U.S. Department of Justice Office of Justice Programs Office of Juvenile Justice and Delinquency Prevention

Jeff Slowikowski, Acting Administrator February 2009

Office of Justice Programs Innovation • Partnerships • Safer Neighborhoods www.ojp.usdoj.gov Characteristics of Juvenile in Confinement A Message From OJJDP Suicide is always tragic, but it is par- ticularly so when the victim is young. The tragedy of young lives cut short Lindsay M. Hayes by suicide poses a significant public health challenge. According to data According to the Surgeon General of the (i.e., juvenile detention centers, reception from the Centers for Disease Control , youth suicide is a national centers, training schools, ranches, camps, and Prevention, suicide is the third tragedy and a major public health prob­ and farms). leading cause of death among youth lem (Carmona, 2005; U.S. Department of 15 to 24 years old. The study identified 110 juvenile Health and Human Services, 1999). The While experts recognize the need to suicide rate of young people (ages 15 to occurring between 1995 and 1999. Data were analyzed for the 79 cases that had intervene on behalf of vulnerable 24) tripled from 2.7 per 100,000 in 1950 to youth, little research has been con- complete survey information. Of these 79 9.9 per 100,000 in 2001 (Arias et al., 2003). ducted on the suicides of youth held suicides, 42 percent occurred in training More teenagers die from suicide than from in detention. To address this deficien- cancer, heart disease, AIDS, birth defects, schools and other secure facilities, 37 per­ cy, the Office of Juvenile Justice and stroke, pneumonia and influenza, and cent in detention centers, 15 percent in Delinquency Prevention has spon- chronic lung disease combined (U.S. residential treatment centers, and 6 per­ sored the first national survey of juve- Department of Health and Human Services, cent in reception or diagnostic centers. nile suicides in confinement. 1999). In addition, a national survey found The survey gathered descriptive data on This Bulletin examines 110 juvenile that more than 3 million youth are at risk the demographic characteristics and suicides that occurred in confinement for suicide each year, with 37 percent social history of each victim, the charac­ between 1995 and 1999. It describes of surveyed youth reporting that they teristics of the incident, and the features the demographic characteristics and attempted suicide during the previous 12 of the juvenile facility in which the suicide social history of victims and examines months (Substance Abuse and Mental took place. Particular attention was paid the characteristics of the facilities in Health Services Administration, 2001). to each facility’s implementation of sui­ which the suicides took place. Draw- ing on this data, the researchers offer Although youth suicide in the general pop­ cide prevention programming. This Bul­ letin presents findings from the survey recommendations to prevent suicides ulation has been identified as a significant in juvenile facilities. public health problem, juvenile suicide in and offers recommendations for address­ confinement has received little attention. ing this tragic problem. The findings reported in these pages The Office of Juvenile Justice and Delin­ present serious challenges for health- quency Prevention (OJJDP) awarded a Victim’s Demographic care and correctional professionals contract to the National Center on Institu­ who work with confined youth and for tions and Alternatives to conduct the first Characteristics administrators charged with ensuring national survey of juvenile suicides in con­ the security and safety of youth in finement. The primary goal of this effort Race, Sex, and Age detention. Preventing juvenile sui- cides in confinement is a critical was to determine the extent and distribu­ More than two-thirds of the suicide victims responsibility. The information provid- tion of juvenile suicides in confinement identified in the survey were Caucasian. ed in this Bulletin is intended to inform such endeavors.

Access OJJDP publications online at www.ojp.usdoj.gov/ojjdp About the Survey In August 1999, the National Center on confinement—juvenile detention centers, housing assignment, (e.g., single or Institutions and Alternatives was award- reception centers, training schools, ranch- multiple occupancy) room confine­ ed a grant from the Office of Juvenile es, camps, and farms—operated by state ment status, method and instrument Justice and Delinquency Prevention to and local governments and private organi- used, time span between last con- conduct the first national survey on zations.d Excluded from the project were tact and finding victim, and possible juvenile suicide in confinement.a The open, physically unrestricted residential precipitating factors of the suicide. primary goal of the project was to deter- programs for juveniles such as shelters, mine the extent and distribution of juve- halfway houses, and group homes. Phase ◆ Facility characteristics including facil­ nile suicides in confinement and to 1 identified 110 juvenile suicides occur- ity type, facility ownership (e.g., gather descriptive data on the demo- ring between 1995 and 1999. The sui- state, county, private), capacity/pop­ graphic characteristics of each victim, cides were distributed among 38 states. ulation at time of suicide, and suicide the characteristics of the incident, and prevention components in use (writ­ the characteristics of the juvenile facility Phase 2 ten policy, intake screening, staff that sustained the suicide. A report of training in and Once facilities that had experienced a sui- the survey’s findings would serve as a cardiopulmonary resuscitation, resource for juvenile justice practition­ cide during the 5-year study period were identified, phase 2 of the survey process observation levels, safe housing, and ers to expand their knowledge and for mortality review). juvenile correctional administrators to was initiated. It included dissemination of create and/or revise policies and train- a seven-page survey instrument to the The phase 2 survey instruments and ing curriculums on suicide prevention. directors of the facilities that sustained cover letters were mailed to directors of Data collection occurred in two phases. suicides. The survey instrument was the 83 facilities that sustained the 110 designed to collect readily available data suicides. Respondents provided com­ on three types of characteristics: Phase 1 pleted surveys on 79 suicides. ◆ During phase 1, a one-page survey Demographic and other victim charac- instrument and cover letter was sent teristics including age, sex, race, living Data Limitations to directors of 1,178 public and 2,634 arrangement, current offense(s), prior Given the epidemiological data regard- private juvenile facilities in the United offense(s), legal status (detained, com- ing youth suicide in the community, States.b Each of the 3,812 facility mitted, other), length of confinement, coupled with the increased risk factors directors was asked to complete the drug/alcohol intoxication at confinement, associated with confinement, the survey if the facility experienced a juve- history of room confinement, substance reported number of suicides in this nile suicide between 1995 and 1999.c abuse history, medical/mental health study would appear low. However, this Similar to OJJDP’s Conditions of Con- history, physical/ history, study identified more deaths per year finement study (Parent et al., 1994), the and history of suicidal behavior. than a contemporary national census project surveyed facilities that housed of juvenile facilities (OJJDP, 2002), ◆ juveniles in more traditional types of Incident characteristics including date, and many experts believe that facility time, and location of suicide, and

This finding is consistent with suicides 38 percent and 2 percent, respectively, of Living Arrangement Before that occur each year in the general popu­ the confined juvenile population through­ Confinement out the country, but 68 percent and 11 lation (Arias et al., 2003). One previous More than a third (38 percent) of suicide percent of the victims in this study. The study found that Caucasian youth held in victims were living with one parent at the causes of these disproportionate relation­ detention attempted suicide at a rate time of their confinement. Slightly less ships are outside the purview of this approximately 3.5 times that of African than one quarter (23 percent) of the vic­ analysis. American youth (Kempton and Forehand, tims were living with both parents. Other 1992). Although African American and A substantial majority (80 percent) of living arrangements included community Hispanic youth comprised approximately the victims were male. Given the fact that placement (11 percent), other relative 39 percent and 18 percent, respectively, of more than 80 percent of all juveniles (9 percent), foster parent or guardian the confined juvenile population through­ confined in the United States are male (8 percent), or adoptive parents (5 per­ out the country (Sickmund and Wan, (Sickmund and Wan, 2001), these findings cent). Two victims were living on their 2001),1 they represented only 11 percent are not surprising. own (3 percent), and the living arrange­ and 6 percent of the victims in this study. ments of the other three victims were Caucasian and American Indian youth, on More than 70 percent of the victims were unknown (4 percent). the other hand, comprised approximately between the ages of 15 and 17 (figure 1). The average (mean) age was 15.7, with one victim as young as 12 and another as 1. For comparative purposes, data collected from old as 20. These findings are consistent OJJDP’s Census of Juveniles in Residential Placement with data from the Census of Juveniles in was limited to the following: gender, age, race, place­ Residential Placement (Sickmund and ment authority, most serious offense charged, and Wan, 2001). adjudication status.

