supe- 94% had depression, 65% had substance rordinate diagnosis in that other Axis I di- abuse history, and 55% had DID diagnosed.24 agnoses tend to remit with successful treat- Von Braunsberg7 reviewed the literature and ment and integration. Ross and Dua19 deter- concluded that having a dissociative experi- mined the lifetime health costs of treatment ence at the moment of trauma is the most sig- for 15 people with DID. The total cost spent niﬁcant long-term predictor for development by the Canadian health system was $4.1 mil- of PTSD. Increased dissociative experiences lion. The average time in the health care sys- were found in 100% of people with trauma tem before diagnosis was 98 months. The av- history. McCauley et al25 found that subjects erage time in care after diagnosis was 32 abused in childhood were four times as likely months. to attempt suicide, three times more likely to be hospitalized, and twice as likely to have DEPRESSION AND ANXIETY suicidal thoughts as nonabused controls.

The single most common presenting symp- SUBSTANCE ABUSE AND EATING tom in MPD patients is depression. A Nation- DISORDERS al Institute of Mental Health (NIMH) survey found depression in 88% of the cases, and Substance abuse is another frequent prob- about 75% of patients reported mood lem and polysubstance abuse was noted in swings.20 Symptoms suggestive of phobic, one third of the NIMH study group. Anal- anxiety, or panic disorder are common in dis- gesic overuse often accompanies the treat- sociative patients. These often accompany ment for headaches, which are common to triggers, or reminders of traumatic experienc- this group. In one study group in Canada, es, and lead to much avoidance behavior. El- Ross and Norton found 43% of a group of 100 lason et al16 found that depression was the subjects with substance abuse met criteria for most common comorbid disorder with DID. a dissociative disorder, and of those, 14 met In a study of subjects presenting for treat- the criteria for DID. In an American group of ment for obsessive-compulsive disorder, 20% subjects, Ross11 found 60% of substance abus- presented with signiﬁcant dissociative symp- ers met criteria for dissociative disorders, in- toms. These persons were more likely to have cluding 18% with DID. Felker et al26 deter- personality disorders and had a greater num- mined that substance abuse alone or in con- ber and severity of obsessive-compulsive junction with other psychiatric disorders has symptoms.21 Goff et al21 did not ﬁnd an over- repeatedly led to increased mortality rates. lap with a diagnosis of DID, but 3 of the sub- Dunn et al27 provided the opinion that per- jects met the criteria for DDNOS. DID has sons with substance abuse complaints should also been associated with phobic anxiety.22 be screened for dissociative disorders based on the relationship of the two disorders. In a TRAUMA AND PTSD group of 265 male chemically dependent mil- itary veterans, 41.5% had a high level of re- Many persons have reported severe physi- ported dissociative experiences. The literature cal and sexual abuse, particularly in child- has produced documentation of 30–70% in- hood. The NIMH survey indicated that 97% cidence of childhood sexual abuse in patients of all MPD patients reported childhood trau- with eating disorders.28,29 Coons et al30 found ma, with 68% reporting incest.23 Combina- a 10% incidence of previously diagnosed eat- tions of sexual abuse and physical abuse were ing disorders in a series of 50 patients with reported in two thirds of those surveyed. MPD. Anecdotally, most people with MPD Ross et al6 found a 95% incidence of physical describe some sort of disrupted eating pat- and/or sexual abuse. terns.29

