Evaluation of a Partial Hospitalization Program for Adolescents

Abstract

This study investigated the magnitude of treatment and client perceptions of change within a partial hospitalization program (PHP) for adolescents with mental health concerns. Participants were 35 adolescents (19 girls; 16 boys) with a mean age of 14.91 years who were predominately Caucasian (n = 31) and experiencing the symptoms of mood disorders (n = 30). Results of a mixed methodology sequential explanatory design indicated that a 6-week PHP effectively decreased symptom severity and increased relational health of participants over time. Qualitative data yielded four themes identified as renewed well-being, relationships, perceptions of effective programming, and areas of dissatisfaction.

Keywords

program evaluation specialty quantitative efficacy treatment children

Adolescence is a critical period for cognitive, social, and emotional maturation during which individuals undergo notable developmental changes that establish behavior patterns that may persist into adulthood. Negotiating the complexities of adolescence to attain a sense of identity and self-esteem can be challenged by the presence of biological or environmentally based mental illness. Some estimates indicate that nearly 20% of adolescents in the United States experience some form of mental health concerns prior to adulthood (Kessler, Berglund, Demler, Jin, & Walters, 2005) and that target populations are difficult for counselors to identify, given the reports that gender differences vary across and within mental health disorders (American Psychiatric Association, 2013; Zahn-Waxler, Shirtliff, & Marceau, 2008). This is troublesome when considering that the presence of mental health concerns such as depression, anxiety, and bipolar disorders are strongly related to one’s professional and personal development over the life span (Schwarz, 2009). There is a positive association between the degree that adolescents experience the symptoms of a mental health disorder and the status of adult career success, romantic involvement, academic achievement, and day-to-day happiness (Breslau, Lane, Sampson, & Kessler, 2008; Lenz, Holman, & Dominguez, 2010; Utay & Utay, 2005). In the case of adolescents, there is some evidence indicating that untreated mental health concerns may predict deleterious outcomes for the future such as involvement with the criminal justice system (Coll, Juhnke, Thobro, Haas, & Robinson, 2008; Coll, Thorbo, & Haas, 2006). Given the high potential for deleterious outcomes associated with unmitigated psychopathology in adolescence, it is a prudent task for counseling professionals to identify and evaluate intervention approaches that may contribute to symptom remediation and relational well-being.

Individuals with a mental illness are treated across several settings in accordance to the severity of their symptoms and level of functioning. Several mental health professionals have indicated that the best model of treating adolescent mental health may be one that provides the appropriate combination of medical, mental health, and case management services treatment within the least restrictive environment (Garfield, Love, & Donohue, 2010; James et al., 2006). One approach to continuum of care has treatment occurring in inpatient hospital settings for the most severe clients, intensive outpatient/partial hospitalization for those experiencing moderately severe distress, outpatient care for those with routine adjustment and distress issues, and aftercare for those maintaining treatment gains. For families searching for the least restrictive environment to adequately address the mental health needs of their adolescent, partial hospitalization program (PHP) may be one helpful approach for acute psychiatric conditions that may not require hospitalization, yet are too severe for routine outpatient care (James et al., 2006); treatment in PHP settings may also be the most effective strategy for containing costs to third-party and private payers (Garfield et al., 2010). Although many of the individual services provided as part of PHP treatment are supported by evidence, there are fewer studies that have evaluated the utility of programs as a whole for promoting mental health and relational development for adolescents.

The purpose of this study was to evaluate the treatment effect of a PHP program providing services to adolescents with mental health diagnoses. Our study was intended to answer the following questions: (a) To what degree is PHP treatment associated with change in the severity of mental health symptoms from admission to discharge? (b) To what degree is PHP treatment associated with changes in relational health from admission to discharge? and (c) What factors do clients and counselors attribute to observed changes from admission to discharge?

Method

To evaluate the treatment effect associated with a PHP for adolescents, we implemented a mixed methodology sequential explanatory design with a single group of participants who were completing treatment. When implementing this approach, we first collected and analyzed quantitative data and then developed an associated qualitative strategy. This approach was applied with the rationale that single-group studies are limited in making causal attributions related to treatment effect; instead, a greater emphasis was placed on understanding the magnitude of treatment effect and identifying participants’ perceptions about why treatment was helpful for them.