2 “self-reporting” of juvenile suicides in regulatory agency should be cause for c. To encourage a high rate of response, the cover custody results in underestimates of great concern within the juvenile justice letter was co-signed by officials of both the National the problem (Sullivan, 1995; Twedt, community. Juvenile Detention Association and the Council of 2001b). Despite concerted efforts by Juvenile Correctional Administrators, and business project staff to locate all possible For More Information reply envelopes were enclosed with the survey juvenile suicides during the 5-year instruments. study period, whether every death For more information about the survey d. By definition, detention centers hold juveniles for was identified remains uncertain. methodology, including copies of the phase 1 and phase 2 survey instruments, short periods of time in a physically restrictive envi- Approximately 13 percent of the see the OJJDP report, Juvenile Suicide in ronment pending juvenile court action, or following reported suicides in this study were Confinement: A National Survey (Hayes, adjudication pending disposition, placement, or trans- identified through nontraditional 2006). The report is available on the fer. Reception centers are short-term facilities that sources (including newspaper articles OJJDP Web site www.ojp.usdoj.gov/ojjdp. hold juveniles committed by courts and conduct and the project director’s consultation screening and assessment to assign them to appro- with facilities sustaining the deaths). Notes priate facilities. Training schools are long-term facili­ In addition, more than one-third of the ties in which treatment and programming are provid­ reported suicides were unknown to a. The National Center on Institutions and Alternatives ed in an environment with strict physical and staff any state agency (e.g., departments was assisted on the project by two prominent national controls. Ranches, camps, and farms are long-term of juvenile corrections or agencies juvenile justice organizations (the National Juvenile residential facilities that do not require the strict responsible for licensing and regulato- Detention Association and the Council of Juvenile Cor- confinement of a training school, often allowing ry services). Most of the deaths that rectional Administrators) and a consultant team com­ offenders greater contact with the community. This were unknown to state agencies posed of four prominent juvenile justice practitioners and last category includes “residential treatment centers” occurred in either county detention researchers (G. David Curry, Ph.D., Robert E. DeComo, and “boot camps.” centers or private residential treat­ Ph.D., Barbara C. Dooley, Ph.D., and David W. Roush, e. Although the study found that 27 percent of the ment centers.e Many of the reported Ph.D.). In addition, Cedrick Heraux, a doctoral student at total number of suicides (N=110) occurred in private suicides in this study were also Michigan State University, provided both data entry and facilities, many of which were residential treatment unknown to many child advocacy data analysis support to the project. centers, two-thirds (67 percent) of private facilities did agencies. The fact that any suicide b. Facilities were identified through OJJDP’s Census of not respond to survey requests. occurring within a juvenile facility Juveniles in Residential Placement (OJJDP, 1999). A throughout the United States could small percentage of facilities were either closed or could remain outside the purview of a not be located, and thus presumed to be closed.

Victim’s Offense and Confinement Status Figure 1: Suicides in Juvenile Facilities 1995–1999, by Age of Victim Most Serious Offense2 35 A significant majority (70 percent) of victims were confined for nonviolent (i.e., 30

25 2. For purposes of this study, offenses were broken down into six categories: property offenses included burglary, grand larceny, petty larceny, auto theft, rob­ bery (other), receiving stolen property, shoplifting, 20 arson, breaking and entering, entering without break­ ing, counterfeiting, forgery, embezzlement, vandalism, and carrying a concealed weapon; person offenses 15 included murder, negligent manslaughter, armed rob­ bery, rape, indecent assault, assault, battery, sexual assault, aggravated assault, and kidnapping; status 10 offenses included running away, truancy, incorrigibility, Victim s of Percent curfew violation, and loitering; probation violation offenses included any technical violation of the terms 5 of probation and/or parole; public order offenses included alcohol-related charges (intoxication, liquor law violation, driving under the influence), resisting 0 arrest, disorderly conduct, prostitution, sex offenses 12 13 14 15 16 17 18 19 20 (other), vagrancy, unauthorized use of a motor vehi­ cle, and minor traffic offenses; and drug offenses Age included possession, use, and distribution of any dangerous controlled substance or narcotic.

3 nonperson) offenses, with property offenses accounting for the highest per­ Figure 2: Suicides in Juvenile Facilities 1995–1999, by Victim’s Most centage of victim confinement (figure 2). Serious Offense In addition, the status, probation viola­ tion, and public order categories com­ 35 bined represented more than a third (34 percent) of the offenses. Person offenses accounted for 30 percent of victim con­ 30 finement; only 3 percent of victims were confined on drug offenses. Approximately 40 percent (13 of 33) of victims housed in 25 a training school or other secure facility were confined for a person offense. 20 With only slight variance, these findings were consistent with data on the confined juvenile population throughout the coun­ 15 try. For example, person offenses account­ ed for 35 percent and property offenses 10 accounted for 29 percent of all confined Victim s of Percent juveniles throughout the country (Sick­ mund and Wan, 2001). However, whereas 5 the status, probation violation, and public order categories combined represented 27 percent of all confined juveniles, these 0 categories represented 34 percent of the Property Person Status Probation Public Drug Offense Offense Offense Violation Order Offense victims in this study. At confinement, 39 percent of victims had a second charge. Property offenses accounted for the majority (52 percent) of the additional charges, followed by school/secure facility victims were com­ person offenses (19 percent). Status, pro­ mitted at the time of death. Figure 3: Suicides in Juvenile bation violation, and public order offenses Facilities 1995–1999, by combined represented 29 percent of addi­ Less than 4 percent of juvenile suicides tional charges. occurred within the first 24 hours of Victim’s Confinement Status confinement (and all of these deaths A substantial majority (79 percent) of sui­ occurred in detention centers). This find­ cide victims had prior offenses. Of the vic­ ing significantly differed from a national tims who had a history of offenses, most study on jail suicides that found more 32.9 percent committed crimes of a nonviolent nature, than 50 percent of suicides took place with property offenses the most common within the first 24 hours, with almost a (50 percent). Status, probation violation, third occurring within the first 3 hours and public order offenses combined repre­ (Hayes, 1989). The deaths reported in this 67.1 percent sented 23 percent of the most serious national survey of juvenile suicide in con­ prior offenses; person offenses accounted finement were distributed fairly evenly for 23 percent of victims’ prior offenses. during a more than 12-month period. For example, the same number of suicides Confinement Status (13) occurred within the first 3 days of Committed Two-thirds (67 percent) of victims were confinement as occurred after more than 3 Detained committed at the time of death (figure 3). 10 months of confinement. Nearly a third This finding was significantly different (32 percent) of suicides occurred within 1 from a national study on jail suicides that to 4 months of confinement. However, all found the overwhelming majority of vic­ detention center suicides occurred within the first 4 months of confinement, with facility suicides occurred 3 months or tims were on detention status at the time more following confinement.4 of death (Hayes, 1989). The finding was, more than 40 percent occurring within the however, somewhat consistent with first 72 hours, while a significant majority (73 percent) of training school/secure national data on confined juveniles that 4. For comparative purposes, although lengths of stay found 74 percent of youth were commit­ within juvenile facilities throughout the country vary ted (Sickmund and Wan, 2001). Not sur­ considerably, earlier OJJDP research has shown the average length of stay in the four facility types to be prisingly, the vast majority (79 percent) of 3. The average length of confinement prior to suicide victims held in detention centers were for the 10 victims who died after more than 12 months as follows: detention center (15 days), training school in custody was 21.8 months. or other secure facility (7.5 months), reception or on detention status and all training diagnostic center (34 days), and residential treatment center (6.5 months) (see Parent et al., 1994).

