Identification of Suicide Risk Among Rural Youth: Implications for the Use of HEADSS

Abstract

Introduction

Nurse practitioners have the power to detect suicide risk and prevent suicide, a problem plaguing rural areas of the United States. Suicide risk assessment can be completed using the HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) interview instrument. The purpose of this study was to determine if HEADSS is appropriate for guiding suicide risk assessment of rural adolescents.

Method

High school students in Southwestern Pennsylvania completed qualitative questions from the Child Behavior Checklist and Coping Response Inventory as part of the Intervention to Promote Mental Health in Rural Youth. Qualitative content analysis was performed.

Results

Prominent themes identified by participants included academic performance, relationships, dislikes about school, friends, death, mental health, and the future. Several minor themes concerned safety. Most known risk factors for suicide were concerns of participants.

Discussion

The expansion of HEADSS to include death and safety should be considered. The modified version—HEADDSSS— can be used to guide suicide risk assessment of youth in rural Pennsylvania, ensuring both thoroughness of assessment and safety.

Nurse practitioners are in a prime position to detect suicide risk and prevent suicide, the third leading cause of death for youth 15 through 19 years of age in the United States (Economic Research Service, 2004). Although youth suicide rates have recently declined in the United States, the declines are not evenly distributed (State and Territorial Injury Prevention Directors Association Rural Youth Suicide Prevention Workgroup, 2008). Western and mountainous states that have rural counties, where the population density is low and residents live in small communities separated by vast landscapes, have the highest youth suicide rates. Nurse practitioners can prevent these suicides because they are often the most prevalent providers, alleviating clinician shortages while providing quality care (National Advisory Committee on Rural Health and Human Services, March 3, 2002). Furthermore, most persons who have died by suicide have had contact with a primary care provider within a month prior to death (Luoma et al., 2002).

One way to prevent suicide is to identify youth at risk and refer them for treatment. A tool that is often used in primary care to guide psychosocial risk assessment is the HEADSS interview instrument (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) (Cohen, Mackenzie, & Yates, 1991). Whether HEADSS is an appropriate tool to use with rural youth, particularly for suicide risk assessment, has never been studied. Therefore, the purpose of this research was to determine if HEADSS is an appropriate tool for nurse practitioners to use to assess rural youth for suicide risk.

Background

Prevalence

The rate of suicide per 100,000 youth 15 through 19 years of age in the United States decreased from a high of 11.14 in 1990 to 7.66 in 2005 (Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, 2009). Despite this decrease, suicide rates per 100,000 population in states having large rural areas remain intolerably high. Iowa, Wisconsin, Minnesota, Louisiana, Arkansas, Nebraska, New Mexico, Arizona, Colorado, Nevada, Wyoming, South Dakota, Alaska, North Dakota, and Montana have suicide rates that vary from 10.6 (Iowa) to 26.6 (Montana) for youth 15 through 19 years of age. While Pennsylvania has vast rural areas, its statewide suicide rate for this age group is less than the national rate (6.85 compared to 7.66 in 2005). However, 2,898 years of life (prior to age 65) were lost (Table 1). Furthermore, the suicide rate for the Commonwealth’s 48 rural counties combined (7.60) is near the national rate (7.66) (Pennsylvania Department of Health, n.d.). More alarming is that the rate for suicide by firearm for 15 through 19 year olds in rural Pennsylvania was 5.32 for 2005 compared to 3.5 for the U.S. as a whole. These rates may indicate that we are failing to recognize those at risk, perhaps because we are not accurately assessing psychosocial concerns.

Table 1.

Age at Death in Years	Number of Suicides	PYLL per Suicide (Based on Age 65 Years)	Total PYLL
15	6	65 − 15 = 50	6 X 50 = 300
16	14	65 − 16 = 49	14 X 49 = 686
17	7	65 − 17 = 48	7 X 48 = 336
18	12	65 − 18 = 47	12 X 47 = 564
19	22	65 − 19 = 46	22 X 46 = 1,012
Total	61		2,898

Characteristics of Rural Life

Understanding how rural life may contribute to suicide risk is a necessity because “ruralness” has important psychosocial implications, including fewer job opportunities, fewer resources in school, and lower levels of available services such as health care (Heck et al., 2004; Robertson & Husenits, 2007). In Pennsylvania, a greater percentage of rural compared to urban residents do not have health insurance or are insured through Medical Assistance. Twenty-two percent live in areas that have too few health care providers (Pennsylvania Rural Health Association, 2003). Rural children are less likely to receive outpatient mental health services compared to urban children (Robertson & Husenits, 2007). Childhood obesity is now more common in rural than urban regions (Center for Rural Pennsylvania, 2005). More rural than urban children live in households with incomes below the poverty level and are eligible for the Free and Reduced School Lunch Program (Center for Rural Pennsylvania, 2007). More rural households receive food stamps, and fewer rural adults have high school diplomas.

Pennsylvania’s rural youth are also placed at risk due to recreation and occupations that pose dangers to their safety and well-being. Risky sports that can injure and kill include snowmobile and all terrain vehicle riding (Department of Conservation and National Resources, 2004). Potential methods for killing one’s self include guns used for hunting and pesticides used on Pennsylvania’s farmlands (Hirsch, 2006; Pennsylvania State University, 2005).

The Pennsylvania Department of Health (2000) created the State Health Improvement Plan: Special Report and Plan to Improve Rural Health Status to address suicide and other issues. The report focuses on Healthy People 2010: Understanding and Improving Health (U.S. Department of Health and Human Services, 2000). Objective 18-1, to reduce the suicide rate for all ages combined from 11.3 per 100,000 in 1998 to 5.0 per 100,000 by 2010, has not been achieved. Suicide rates for all ages combined (11.0) and youth 15 through 19 years of age (6.85) continue to be greater than the target rate of 5.0 (Pennsylvania Department of Health, n.d.).