Participant Characteristics

Participants were 35 adolescents (19 girls, 54%; 16 boys, 46%) with a mean age of 14.91 years (standard deviation [SD] = 1.35) who were completing treatment at an intensive outpatient mental health program in the Mid-Southern region of the United States. Participants were predominately Caucasian (n = 31; 88%) with others identifying as either African American (n = 1; 3%), Asian American (n = 1; 3%), or other (n = 2; 6%) ethnic identities. Among participants, the majority of Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR) primary diagnoses at admission were mood disorder, not otherwise specified (n = 24; 69%) and major depressive disorder (n = 6; 17%) with other variations in anxiety disorders and substance abuse disorders also represented.

Measures

Symptom checklist-90–revised (SCL-90-R)

The SCL-90-R (Derogatis, 1994) was developed to assess an individual’s current psychological symptom status across community, outpatient, and inpatient psychiatric populations. Although developed with clinical use in mind, the SCL-90-R has been used by several researchers to detect therapeutic variables associated with the treatment of adolescents (Biegel, Shapiro, Brown, & Schubert, 2009; Groves, Backer, van de Bosch, & Miller, 2012; Self-Brown et al., 2006).

The SCL-90-R is a 90-item self-report questionnaire that yields nine symptom dimensions: anxiety (ANX), depression (DEP), hostility (HOS), interpersonal sensitivity, (INT), obsessive–compulsive characteristics (O–C), paranoid ideation, (PAR), phobic anxiety (PHO), psychoticism (PSY), and somatization (SOM). The response format on the SCL-90-R is a 5-point Likert-type scale requiring individuals to assess symptom characteristics along a continuum of frequency and magnitude ranging from 0 (not at all) to 4 (extremely). Higher scores on the SCL-90-R indicate a greater degree of symptom severity, with the nine dimensions contributing to a Global Severity Index that represents the current seriousness of an individual’s psychological distress. Examples of scale items include “the feeling that something bad is going to happen to you” (ANX), “feeling hopeless about the future” (DEP), “having urges to break or smash things” (HOS), “feeling very self-conscious with others” (INT), “having to check and double-check what you do” (O-C), “feeling most people cannot be trusted” (PAR), “feeling afraid to go out of your house alone” (PHO), “hearing voices that other people do not hear” (PSY), and “hot or cold spells” (SOM). Derogatis (1994) reported Cronbach’s αs for scores across all dimensions within the good range (.77 to .88) with test–retest reliability within the acceptable to excellent range (.68 to .90).

Relational health indices-youth (RHI-Y)

The Relational Health Indices–Youth (RHI-Y; Liang, Tracy, Kenny, Brogan, & Gatha, 2010) was developed to evaluate the degree that adolescents believe they are participating in growth-fostering relationships characterized by authenticity, empowerment, and engagement. The theoretical structure underlying the RHI-Y has been demonstrated as reliable and valid for use with both women and men (Liang, Tracy, Glen, Burns, & Ting, 2007; Liang, Tracy, Taylor, Williams, & Jordan, 2002) and with adolescents (Liang et al., 2010). Although a newer instrument, preliminary research has revealed the utility of the RHI-Y and its related constructs for evaluating the influence of psychosocial variables on the relational health of adolescents (Grossman & Liang, 2008; Lenz, Speciale, & Aguilar, 2012; Liang et al., 2010).

The youth version of the RHI (Liang et al., 2010) is an 18-item self-report questionnaire that represents three qualities of growth-fostering relationships (authenticity, empowerment, and engagement) evaluated between Friends, Mentors, and the Community at large. The response format for the RHI is a 5-point Likert-type scale ranging from 1 (never) to 5 (always) and requires participants to assess variables associated with relationships. Higher scores on the RHI-Y represent higher quality of relational health. Three subscales can be calculated to assess relational domains of Authenticity, Empowerment, and Engagement within each of the composite domains. Authenticity is conceptualized as “the process of acquiring knowledge of self and others and feeling free to be genuine in the context of the relationship” (Liang et al., 2002, p. 26) and is represented by statements such as “Even when there is something hard to talk about, I can be real with this friend” as depicted in the Friend subscale. Empowerment is defined as “the experience of feeling personally strengthened and inspired to take action” (Liang et al., 2002, p. 26) and is represented in the Mentor subscale by items such as “I feel happy after being with my mentor.” Engagement is described as the “perceived mutual involvement, commitment, attunement to the relationship” (Liang et al., 2002, p. 26) and is depicted in the Community subscale by items such as “People in this group believe and support me.” Liang, Tracy, Kenny, Brogan, and Gatha (2010) reported Cronbach’s αs for Friend, Mentor, and Community subscale scores within in the good range (.83, .81, .86 respectively).