4 Victim’s History Slightly more than one-third of victims Figure 4: Suicides in Juvenile Facilities 1995–1999, by Victim’s History had a history of physical abuse (figure 4), with an immediate family member (e.g., 100 father or stepfather) as the perpetrator in the majority of cases (20 of 27). The 90 victim’s history of physical abuse was unknown in approximately 20 percent 8 0 of cases (see “Unknown Responses and

s 70 Detention Centers”). More than one-quarter (28 percent) of vic­ 60 tims had a history of sexual abuse, with an equal number of victims whose history 50 of sexual abuse was unknown. For those who were abused, an immediate family 40

member (e.g., father or stepfather) was Victim ercent of P the perpetrator in many cases. 3 0 Somewhat less than half the victims (44 20 percent) had a history of emotional abuse. The most frequent types of abuse were 10 excessive punishment, neglect/abandon­ 0 ment, verbal abuse, and other types of ual Prior Other family dysfunction. The victim’s history Physicael Sex e e e Mental Emotional Substance Suicidal of emotional abuse was unknown in Abus Abus Abus Abus Illness vior Mediclemal s almost one-quarter of the cases. Beha Prob Component A significant majority (73 percent) of vic­ tims had a history of substance abuse. Ye s No Unknown Approximately one-third of victims with a substance abuse history used alcohol, marijuana, or cocaine before their confine­ ment. This finding was consistent with percentage of psychiatric disorders) was Suicide Incident other recent data suggesting that two- considered separately from mental Characteristics thirds of confined youth have one or more disorders. alcohol, drug, or mental disorders (Teplin More than two-thirds (70 percent) of Date and Time et al., 2002). victims had a history of suicidal behavior. Contrary to common belief, certain sea­ Many suicide victims (66 percent) had a The most frequent type of suicidal behav­ sons of the year and holidays did not cor­ history of mental illness, and a majority ior was (46 percent), relate with a higher number of suicides. (65 percent) of victims with a history of followed by /threat (31 Although more than 30 percent of all mental illness were suffering from depres­ percent), and suicidal gesture/self-mutila­ deaths occurred during the months of Jan­ sion at the time of death. Other mental tion (24 percent). Although other research uary and May, suicides were distributed illnesses reported included attention shows that between 8 percent and 52 per­ throughout the year. Further, no statisti­ deficit/hyperactivity disorder, conduct dis­ cent of confined youth have a history of cally significant difference existed regard­ order, post-traumatic stress disorder, and suicidal behavior (see “Self-Injurious ing the day of the week on which suicides psychotic disorder (54 percent of victims Behavior in Juvenile Facilities,” on page occurred. were taking psychotropic medication at 7), findings from this national survey Research on jail suicide has found that time of death).5 Although other research suggest that confined youth who died by deaths are more prevalent when staff also indicates that a high percentage of suicide have a higher incidence of prior supervision is reduced. For example, less youth in the juvenile justice system suffer suicidal behavior than those confined than 20 percent of deaths in a national from at least one mental disorder and youth who engage in suicidal behavior study of jail suicides occurred during the have higher rates of mental disorders than but do not die. Suicide victims in deten­ 6-hour period between 9 a.m. and 3 p.m., youth in the general population (Cocozza tion centers, compared with other facility a major portion of the day shift (Hayes, and Skowyra, 2000), it should be noted types, were less likely to have a known 1989). As shown in figure 5, findings from that, in the current study, substance abuse history of suicidal behavior (52 percent this study indicate that 71 percent of sui­ disorder (which accounts for a sizable versus 80 percent in all other facilities). cides occurred during traditional waking Few (19 percent) victims had a known his­ hours (6 a.m. to 9 p.m.), whereas 29 per­ tory of other medical problems. Allergies cent of suicides occurred during tradition­ 5. For the most part, survey respondents did not and asthma were common types of medical report victims’ mental illnesses according to the Diag­ al nonwaking hours (9 p.m. to 6 a.m.). In nostic and Statistical Manual (DSM) III or IV editions problems found in the few victims with addition, approximately half (51 percent) (American Psychiatric Association, 2000). documented medical conditions. of suicides occurred during the 6-hour

5 Unknown Responses and Detention Centers A high percentage of unknown respons- Detention Association, 1990:1). Because appropriately trained and licensed spe­ es to survey questions relating to sever- of the lack of available community cialists” (National Juvenile Detention al personal characteristics of the victims resources, and due to their unique ability Association, 2001). More importantly, (including histories of substance abuse, to provide physical custody, detention the findings suggest that the significant medical problems, emotional abuse, centers often bear the responsibility for deficiencies in intake screening and physical abuse, sexual abuse, and men- troubled youth. However, these centers overall suicide prevention programming tal illness) came from detention center are both ill-equipped and underresourced within detention centers experiencing respondents.* In addition, suicide vic- to provide anything more than basic suicides warrant immediate attention. tims housed in detention centers had a healthcare services on a short-term basis. Resources need to be channeled to all lower percentage of reported histories Although the temporary nature of the juvenile facilities throughout the country, of suicidal behavior, suggesting that detention center experience may help to particularly detention centers, to ensure perhaps these facilities fail to inquire explain some of the survey findings that any agency housing a juvenile pro- about such history. Finally, although the regarding these types of facilities, such a vides basic, yet comprehensive, suicide study found that many facility types distinction should not be viewed as a miti- prevention programming. lacked comprehensive suicide preven- gating factor for suicide prevention. All tion programming at the time of the sui- juvenile facilities, regardless of size and cide, detention centers had the lowest mission, have a responsibility for the safety * Communication among agencies also appeared to percentage of such programming of all their youth, including those at risk be a problem in many cases. Surveys were received (approximately 10 percent). for self-harm. from several detention centers in which respondents According to the National Juvenile The findings from this study support the complained that they had been temporarily “holding” Detention Association, juvenile deten- National Juvenile Detention Association’s the victim for another jurisdiction (e.g., state correc- tion is defined as “the temporary and position that youth with severe mental ill- tional facility, probation office) and knew little, if any­ safe custody of juveniles who are ness should be provided services in “the thing, about the youth. As one facility director stated, accused of conduct subject to the juris- appropriate therapeutic environment . . . “I do not know the answers to some of these ques­ diction of the court who require a when juvenile detention facilities are tions because the child was not from our county. He restricted environment for their own and forced to house youth with severe mental was being housed here in a state-contract bed.” the community’s protection while pend- health issues, [the Association] promotes ing legal action” (National Juvenile the provision of adequate services by