Risk Factors for Suicide

Problems and concerns of rural youth must be explored and understood in order to detect and prevent suicide. Risk factors--conditions or agents that increase the likelihood of problematic behavior and negative outcomes (Jessor, 1991; Jessor, Turbin, & Costa, 2003)—may be associated with or lead to suicide (Education Development Center, n.d.). According to Pfeffer (2002), the recognition of risk factors is the most important endeavor in reducing risk and preventing suicidal behavior. Factors range from self- and parental psychiatric problems and history of suicide attempts to dysfunctional cognition, stressful life events, parental divorce, relationship problems, and socioenvironmental factors that include problems at school or work, sexual orientation, and biological risk factors (Gould & Kramer, 2001) (Table 2). Many studies have investigated risk factors, but none have focused solely on Pennsylvania’s rural youth.

Table 2.

Risk Factors for Suicide Based on HEADDSSS Domains

HEADDSSS Domains and Corresponding Risk Factors for Suicide	Concern in Prominent Theme	Concern in Minor Theme
Home
• Less parental closeness (Zweig, Phillips, & Lindberg, 2002)	*	-
• Loss of a parent through divorce (Neiger & Hopkins, 1988) or living in a non-intact family (Brent et al., 1999)	*	-
• Few activities with parents (Bearman & Moody, 2004)	*	-
• Parent-child conflict (Brent et al., 1999) or sibling conflict (Brent, Moritz, Bridge, Perper, & Canobbio, 1996)	*	-
• Family history of conduct or affective problems/psychopathology or substance abuse (Brent et al., 1999; Brent et al., 1996)	*	-
• Lack of family support/availability or no one to count on (O’Donnell, Stueve, Wardlaw, & O’Donnell, 2003; Perkins & Hartless, 2002)	*	-
• Basic needs not being met (O’Donnell, O’Donnell, Wardlaw, & Stueve, 2004)	*	-
Education
• Poor school performance (Slap, Vorters, Chaudhuri, & Centor, 1989; Watt & Sharp, 2001)	*	-
• Lower GPA than students not at risk for suicide (Field, Diego, & Sanders, 2001)	-	*
• Self- and parental dissatisfaction with grades (Lewinsohn, Rohde, & Seeley, 1993)	-	*
• Failing a grade (Caucasian youth) (Gould, Fisher, Parides, Flory, & Shaffer, 1996) or repeating a grade (Hispanic females) (Borowsky, Ireland, & Resnick, 2001)	*	-
• Missed school days (Lewinsohn et al., 1993)	*	-
• Greater likelihood of dropping out of school compared to other students (Thompson, Moody, & Eggert, 1994) or having already dropped out of school (Gould et al., 1996)	-	-
• More unmet school goals compared to other students (Thompson et al., 1994)	*	-
• Spend fewer afternoons/evenings spent doing homework compared to others (Mazza & Eggert, 2001)	*	-
• Suspended from school (Gould et al., 1996)	*	-
• Disciplinary crises (Gould et al., 1996)	*	-
• Conduct disorder combined with disruptive behavior at school (Shah, Hoffman, Wake, & Marine, 2000)	-	*
• Carrying a weapon at school (Borowsky et al., 2001)	-	-
• Not being in school or not working (“drifting”) (Gould et al., 1996)	-	-
• Not going to college (Gould et al., 1996), being female and not expecting to attend college when one’s mother attended college, or being male and not expecting to attend college (Watt & Sharp, 2001)	*	-
• School climate (whether teachers really cared about the subject, teachers paying attention to the subject, receiving a lot of encouragement at school, and liking school) (males) (Perkins & Hartless, 2002)	*	-
• Socially disconnected at school	*	-
• School related problems in general (Borowsky et al., 2001; Brent et al., 1999; Centers for Disease Control and Prevention, 1995)	*	-
Activities
• Arrest for and/or involvement in a crime (Windle, 2004) or disciplinary/legal problems (Brent et al., 1999; Brent et al., 1996)	-	*
• Perpetration of violence/carrying a weapon (Borowsky et al., 2001)	-	*
• Fighting (females) (Bearman & Moody, 2004; Zweig et al., 2002)	-	*
• Socially isolated from peers (Bearman & Moody, 2004)	*	-
• Work problems (Gould et al., 1996)	-	*
• Friendship networks among females that encourage expression of emotional distress, heighten perceived severity, and encourage drastic responses (Watt & Sharp, 2001)	*	-
Drug Use and Abuse
• Substance abuse (alcohol, marijuana, etc.) (Borowsky et al., 2001; Brent et al., 1999; Perkins & Hartless, 2002; Zweig et al., 2002)	-	*
• Comorbidity of substance abuse and mood disorder (Brent et al., 1999)	*	-
• Binge drinking (Windle, 2004)	-	-
Death
• Friends died by suicide (Borowsky et al., 2001)	*	-
• Loss of a preexisting relationship due to death (Brent, Perper, Moritz, Baugher, Roth et al., 1993)	*	-
Sexual Behavior
• Homosexuality; homosexual attraction; gay/lesbian/bisexual/unsure (GLBU) status; or being both homeless and gay, lesbian, or bisexual (Bearman & Moody, 2004; Borowsky et al., 2001; Noell & Ochs, 2001; O’Donnell et al., 2004; Russell & Joyner, 2001)	-	-
• Sexual abuse (Brent et al., 1996; Lipschitz et al., 1999; Perkins & Hartless, 2002)	-	*
• Forced sexual relations (females) (Bearman & Moody, 2004)	-	*
Suicidality and depression
• Past suicide attempts; family history of suicide attempts, including parental attempts and ideation; and suicide attempts by friends (Bearman & Moody, 2004; Borowsky et al., 2001; Brent et al., 1999; Rueter & Kwon, 2005; Russell & Joyner, 2001)	*	-
• Mood/affective disorders, sadness or depression, hopelessness, or problems with conduct/conduct disorder (Bearman & Moody, 2004; Brent et al., 1999; Brent et al., 1996; O’Donnell et al., 2004; O’Donnell et al., 2003; Perkins & Hartless, 2002; Shah et al., 2000)	*	-
• Self-esteem issues (females) (Bearman & Moody, 2004)	*	-
• Stressful events (Windle, 2004)	*	-
• Emotional neglect/abuse (Lipschitz et al., 1999)	*	-
• Interpersonal conflict (Brent et al., 1996)	*	-
• Disruption of a romantic relationship (Brent et al., 1999) or preexisting relationship (Brent, Perper, Moritz, Baugher, Roth et al., 1993)	*	-
• Intransitive relationships (Bearman & Moody, 2004)	*	-
• Emotional or learning disabilities (Blum, Kelly, & Ireland, 2001)	-	-
• Mobility impairments (Blum et al., 2001)	-	-
• Low psychosocial adjustment (Zweig et al., 2002)	*	-
• Extreme perceptions of weight (underweight and overweight)/BMI (Bearman & Moody, 2004)	-	*
• Physical illness (Brent et al., 1996)	-	*
• Less likely to have accessed formal network of contacts (teacher, minister/priest, medical doctor/nurse, mental health professional) alone (O’Donnell et al., 2003)	-	-
Safety
• Being a victim of violence (Borowsky et al., 2001)	-	*
• Witnessing violence (Brent et al., 1996)	-	*
• Physical abuse (Brent et al., 1996; Kaplan et al., 1999; Lipschitz et al., 1999; Perkins & Hartless, 2002) or lifetime history of abuse (Brent et al., 1999)	-	*
• Being bullied or frequent bullying of others (Klomek, Marrocco, Kleinman, Schonfeld, & Gould, 2007)	-	*
• Firearm/gun in the home (Bearman & Moody, 2004; Brent, 2001; Brent et al., 1999)	-	*