73 JOURNAL OF INSURANCE MEDICINE

HALLUCINATIONS AND PSYCHOTIC Suicide is often underreported for many rea- DISORDERS sons, sometimes because of lack of suicide The majority of the patients report auditory identification in statistics regarding cause of 18 and/or visual hallucinations. Ellason and death. Ross and Norton identified that 72% Ross17 noted that 24–50% of patients with dis- of subjects in their study group of 100 per- sociative disorder have been previously di- sons with DID diagnoses had attempted sui- agnosed and treated for schizophrenia. The cide. The group that had attempted suicide hallucinations primarily consisted of com- had spent more time in the mental health sys- mentary from internal parts sometimes sup- tem, had more reports of physical abuse and portive but also abusive in tone. The voices rape, had more psychiatric diagnoses, had are described as loud thoughts differentiated more inpatient treatment, and more Schnei- from the voices typical in psychotic disorders, derian symptoms. The group that completed which are more often perceived as outside the suicide (suicide resulting in death) had more person. Ross et al6 found that patients with time in jail and may have had more antisocial DID report a greater number of Schneiderian personality features but were not different first-rank symptoms than do schizophrenics. from the parasuicidal group in other dimen- Schneiderian first-rank symptoms are the sions. 23 positive symptoms thought to identify Putnam et al noted suicidality as a pre- schizophrenia, such as auditory hallucina- senting symptom in nearly 70% of 100 cases 18 tions (hearing voices in your head), intrusive reported. Ross and Norton reported 236 thoughts, thought withdrawal, thought inser- cases of dissociative identity disorder, in tions, thought broadcasting, and delusional which 72% had attempted suicide. Of this thinking. One study of schizophrenic patients group, 2% actually were successful in killing found that this group reported 1.3 Schneider- themselves. The methods of the suicide at- ian symptoms while patients with DID report tempt included drug overdose (68%), burns an average of 4.9 Schneiderian symptoms.17 or other injury (57%), and wrist slashing This suggests that Schneiderian symptoms (49%). Anderson et al studied 51 sexual abuse are more indicative of DID than schizophre- survivors to address dissociative disorders. nia. In these patients, when the personality Of this group, 49% had attempted suicide at fragments, or alters, integrate, these symp- least once by drug overdose, cutting with a toms disappear. Fink and Golinkoff31 found knife or other weapon, with a gun, burning, 24 that persons with DID had a greater number and hanging. Kaplan et al found that in- of somatic symptoms, more prevalent major creased suicide ideation was related to both depression, more dissociative symptoms, and previous and current abuse. This group noted a higher level of childhood trauma than pa- significantly higher levels of dissociation, de- 33 tients diagnosed with schizophrenia. Disso- pression, and somatization. ciative patients were found to have greater levels of anxiety and depression than persons PERSONALITY DISORDERS with schizophrenia.17 Self mutilation occurs in at least a third of patients and is differentiated from suicidal SUICIDE behavior.20 This behavior is common in peo- One of the most common presenting fea- ple diagnosed with borderline personality tures of DID consists of suicidal ideation with disorder and serves in a soothing capacity, additional suicide attempts. Kung et al32 possibly relating to an emotional release or found that social isolation, lack of social in- release of endorphins as a reaction to the tegration, depression, addiction, somatic ill- physical act. Ross11 indicated a belief that ness, and stressful life events all have been both DID and borderline personality disorder identified as causes or risk factors for suicide. exist as comorbid condition. Studies have in-

74 GALBRAITH ET AL—UNDERWRITING CONSIDERATIONS FOR DISSOCIATIVE DISORDERS

dicated that more than 50% of persons with OTHER DID meet criteria for borderline personality The most common neurological symptom 11 disorder. Ellason et al found that borderline, reported is headaches, which are often de- avoidant, and self-defeating personality dis- scribed as blinding and resistant to standard orders were the most frequent comorbid Axis analgesics. Seizures or seizurelike activity II diagnoses. Only 8.7% of the subjects met was reported in 21% of cases. Unexplained the criteria for histrionic personality disor- nausea and abdominal and pelvic pain are der.16 common and found in about a third of Herman found that persons with PTSD cases.23 Complex partial seizures (also called have a problem with affect regulation. Behav- temporal lobe epilepsy) are listed in the iors identified in the category of affect dys- DSM-IV as an exclusion criteria for DID. In regulation included difficulty modulating an- Ross’ opinion and his review of the literature, ger, chronic self-destructive and suicidal be- no confirming data existed that would con- haviors, difficulty modulating sexual involve- firm overlap or interaction between partial ment, and impulsive risk-taking behaviors.12 complex seizures and DID. Thus, persons Van der Kolk et al13 found that subjects with could have both disorders, although differ- PTSD have significantly higher rates of dis- entiation may be tricky without evidence on sociation, somatization, and affect dysregu- electroencephalogram. In the van der Kolk et lation. Dell found identical personality pa- al review of the literature, a close association thology in DID and chronic PTSD patients. between somatization and dissociation and He provided the opinion that the psychopa- PTSD was found.13 In women with somati- thology of dissociative patients was more re- zation disorder, more than 90% reported lated to severity of PTSD symptoms than dis- some type of abuse and 80% reported sexual sociative symptoms. The core personality pa- abuse as either a child or an adult. thology is avoidant and self defeating. With increased severity of illness, more borderline RECOVERY AND TREATMENT features are noted.34 Another issue complicating the under- In the Ross and Norton study group, DID standing of risk with persons with DID is the patients were noted to have engaged in an- lack of agreement regarding appropriate tisocial activities. These activities included treatment and what constitutes recovery. the following: 12% patients had been con- Kluft36 notes that approximately 90% of pa- victed of a crime, 19% had worked as a pros- tients who remain in treatment achieved in- 18 titute, and 12% had been in jail. Spanos et tegration of personality states. Initial study 35 al postulate that a scheme of presenting indicated that patients with clearly defined multiple personalities could be a vehicle for treatment dimensions responded with mea- negotiating a difficult personal dilemma or sured improvement. Barach reported the avoiding consequences of criminal behavior, treatment recommendations for DID that such as that presented by Kenneth Bianchi, were identified by the International Society accused of being the Hillside Strangler. Wom- for the Study of Dissociation (ISSD) stan- en make up approximately 87% of diagnosed dards of practice committee. For a successful cases with DID.10 As men have similar abuse outcome, treatment recommendations range rates as women, this finding has led to spec- from 2–3 years of psychotherapy for stable ulation. Men with DID may show more an- clients and up to 6 years for complex patients. tisocial traits or seek treatment less frequent- Axis II disorders and comorbid mental dis- ly. Researchers have questioned whether men orders would increase the length of treat- with DID would be more likely to be identi- ment. Common comorbid disorders included fied in prison populations than clinical pop- addictions, eating disorders, sexual disorders, ulations.10 mood disorders, and anxiety.37 Hospital stays