period between 6 p.m. and midnight, and a national study on jail suicides that found circumstances that led to room confine­ almost a third (29 percent) occurred more than 60 percent of adult suicide vic­ ment included threat or actual physical between 6 p.m. and 9 p.m. tims were intoxicated at the time of their abuse of staff or peers (41 percent), ver­ suicide (Hayes, 1989). bal abuse of staff or peers (26 percent), Method, Instrument, and failure to follow program rules or inappro­ Anchoring Device Room Assignment priate behavior (26 percent), and other (7 percent), which included two cases of The survey found that all but one victim Three-quarters (75 percent) of victims youth involved in gang activity and one used hanging as the method of suicide. were assigned to single occupancy rooms case of a standard protocol for new intake. (The sole victim of other means abscond­ at the time of suicide. The remainder (25 ed from the facility and ran in front of a percent) were assigned to multiple occu­ Approximately half the suicide victims were passing train.) The vast majority (72 per­ pancy rooms. No significant differences on room confinement status at the time of cent) of these victims (n=78) used bedding existed between room assignments and death.6 Compared to other facility types, a (e.g., sheet, blanket) as the instrument to the types of facilities where the suicides much smaller percentage (17 percent) of hang themselves. Clothing (13 percent), occurred. suicide victims housed in residential treat­ belts (5 percent), and shoelaces (5 per­ ment centers were on room confinement cent) were used to a lesser degree. Other Room Confinement status at the time of death. instruments included a towel and a bag. For purposes of this study, room confine­ Suicide victims used a variety of anchor­ In addition, 85 percent of victims who ment was defined as a “behavioral sanc­ died by suicide while on room confine­ ing devices, including door hinges/knobs, tion imposed on youth that restricted air vents, bedframes, window frames, clos­ ment status died during waking hours (6 movement for varying amounts of time.” It a.m. to 9 p.m.), a percentage found to be et rods, sprinkler heads, toilets, sinks, and included, but was not limited to, isolation, television stands. segregation, time-out, or a quiet room. 6. Room confinement did not include youth The circumstances that led to room confinement Intoxication included failure to follow program rules or inappropri­ assigned to their room during traditional ate behavior (47 percent), threat of or actual physical None of the 79 victims was under the nonwaking hours (9 p.m. to 6 a.m.). abuse of staff or peers (42 percent), and other (11 per­ influence of alcohol or drugs at the time of cent), which included two cases of standard proce­ Most (62 percent) suicide victims had suicide. This finding is in stark contrast to dure for new intake, one case of court-ordered con­ a history of room confinement. The finement, and one case of group confinement during a shift change.

6 Self-Injurious Behavior in Juvenile Facilities Although little research has been con­ Chowanec et al. (1991) found higher rates behavior in incarcerated adolescents” ducted regarding youth suicide in cus­ of self-harm behavior among incarcerated (Esposito and Clum, 2002:145). tody, the information available suggests male youth than in the general adolescent a high prevalence of self-injurious community population. Findings from a 2002 study indicate behavior in juvenile correctional facili­ that more than half (52 percent) of all ties. For example, according to one Caucasian youth appear to attempt sui­ detained youth self-reported current national study, more than 11,000 juve­ cide in confinement at a higher rate than suicidal ideation, with 33 percent having niles are estimated to engage in more African American youth (Kempton and a history of suicidal behavior (Esposito than 17,000 incidents of suicidal behav­ Forehand, 1992; Alessi et al., 1984). and Clum, 2002). In addition, a study of ior in juvenile facilities each year (Par­ Morris and colleagues (1995) found that youth confined in a juvenile detention ent et al., 1994). In another national American Indian and Caucasian youth facility found that suicidal behavior in survey, conducted in 1991, a modified reported higher rates of suicidal ideation males was most significantly associated version of the Centers for Disease Con­ (29 percent and 25 percent, respectively) with depression, major life events (such trol’s Youth Risk Behavior Surveillance than Hispanic, Asian, and African Ameri­ as court involvement, death of a family System survey was administered to can youth (15 percent, 12 percent, and 8 member, etc.), poor social connections, more than 1,800 confined youth in 39 percent, respectively). Other researchers and past suicide attempts, whereas juvenile institutions throughout the have reported similar findings of high suicidal behavior in females was associ­ country (Morris et al., 1995). The study rates of suicidal behavior among Ameri­ ated with impulsivity, current depression, found that almost 22 percent of confined can Indian youth confined in juvenile facili­ instability, and younger age (Mace, youth seriously considered suicide, 20 ties (Duclos, LeBeau, and Elias, 1994). Rohde, and Gnau, 1997; Rohde, Seeley, and Mace, 1997). The most common percent made a plan, 16 percent made Several studies consistently found that at least one attempt, and 8 percent correlate among both males and females certain risk factors point to high rates of was not living with a biological parent were injured in a suicide attempt during suicidal behavior for incarcerated youth. the previous 12 months. before detention. Suicidal behavior of a For example, researchers have reported friend was significantly associated with Other studies found that large percent­ that confined youth with either a major past and current suicidal ideation among ages of detained youth had histories of affective disorder or borderline personality boys, but not girls (Rohde, Seeley, and suicide attempts (Dembo et al., 1990) disorder have a higher degree of suicidal Mace, 1997). and current suicidal behavior (Robert­ ideation and more suicide attempts than son and Husain, 2001; Shelton, 2000; adolescents in the general population Finally, a recent study of confined youth Davis et al., 1991; Woolf and Funk, (Alessi et al., 1984); incarcerated male referred for psychiatric assessment 1985). In fact, Robertson and Husain youth whose parents had affectionless found that 30 percent reported suicidal (2001) found that 31 percent of con­ bonding styles also reported more suicidal ideation/behavior and 30 percent report­ fined youth self-reported a suicide ideation and/or attempts (McGarvey et al., ed self-mutilative behavior while incar­ attempt, and 9 percent were currently 1999). Other researchers concluded that a cerated (Penn et al., 2003). These youth suicidal with either ideation and/or a history of sexual abuse “directly affects the reported more depression, anxiety, and plan to act on suicidal thoughts. Finally, development of suicidal ideation and anger than nonsuicidal confined youth.

higher than that of victims not on room actual prevalence of suicide in the types that overcrowding was a contributing confinement who died by suicide during of facilities that had the highest rates of factor to juvenile suicide. In fact, the the same hours (71 percent). nonresponses to the survey (e.g., private majority (68 percent) of suicides took programs). The 79 suicides were distrib­ place in facilities that were either at or uted among 70 juvenile facilities: 65 facili­ below bed capacity; an additional 10 per­ Facility Characteristics ties sustained a single suicide, 3 facilities cent of suicides occurred in facilities that and Response had 2 suicides each, 1 facility had 3 sui­ were slightly (less than 10 percent) over cides, and 1 facility had 5 suicides during capacity. Facility Type and Population the survey period. As previously indicated, this national sur­ Two-thirds (67 percent) of suicides Assessment by Qualified vey of juvenile suicides in confinement occurred in facilities with populations of Mental Health Professional found that 42 percent of juvenile suicides 200 or fewer youth, and 42 percent in facil­ National juvenile correctional standards took place in training schools and other ities with 50 or fewer youth7 (figure 6). The and standard correctional practice indi­ secure facilities, 37 percent in detention study did not find any evidence to suggest cate that confined youth should be centers, 15 percent in residential treat­ assessed as soon as possible by a quali­ ment centers, and 6.3 percent in recep­ 7. fied mental health professional (National tion or diagnostic centers. Almost half (48 The direction of this finding is somewhat consistent with earlier OJJDP research finding that approximately Commission on Correctional Health Care, percent) of the suicides occurred in facili­ 72 percent of juveniles are housed in facilities with 250 1995, 1999, 2004; Roush, 1996; Underwood ties administered by state agencies, 39 or fewer beds, although only 21 percent are housed in and Berenson, 2001), with Performance- percent in county facilities, and 13 per­ facilities with 50 or fewer beds (see Parent et al., based Standards requiring an assessment cent in private programs. These percent­ 1994). within 7 days of entry into the facility ages, however, may underestimate the