^*Note. = risk factor identified in this study. - = risk factor not identified in this study.

Psychosocial issues of Southwestern Pennsylvania’s rural high school students were studied by Puskar and colleagues, who surveyed students using quantitative self-report questionnaires. Psychosocial concerns included stressors and coping methods, anxiety, anger, somatic complaints, depression, health, and risk behaviors (Puskar & Martsolf, 1993; Puskar, Ren, Bernardo, Haley, & Stark, 2008; Puskar, Sereika, & Haller, 2003; Puskar, Tusaie-Mumford, Sereika, & Lamb, 1999a and 1999b). Common stressors identified by students were school (exams, tests, and grades), family (fighting with parents and loss of a family member), and friendship/social concerns (girlfriends, boyfriends, and dating) (Puskar & Martsolf). Problems identified in another study included confusion about the future, depression, feeling lonely, trouble at home, suicidal thoughts, lack of a best friend, tobacco and alcohol use, sexual abuse, and general health concerns of both a physiological and psychosocial nature (feeling tired, being overweight, and headaches) (Puskar et al., 1999a). Puskar and colleagues also found that depression, negative life events, anxiety, and drug use were related to anger (Puskar et al., 2008). While these studies contributed greatly to our knowledge about psychosocial problems of rural youth, none concerned suicide risk.

Other studies that addressed psychosocial risk factors used psychological autopsies—tools used by researchers to gather information about persons who died by suicide. Sources include forensic reports, psychiatric and health care records and other documents, and structured interviews of family members, relatives, and friends. The purpose is to synthesize information and identify factors that may have been related to a person’s suicide (Isometsa, 2001).

Brent and colleagues conducted psychological autopsies that included both rural and urban youth from the 28 counties in western Pennsylvania (Brent, Baugher, Bridge, Chen, & Chiappetta, 1999; Brent, Perper, Moritz, Baugher, Schweers & Roth, 1993). All but four of these counties (Allegheny, Beaver, Erie, and Westmoreland) are classified as rural. Risk factors for completed suicides included conduct disorder, substance abuse, mood disorders with nonaffective comorbidity or comorbid with substance abuse, major depression, mixed bipolar state, past suicide attempts, family history of psychopathology, stressors (parent-child conflict, boy/girlfriend conflict, disruption of a romantic relationship, legal/disciplinary problems), and guns in the home (Brent et al., 1999; Brent, Perper, Moritz, Allman et al., 1993; Brent, Perper, Moritz, Baugher et al., 1993). These studies greatly increased our knowledge of risk factors for suicide, but only families and friends of the deceased could be surveyed. Due to the nature of the studies (psychological autopsies), subjects (deceased youth) could not participate.

A literature search of MEDLINE, CINAHL, and PsycINFO revealed only a few studies concerning suicide risk factors among rural youth who live in other states. Quantitative self- report surveys addressed perceived risk factors for suicide among rural youth in Nevada (Albers & Evans, 1994; Evans, Smith, Hill, Albers, & Neufeld, 1996). Students responded to questions concerning suicidal ideation (Albers & Evans, 1994; Evans et al., 1996) and attempts (Evans et al., 1996), as well as lists of common concerns. The lists were created by focus groups of students. Top concerns identified in both studies included abuse, AIDS, family conflict/relationships with parents and step-parents, alcohol/substance abuse, pregnancy, and what to do after high school (Albers & Evans, 1994; Evans et al., 1996). Other concerns included quality of education, paying for education after high school, making decisions, obtaining a satisfying job, family bills (Albers & Evans, 1994), psychological/emotional problems, relationship/dating issues, and sexuality issues other than pregnancy (Evans et al., 1996).

A secondary analysis of female adolescent smokers in rural Virginia was conducted using quantitative data from the Youth Risk Behavior Survey and the Teen Assessment Project (Huebner et al., 2005). Female adolescents who previously smoked but had not smoked during the past 30 days had fewer suicide attempts, fewer suicidal thoughts, and less depression compared to those who smoked during the past 30 days.

None of the studies reviewed provided adolescents with the opportunity to express concerns in their own words. This is important because self-descriptions can help clinicians understand the perspectives of youth and become more sensitive to their problems and concerns. Furthermore, causes of psychosocial issues, including suicidal ideation, could not be determined.

Although the types of psychosocial concerns discussed in the literature review can be viewed as problematic, one might be inclined to argue that they are part of normal adolescence. While adolescence can be an exciting time, too many new experiences can be overwhelming, especially to those who are already vulnerable in any way (Petersen, Leffert, & Graham, 1995). Furthermore, adolescence is a phase that includes a predominance of risk-taking behavior and special dangers (Remschnidt, 1994). The outcome and timing of pubertal processes can be affected by psychosocial factors. Therefore, identification of problems and concerns is essential.