Context of Study

This evaluation was completed at a PHP facility for children and adolescents with moderate to severe emotional and behavioral disorders through a time-limited program that offers therapeutically intensive, coordinated, and structured services. Clients range in age from 12 to 18 years, with a daily mean census of 12, and are representative of a cross section of diagnoses typical of this age group. The program is designed to be 6 weeks in duration, followed by 12 weekly aftercare groups. The PHP model being evaluated is informed by the psychoanalytic perspective of milieu treatment as a therapeutic holding environment (Stramm, 1985; Winnicott, 1965) and systems-based structural family therapy (Minuchin, 1974). The PHP setting was organized to meet the five functions of milieu therapy articulated by Gunderson (1978): containment, support, structure, involvement, and validation. Containment is achieved through assuring the physical and interpersonal safety of each client. Providing a safe setting is intended to promote perceptions of trust and security, which is the basic foundation that all other treatment efforts are contingent upon. Support refers to the staff’s ability to stimulate a feeling of hope in clients and foster adaptive, growth-oriented attitudes and behaviors. Structure is established through the daily schedule, implementation of rules/consequences, and behavior management practices. Involvement refers to features that prompt clients and family members to attach value to therapeutic community membership and reaching goals established during individual care planning. Finally, validation of each client’s unique treatment needs is accomplished by the primary therapist who develops and implements a problem-oriented individual treatment plan.

Specific intervention modalities implemented within this PHP are designed to support success in decreasing symptom severity, acquiring coping skills, interpersonal relationships, abstinence from substance abuse, and academic achievement. This PHP featured opening and closing goal setting meetings, individual counseling, interpersonal process groups, alternating group therapy modalities, and school attendance (see Table 1). Program requirements provided weekly parent support groups that augment weekly individual and family therapies provided by master’s level counselors. During daily goal setting meetings, clients are assisted in identifying achievable objectives related to their treatment plan as well as problems in the program and home/family setting. Individual counseling is provided utilizing an eclectic approach to help clients maintain a commitment to change and modify their emotional, behavioral, and/or personal interactions with others in a manner that improves overall adaptive functioning. Process groups facilitate discussions and experience that promote client insight into interpersonal issues, learning to provide feedback to others, collaborate for problem solving, and prepare for effective utilization of other treatment modalities. Alternating group modalities are provided on a rotating basis and include material related to psychoeducation, social skills, expressive arts, recreation, and stress management. Family counseling is based on principles of structural family therapy and designed to address contextual determinants of dysfunction within the family structure while providing a forum to maintain family commitment to care and to address aftercare and discharge planning needs. Clients also attend an in-house state approved school program where an educational plan is prepared for each client which includes academic, emotional, and behavioral objectives.

Table 1.

Curriculum Template for a Partial Hospitalization Program for Adolescents.

Time	Monday–Friday
7:45 a.m.	Arrival and preparation for morning meeting
8:00 a.m.	Morning meeting and goal setting
8:30 a.m.	Educational therapy and academic preparation
11:00 a.m.	Interpersonal process group
12:00 p.m.	Lunch and personal duties
12:30 p.m.	Group therapy^a,b,c
1:45 p.m.	Closure group and goal setting
2:00 p.m.	Dismissal

Note. ^aParent group alternating with multifamily therapy once weekly. ^bIndividual and family sessions are scheduled a minimum of twice weekly. ^cGroups alternate between art, music, and recreational therapies.

Procedure

All participants in the quantitative evaluation completed the 6-week PHP as described previously. Prior to their admission interview, each participant completed a small battery of self-report formal assessments that included the SCL-90-R and the RHI-Y; the same assessment battery was administered to clients at discharge from the program. All data and other personal health information were entered into a database by PHP staff and maintained on location in an encrypted, secured format consistent with the Health Insurance Portability and Accountability Act and The Joint Commission accreditation standards. A secondary analysis of the deidentified quantitative data collected by the PHP was completed by the first author.