75 JOURNAL OF INSURANCE MEDICINE may be necessary for patients with combi- hospital staff identified dissociative episodes, nations of DID and major depression or eat- and the claimant began to relate her history ing disorders. Medications have not been of internal voices that she had never reported found to be generally effective, but antide- on prior hospitalizations. The claimant iden- pressants and anxiolytics were rated by a tified 12 alters with the help of the staff and group of treatment providers as having mod- left the hospital with a referral to a therapist erate effectiveness.38 who specialized in care for patients with DID. With several months of therapy aimed at METHODS helping her identify her system of alters and address experiences related to severe child- Claimants with DID diagnoses were iden- hood physical, emotional, and sexual abuse, tified from disability claims filed in 1997–98. the claimant was now chronically depressed, Areviewof6files indicated that these claims suicidal, and unable to return to work. The reached the maximum duration for mental/ therapist had recommended that the claimant nervous benefits despite active case manage- apply for Social Security disability and ex- ment and return to work activities. These files pected 2–3 years for recovery. The claimant were reviewed and a compilation of features no longer had insurance through her com- was utilized to describe a common clinical pany and did not know how long she would picture of a person with DID who filed a dis- be able to maintain her health insurance ability claim. The compilation served to ob- through Cobra. scure the identity of any one person with DID who has filed a disability claim. CONCLUSIONS

RESULTS Morbidity Susan P. is a 36-year-old technical writer Attempted suicide would logically be a sig- for a computer-related business. She reported niﬁcant risk factor for morbidity and an un- a history of chronic depression and suicide acceptable risk for both short-term and long- attempts, dating to her teenage years. She term disability and for waiver premium often had been hospitalized on 3 occasions prior to sold with life insurance. These individuals the present illness and completed 2 inpatient are frequently very emotionally, fragile indi- detoxiﬁcations for alcohol and polysubstance viduals and when put in stressful situations abuse. She had previously been diagnosed in the workplace will decompensate so as to with both dysthymia and major depression, be unable to perform the duties of their own PTSD, bipolar II, borderline personality dis- or any occupation. Many may become re- order, and mixed personality disorder. She stricted from the workplace for fear of exces- has been addressing her substance abuse and sive risk of relapse of major symptomatology. anorexia/bulimia through Alcoholics Anon- The confounding comorbid mental conditions ymous and Overeater’s Anonymous during such as bipolar disorders, affective disorders, her previous year of sobriety. The current ep- and personality disorders can all be quite isode of major depression was precipitated by limiting. Memory lapses, alter personalities a recent divorce and custody battle with her with hostile or destructive tendencies, and third husband. She was described as very avoidance of interpersonal contact would bright and capable by her policyholder, but present problems for employment in the short they had grave concerns about continued em- and long term. Rossini et al found that pa- ployment because of her mood swings, irri- tients with DID fell into the average range of tability, and uneven attendance. intellectual abilities on the Wechsler Adult In- During the hospitalization for major de- telligence Scale-Revised test but had greater pression at the onset of her current illness, the scatter than would be expected on verbal