Status of Evidence Based Practice

No studies exist concerning the value of suicide risk assessment in primary care. Research concerning screening instruments, impact of screening on suicide completions and attempts, and treatment outcomes for primary care patients is needed. Thus, the United States Preventive Services Task Force (2004) does not recommend for or against screening.

The American Academy of Pediatrics (AAP), however, recommends that pediatricians ask questions to elicit known risk factors for suicide, inquiring into suicidal thoughts or concerns (Shain, 2007). The American Medical Association’s (AMA) Guidelines for Adolescent Preventive Services (GAPS) include asking all adolescents annually about emotions or behaviors that indicate severe or recurrent depression or suicide risk (Department of Adolescent Health, 1997). The Institute of Medicine’s Committee on Pathophysiology and Prevention of Adolescent and Adult Suicide (Goldsmith, Pelmar, Kleinman, & Bunney, 2002) recommends that risk factors for suicide be uncovered during primary care visits. According to the committee, both primary care providers and those who deal with special high-risk populations such as adolescents have an important role in the identification and referral of patients at risk for suicide.

Conceptual Framework

The theoretical basis for suicide risk assessment using HEADSS is the model proposed by Kral and Sakinofsky (1994). In this model, suicide risk assessment is performed by assessing two tiers of risk factors: (1) background risk factors, and (2) subjective risk factors.

Background risk factors include sociodemographic and related indices correlated with increased risk of suicide (Kral & Sakinofsky, 1994). They are based on different populations, cultures, and cohorts, and inform clinicians about a client’s general level of risk. Many of the indices change over time. Therefore, they should be assessed at every visit. One background factor alone may not be very meaningful, but, as background factors accumulate, the risk for suicide increases. The assessment of background factors (e.g. recent loss and alcohol abuse) can provide the means for assessing the second tier of risk factors—subjective factors.

The identification of subjective factors that concern psychological state includes the assessment of mental state, experience, thinking, emotions, and personal meaning and experience of distress (Kral & Sakinofsky, 1994). Clinicians must assess perturbation (degree of upset, disturbance, anguish, discomfort, turmoil, hopelessness, dread, tension, discomfort, or other excessive psychological pain), lethality (consciously selecting suicide as a viable option, then as a specific option, and eventually as the only option to alleviate perturbation), and cognitive constriction (tunnel vision or a narrowing of options to two and ultimately one—suicide).

HEADSS Interview Instrument

We recommend psychosocial assessment of background and subjective risk factors using HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) (Table 2). HEADSS is a psychosocial interview designed to prompt health care professionals to ask questions about areas they might otherwise ignore or forget (Cohen et al., 1991). It helps clinicians focus on stressors that may contribute to mental and physical illness. The process involved in conducting the interview and ordering the questions allows the provider to develop a relationship with the client. The interview begins with less emotionally charged concerns, such as home, and progresses to sensitive concerns, such as sexuality. Youth welcome the opportunity to discuss topics that they may think are outside of the provider’s interests and often appreciate the extra interest taken in them. In order for the interview to be successful, clinicians must set the stage by making it safe to talk about charged issues such as suicide (Clark & Ginsburg, 1995, Winter; Cohen et al., 1991).

Questionnaires such as the Suicidal Behaviors Interview, Suicide Probabilities Scale, and Multi-Attitude Suicide Tendency Scale (Cull & Gill, 1988; Orbach et al., 1991; Reynolds, 1990, 1991) may complement suicide risk assessment using HEADSS. However, they should not substitute for assessment by providers (Simon, 2002).

Assessments that were completed using HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) for 1,015 new homeless/runaway clients at a High Risk Youth Clinic in Los Angeles were compared to assessments for non-homeless youth (Cohen et al., 1991). Homeless teens were far more likely to be depressed, to experience current suicidal ideation, and to have dropped out of school. They were six times more likely to be at risk for HIV and to have engaged in their first sexual intercourse at a younger age. They also had a higher incidence of prostitution and sexual abuse.

Although further research concerning HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) is needed, a prospective interventional investigation showed that documentation of HEADSS categories was increased when the acronym was stamped on charts prior to medical interview (Van Amstel, Lafleur, & Blake, 2004). However, a review of 100 adolescent inpatient records showed that screening using HEADSS was complete in only 7 charts (all 7 domains were screened), thorough in only 3 charts (5–6 domains were screened), inadequate in 29 charts (1–4 domains were screened), and absent from 62 charts (0 domains were screened) (Yeo, Bond, & Sawyer, 2005). Perhaps HEADSS is not an appropriate tool for inpatient settings or perhaps clinicians must be more diligent in assessing for psychosocial risk. No studies have determined if HEADSS is an effective acronym to use to assess psychosocial risk of rural youth. This is the first study to do so.

Purpose

The purpose of this study was to determine if HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) is an appropriate tool to use for suicide risk assessment of rural youth. The research questions included the following: (1) What are the self-identified problems and concerns of rural high school students in Southwestern Pennsylvania? (2) Do identified problems and concerns include known risk factors for suicide? and (3) Do the problems and concerns of participants in this study fit the domains of the HEADSS assessment or should a modified version of HEADSS be used when performing risk assessment of rural youth?

Methodology

This study was a secondary analysis of qualitative data collected during Phase I of the Intervention to Promote Mental Health in Rural Youth, a two-phase study that was funded by the National Institute of Nursing Research (NINR) of the National Institutes of Health (NIH). Purposes of Phase I of the study were to: (1) determine the prevalence of symptoms of anxiety in the sample of rural youth, (2) identify relationships among anxiety symptoms, depressive symptoms, and somatic complaints, and (3) determine how somatic complaints predict depression and anxiety. The goal of Phase II was to test the effectiveness of a nurse-managed intervention in increasing coping, social support, and self-esteem of rural students.