7 (Council of Juvenile Correctional Adminis­ trators, 2003).8 For the purposes of this Figure 5: Suicides in Juvenile Facilities 1995–1999, by Time of Suicide study, and consistent with national stan­ dards, a qualified mental health profes­ 35 sional was defined as an individual who by virtue of his or her education, creden­ tials, and experience is permitted by law 30 to evaluate and care for the mental health needs of patients. This definition includes, but is not limited to, psychiatrists, psy­ 25 chologists, clinical social workers, and psychiatric nurses. This examination by s 20 a qualified mental health professional is separate from an initial intake screening. A large majority (70 percent) of suicide 15 victims were assessed by a qualified men­ tal health professional. Compared to 10 other facility types, a significantly smaller Victim ercent of P percentage (35 percent) of suicide victims housed in detention centers received 5 mental health assessments. However, slightly more than half (52 percent) of the detention center victims died within the 0 first 6 days of confinement, thus reducing 3 a.m 6 a.m 9 a.m Noon 3 p.m. 6 p.m. Midnight p.m. 9 p.m. the opportunity for assessment. to 3 a.m. to 6 a.m. to 9 a.m. to Noon to 3 to 6 p.m. to 9 p.m. to Midnight Of those victims receiving a mental health assessment, almost half (49 percent) had Time a contact visit with a qualified mental health professional within 6 days of their Time Span death. However, 20 percent of assessed Figure 6: Suicides in Juvenile victims had not been assessed within 30 Approximately 41 percent of respondents days of their death and slightly less than stated that staff found the victim in less Facilities 1995–1999, by Facility half (44 percent) of all victims either had than 15 minutes following the last obser­ Population never been assessed by a qualified mental vation of the youth. However, slightly health professional or had not been more than 15 percent of victims were assessed within 30 days of their death. reported to be found more than an hour 17 percent 9 percent following the last observation, including 1 percent Suicide Precaution Status several victims found after 3 hours. In one case, the time span between the last obser­ 6 percent 25 percent A small percentage (17 percent) of youth vation and the suicide was not known. were on suicide precaution status at the time of death. Of these 13 victims, 10 were required to be observed at 15-minute Suicide Prevention 42 percent intervals; the 3 remaining youth were to be observed at continuous, 5-minute, and Programming 60-minute intervals, respectively. Despite For this national survey of juvenile suicide their identified risk of suicide, almost half in confinement, data were collected to 50 or fewer youth (6 of 13) of these victims were found to be determine whether facilities sustaining a last observed in excess of 15 minutes suicide had comprehensive suicide pre- 51–200 youth before their suicide. vention programming in place at the time 201–500 youth of the death. Consistent with national juvenile correctional standards, compre­ 501–1,000 youth 8. In 1995, OJJDP contracted with the Council of Juve­ hensive suicide prevention programming nile Correctional Administrators to develop, field test, More than 1,000 youth and implement performance-based standards for juve­ included the following seven critical com­ Unknown nile correctional and detention facilities. The Perfor­ ponents: written policy, intake screening, mance-based Standards Project offers a systematic training, CPR certification, observation, method for facilities to measure outcomes and pro­ safe housing, and mortality review (Hayes, vides guidance for facilities to review their practices 1999). and to take corrective action. As indicated in figure 7, a substantial majority (90 percent) of suicides occurred in facilities that implemented one or more suicide prevention components. However,

8 as shown in figure 8, only 20 percent of suicides occurred in facilities that imple­ Figure 7: Suicides in Juvenile Facilities 1995–1999, by Facility’s mented all seven suicide prevention com­ Implementation of Suicide Prevention Components ponents. The degree to which suicides occurred in a facility that had all seven 100 suicide prevention components varied con­ siderably by facility type: detention centers 90 (10 percent), training schools and other secure facilities (24 percent), reception or 80 diagnostic centers (40 percent), and resi­ dential treatment centers (25 percent). 70

60 Written Suicide Prevention Policy 50 Standard correctional practice and nation­ al juvenile correctional standards indicate 40 that juvenile facilities should have a writ­ ten suicide prevention policy that details 30 the identification and management of 20 suicidal youth (American Correctional Implementing Component

Association, 1991; Council of Juvenile acilities ercent of Suicides in F P 10 Correctional Administrators, 2003; Hayes, 1999; National Commission on Correction­ 0 al Health Care, 1995, 1999, 2004; Roush, Intake Staff Safe 1996). A significant majority (79 percent) Suicide Suicide Mortality Housing Review of suicides occurred in facilities that main­ Written Suicide Screening Prevention Certification Precaution Prevention Training CPR Protocol tained a written suicide prevention policy Policy at the time of the suicide, although this Component was less true for suicides that took place in detention centers (62 percent).

Intake Screening for Suicide Risk Figure 8: Suicides in Juvenile Facilities 1995–1999, by Number of Suicide Prevention Components Implemented by Facility Most (71 percent) suicides took place in facilities that maintained an intake screen­ 100 ing process to identify the suicide risk of youth entering the facility, although this 90 was less true (48 percent) of suicides in detention centers. This finding is consis­ 80 tent with OJJDP research suggesting that approximately 70 percent of confined 70 youth are screened for suicide risk (OJJDP, 2002). 60 50 Suicide Prevention Training More than 4 in 10 juvenile suicides (43 40 percent) occurred in facilities that did not provide any type of suicide prevention 30 training (pre-service, annual, and/or peri­ 20

odic) to their direct care staff. Implementing Component

Of the 45 suicides that occurred in facili­ of Suicides in Facilities Percent 10 ties that provided suicide prevention 0 training, two-thirds (67 percent) were in 0 1 2 3 4 5 6 7 facilities that provided annual instruction, with training schools and other secure Number of Suicide Prevention Components facilities providing the lowest percentage (42 percent) of annual training. Only 38 percent of juvenile suicides (30 of 79) took place in facilities that provided annual