76 GALBRAITH ET AL—UNDERWRITING CONSIDERATIONS FOR DISSOCIATIVE DISORDERS

subtests. The deficit was manifested in diffi- group of individuals.44 Because of the number culty with concentration and distractibility.39 of comorbid conditions, persons with DID This pattern of deficits would have a great would be categorized as very high to unac- impact on the patient’s ability to perform job ceptable substandard mortality risks. responsibilities requiring focus and attention. The lack of accepted treatment and prolonged REFERENCES treatment periods to integration and clinical improvement would make this group high 1. American Psychiatric Association. Diagnostic and risk for prolonged periods of work absences. Statistical Manual of Mental Disorders, Fourth Edition. Washington DC: American Psychiatric Association; 1994. Mortality 2. Saxe GN, van der Kolk BA, Berkowitz R, et al. Dis- sociative disorders in psychiatric inpatients. Am J Despite deinstitutionalization and more ef- Psychiatry. 1993;150:1037–1042. fective diagnosis and treatment of various 3. Murray JB. Dimensions in multiple personality psychiatric conditions, several recent studies disorder. J Gen Psychol. 1994;155:233–246. have shown an excess number of natural 4. Horen SA, Leichner PP, Lawson JS. Prevalence of deaths in psychiatric patients. This may be dissociative symptoms and disorders in an adult because patients with serious but undetected psychiatric inpatient population in Canada. Can J Psychiatry. 1995;40:185–191. physical disorders are selectively referred to 5. Latz TT, Kramer SI, Hughes D. Multiple person- 40 psychiatric clinics. There is still marked ex- ality disorder among female inpatients in a state cess in unnatural deaths, primarily due to su- hospital. Am J Psychiatry. 1995;152:1343–1348. icides, accidents, and homicides. In a follow- 6. Ross CA, Anderson G, Fleisher WP, Norton GR. up of 500 psychiatric outpatients over a 6–12- The frequency of multiple personality disorder year period, death from unnatural causes oc- among psychiatric inpatients. Am J Psychiatry. curred at three and a half times the expected 1991;148:1717–1720. 40 7. Von Braunsberg MJ. Multiple personality disorder: rate, which is a significant increase. The an investigation of prevalence in three populations. rates of suicide were nearly 15 times that ex- In: Dissertation Abstracts International 1994;54: pected, and homicide nearly 5 times that ex- 5955B. pected. Accidents, including overdoses not 8. Rifkin A, Ghisalbert G, Dimatou S, Jin C, Sethi M. ruled suicides by the coroner, were seen at Dissociative identity disorder in psychiatric inpa- more than twice the rate expected in the nor- tients. Am J Psychiatry. 1998;155:844–845. mal population. Excessive rates of unnatural 9. Ross CA. Epidemiology of multiple personality disorder and dissociation. Psychiatr Clin North Am. deaths, particularly suicide, have been ob- 1991;14:503–517. served in virtually every study of mortality 10. Ross CA, Norton GR, Wozney K. Multiple person- in psychiatric patients.41 ality disorder: an analysis of 236 cases. Can J Psy- Suicide is the most important reason for ex- chiatry. 1989;34:413–418. cess mortality in the psychiatric population. 11. Ross CA. Diagnosis of dissociative identity disor- Completed suicide in the psychiatric popula- der. In: Cohen LM, Berzoff J, Elin M, eds. Dissoci- ative Identity Disorder: Theoretical and Treatment Con- tion tends to occur at younger ages. A com- troversies. Northvale, NJ: Jason Aronson, Inc.; 1995: parison of psychiatric patient suicides and 261–284. those in the general population gives a ratio 12. Herman, J. Complex PTSD: a syndrome of pro- of 5:1.42 The rate of suicide among psychiatric longed and repeated trauma. J Trauma Stress. 1992; patients is 30 times that of the general pop- 5:377–391. ulation in those suffering with affective dis- 13. van der Kolk BA, Pelcovitz D, Roth S, Mandel FS, orders.43 Overall, suicide may range from 10– McFarlane A, Herman JL. Dissociation, affect dys- regulation and somatization: the complex nature of 15% in various mental disorders, and it is adaptation to trauma. Available at: http:// very difficult to predict which patients will www.trauma-pages.com/vanderk5.htm. 1996. attempt or commit suicide. Thus, statistical 14. Sidran Foundation. (1998). Dissociative identity measures are used in underwriting this disorder (multiple personality disorder). Available