Many beliefs, understandings, and life experiences that place youth at risk are difficult to quantify, but they must be assessed and evaluated (Rich & Ginsburg, 1999). Health care providers use qualitative techniques every day when they ask open-ended questions and analyze responses. In a similar manner, open-ended questions were asked of participants in this study and a secondary analysis of the data was performed.

Participants and Setting

The sample for Phase I of the study included 466 adolescents drawn from four rural high schools in Southwestern Pennsylvania. Students were predominantly Caucasian (97%), and included 9th (42.3%), 10th (27.7%), 11th (27.9%), and 12th graders (2.1%). Physical and mental health were assessed using a baseline battery of self-report instruments including the Screen for Child Anxiety Related Emotional Disorders (SCARED), the Reynolds Adolescent Depression Scale (RADS), the Child Behavior Checklist-Youth Self-Report (CBCL-YSR), and the Coping Response Inventory-Youth (CRI-Y).

Data Collection

Consent was granted for the quantitative and qualitative portions of the study by the Institutional Review Board of the University of Pittsburgh. Letters of support for the Intervention to Promote Mental Health in Rural Youth were sent by high school administrators to the researchers. Parents, students, and teachers were given an informational letter about the study. The letter and consent forms were distributed during health class and assemblies. Criteria for participation included enrollment in regular, college preparatory, or honors classes, and ability to read and write in English. After consent was obtained, research associates (master level nurses) administered nine questionnaires to students for one and one-half hours during school time.

Data was obtained from written qualitative narratives included on the Child Behavior Checklist-Youth Self-Report (CBCL-YSR) and the Coping Response Inventory-Youth (CRI-Y). The CBCL-YSR can be completed by 11 through 18-year-old adolescents to describe their own problems and competencies (Achenbach & Ruffle, 2000). The following open-ended qualitative questions that were used in this study were designed to provide background information and starting points for interviews (Achenbach, 1991): (a) “Please describe any concerns or problems you have about school,” and (b) “Please describe any other concerns you have.”

The Coping Response Inventory-Youth (CRI-Y) is a 48-item instrument that assesses how adolescents cope with problems (Moos, 1993). The first 10 questions concern the problem described in the open-ended qualitative question that was used in this study. That question—”Describe the problem or situation”—is located at the beginning of the questionnaire.

Data Analysis

Content analysis was the qualitative technique used to examine responses and inductively derive themes from the data (Denzin & Lincoln, 1998). Analyses followed a summative approach in which key words were identified and quantified (Hsieh & Shannon, 2005). The following process involved isolating, interpreting, and counting themes (Denzin & Lincoln, 1998).

1. All problems and concerns of participants were entered into Microsoft Excel.

2. Codes, units of meaning, were assigned to individual problems and concerns.

3. Themes were then identified by comparing the codes to each other and grouping those that were most similar. Fifty-one themes were identified, some by a large number of participants (47.9%, n = 223) and others by as few as one participant (0.2%). Because of this variation, we needed to determine criteria for classifying themes as prominent and non-prominent (minor). We decided to look at the number of participants who identified the six most common themes (top 10% of themes). That number ranged from 56 (12.0%) to 223 (47.9%). Therefore, we decided to consider all themes that were identified by at least 56 participants to be prominent themes. Minor (non-prominent) themes were identified by 55 or fewer participants.

4. Themes were also mapped to the domains of the HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) acronym. We assessed whether the HEADSS acronym is comprehensive enough for assessing all prominent themes identified in this study and whether modification of HEADSS would help to ensure that all themes are assessed during a HEADSS interview.

Strategies for Validating Findings

Findings were validated by having a second coder code the data. The second coder was a doctorally prepared psychiatric clinical nurse specialist and qualitative researcher who was a member of the research team. Codes to be used were discussed, and each coder coded the data independently. Results were compared. Inter-rater reliability was 95%. Differences were discussed, and final coding was completed.

Findings

Prominent themes identified in the study included academic performance, relationships, dislikes about school, friends, death, mental health, and the future (Table 3). Several minor themes were also identified (Table 4). Direct quotations are contained in Tables 3 and 6.

Table 3.

Grouping of Codes to Form Prominent Analytical Themes

Qualitative Statements	Codes Assigned to Statements	Prominent Analytical Themes	n	Percentage of Participants
• “I would like to do as well in school as I can. Mostly to make my family proud. But either I can’t grasp certain subjects or I’m not dedicated enough.”	- Performance	Academic Performance	223	47.9%
• ”Not being able to get my English grade up enough to pass the class.”	- Passing a class
• “Bringing my Grade Point Average up.”	- Grades
• “I sometimes argue with my parents. It doesn’t get really bad, but we just can’t always see the other’s point of view.”	- Parental Relationship	Relationships	138	29.6%
• “Having problems with my brother. He doesn’t seem to want anything to do with me.”	- Sibling Relationship
• “Last year I thought a lot that my girlfriend was cheating on me and she almost did and we used to get into big fights but that changed.”	- Relationship with Girl/Boyfriend
• “I feel my school does not have adequate college prep courses. I feel I may not be prepared when I go to college for the classes I will need to take.”	- Dislike Curriculum	Dislikes about school	133	28.5%
• “The food is not very good and they don’t give you enough of it.”	- Dislike School Food
• “Anyone with a popular last name automatically makes any sport, etc. Favoritism really hurts everyone.“	- Dislike Favoritism at School
• “I worry about losing my friends. Last year, I had a fight with my old friends and we still are not speaking. I worry the same thing will happen with my new friends.”	- Loss of Friends	Friends	74	15.9%
• “I have a concern about a friend who keeps letting another friend control her life.”	- Boundaries of Friends
• “My most difficult situation was having to deal with the many changes of high school. My friends all seemed to be different and nothing seemed the same as it was last year.”	- Friends Changing
• “A couple of months ago my best friend’s dad died. He was very close to me. He was like my father. We used to go camping all the time, tubing, and other stuff that was fun. It was very hard for me to cope with.”	- Death of Father Figure	Death	65	13.9%
• “My grandfather died two months ago of lung cancer. I had a hard time because at the same time last year, my grandmother died of the same cancer.”	- Death of Grandparent
• “My dog died.”	- Death of Pet
• “Pretty severe depression. I thought about suicide a lot. I usually would come to my senses after awhile, but then it would happen again. Some months were okay and I thought I was fine. Other months were awful. I would cry a lot and scream and yell. God has helped me get through, though.”	- Suicidal Ideation	Mental health	58	12.4%
• “Got threw being really depressed and having anxiety attacks by getting help.”	- Depression
• “I am very jealous of others. I am very self conscience and I always worry about what others think.”	- Jealousy
• “I get nervous about what I’m going to do after school.”	- Future Career	Future	56	12.0%
• “I’m afraid that after I graduate I won’t be able to handle it in the real world.”	- Future in Real World
• “I am afraid I won’t get a good job after college.”	- Future Job

Table 4.

