suicide by cop

PAPER

PSYCHIATRY & BEHAVIORAL SCIENCES

Kris Mohandie,1 Ph.D. and J. Reid Meloy,2,3 Ph.D.

Suicide by Cop Among Female Subjects in Officer-Involved Shooting Cases

ABSTRACT: An analysis of suicide by cop (SBC) among female subjects in a large sample (n = 707) of officer-involved shootings (OIS) is reported. Women represented 3% of the total sample (n = 21) and 5% (n = 12) of the 256 SBC cases. Therefore, 57% of the women (n = 12) were classified as SBC, and 81% of the women (n = 17) behaviorally evidenced some suicidal motivation. The results underscore that when a female sub- ject becomes involved in a situation leading to an OIS, there is a very high likelihood that she is motivated to be intentionally killed. She is almost guaranteed to have a diagnosed mental disorder, has been prescribed psychiatric medications, and is abusing a substance at the time, most likely alco- hol. Nevertheless, she also poses a serious risk of injury to both police officers and civilians during the encounter and will possess a firearm half the time. The wish for suicide is often precipitated by serious relationship issues. Comparisons are made to the male subjects in our previous study.

KEYWORDS: forensic science, suicide, lethal force, law enforcement, suicide by cop, officer-involved shootings, gender and suicide

Suicide by cop (SBC) is a method of suicide that occurs when a subject engages in behavior which poses an apparent risk of serious injury or death, with the intent to precipitate the use of deadly force by law enforcement against the subject (1,2). Past research has demonstrated that these situations are overwhelmingly caused by male subjects (1–7).

Hutson et al. (6) found that 98% of their SBC subjects were men, and Homant and Kennedy (3) reported that 92% of their sam- ple were men; while both Wilson et al. (5) and Haruff et al.’s (7) groups were exclusively men. Mohandie et al. (2) reported that 97% of their 707 officer-involved shooting (OIS) subjects and 95% of their 256 SBC subjects were men. Below is an example of a female SBC case from the study:

A 34-year-old Caucasian woman, despondent over her estranged boyfriend’s affair and rejection, called him demanding a lunch meeting to bring ‘‘closure’’ to their relationship. He agreed to the meeting hoping that she would stop calling and pursuing him. Unbeknownst to him, she had visited her brother, borrowed his car, accessed and stolen his personal handgun and several rounds of ammunition. She picked up her estranged boyfriend from his office, obtained some takeout food, and drove to a nearby vacant parking lot ‘‘to talk.’’ Their conversation about the relationship quickly escalated into a heated argument. The woman produced the handgun, threatened to shoot him, and commit suicide, while he attempted to calm her. Ultimately, he was able to get the passenger door open, but not before she fired her weapon, striking him once in the shoulder, and missing him with other rounds. He fled to a nearby restaurant, called 911, while she

remained in the car. When the police arrived, she was noncom- pliant, and refused to surrender. Negotiators were deployed and attempted to communicate with her, but she would not engage in meaningful dialogue. She yelled obscenities, gestured to the offi- cers with her middle finger several times, and was observed to be drinking hard alcohol straight out of a bottle. She dropped what was later determined to be a suicide note out the window, brandished her weapon through the sunroof and out the driver’s window. She was shot to death by a police sniper after she pointed her weapon directly at nearby officers. The gun was later determined to be empty of additional cartridges.

Research into suicide by other means consistently demonstrates that men are more likely to complete their suicide while women attempt it more frequently (8). Men typically choose more violent means (jumping, hanging, and self-inflicted gunshot), while women typically choose less lethal methods (overdosing, slashing wrists), often engaging in self-injury as an expression of distress and as a cry for help. Thus, it is no surprise that SBC—an extension of the lethal means preferred by men—would usually involve male sub- jects. The reported results of these studies may therefore be over- shadowing aspects unique to the female SBC subject. The present study attempts to shed some light on the question: are there issues unique to female subjects who choose SBC as their method of suicide?

The goals of this study are to examine SBC among female OIS cases, specify the occurrence rate among female OIS subjects, and discern historical, incident, and behavioral characteristics that might be helpful in understanding this small but important demographic. To highlight apparent gender differences, results of male SBC sub- jects are also reported on key issues.

Methodology

As described in Mohandie et al. (2009), the OIS files of partici- pating police and criminal justice agencies were analyzed over an

1Operational Consulting International, Inc., P.O. Box 88, Pasadena, CA 91102.