Data informed qualitative interviews were completed with two focus groups of participants and PHP service providers by the third author to identify perceptions of factors that contributed to therapeutic changes. The purpose of the focus groups was to elicit the clients’ perceptions of therapeutic experiences within the PHP and develop the context for understanding the quantitative results. Data were collected through semistructured interviews to permit the researcher to adapt questions as our understanding developed yet create consistency across interviews (Lincoln & Guba, 1985). Participants for the focus groups were selected based on their age (12 years and older) and proximity to program completion; children 12 years and under were excluded from the focus groups due to developmental tendencies that preclude them from articulating abstract thoughts related to complex experiences (Piaget, 1963). No boys met the criteria for minimum age or time in the program and therefore were not represented in the focus groups. The two focus groups consisted of four girls per group who were asked to respond to the following questions: (a) Tell me about the counseling programs you participate in? (b) Tell me what you like about (therapeutic modalities)? (c) Is one type of therapy that you believe is more helpful than another? (d) Have you noticed any changes in yourself since you started? (e) Have you noticed any changes in the way people are responding to you? (f) Are you using anything that you’ve learned in therapy in your life? (g) Is there anything that you do not like or find hard about therapy?

Data Analyses

Quantitative

Quantitative analyses were implemented to demonstrate significant and practical changes within a single group of participants while receiving intensive outpatient counseling program.

Statistical power analysis

An a priori power analysis was conducted using G*Power 3.1.3 statistical power analysis program (Faul, Erdfelder, Lang, & Buchner, 2007) to determine the number of participants needed to establish statistical power for this research design at the .80 level given α = .05. The power analysis indicated that a sample size of 27 was necessary to detect a moderate effect of change over time for the outcome variables. A post hoc sensitivity power analysis revealed that given our sample of 35, a critical t value of 1.67 was required to reject the null hypothesis that no meaningful changes occurred during treatment.

Preliminary analysis

Raw scores on the SCL-90-R were calculated using the procedure described in the instrument manual and converted to age and gender-referenced T scores; RHI total scale scores were calculated using the protocol provided by the instrument’s authors. Descriptive statistics, means scores, SDs, and coefficient αs were computed for each measure and are presented in Table 1. Reliability coefficients for the SCL-90-R dimensions were within the good to excellent range (.75 to .94); α coefficients for the RHI-Y Friend, Mentor, and Community subscales were within the acceptable to excellent range (.93, .88, and .73, respectively).

Primary analysis

Paired-samples t-tests were conducted to assess statistically significant changes from intake assessment to discharge. To obtain a more accurate depiction of PHP effectiveness, Cohen’s d effect sizes estimates were computed and interpreted using guidelines suggested by Cohen (1988) for regarding magnitudes of effect sizes as small (d ≥ .20), medium (d ≥ .50), and large (d ≥ .80).

Qualitative

Data from the focus group were analyzed utilizing phenomenological procedures developed by Moustakas (1994). Transcripts from the focus groups were read repeatedly to identify significant statements to establish an understanding of how the participants experienced the phenomenon. Significant statements were subsequently clustered into meaning units, which were reworked into emergent themes. Data analysis was further distilled through the written description (textured descriptions) of each theme describing the essence of participants’ experience (Moustakas, 1994). Triangulation of data was achieved through member checking (Patton, 2002), a process that involved the presentation of themes to participants to provide feedback on their accuracy and relevance. An expert auditor who was not involved in the current research project analyzed the data to develop an independent set of themes. The convergence of themes derived from the audit to the original themes provided the second source of triangulation. Finally, the close fit between this study’s qualitative and quantitative data sources offers additional verification to support the themes presented herein.