Top 10 Minor Themes

Minor Themes	Frequency	Percentage of Participants
1. Parental Approval or Disapproval	35	7.5%
2. Classmates	32	6.9%
3. Organized Sports	32	6.9%
4. Physical Health	24	5.2%
5. Labeled/Rumors	18	3.9%
6. Having Enough Time	18	3.9%
7. Moving	17	3.6%
8. Relationship between Parents	16	3.4%
9. Peer Pressure	14	3.0%
10. Bullying	11	2.4%

Table 6.

Codes that Form the Safety Theme

Qualitative Statements	Codes Assigned to Statements	n	Percentage of Participants
• ” The 2 girls want to beat me up. Everywhere I see them they swear at me and call me all of these names. They always say that they are going to jump me. I can’t go anywhere without getting threatened. The bad part is my cousin is the one that is telling them to do it.”	- Bullying	11	2.4%
• “I was driving a snowmobile and I wrecked. I broke my leg.”	- Motor vehicle accidents	8	1.7%
• “I received $10 from a stolen check and got a $1200 fine and 18 months probation and a felony 3 charge.”	- Trouble with the law	8	1.7%
• “Father being charged with many types of abuse against me. Not being able to see my brother ‘cause of it.”	- Alleged Abuse	3	0.6%
• “Too many of my close friends are in gangs..	- Friends in gangs	3	0.6%
• “Taking my ex-boyfriend to court to get him to stop bothering me since June 5—months of stalking, harassment, threats and everything else, trying to break in, etc..”	- Harassment	1	0.2%
• “I almost drown in the ocean while on vacation.”	- Drowning	1	0.2%
• “Shot gun was stolen.”	- Gun	1	0.2%
• “Violence (school).”	- Violence	1	0.2%
• “I got drunk and ran away from home and the police had to find me.”	- Running away	1	0.2%
TOTAL		38	8.0%

Academic performance

A number of participants were insightful about their academic achievement, taking problems with school performance seriously. Some were concerned about earning high grades while others were concerned about passing grades. Problems ranged from trouble understanding particular subjects to having a job that interfered with study time, negatively impacting grades. Concerns were also raised about the level of difficulty of classes and people not understanding what a difficult time they were having grasping material. Some students who were likely performing well in school were concerned about grades being high enough to make the honor roll or about being at the top of their class.

Relationships

The second prominent theme involved relationships. Participants had insight into their own relationship problems, including those with family members and significant others. Concerns about parents included divorce or separation, affairs (“cheating” on the other parent), not getting along with step-parents, not getting enough love at home, arguing with parents, not seeing each other’s point of view, and not spending enough time with parents. Other problematic situations included living with grandparents, fighting with family members over “stupid” things, and fighting or breaking up with girlfriends or boyfriends.

Dislikes about school

Participants identified several things that they disliked about school, ranging from quality of the education to quality of the food. Some were concerned about preparation for college, including whether or not they were taking enough college prep classes. Other students thought that their classes were not challenging enough or “moved too slowly.” Additional concerns included the choice of electives and lack of athletics. A few participants disliked particular teachers or coaches, and some accused school personnel of favoritism. Others made suggestions that would improve their educational experience, such as upgrading computers. Complaints were also made about rude people, long hours, small schools, the dress code for physical education, short lunch periods, vending machines purposely turned off, uncomfortable temperatures (too cold or too hot), and the poor condition of restrooms.

Friends

Participants had several concerns about friends. These concerns may have had a negative effect on the participants themselves, their friends, or both. Students discussed fighting with friends and losing them as a result, or fighting with a friend over a girl. In some cases, close contact would be lost because friends were moving away or would graduate soon. Some students struggled to make new friends. Other concerns about friends included depression, suicide attempts, trying to control each other’s lives, not getting along with parents, pregnancy, making poor choices, drinking and using drugs, overdosing, allegedly being abused, illness, anorexia, and belonging to gangs. Others were concerned about the way they were treated by friends, reporting that they had been “stabbed in the back” or lied to. Some participants were also concerned about not having enough time to spend with friends and trying to please friends.

Death

Another prominent theme was death. Concerns included the deaths of relatives (parents, grandparents, aunts, uncles, and cousins), friends, and a pet. One student discussed how his mother developed cancer and died. Another felt sad because her grandmother died when she was a child, and her grandfather died just one year ago. One participant lost a close aunt over Fourth of July weekend. Another discussed losing a “blood brother” as well as a best friend who died in gang violence. A male student suffered a significant loss when his friend’s father died, and another was concerned about the death of his 19 year old neighbor who killed himself. An additional concern was “What happens to you when you die?”

Mental Health

A variety of mental health problems were discussed. Depression and suicide attempts were the most concerning problems of participants. One student experienced depression combined with anxiety. Jealousy, self-consciousness, worrying about what others think, worry in general, family dysfunction, rebelling against parents, stress, and being angry over little things also had negative impact on the mental health of participants and their ability to function. Others were concerned about being happy or finding reasons to be happy.

Future

Students who participated in the current study also expressed concerns about the future. Some participants wondered what life will be like after they graduate from high school, whether jobs will be available, and how they will earn a living. Related concerns included money and the possibility of not succeeding in life. Another student was concerned about not being able to play pro basketball. College was an important future concern, with some participants wondering if they would be able to go to college. Some students had not decided upon a career. Others were concerned about their personal lives. One wondered if marriage would lead to divorce and another wondered if he could handle life in the real world.