2Forensis, Inc., P.O. Box 90699, San Diego, CA 92169. (http://www. forensis.org)

3University of California, San Diego, CA. Received 18 Dec. 2009; and in revised form 28 Feb. 2010; accepted 6

Mar. 2010.

J Forensic Sci, May 2011, Vol. 56, No. 3 doi: 10.1111/j.1556-4029.2010.01686.x

Available online at: onlinelibrary.wiley.com

664 � 2011 American Academy of Forensic Sciences

11-month time period between March 2006 and January 2007 by three trained researchers (a primary researcher and two assistants). These OIS files were from eight invited law enforcement sources representing more than 90 North American police departments in the United States and Canada. OIS files consisted of every single deadly force and less lethal incident investigated as an OIS by the participating agency from 1998 through 2006. All data were archi- val, and therefore, subject permission for inclusion in the study was not required. Seven hundred and seven cases were included in the final sample. Cases were excluded if officers did not discharge their weapons (lethal or less lethal), if officers only fired at animals, or if the officer accidentally discharged his or her firearm. Lethal weapons were handguns, shotguns, rifles, and MP-5s; less lethal weapons included the officer’s hands, baton, taser, K-9, bean bag shotgun, Arwen impact weapons, and Oleoresin Capsicum (OC) pepper spray.

Data reviewed included primary source material in the OIS inves- tigative files. These materials were usually extensive and included police reports, witness statements, criminal histories on subjects, photographs, videotapes, and external review reports. Additional support material was sought as needed and included interviews with investigating detectives and occasionally direct contact with involved officers. This occurred in c. 10% of the SBC cases.

Data for each of the included cases were recorded on a six-page, 110-variable codebook developed by the authors which covered the following areas: (i) Incident Characteristics included the type of shooting (deadly force and ⁄ or less lethal), fatalities, number of responding officers, number of rounds fired by officers, use of alternatives to deadly force including verbal strategies and their reported effects, call type, setting and location of incident, whether the event was spontaneous or deliberate, and type of crime (major or minor); (ii) Subject Data included demographics and behavioral information about the subject such as communication of suicidal ideation (any communications with suicidal content, including state- ments of intent or plans) 2 months or less preceding incident, sui- cide notes, weapon possession and simulation, weapon status, violence against others during the incident, threats, escape behavior, resistance, known psychological history such as prior suicidal idea- tion (2 months or more preceding incident) or attempts, mental health diagnoses and treatment, the presence of psychosis, sub- stance use and prior treatment, intoxication, health problems, recent relationship problems, criminal history, and current criminal justice status (on parole or probation); and (iii) Outcomes, most notably whether injury or death occurred to anyone involved in the inci- dent—subject, law enforcement, or others—as well as category of overall tactics deployed by law enforcement during the incident. A short narrative overview of each case was recorded in the code- book, along with any spontaneous statements made by surviving subjects after the incident. A variable was coded as ‘‘unknown’’ if data for the particular case variable were unavailable (the codebook is available from the senior author KM).

Cases were categorized as SBC when the subject engaged in actual or apparent risk to others with the intent to precipitate the use of deadly force by law enforcement personnel. An initial deter- mination of SBC status was made by the primary researcher; how- ever, narrative summaries were independently reviewed by the first author (KM) to verify each determination. The first and second (JRM) authors blindly and independently scored a representative 8% (n = 53) of the overall sample to formally assess the reliability of these determinations of OIS or SBC. An intraclass correlation coefficient (ICC) was calculated.

Because three researchers were employed to code the overall data and to assess the accuracy of data recording, an interrater

reliability correlation coefficient was calculated on a representative 10% (n = 73) of the overall sample for all coding categories.

Results

Reliability

Coefficient alpha for interrater reliability on overall variables was 0.88. Eighty-eight percent of the time, the two coders agreed on all the variables in each case from the entire code sheet (except for those excluded from analysis). Certain variables were excluded in this analysis: shooting distance, where rounds hit, other call for service, date of birth (age was used instead), number of children, recency of job loss, length of gun ownership, manner of weapon acquisition, survivor statements, and case narrative. The ICC for assignment to SBC or OIS groups was 0.93.

SBC, Suicide, and OIS Statistics

Twenty-one of the 707 cases in the entire database involved female subjects, representing 3% of the total database. Of these, 12 (57%) were determined to be SBC, five (24%) were classified as suicidal or attempted suicide by their own hand, and four (19%) were determined to be regular OIS.