Results

Psychological Symptoms

The results of paired samples t-tests revealed a number of statistically significant findings related to the changes in psychological symptoms over time for participants completing the PHP (see Table 2). The t-tests revealed that participants reported significantly fewer symptoms of anxiety (ANX), t(34) = 2.48, p < .01, d = .55 and depression (DEP), t(34) = 2.79, p < .01, d = .77 indicative of medium and medium to large effect sizes respectively. This finding suggests that participants completing the PHP tended to report a decrease in severity or frequency of the symptoms representing these constructs such as tension and apprehension or dysphoric mood. Statistically significant changes over time were also detected for paranoid ideation (PAR), t(34) = 3.61, p < .01, d = .78 and obsessive–compulsive (O–C), t(34) = 2.51, p < .01, d = .70 indicative of medium to large effect sizes. These findings indicate that not only were participants less likely to engage in suspicious, projective thought about others, but that they were also less likely to endorse items related to unremitting or irresistible thoughts. Although participants reported statistically significant changes in hostility (HOS), t(34) = 2.77, p < .01, d = .60 indicative of a medium effect size, rating of interpersonal sensitivity (INT) did not reveal meaningful change over time t(34) = 1.92, p = .06, d = .47. This suggests that despite endorsing fewer items related to aggression, resentment, or rage, participants continued to report similar levels of acute self-consciousness and self-directed resentment at discharge as they did when admitted to the PHP. The t-tests revealed that although marked differences were noted in items related to psychoticism and a schizoid lifestyle (PSY), t(34) = 2.73, p < .01, d = .76 indicative of a medium to large effect size, meaningful changes in phobic anxiety (PHO) item endorsement was not detected t(34) = 1.00, p = .32, d = .25. These findings suggest that despite depicting fewer experiences of a withdrawn, isolated lifestyle, participants demonstrated similar levels of avoidance to certain places or situations. Finally, the results of a t-test revealed that participants experienced significantly fewer somatic complaints (SOM) at discharge when compared to admission to the PHP t(34) = 2.18, p < .05, d = .56 indicative of a medium effect size.

Table 2.

Reliability Coefficients and Comparisons of Admission and Discharge Assessments for Symptom Checklist-90–Revised Dimensions and the Relational Health Indices–Youth Subscales.

	α	Admission M (SD)	Discharge M (SD)	t	ES
SCL-90-R
ANX	.86	41.82 (8.22)	37.89 (5.77)	2.48**	.55
DEP	.94	45.02 (9.65)	38.03 (8.38)	2.79**	.77
HOS	.78	43.36 (6.30)	39.84 (5.26)	2.77**	.60
INT	.87	43.84 (7.86)	40.46 (6.36)	1.92	.47
O-C	.82	45.59 (7.39)	40.89 (5.88)	2.51**	.70
PAR	.75	44.16 (5.20)	40.08 (5.14)	3.61**	.78
PHO	.79	44.23 (5.68)	42.92 (4.38)	1.00	.25
PSY	.90	45.86 (9.74)	39.19 (7.55)	2.73**	.76
SOM	.87	48.30 (9.54)	43.32 (8.35)	2.18*	.56
RHI-Y
Friend	.93	17.73 (4.22)	19.88 (3.50)	2.31*	.55
Mentor	.88	16.01 (3.90)	19.50 (8.61)	2.10*	.52
Community	.73	16.37 (3.26)	16.64 (4.67)	.31	.06

Note. SCL-90-R = Symptom checklist-90–Revised; RHI-Y = Relational health indices-youth; SD = standard deviation; ES = Cohen’s d effect size interpreted as small (ES ≥ .20), medium (ES ≥ .50), and large (ES ≥ .80).

*p ≤ .05. **p ≤ .01.

Relational Health

The results of paired samples t-tests revealed statistically significant findings for two of the three RHI-Y subscales related to changes in relational health over time for participants completing the PHP (see Table 2). The t-test results demonstrated that participants reported significantly greater perceptions of relational health with peers t(34) = 2.31, p < .05, d = .55 indicative of a medium effect size. This finding suggests that participants endorsed items related to a greater sense of relational authenticity, empowerment, and engagement with meaningful peers at discharge than during admission. Statistically significant changes over time were also detected for relational health with mentors t(34) = 2.10, p < .05, d = .52 indicative of a medium effect size. Similar to the experience with peers, this finding suggests that participants endorsed items related to experiences of relational authenticity, empowerment, and engagement with a mentor to a greater degree at discharge than admission. Finally, the results of a t-test revealed that there was no significant change in the perceptions of relational health with their community at large following completion of the PHP t(34) = .31, p = .75, d = .06.

Perceptions From Participants

Analysis of focus interviews with participants yielded four themes related to perceptions of treatment within a PHP: renewed well-being, relationships, perceptions of effective programming, and areas of dissatisfaction.