Discussion

The qualitative questions asked in this study added to our knowledge about the use of HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) for suicide risk assessment of rural youth. We explored self-reported psychosocial problems and concerns of rural adolescents in Southwestern Pennsylvania, identified themes in the data, compared those themes to known risk factors for suicide, and determined if HEADSS is comprehensive enough to use as a guide for suicide risk assessment of Southwestern Pennsylvania rural youth based on identified themes. Three questions were addressed.

What are the self-identified problems and concerns of rural high school students in Southwestern Pennsylvania?

Using qualitative content analysis, we identified the prominent themes of academic performance, relationships, dislikes about school, friends, death, mental health, and the future (Table 3). Several minor themes were also identified (Table 4). Most problems and themes were discussed in prior Puskar studies of Southwestern Pennsylvania rural youth, including stressors (school, family, loss of a family member, friends, dating), the future, loneliness, friends, suicide, drugs and alcohol, depression, anxiety, anger, somatic complaints, and risk behaviors (Puskar & Martsolf, 1993; Puskar et al., 2003; Puskar et al., 1999a and 1999b). The prominent themes of dislikes about school and death of friends were identified in this study only. Perhaps the qualitative nature of the questions enabled participants to discuss particular concerns such as dislikes about school. These dislikes may have been due to the inability of school districts to provide particular services and amenities because of limited tax revenue (Center for Rural Pennsylvania, 2007, September/October). Concerns about the death of friends were also identified in this study only. At least two friends who died were adults (the friend of a father and a neighbor) rather than peers. Perhaps this demonstrates features of rural culture—that residents know one another and one another’s families, and that they have a strong sense of loyalty to each another, especially during crises (State and Territorial Injury Prevention Directors Association Rural Youth Suicide Prevention Workgroup, 2008).

Several psychosocial issues identified by rural youth in states other than Pennsylvania were also identified by participants in this study, including mental health concerns, stressors, relationship and family issues, illness and deaths of significant others, and future plans. Sexuality issues, abuse, and drug and substance use were only discussed by a small number of participants (minor themes). Perhaps participants were concerned about confidentiality, despite reassurances. AIDS was discussed by just one student as the cause of death of a cousin. Perhaps AIDS was not a prominent concern because only seven percent of AIDS cases in Pennsylvania have occurred in rural areas (Center for Rural Pennsylvania, 2000, July/August). Another concern, paying for education after high school, was not discussed. Participants may have expected to obtain financial aid or scholarships. A few were concerned about being prepared for college or being accepted into a “good” college. Others wondered if they would attend college at all.

Do identified problems and concerns include known risk factors for suicide?

Most risk factors for suicide were themes in this study (Table 2). These factors included problems with family relationships and other interpersonal conflicts, lack of support, school problems, legal issues and violence, substance abuse, neglect and abuse, forced sexual relations, history of suicide attempts and attempts by family and friends, family psychopathology, conduct problems, self-esteem issues, hopelessness, stressful life events, losses, extreme perceptions of weight, physical illness, and low psychosocial adjustment. Risk factors that were not identified by rural participants included not being in school (since only active high school students participated in this study). In addition, no participants discussed dropping out of school, although some were concerned about earning passing grades. None discussed carrying a weapon at school, although one participant was concerned about a gun being stolen. Perhaps the lack of concern about firearms is an indication of the acceptability of guns in rural areas. Other concerns that were not discussed in this study included binge drinking, gay/lesbian/bisexual/unsure status and mobility impairments. This may have been due to concerns about confidentiality, despite reassurances. Whether any participants had mobility impairments is unknown. While physical/emotional/sexual abuse was an identified problem, it is unknown whether any participants experienced a lifetime history of abuse, another risk factor for suicide. No information was available about the use of formal (teacher, minister/priest, clinician) or family networks.

Do the problems and concerns of participants in this study fit the domains of the HEADSS assessment or should a modified version of HEADSS be used when performing suicide risk assessment of rural youth?

Except for death and safety, all themes identified in this study are an obvious part of existing HEADSS domains (Table 5). Death is important because suffering the loss of a preexisting relationship due to death has been identified as a risk factor for suicide (Brent, Perper, Moritz, Baugher, Roth et al., 1993). Several minor themes identified in this study concerned safety, an important topic addressed by nurse practitioners via counseling (Anonymous, 1998) (Table 6). Since threats to safety may result in morbidity and mortality, the addition of Safety to HEADSS must be considered. In their study of adolescents who presented to an emergency department with non-fatal firearm injuries, Paris et al. (2002) added Safety. Goldenring and Rosen (2004) added Eating and Safety. Mattel Children’s Hospital of UCLA (Counsel Teens About Drugs and Alcohol, 2001, September) suggested that the Ss represent Sexual identify or activity, Suicide or depression, and Safety.

Table 5.

Proposed HEADDSSS Domains, Corresponding Themes, and Sample Interview Questions for Assessment