Mortality

Three of the female SBC subjects died from their wounds (two by police gunfire, one committed suicide while provoking the police to kill her), and eight (67%) of the SBC attempters were injured during the encounter with police—only one was uninjured. This represents a 25% fatality rate, but a 92% casualty rate in the SBC group. One of the suicide ⁄ attempted suicide group died from her wounds, three were injured (60%), and one was uninjured, yielding a 20% fatality rate, but an 80% casualty rate. One of the women in the OIS group was killed, and one (25%) was injured; the other two were unharmed. There was a 25% chance of death to the subject, but a 50% casualty rate among female OIS subjects.

Demographics

The mean age of the female SBC subjects was 40, with a range of 19–54. Fifty-eight percent (n = 7) of the SBC subjects were Caucasian, 25% (n = 3) were Hispanic, 8% (n = 1) were Native American, and 8% (n = 1) were unknown. Forty-two percent (n = 5) were single, 17% (n = 2) separated or divorced, 17% (n = 2) cohabiting, 17% (n = 2) married, and 8% (n = 1) were of unknown marital status. Seventy-five percent (n = 9) were deter- mined to be heterosexual, 8% (n = 1) homosexual, and 8% (n = 1) were of unknown sexual orientation.

Fifty percent (n = 6) of the subjects had children, 33% (n = 4) did not, and this factor was unknown in 17% (n = 2) of the cases. Of those known to have children, 20% (n = 2) had custody at the time, 30% (n = 3) had adult children, and 10% (n = 1) did not have custody. In 17% (n = 2) of the subjects with children, it was determined that issues pertaining to the children (custody and child support frustrations, etc.) were related to the situation.

Weapon Possession

Female SBC subjects were armed with weapons during 100% (n = 12) of the incidents—none feigned or simulated weapon pos- session. Fifty percent (n = 6) possessed a firearm, which was

MOHANDIE AND MELOY • SUICIDE BY COP AMONG FEMALE SUBJECTS 665

loaded and operational 33% (n = 4) of the time, unloaded 17% (n = 2) of the time, and inoperable 0% (n = 0) of the time. One third (n = 2) of those who possessed a firearm (n = 6) actually fired their weapon at the police. The other six female SBC subjects (50%) possessed a knife, while one of those six women with a knife also had a blunt force object as a weapon as well.

Danger to Others

Fifty-eight percent (n = 7) of the female SBC subjects attempted or engaged in violence against other civilians in conjunction with the incident, and 50% (n = 6) directed violence toward the police, including attempting to kill the officers (n = 2), or assault with a deadly weapon (n = 4). Two of the female SBC subjects actually harmed civilians, and one harmed a police officer, yielding a 25% likelihood of injuries to others during the incident. There were no homicides of others.

Suicidal Communications

Suicidal communications by the subject at any point prior to or during the incident occurred in 92% (n = 11) of the cases. For those who communicated a suicidal wish or intent prior (1 month or less) to the incident (n = 7), 57% (n = 4) did so in the minutes prior to the event, 14% (n = 1) within a week, and 29% (n = 2) within a month. Only one of the 12 female SBC subjects did not communicate her suicidal ideation prior to, or during the event to anyone (she survived the attempt and admitted it afterward). Those who did communicate prior (n = 7) told their significant other 43% (n = 3) of the time, a family member 14% (n = 1), friends 29% (n = 2), and others 14% (n = 1). These communications referenced the SBC method 14% (n = 1) of the time—86% (n = 6) of the female subjects who talked about suicide prior did not specifically mention SBC.

Seventy-five percent (n = 9) of the female SBC subjects talked about their suicidal ideation during the incident while 25% (n = 3) did not. Of these communications (n = 9), 78% (n = 7) of them did refer to SBC specifically while 22% (n = 2) did not. Seventy- eight percent (n = 7) communicated their suicidal ideation to police officers, 11% (n = 1) to family members, and to others in 11% (n = 1) of the cases. Among female SBC subjects who survived the incident (n = 9), 78% (n = 7) admitted afterward that they were suicidal during the incident. Fourteen percent (n = 14) of the male subjects who survived their SBC incident (n = 98) admitted after- ward that they were suicidal during the incident.

Suicide notes were reported in 25% (n = 3) of the women- involved SBC cases; 75% (n = 9) of the subjects apparently left no note. None of the suicide notes made specific mention of SBC as a method.