9 Comprehensive Suicide Prevention Programming The findings from this survey suggest and mental health personnel); and (3) personnel, and begin life-saving that, although the rate of compliance between facility staff and the suicidal measures. with individual suicide prevention com­ youth. ◆ ponents was high, few juvenile facilities Staff should never presume that the that sustained a suicide had all the Housing. Isolation should be avoided. youth is dead, but rather initiate and components of a comprehensive suicide Whenever possible, suicidal youth should continue appropriate life-saving prevention program. Consistent with be housed in the general population, measures until relieved by medical national correctional standards and mental health unit, or infirmary, in close personnel. practices, all juvenile facilities, regard­ proximity to staff. Youth should be housed All housing units should contain a first less of size and type, should have a in suicide-resistant, protrusion-free rooms. aid kit, pocket mask or mouth shield, detailed written suicide prevention poli­ Removal of clothing (excluding belts and Ambu bag, and rescue tool (to quickly cy that addresses each of the following shoelaces) and use of restraints should cut through fibrous material). critical components (Council of Juvenile be avoided whenever possible, and should only be used as a last resort for Correctional Administrators, 2003; Reporting. In the event of an attempt­ short periods of time when the youth is Hayes, 1999, 2000; National Commis­ ed or completed suicide, all appropriate engaging in self-destructive behavior. sion on Correctional Health Care, 1999, facility officials should be notified 2004; Roush, 1996). Levels of supervision. Two levels are through the chain of command. All staff who came in contact with the victim Training. All facility, medical, and men­ normally recommended for suicidal youth: before the incident (or while responding tal health staff should receive 8 hours of ◆ Close observation—reserved for youth to the incident) should submit a state­ initial suicide prevention training, followed who are not actively suicidal, but ment as to their full knowledge of the by a minimum of 2 hours of annual express suicidal ideation and/or have youth and the incident. refresher training. Training should pro­ recent histories of self-destructive vide information about predisposing behavior and are now viewed as Followup/mortality review. All staff factors, high-risk periods, warning signs potentially suicidal—requires supervi­ (and youth) involved in the incident and symptoms, identifying suicidal should be offered critical incident stress sion at staggered intervals not to behavior despite the denial of risk, and debriefing. If resources permit, a psy­ exceed every 15 minutes. In addition, components of the facility’s suicide pre­ chological autopsy is recommended. a youth who denies suicidal ideation vention policy. Every completed suicide and serious or does not threaten suicide, but suicide attempt (i.e., requiring hospital­ Identification/screening. Intake demonstrates other characteristics of ization) should be examined by a mor­ screening for suicide risk should take concern (through actions, current cir­ tality review process. Ideally, the review place immediately upon confinement cumstances, or recent history) indicat­ should be coordinated by an outside and prior to housing assignment, and ing the potential for self-injury, should agency or facility to ensure impartiality. include inquiry regarding current and be placed on close observation. The mortality review—separate and past suicidal behavior, current suicidal apart from other formal investigations ideation, earlier mental health treat­ ◆ Constant observation—reserved that may be required to determine the ment, recent significant loss, suicidal for youth who are actively suicidal cause of death—should be multidiscipli­ behavior by a family member or close (threatening/engaging in the act)— nary (i.e., involve correctional, mental friend, suicide risk during prior contact requires supervision on a continuous, health, and medical personnel) and with or confinement in agency, and uninterrupted basis. include a critical inquiry of the following: arresting or transporting officers’ opin­ ion regarding whether youth is currently In addition, an intermediate level of super­ ◆ The circumstances surrounding the at risk. The policy should include proce­ vision can be used with observation at incident. dures for referral to mental health per­ staggered intervals not to exceed every 5 sonnel for further assessment. (Several minutes. Other supervision aides (e.g., ◆ Facility procedures relevant to the intake screening and assessment forms closed-circuit television, companions, or incident. are available for the identification of sui­ watchers) can be used as a supplement ◆ All relevant training received by cide risk, including the “Intake Screen­ to, but not as a substitute for, these obser­ ing Form/Suicide Risk Assessment” vation levels. involved staff. [Hayes, 1999], the “Juvenile . A facility’s policy regarding ◆ Pertinent medical and mental health Assessment” [Galloucis and Francek, intervention should be threefold: services/reports involving the victim. 2002], and the Massachusetts Youth Screening Instrument-MAYSI-2 [Grisso ◆ All staff should be trained in standard ◆ Possible precipitating factors leading and Barnum, 2000].) first aid and cardiopulmonary resusci­ to the suicide. tation (CPR). Communication. At a minimum, facility ◆ Recommendations, if any, for procedures should enhance communi­ ◆ Any staff member who discovers a changes in policy, training, physical cation (1) between the arresting/ youth attempting suicide should plant, medical or mental health serv­ transporting officer(s), family members, immediately respond, survey the scene ices, and operational procedures. and facility staff; (2) between and to ensure the emergency is genuine, among facility staff (including medical alert other staff to call for medical

10 suicide prevention training to direct care committee process that examines the staff. events surrounding the death to determine Precipitating Factors if the incident was preventable. The review Approximately half (51 percent) of sui­ for the Suicide process might include recommendations cides in facilities that provided suicide aimed at reducing the opportunity of future Of the suicides that occurred in facili­ prevention training were in facilities that deaths.” The process also attempts to iden­ ties that conducted mortality reviews provided the training in a 1- or 2-hour tify any possible precipitating factors that (n=51), precipitating factors were block. Only three suicides took place in a may have caused the suicide (see “Precipi­ identified for more than half (59 per­ facility that provided a full day (7–8 tating Factors to the Suicide” above). Near­ cent). These factors included: hours) of instruction. ly two-thirds (65 percent) of respondents ◆ Fear of waiver to adult system, reported that a mortality review was con­ transfer to a more secure juvenile Certification in ducted following the suicide, although Cardiopulmonary facility, or pending undesirable deaths in detention centers were reviewed placement (including home) Resuscitation to a lesser degree (52 percent). (10 cases). More than two-thirds (68 percent) of sui­ ◆ Recent death of a family member cides occurred in facilities where all direct Recommendations care staff had received certification in (6 cases). The findings from this survey reveal sev­ cardiopulmonary resuscitation (CPR), ◆ Failure in the program (5 cases). although this was true to a lesser degree eral key issues that merit consideration. (55 percent) in training schools and other Recommendations arising from the study ◆ Contagion (from another recent secure facilities. are presented below: suicide in facility) (3 cases). ◆ Consistent with national corrections ◆ Parent(s) threat of/failure to visit Suicide Precaution Protocol standards and practices, juvenile (2 cases). The overwhelming majority (90 percent) facilities, regardless of size and type, ◆ Other (i.e., loss of relationship, of victims were located in facilities that should have a detailed written suicide close proximity to birthday, suicide maintained a suicide precaution protocol prevention policy that addresses each pact with peer, ridicule from peers) for the observation (excluding closed- of the following critical components: (4 cases). circuit television monitoring) of youth. Of training, identification/screening, these 71 victims, fewer than half (48 per­ communication, housing, levels of In several cases, more than one pre­ cent) were in facilities where constant supervision, intervention, reporting, cipitating factor was identified—only observation was the highest level of and followup/mortality review (see the perceived primary factor is listed suicide precaution, including only 28 “Comprehensive Suicide Prevention above. However, of the 79 suicides percent of suicides in detention centers. Programming” on page 10). reported in this study, possible precip­ itating factors for the deaths were More than one-third (37 percent) were in ◆ Juvenile facility administrators should offered by respondents in only 30 (or facilities that reported observation at 15­ create and maintain effective training 38 percent) of the cases. minute intervals as the highest suicide programs and ensure that direct care, precaution level. medical, and mental health personnel receive both pre-service and annual Safe Housing instruction in suicide prevention. prevention programming, including ◆ Less than half (46 percent) the suicides Suicide prevention training curriculums intake screening for suicide risk. occurred in a facility that had a housing used in juvenile facilities have histori­ ◆ process by which a suicidal youth could cally relied on information gathered More than one-third of suicides identi­ be assigned to a safe and protrusion-free about adult inmate suicide and youth fied in this study were unknown to gov­ room. Although the majority (61 percent) suicide in the general population. ernment agencies responsible for the of suicides in training schools and other Given the findings from this study, care and advocacy of confined youth. secure facilities and reception/diagnostic which demonstrate significant differ­ The fact that any suicide occurring centers took place in a facility that provid­ ences between adult inmate suicide within a juvenile facility could remain ed safe and protrusion-free housing for and juvenile suicide, development of outside the purview of regulatory agen­ suicidal youth, this was true for only 35 separate training curriculums targeted cies is disturbing. At a minimum, each percent of suicides in detention facilities to suicide prevention within juvenile death within a juvenile facility should and 25 percent of suicides in residential facilities is warranted. be accounted for, comprehensively reviewed, and provisions made for treatment centers. ◆ Significant deficiencies in intake appropriate corrective action. screening and inadequate suicide pre­ Mortality Review vention programming within detention National juvenile correctional standards centers experiencing suicides warrant Conclusion recommend that a mortality review be con­ immediate attention. Resources need This study was the first attempt to collect ducted following each suicide (Hayes, 1999; to be channeled to juvenile facilities data on the extent, characteristics, and National Commission on Correctional throughout the country, particularly distribution of suicides within juvenile Health Care, 1995, 1999, 2004; Roush, 1996). detention centers, to ensure that any facilities throughout the country. More For the purposes of this study, mortality agency housing a juvenile provides research is clearly needed. For example, review is defined as “a multidisciplinary basic, yet comprehensive, suicide