Renewed well-being

The majority of girls interviewed (7 of 8) expressed that they were experiencing a greater sense of internal well-being since commencing the program 5 weeks earlier. The member who suggested that she felt less adjusted had experienced a breach in her stay, due to surgery that caused a 2-week absence, and struggled to “fit in” with a new set of peers she encountered upon her return. Nonetheless, the girls who experienced an uninterrupted stay noted that they felt more in control of their emotions, reflected by comments such as, “I’m less up and down,” “I’m acting more mature … my mood is a little wonky but much better.” Girls suggested that the shift occurred around the second to third week into the program when they began to see a difference resulting from the application of skills learned in therapy to various contexts of their lives. “After about 2 week I felt better,” “Once I learned to control my behavior things started turning around the end of the second, no the third week I was here.” Through therapy they had learned to recognize triggers and traps that escalated their behavior, which in turn had permitted them to modify their responses. “I noticed that my Dad likes to escalate things—I used to escalate with him, but I learned not to respond the same way.” Another participant commented that “I changed my responses to them (family members) and that changed how they reacted.” All the girls were scheduled to return to their regular schools the following week, and while they felt both nervous to return and sad terminating the program, they did express feeling equipped with the resources to cope with the transition.

Relationships

The theme of “relationships” provides the catalyst for understanding the girls’ renewed sense of well-being and factors of the counseling programs identified as most curative. Underlying the girls’ reticence to depart the PHP were the close relationships forged with one another and their counselors over the duration of their stay. The girls’ interactions during the focus groups were marked by a familiarity, mutual support, and respect that belied the brevity of their relationships, and the program’s ban on communication between clients outside the center during their stay. The girls indicated that they had entered the treatment center within the same week and completed the program in unison, effectively forming an ad hoc cohort. “We’ve been here together pretty much from the beginning,” “We’ve grown together,” “We listen,” “We give advice and encouragement,” and, “It has helped to keep me motivated when I’m having a bad day.” The close peer relationships were paralleled by girls’ relationships to their counselors, as both focus group participants pronounced their attachment to their therapists. Relationships with counselors were described as, “maybe the best thing about being here.” Members of one focus group emphatically agreed with a members’ statement that “Our therapists really helped us to develop … and see ourselves, and situations differently—I wish I could keep Ms. (names therapist).” In essence, the peer and counselor relationships created a therapeutic milieu that transcended individual programming and underscored the positive change experienced by the girls in the focus groups.

Perceptions of effective programming

Focus group members suggested that they participated in two daily groups (morning check in and afternoon process group), psychoeducational presentations, weekly family and individual counseling, and art therapy. The ameliorative effects of relationships developed within the PHP reemerged in the girls’ identification of programs they perceived as most helpful. Process group and individual therapy were overarchingly cited as their favored counseling modalities. When asked to identify therapeutic aspects of these programs, the girls suggested the process group created an atmosphere of freedom wherein anything could be explored. “I like to vent and share in the process groups-sometimes that’s enough.” Girls appreciated the insight gained through peer feedback and the opportunity to challenge and “confront” group members. “I’ve learned a lot in process group about myself—sometimes it can be tough—but it’s helped and I’ll miss it the most.”

While the process group provided a forum that strengthened peer support, individual counseling helped individualize client experience. All clients receive an hour of individual therapy each week that is often supplemented with an additional hour of therapy with a counseling intern. The girls in the focus group indicated that individual therapy had helped them to “work through stuff” and was instrumental in their growth. In particular, the girls enthused at their counselor’s ability to listen and get to the root of the issue. “My counselor—he’s amazing—he knows exactly the right questions to ask.” Girls commented that they felt understood and accepted by their therapists. In addition to individual therapy, the focus group members highlighted their affinity for art therapy. Art therapy was identified as providing a relaxing, creative, setting that helped them to de-stress. At least two girls mentioned using art therapy at home as a tool of stress management “I’ve been beading at home too—it helps me unwind and block out negative stuff.”

Areas of dissatisfaction

Focus groups were asked to comment on areas of the programming that they perceived as less helpful. The consensus of both groups was that the program’s token economy could be unfairly punitive. The token economy was organized by a point system, wherein points could be constantly gained or lost through engaging, or failing to engage, a target behavior. Hence, desired behaviors such as arriving to school prepared and on time, and participating in school and counseling activities, were reinforced through the accrual of points. Although the girls recognized that the point system provided for order and structure, they believed that points could be lost at a rate disproportionate to points earned. “My dad didn’t sign the sheet and I lost points,” “You can never get a 5 score (total points) in class … even if you really really try you can just get a 4 … but then if you mess up and end up in timeout you instantly lose 10 points.” In addition to perceptions that the point system was unfairly weighted toward attrition of points, girls complained that at times it compromised the therapeutic environment. Because points could be earned through participation in therapy, girls drew attention to instances when, “Everyone talks over one another [in process group] … cos everyone’s trying to get their points.”