Mental Health History

One hundred percent (n = 12) of the female SBC subjects had a confirmed or probable mental health history. Sixty-seven percent (n = 8) of the confirmed mental health subjects were clinically judged by the researchers to be suffering from depression or some form of mood disorder, and 50% (n = 6) had a prior substance abuse history. The limitations of the absence of direct clinical inter- views by the authors should be known, and all mental health find- ings should be considered tentative.

Seventeen percent (n = 2) of the SBC subjects had a prior known suicide attempt, but none had ever attempted SBC on a prior occasion. Fifty percent (n = 6) had prior documented

suicidality. Fifty percent (n = 6) had a prior reported psychiatric hospitalization. Forty-two percent (n = 5) of the subjects were spe- cifically described as psychotic (utterances that indicated delusions and ⁄ or hallucinations) at the time of the event, 58% (n = 7) were apparently under current mental health care, and 67% (n = 8) had prescribed psychotropic medications. There was no way to deter- mine whether those on medications were compliant nor whether they were being prescribed the proper medication for their particu- lar condition.

Intoxication and Use of Substances

Fifty-eight percent (n = 7) of the 12 SBC female subjects were under the influence of any substance at the time of the incident, and 42% (n = 5) of the female SBC subjects were under the influ- ence of alcohol. Seventy-five percent (n = 3) of the four women involved in regular OIS were alcohol intoxicated. Sixteen percent of the SBC female subjects (n = 2) were under the influence of methamphetamine or cocaine during the incident, compared with one of those subjects deemed regular OIS who had ingested cocaine.

Subject’s Status at the Time of the Incident

At the time of the incident, 100% (n = 12) of the female SBC subjects reportedly experienced recent behavioral changes, 75% (n = 9) relationship problems, 58% (n = 7) had just experienced a relationship ending, 33% (n = 4) had health issues, 17% (n = 2) had current criminal problems, and 17% (n = 2) were involved in child custody issues. Thirty-three percent (n = 4) had a prior crimi- nal history, 17% (n = 2) had been incarcerated before, and 33% (n = 4) had a prior violence history.

Incident Context

Fifty-eight percent (n = 7) of the incidents were apparently unplanned and spontaneous (subject did not apparently choose to initiate the incident that day, but rather became acutely suicidal in response to intervention and circumstances), while 42% (n = 5) were deliberate. Cases were further categorized using a modified version of Homant and Kennedy’s (2000) typology that splits the criminal category into three subtypes. Forty-two percent (n = 5) were Criminal Intervention Domestic Violence, 33% (n = 4) were Disturbed Intervention, 17% (n = 2) were Direct Confrontation, 8% (n = 1) of the subjects were involved in Criminal Intervention Major Crime, and 0% (n = 0) were Criminal Intervention Minor Crime.

Outcomes

Seventy-seven percent (n = 7) of the surviving women (n = 9) were arrested and ⁄ or convicted of a criminal offense after the inci- dent. Twenty-two percent (n = 2) received mental health intervention.

Table 1 compares the male and female samples of SBC cases on select variables, a few of which indicate significant differences. The effect size of these differences is also noted.

Discussion

The fact that 57% of the female subjects involved in OIS cases were determined to be SBC and an additional 24% were suicidal during their law enforcement confrontation indicates that there is a

666 JOURNAL OF FORENSIC SCIENCES

very high likelihood of a suicidal motivation among women who become involved in police shootings. Eighty-one percent of women who ended up in an OIS case evidenced suicidal motivation. While the substantial frequency of SBC among OIS cases was noted in Mohandie et al. (2009), these data suggest that police shootings involving women need even closer scrutiny for the issue of suicidal motivation.

Like the male subjects involved in SBC, the intent to be killed by the police officer occurs during the incident in most cases and then appears to be deliberately carried out. As we wrote earlier, ‘‘The paradox among SBC cases is that unplanned, acute suicidality becomes, within moments, a resolute intentionality to be killed by the police once the engagement begins’’ ([1], p. 461). Women, like men, usually do not plan a lethal encounter with the police before- hand, but given their preexisting suicidal ideation, and sometimes intent, the decision to utilize the police as their weapon of choice appears to be made, and often articulated, during the confrontation. However, while women are usually spontaneous in their SBC intent, it is noted that women were in fact more likely than men to deliberately create an SBC situation (see Table 1).