11 possible precipitating factors were identi­ area of juvenile detention and corrections Council of Juvenile Correctional Adminis­ fied for only slightly more than one-third is to establish a continuum of comprehen­ trators. 2003. Performance-based Standards of the suicides reported in this study. This sive suicide prevention services aimed at (PbS) for Youth Correction and Detention indicates uncertainty of the concept of the collaborative identification, continued Facilities: PbS Goals, Standards, Outcome precipitants, inadequate review of the cir­ assessment, and safe management of Measures, Expected Practices and Processes. cumstances surrounding the death, limit­ youth at risk for self-harm. Braintree, MA: Council of Juvenile Correc­ ed knowledge of the victim’s background, tional Administrators. or all the preceding. Further inquiry Davis, D., Bean, G., Schumacher, J., and regarding possible precipitating factors is For Further Information Stringer, T. 1991. Prevalence of emotional essential to enhancing understanding of This Bulletin presents information taken disorders in a juvenile justice institutional this problem. from the OJJDP Report, Juvenile Suicide population. American Journal of Forensic in Confinement: A National Survey (NCJ Although approximately half the victims in Psychology 9:1–13. 213691). The full report is available on this study were under room confinement OJJDP’s Web site (www.ojp.usdoj.gov/ at the time of death, further research is Dembo, R., Williams, L., Wish, E., Berry, E., ojjdp). needed to explore the relationship Getreu, A., Washburn, M., and Schmeidler, J. 1990. Examination of the relationships between isolation and suicide. Despite the among drug use, emotional/psychological fact that youth were alone in their rooms References problems, and crime among youths enter­ between the hours of midnight and 6 a.m., Alessi, N., McManus, M., Brickman, A., and ing a juvenile detention center. The Inter­ with ample opportunity and privacy to Grapentine, L. 1984. Suicidal behavior national Journal of the Addictions engage in self-injurious behavior, few sui­ among serious juvenile offenders. Ameri­ 25:1301–1340. cides took place during this 6-hour period. can Journal of Psychiatry 141(2):286–287. Instead, approximately half the deaths Duclos, C., LeBeau, W., and Elias, G. 1994. occurred during the 6-hour period American Correctional Association. 1991. American Indian suicidal behavior in between 6 p.m. and midnight—with Standards for Juvenile Detention Facilities detention environments: Cause for contin­ almost a third occurring between 6 p.m. and Standards for Juvenile Correctional ued basic and applied research. Jail Suicide and 9 p.m. Perhaps most importantly, the Facilities. Laurel, MD: American Correc­ Update 5(4):4–9. majority of victims who died by suicide tional Association. while on room confinement status died Esposito, C., and Clum, G. 2002. Social during waking hours. These are periods American Psychiatric Association. 2000. support and problem-solving as modera­ when youth are normally either involved Diagnostic and Statistical Manual of Mental tors of the relationship between child­ in programming or back on their housing Disorders (DSM-IV-TR). Washington, DC: hood abuse and suicidality: Applications units, interacting with staff and peers— American Psychiatric Association. to a delinquent population. Journal of Traumatic Stress 15(2):137–146. perhaps more likely to become involved in Arias, E., Anderson, R., Kung, H., Murphy, confrontations and/or behavior that S., and Kochanek, K. 2003. Deaths: Final Galloucis, M., and Francek, H. 2002. The results in room confinement. Further data for 2001. National Vital Statistics juvenile suicide assessment: An instru­ research is needed to explore this issue. Report 52(3). Hyattsville, MD: National ment for the assessment and management Although only a small percentage of vic­ Center for Health Statistics. of suicide risk with incarcerated juveniles. International Journal of Emergency Mental tims died by suicide following more than Carmona, R.H. 2005. Suicide Prevention Health 4(3):181–199. 12 months of custody, the average length Among Native American Youth. Prepared of confinement before suicide for these Remarks of Richard H. Carmona, M.D., Grisso, T., and Barnum, R. 2000. The Mass­ youth was quite high (approximately 22 M.P.H., F.A.C.S., Surgeon General, U.S. Public achusetts Youth Screening Instrument-2: months), suggesting that prolonged con­ Health Service, U.S. Department of Health User’s Manual and Technical Report. finement might have been one of the pre­ and Human Services. Testimony Before the Worcester, MA: University of Massachu­ cipitating factors in the suicides. This Indian Affairs Committee, U.S. Senate, June setts Medical Center. issue merits further study. 15, 2005. AQ: Retrieved May 13, 2008 from Hayes, L. 1989. National study of jail Findings from this study pose formidable the Web: www.surgeongeneral.gov/news/ testimony/t06152005.html. suicides: Seven years later. Psychiatric challenges for juvenile correctional and Quarterly 60(1):7–29. healthcare officials and direct care staff. Chowanec, G., Josephson, A., Coleman, C., For example, although room confinement and Davis, H. 1991. Self-harming behavior Hayes, L. 1999. Suicide Prevention in Juve­ remains a standard procedure in most in incarcerated male delinquent adoles­ nile Correction and Detention Facilities: A juvenile facilities, its use should be judi­ cents. Journal of the American Academy Resource Guide. South Easton, MA: Council cious and closely scrutinized. Since sui­ of Child and Adolescent Psychiatry of Juvenile Correctional Administrators. cides can occur at any time during a 30(2):202–207. Hayes, L. 2000. Suicide prevention in juve­ youth’s stay in a facility, with the same nile facilities. Juvenile Justice 7(1):24–32. number of deaths occurring within the Cocozza, J., and Skowyra, K. 2000. Youth first few days of custody as after almost a with mental health disorders: Issues and Hayes, L. 2006. Juvenile Suicide in Confine­ year of confinement, intake screening for emerging responses. Juvenile Justice ment: A National Survey. Report. Washing­ the identification of suicide risk should be 7(1):3–13. ton, DC: U.S. Department of Justice, Office viewed as time-limited. Because youth can of Justice Programs, Office of Juvenile be at risk at any point during confinement, Justice and Delinquency Prevention. the challenge for those who work in the