77 JOURNAL OF INSURANCE MEDICINE

at Trauma Resource Area at: http://www.sidran. tiple personality in eating disorder patients. Int J org/didbr.html. Eat Disord 1993;13:235–239. 15. Pribor EF, Dinwiddie SH. Psychiatric correlates of 30. Coons PM, Bowman ES, Milstein V. Multiple per- incest in childhood. Am J Psychiatry. 1992;149:52– sonality disorder: a clinical investigation of 50 56 cases. J Nerv Ment Dis. 1988;176:519–527. 16. Ellason JW, Ross CA, Fuchs DL. Lifetime Axis I 31. Fink D, Golinkoff M. Multiple personality disor- and II comorbidity and childhood trauma history der, borderline personality disorder, and schizo- in dissociative identity disorder. Psychiatry. 1996; phrenia: a comparative study of clinical features. 59:255–266. Dissociation. 1990;3:127–134. 17. Ellason JW, Ross CA. Positive and negative symp- 32. Kung HC, Liu X, Juon HS. Risk factors for suicide toms in dissociative identity disorder and schizo- in caucasions and african-americans: a matched phrenia: a comparative analysis. JNervMentDis. case-control study. Soc Psychiatry Psychiatr Epide- 1995;183:236–241. miol. 1998;33:155–161. 18. Ross CA, Norton GR. Suicide and parasuicide in 33. Kaplan ML, Asnis GM, Lipschitz DS, Chorney P. multiple personality disorder. Psychiatry. 1989;52: Suicidal behavior and abuse in psychiatric outpa- 365–371. tients. Compr Psychiatry. 1995;36:229–235. 19. Ross CA, Dua V. Psychiatric health care costs of 34. Dell PF. Axis II pathology in outpatients with dis- multiple personality disorder. Am J Psychother. sociative identity disorder. J Nerv Ment Dis. 1998; 1993;47:103–112. 186:352–356. 20. Putnam FW. Diagnosis and Treatment of Multiple Per- 35. Spanos NP, Weekes JR, Bertrand LD. Multiple personality Disorder. New York, NY: Guilford Press; sonality: a social psychological perspective. JAb- 1989. norm Psychol. 1985;94:362–376. 21. Goff DC, Olin JA, Jenike MA, Baer L, Buttolph ML. 36. Kluft RP. Treatment trajectories in multiple per- Dissociative symptoms in patients with obsessive- sonality disorder. Dissociation. 1994;7:63–76. compulsive disorder. J Nerv Ment Dis. 1992;180: 37. Barach PM. Guidelines for treating dissociative 332–337. identity disorder (multiple personality disorder) in 22. Ellason JW, Ross CA. Two-year follow-up of in- adults. International Society for the Study of Dis- patients with dissociative identity disorder. Am J sociation. Available at: http://www.issd.org/ Psychiatry. 1997;154:832–839. isdguide.htm. 23. Putnam RW, Guroff JJ, Silberman EJ, Post RM. The 38. Putnam FW, Loewenstein RJ. Treatment of multi- clinical phenomenology of multiple personality ple personality disorder: a survey of current prac- disorder: review of 100 recent cases. J Clin Psychi- tices. Am J Psychiatry. 1993;150:1048–1052. atry. 1986;47:285–293. 39. Rossini ED, Schwartz DR, Braun BG. Intellectual 24. Anderson G, Yasenik L, Ross CA. Dissociative ex- functioning of inpatients with dissociative identity periences and disorders among women who iden- disorder and dissociative disorder not otherwise tify themselves as sexual abuse survivors. Child speciﬁed. Cogntive and neuropsychological as- Abuse Negl. 1993;17:677–686. pects. J Nerv Ment Dis. 1996;184:289–294. 25. McCauley J, Kern DE, Kolodner K, et al. Clinical 40. Martin RL, Cloninger CR, Guze SB, Clayton PJ. characteristics of women with a history of child- Mortality in follow up of 500 psychiatric outpa- hood abuse. JAMA. 1997;277:1362–1368. tients. Total mortality. Arch Gen Psychiatry. 1985; 26. Felker B, Yazel JJ, Short D. Mortality and comor- 42:47–54. bidity among psychiatric patients: a review. Psy- 41. Eastwood MR, Stiasny S, Meier HM, Woogh CM. chiatr Serv. 1996;47:1356–1363. Mental illness and mortality. Compr Psychiatry. 27. Dunn GE, Paolo AM, Ryan JJ, Van Fleet J. Disso- 1982;23:377–385. ciative symptoms in a substance abuse population. 42. Morrison JR. Suicide in a psychiatric practice po- Am J Psychiatry. 1993;150:1043–1047. plulation. J Clin Psychiatry. 1982;43:348–352. 28. Goldberg LA. Sexual abuse antecedent to anorexia 43. Guze SB, Robins SB. Suicide and primary affective nervosa, bulimia, and compulsive overeating: disorders. Br J Psychiatry. 1970;117:437–438. three case reports. Int J Eat Disord 1987;6:675–680. 44. Brackenridge RDC, Elder WJ. Medical Selection of 29. Levin AP, Kahan M, Lamm JB, Spauster E. Mul- Life Risks. 4th ed. Indianapolis: Macmillan; 1998.