Women were half as likely as men to die in their SBC events, replicating the finding from other more general suicide studies that men are more likely to complete their suicide. They were quite likely to experience injuries, reflecting the dangerous nature of these encounters for them. While no women in this sample actually killed a police officer or civilian, 25% actually did inflict injuries upon others (police or civilians), half of them attempted violence toward police or civilians, all of them possessed weapons, and a third of those who had a firearm actually fired it at the police. These preliminary findings are consistent with the body of research demonstrating that men are more violent, but underscore that a sui- cidal woman can and will pose a risk to others in her quest to commit suicide.

Female SBC subjects appeared to communicate more about their suicidal intentions before, during, and after the incident. They appeared more likely to leave a suicide note than males. This likely reflects gender differences in willingness to discuss personal feel- ings, and may also represent a greater involvement in personal rela- tionship issues at the time of the event. Relationship issues could provide a potential dialogue strategy for intervention by first responders or police negotiators. However, the average duration of an incident was 1 h but the most frequently observed duration was 2 min. Time is a necessary requirement for effective dialogue by first responders and negotiators (negotiation teams usually take at least an hour to be deployed). Many of these subjects remove this factor from the equation by quickly raising the stakes and appear- ing to become an imminent threat to others. Relationship, health issues, and recent behavioral changes seemed to be common pre- ludes to the SBC event. Contextually, many of the events repre- sented an escalating culmination of domestic violence-related criminality on the part of the woman. More women than men were currently being treated for mental health issues at the time of their encounter with law enforcement, most often for a mood disorder with psychotic features. It is noted that diagnostic accuracy is a potential limitation because of the lack of direct evaluation of the subjects by the researchers and the exclusive reliance upon the available data in the law enforcement files. Future research should further explore these issues in a more reliable and valid manner.

Table 1 indicates the similarities and differences between male and female SBC subjects. The general findings are notable for two significant differences: women are more likely to be suicidal in a lethal or potentially lethal force encounter with the police; and despite the fact that most subjects of either gender arrive at their

TABLE 1—Comparison of male and female suicide by cop (SBC) subjects on select variables.

Total Sample Male Subjects

(n = 685) Female Subjects

(n = 21)

General findings SBC percentage by gender Male SBC Female SBC

36% (n = 243) 57% (n = 12) SBC ⁄ suicidal combined* Male SBC ⁄ suicidal Female SBC ⁄ suicidal

40% (n = 274) 81% (n = 17) SBC fatality Male SBC Female SBC

59% (n = 144) 25% (n = 3) SBC injury Male SBC Female SBC

38% (n = 93) 67% (n = 8) Total casualty Male SBC Female SBC

97% (n = 237) 92% (n = 11) Mean number of rounds fired by police

Male SBC Female SBC 16 (range 0–614) 4 (1–16)

Average age 34 (range 16–76) 40 (range 19–54) Has children 33% (n = 81) 50% (n = 6) Spontaneous vs. deliberate event�

82% (n = 198) 58% (n = 7) 16% (n = 38) 42% (n = 5)

SBC type Direct confront 16% (n = 40) 17% (n = 2) Criminal 48% (n = 116) 8% (n = 1) Disturbed 19% (n = 47) 33% (n = 4) Domestic 16% (n = 39) 42% (n = 5)

Incident duration Mean 3.2 h 1 h Mode 10 min 2 min Range 1 min–9 days 2 min–5 h

Suicidal communication Any suicidal communication 86% (n = 210) 92% (n = 11) Prior 55% (n = 133) 58% (n = 7) Prior w ⁄ SBC reference 39% (n = 52) 14% (n = 1) During 60% (n = 147) 75% (n = 9) During w ⁄ SBC reference 79% (n = 116) 78% (n = 7) Postincident admission 32% (n = 32)

(99 survivors) 78% (n = 7) (9 survivors)

Suicide note 14% (n = 34) 25% (n = 3) Behavior

Suicidal behavior during 31% (n = 75) 58% (n = 7) Weapon possession 79% (n = 192) 100% (n = 12) Gun possession 48% (n = 116) 50% (n = 6) Fake ⁄ simulated weapon 11% (n = 26) 0 Gun fired at police 49% (n = 57) 33% (n = 2) Violence toward civilians 49% (n = 118) 50% (n = 6) Violence toward police 63% (n = 154) 50% (n = 6)

Mental health issues Current meds 26% (n = 64) 67% (n = 8) Prior suicidality 29% (n = 70) 50% (n = 6) Prior SBC attempt n = 10 0 Probable ⁄ confirmed 60% (n = 146) 100% (n = 12) Mood disorder 28% (n = 68) 67% (n = 8) Psychotic 19% (n = 46) 42% (n = 5) Under the influence 64% (n = 155) 58% (n = 7) Current treatment 19% (n = 46) 58% (n = 7)