12 Kempton, T., and Forehand, R. 1992. Parent, D., Leiter, V., Kennedy, S., Livens, Twedt, S. 2001b. Lack of options keeps Suicide attempts among juvenile delin­ L., Wentworth, D., and Wilcox, S. 1994. mentally disturbed youth locked up. quents: The contribution of mental health Conditions of Confinement: Juvenile Deten­ Pittsburgh Post-Gazette (July 15):A1. factors. Behaviour Research and Therapy tion and Corrections Facilities. Washington, Underwood, L., and Berenson, D. 2001. 30(5):537–541. DC: U.S. Department of Justice, Office of Mental Health Programming in Youth Cor­ Justice Programs, Office of Juvenile Jus­ Mace, D., Rohde, P., and Gnau, V. 1997. rection and Detention Facilities: A Resource tice and Delinquency Prevention. Psychological patterns of depression and Guide. South Easton, MA: Council of Juve­ suicidal behavior of adolescents in a juve­ Penn, J., Esposito, C., Schaeffer, L., Fritz, nile Correctional Administrators. nile detention facility. Journal for Juvenile G., and Spirito, A. 2003. Suicide attempts U.S. Department of Health and Human Justice and Detention Services 12(1):18–23. and self-mutilative behavior in a juvenile Services. 1999. The Surgeon General’s Call correctional facility. Journal of the Ameri­ To Action To Prevent Suicide, 1999. McGarvey, E., Kryzhanovskaya, L., Koop­ can Academy of Child and Adolescent Psy­ Wash­ man, C., Waite, D., and Canterbury, R. ington, DC: U.S. Department of Health and chiatry 42(7):762–769. 1999. Incarcerated adolescents’ distress Human Services. and suicidality in relation to parental Robertson, A., and Husain, J. 2001. Preva­ Woolf, A., and Funk, S. 1985. Epidemiology bonding styles. Crisis 20(4):164–170. lence of Mental Illness and Substance Abuse Disorders Among Incarcerated Juvenile of trauma in a population of incarcerated Morris, R., Harrison, E., Knox, G., Troman­ youth. Pediatrics 75(3):463–468. Offenders. Jackson, MS: Mississippi Depart­ hauser, E., Marquis, D., and Watts, L.L. ment of Public Safety and Mississippi 1995. Health Risk Behavioral Survey from Department of Mental Health. 39 juvenile correctional facilities in the United States. Journal of Adolescent Health Rohde, P., Seeley, J., and Mace, D. 1997. 17(6):334–344. Correlates of suicidal behavior in a juve­ nile detention population. Suicide and Life- National Commission on Correctional Threatening Behavior 27(2):164–175. Health Care. 1995. Standards for Health Services in Juvenile Detention and Confine­ Roush, D. 1996. Desktop Guide to Good ment Facilities. Chicago, IL: National Com­ Juvenile Detention Practice. Washington, mission on Correctional Health Care. DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Jus­ National Commission on Correctional tice and Delinquency Prevention. Health Care. 1999. Standards for Health Services in Juvenile Detention and Confine­ Shelton, D. 2000. Health status of young ment Facilities. Chicago, IL: National Com­ offenders and their families. Journal of mission on Correctional Health Care. Nursing Scholarship 32(2):173–178. National Commission on Correctional Sickmund, M., and Wan, T. 2001. Census Health Care. 2004. Standards for Health of Juveniles in Residential Placement Services in Juvenile Detention and Confine­ Databook. Washington, DC: U.S. Depart­ ment Facilities. Chicago, IL: National Com­ ment of Justice, Office of Justice Programs, mission on Correctional Health Care. Office of Juvenile Justice and Delinquency Prevention. National Juvenile Detention Association. 1990. Position Statement: Definition of Juve­ Substance Abuse and Mental Health Ser­ nile Detention. Richmond, KY: National vices Administration. 2001. Summary of Juvenile Detention Association. Findings From the 2000 National Household Survey on Drug Abuse. NHSDA Series: H-13, National Juvenile Detention Association. DHHS Publication No. SMA 01-3549. 2001. Position Statement: Use of Juvenile Rockville, MD: U.S. Department of Health Detention Facilities for Youth With Severe and Human Services, Substance Abuse and Mental Health Issues. Richmond, KY: Mental Health Services Administration. National Juvenile Detention Association. Sullivan, C. 1995. Juvenile custody sui­ Office of Juvenile Justice and Delinquency cides blamed on apathy, impulse, gaps in Census of Juveniles in Prevention. 1999. care. Los Angeles Times (March 12):A1. Residential Placement. Washington, DC: U.S. Department of Justice, Office of Jus­ Teplin, L., Abram, K., McClelland, G., tice Programs, Office of Juvenile Justice Dulcan, M., and Mericle, A. 2002. Psychi­ and Delinquency Prevention. atric disorders in youth in juvenile detention. Archives in General Psychiatry Office of Juvenile Justice and Delinquency 59:1133–1143. Prevention. 2002. 2000 Juvenile Residential Facility Census. Unpublished data. Wash­ ington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juve­ nile Justice and Delinquency Prevention.

13 This Bulletin was prepared under grant number Acknowledgments 1999–JN–FX–0005- from the Office of Juvenile Lindsay M. Hayes, M.S., is Project Director of the National Center on Institutions Justice and Delinquency Prevention (OJJDP), and Alternatives. For their support and assistance in designing the data collection U.S. Department of Justice. instruments, analyzing the data, and reviewing the draft report, the author would Points of view or opinions expressed in this like to thank G. David Curry, Ph.D., Department of Criminology, University of document are those of the author(s) and do Missouri-St. Louis; Robert E. DeComo, Ph.D., Director of Research, National Council not necessarily represent the official position on Crime and Delinquency; Barbara C. Dooley, Ph.D., former Director, Madison or policies of OJJDP or the U.S. Department County (TN) Juvenile Court Services; Cedrick Heraux, Ph.D. candidate, School of of Justice. Criminal Justice, Michigan State University; and David W. Roush, Ph.D., Director, Center for Research and Professional Development, Michigan State University. The Office of Juvenile Justice and Delinquency Thanks also to Alice Boring of the National Center on Institutions and Alterna­ Prevention is a component of the Office of tives, who brought the report together to its final form. Justice Programs, which also includes the In addition, two consulting agencies, the Council of Juvenile Correctional Adminis­ Bureau of Justice Assistance; the Bureau of trators (CJCA) and the National Juvenile Detention Association (NJDA), provided Justice Statistics; the Community Capacity invaluable assistance to the author and the team in both endorsing the project Development Office; the National Institute and encouraging juvenile facility directors to participate in the survey process, of Justice; the Office for Victims of Crime; as well as in reviewing the draft report. Special thanks are extended to CJCA’s and the Office of Sex Offender Sentencing, Edward J. Loughran, Kim Godfrey, and Robert Dugan, and NJDA’s Earl L. Dunlap Monitoring, Apprehending, Registering, and and Michael A. Jones. Tracking (SMART).

Share With Your Colleagues Unless otherwise noted, OJJDP publications are not copyright protected. We encourage you to reproduce this document, share it with your colleagues, and reprint it in your newsletter or journal. However, if you reprint, please cite OJJDP and the authors of this Bulletin. We are also interested in your feedback, such as how you received a copy, how you intend to use the information, and how OJJDP materials meet your individual or agency needs. Please direct your comments and questions to: Juvenile Justice Clearinghouse Publication Reprint/Feedback P.O. Box 6000 Rockville, MD 20849–6000 800–851–3420 301–519–5600 (fax) Web: tellncjrs.ncjrs.gov

14 U.S. Department of Justice PRESORTED STANDARD Office of Justice Programs POSTAGE & FEES PAID Office of Juvenile Justice and Delinquency Prevention DOJ/OJJDP *NCJ~214434* PERMIT NO. G–91 Washington, DC 20531 Official Business Penalty for Private Use $300

Bulletin NCJ 214434