	Corresponding Themes
HEADDSSS Domains	Prominent	Minor	Sample Interview Questions
Home	Relationships	Parental approval or disapproval	• How do you get along with your mother/father? him/her? What do you think of
		Relationship between parents	• How do you get along with your sister and brother?
		Parent’s significant other or spouse	• Describe your relationship with your stepfather.
			• Do you live with your parents?
		Home life	• Are your parents divorced? What has it been like for you?
		Caring for own children	• Being a parent and going to school can be tough. How has it been for you?
Education	Academic performance	Classmates	• How have your grades been this year? What were they?
		Punishment at school	• What do you like about school? Dislike about school?
	Dislikes about school	Favoritism at school	• What concerns do you have about school?
		Attending high school	• How do get along with your classmates?
Activities	Friends	Organized Sports Time	• Sometimes people do things for fun that aren’t safe. What about your friends? What about you?
		Peer pressure	• What do you like to do in your spare time?
		Organized activities	• What kinds of things do you like to do with friends?
		Current job	• Where do you work? What do you like about it? Dislike about it?
		Religion	• Do you have any religious beliefs you would like me to be sensitive to?
Death	Death	Pet	• It can be difficult to lose someone we love. Have you lost any members of your family? A grandparent? A pet?
			• Have you lost any friends?
			• Have you had any other losses in your life?
Drug use and abuse	None	Alcohol/drug use	• Sometimes kids drink or use drugs to fit in. How about your friends?
		Drug problems of classmates	How about you?
		Exposure to alcohol/drug use by others	• Have you ever drunk alcohol? Smoked marijuana? What other drugs?
		Smoking weed	• What do you think about drinking and driving?
		Chewing tobacco	• What if the driver of the car in which you were riding was drinking?
			How would you keep yourself safe?
Sexual behavior	None	Sexual activity and pregnancy	• Are you sexually active with males, females, or both?
			• Do you use any type of contraception or methods to prevent pregnancy? What do you use?
			• Do you use any barriers? What type(s)?
Suicidality and depression	Mental health	Physical health	• Is there anything you would change about yourself if you could?
	The Future	Labeled/rumors	• Sometimes when people are depressed they think about hurting or killing themselves or someone else. Have you thought about hurting or killing yourself? Someone else?
		Moving
		Prejudice/hate
		Expectations	• What do you plan to do after graduation?
		Concerns about the world, environment, or people	• What are your plans for the future?
		Suicidal ideation	• Have you lost interest in things that you use to enjoy? Tell me about them.
		Alleged abuse
		Concerns about weight	• Have you ever been physically hurt by someone?
		Behavior	• If you had three wishes, what would they be?
		Making choices	• Do you worry most days? What kinds of things do you worry about?
		Boredom	• How is your sleep? Do you have trouble getting to sleep? Wake up in the middle of the night and can’t get back to sleep? Wake up earlier than you need to in the morning and can’t get back to sleep?
Safety	Safety	Bullying	• Are you teased or bullied by anyone?
		Motor vehicle accidents	• Has anyone ever touched you without your permission?
		Trouble with the law	• Have you ever gotten into physical fights? Were you or anyone else hurt?
		Alleged abuse
		Harassment	• Have you ever gotten into trouble with the law?
		Drowning	• Have you ever been part of a gang?
		Gun	• Have you ever committed a crime? Ever had a crime committed against you?
		Violence
		Friends in gangs	• Have you ever run away? Thought about running away?
		Running away	• Are there any guns in your home?

HEADSS content from Cohen, Mackenzie, and Yates (1991).

Note. Interview questions are based on the clinical experience of the authors.

Perhaps Viner’s (2003) view of HEADSS (Home, Education, Activities, Drug use and abuse, Sexual behavior, and Suicidality and depression) is correct. He referred to the Suicide domain as “shorthand” for self-harming behavior, depression, and other mood disturbances. The meaning of HEADSS “shorthand” may depend upon the interpretation and experience of the individual clinician. Thus, whether to expand HEADSS may depend upon clinician preference. However, we believe that the safest and most thorough approach based on the results of this study may be to expand HEADSS to HEADDSSS--Home, Education, Activities, Death, Drug use and abuse, Sexual behavior, Safety, and Suicidality and depression. The remaining themes identified in this study could be incorporated into existing domains (Table 5).

Limitations and Future Considerations

The current study had several limitations. Because it was a secondary analysis of qualitative data, it did not provide an opportunity to study the use of HEADSS in clinical practice. The study was also limited to adolescents in rural Southwestern Pennsylvania. Results may not apply to other rural areas or cultures. Questionnaires were administered to participants sequentially, and some responses may have been influenced by previous questions or answers. Since data was analyzed after data collection ended, there were no opportunities to clarify meaning with participants. An additional limitation was the reliance on student self-report. Problems and concerns were based on participant perceptions. Problems such as drug use were discussed, but frequency of use was not obtained. Despite reassurances, students may not have reported concerns about drug/substance use, safety, sexuality, and mental health due to concerns about confidentiality. In addition, some students chose not to participate in this study. Their problems and concerns are unknown. Despite these limitations, themes identified in this study included most risk factors for suicide and several psychosocial problems identified in the literature review.

Future research should include duplication of this study with rural populations having high adolescent suicide rates. This will help to determine if death and safety or any other domains need to be added to HEADSS. Clinical charts must be audited to determine if risk assessments using HEADSS are documented and thorough, that appropriate mental health referrals have been made, that youth are participating in treatment, and that rates for suicides and suicide attempts are declining. In order for referrals to occur, mental health outpatient services must be available to rural youth. Telepsychiatry, a method of providing mental health services via phone, the Internet, and one- or two-way televideo, is currently being piloted in Greene County in Pennsylvania (Office of Mental Health and Substance Abuse Services, 2006).

Summary and Implications

The purpose of this qualitative secondary analysis was to determine if HEADSS is an appropriate interview instrument to use for suicide risk assessment of rural youth. The study added to our knowledge about HEADSS by providing initial evidence that two new domains, death and safety, may need to be added to HEADSS. Expanding HEADSS would help to ensure that risk assessments are thorough and that clients remain safe. The modified version, HEADDSSS (Home, Education, Activities, Death, Drug use and abuse, Sexual behavior, Safety, and Suicidality and depression), is the expanded acronym that can be used for suicide risk assessment of rural youth in Southwestern Pennsylvania. The remaining prominent themes from this study—academic performance, relationships, dislikes about school, friends, mental health, and the future—can be assessed using the original HEADSS domains (Home, Education, Activities, Death, Drug use and abuse, Sexual behavior, and Suicidality and depression).

Many known risk factors for suicide were concerns of participants in this study. The study showed that these factors can be incorporated into domains of the modified HEADDSSS acronym, illustrating its comprehensiveness in the identification of suicide risk.

This study must be duplicated with other rural adolescent populations to determine if HEADDSSS is the most appropriate interview instrument for suicide risk assessment of America’s rural youth. The goal is to prevent the lost years of life, grief, guilt, and devastation that suicide brings to individuals, families, and communities.