Recent issues Recent behavioral changes 67% (n = 164) 100% (n = 12) Relationship conflict 58% (n = 141) 75% (n = 8) Relationship ended 33% (n = 79) 58% (n = 7) Health issues 16% (n = 38) 33% (n = 4) Criminal problems 35% (n = 85) 17% (n = 2)

Outcome Death to others (police ⁄ civilians)

5% (n = 12) 0

Injury to others (police ⁄ civilians)

27% (n = 66) 25% (n = 3)

Arrest ⁄ convicted (survivors) 70% (n = 69) 78% (n = 7) Mental health 5% (n = 5) 22% (n = 2)

Because of the small sample size of the female SBCs and assumed heter- ogeneity of variance between the two groups, no additional statistical analy- ses were attempted.

*Sign v2 = 14.10, p < 0.001, / = 0.14. �Sign v2 = 5.207, p < 0.05, / = 0.15.

MOHANDIE AND MELOY • SUICIDE BY COP AMONG FEMALE SUBJECTS 667

SBC intent quite spontaneously during the encounter, women are more significantly likely to deliberately plan a suicidal encounter with the police beforehand. However, both of these significant find- ings have a very small effect size.

Suicidal behavior and communications are strikingly similar, and gender does not appear to affect these variables in ways that would have operational significance for the police. Again, we emphasize the positive relationship between homicidal risk and suicidal risk, and the caution that should be exercised when managing such a field encounter. The mental health differences, although not signifi- cant, generally reflect the finding that women are more amenable to mental health care and typically will be more compliant with treatment than men. Precipitants are more heavily weighed toward relationships for the women than the men.

Our study only has a total of 21 female subjects which cautions against the use of inferential statistical comparisons. Nonetheless, this is the first exploratory study examining the issue of SBC among female subjects, as well as women who become involved in OIS cases. Further study with larger samples may help generalize these findings or identify other important issues we have not addressed. However, despite our small sample, there is some exter- nal validity to our findings given the many police jurisdictions that were utilized to gather the data. Suicidal females who choose such a lethal method to attempt or commit suicide are extremely rare, but when a mentally disturbed female subject with a weapon con- fronts a police officer, her likely intent is to die at the hands of another.

Acknowledgments

We would like to thank the following individuals for their valuable contributions to this research: Detective Paul Delhauer,

Captain Clay Renard, Commander Armando Guzman, Lieute- nant Mike Hurley, Trish Walters, Jenn Williams, Jenna Blashek, Lieutenant Lewis Roberts, Sergeant Matt Youngblood, Sergeant John Yarbrough (retired), Russ Moore, and William Kidd. We owe special thanks and gratitude to Mila Green, Ph.D., without whom this research could not have been completed.

References

1. Mohandie K, Meloy JR. Clinical and forensic indicators of ‘‘suicide by cop.’’ J Forensic Sci 2000;45:384–9.

2. Mohandie K, Meloy JR, Collins P. Suicide by cop among officer involved shooting cases. J Forensic Sci 2009;54:1–7.

3. Homant RJ, Kennedy DB. Suicide by police: a proposed typology of law enforcement officer assisted suicide. Policing: Int J Police Strategies & Management 2000;23:339–55.

4. Homant RJ, Kennedy DB, Hupp RT. Real and perceived threat in police officer assisted suicide. J Crim Justice 2000;28:43–52.

5. Wilson EF, Davis JH, Bloom JD, Batten PJ, Kamara SG. Homicide or suicide: the killing of suicidal persons by law enforcement officers. J Forensic Sci 1998;43:46–52.

6. Hutson HR, Anglin D, Yarbrough J, Hardaway K, Russell M, Strote J, et al. Suicide by cop. Ann Emerg Med 1998;32:665–9.

7. Haruff RC, Llewellyn AL, Clark MA, Hawley DA, Pless JE. Firearm sui- cides during confrontations with police. J Forensic Sci 1994;39:402–11.

8. Moscicki EK. Epidemiology of suicidal behavior. Suicide Life Threat Behav 1995;25:22–35.

Additional information and reprint requests: Kris Mohandie, Ph.D. Operational Consulting International, Inc. P.O. Box 88 Pasadena, CA 91102 E-mail: [email protected]

668 JOURNAL OF FORENSIC SCIENCES