Girls in the focus groups were vocal in their dissatisfaction toward punitive aspects of the point system. Other areas of frustration related to the structure of morning routine, which was dominated by a 2.5-hour period of classroom instruction. “I have ADHD and get distracted a lot, I feel like my effort is not rewarded as I try hard then I lose point … I could just do with a break.” Although the process group was heralded as a preferred modality, the girls suggested that the changing membership, resultant from the open group format, had at times interfered with group cohesion and trust. Furthermore, new members increased the group size to the extent that inadequate time was afforded to sharing by individual members. Nonetheless, it was noted that their therapists would divide the group once membership reached a threshold of six. Art therapy was another recognized favorite among the focus groups, yet members noted a similar caveat related to insufficient time allotted to project completion, “We like it but we never have enough time and seems like we never get to finish.” Taken together the point system, changing group size and membership, and time limitations of art therapy comprised the entirety of frustrations expressed by members of the focus groups. When asked to comment on the overall impact of these negative experiences, the girls indicated that the net effect was perhaps peripheral to the overall positive aspects related to experiences within the PHP.

Discussion

The purpose of this study was to evaluate the treatment effect of a PHP program providing services to adolescents with mental health diagnoses. The PHP program that we evaluated was intended to provide therapeutic relief of psychological symptoms using a time-limited, active treatment that was intensive, coordinated, and structured within a stable milieu. Our findings have indicated that over time, participants demonstrated improvement across psychological symptom domains that were both statistically significant and suggestive of moderate to large treatment effect sizes.

Qualitative interviews revealed that participants believed that prolonged stays in the PHP (i.e. more than 3 weeks) provided time to become stable, learn coping strategies, and practice them in a safe context. Participants also reported that although the context that they live in had not changed drastically, their response and associated affective regulation was distinctly more functional. Participants tended to credit group counseling experiences as a catalyst for accountability of using coping strategies and the varied modalities as a creative mechanism for exploring intrapersonal dynamics; individual counseling was attributed to promoting the depth understanding and self-acceptance that reduced depression, anxiety, and other distressing symptoms.

These findings are encouraging when considering the supposition by authors such as Schwarz (2009) that early onset psychological disorders may be related to the degree of adaptive personal and professional development over the life span. Given the assumption that degree of adolescent mental health distress is negatively associated with status of adult career success, romantic connection, academic achievement, and happiness (Breslau et al., 2008; Lenz et al., 2010; Utay & Utay, 2005), it may be possible that PHP interventions can provide an important mitigating experience that promotes later functioning. Furthermore, findings by Coll, Juhnke, Thobro, Haas, and Robinson (2008) and Coll, Thorbo, and Haas (2006) suggest that early remediation of mental health symptoms may be a protective factor associated with fewer criminal activities when compared to untreated adolescents. Therefore, it is reasonable to conjecture that when clients are not experiencing these issues and have learned associated coping skills, they may be less likely to experience negative life outcomes.

Our results indicated that while receiving treatment in a PHP, participant perceptions of relational health with peers and mentors increased significantly; however, increases in relational connection with their community at large were not observed. In particular, participants reported that the familiarity, mutual support, and respect that characterized relationships with peers and counselors were central to all changes and perceptions of resilience noted during treatment. This is complimentary to previous researchers’ indication that the preponderance of therapeutic change during counseling is contingent upon the quality of relationships experienced therein (Horvath, Del Re, Fluckiger, & Symonds, 2011; Shirk, Karver, & Brown, 2011). One explanation for this phenomenon is that when enrolled in a PHP, new styles of coping and perceptions of self-in-relation occur among peers and counselors. As a consequence, learning is generalized to the therapeutic milieu and community as a whole at a less accelerated rate. It may be possible that an emphasis on transferring acquired skills to the community by having task home tasks situated in a broader context would provide movement along this relational dimension. Conversely, programs that detect a similar finding may want to include aspects of community-oriented relational development as a part of their aftercare programming. Additionally, qualitative reports from our participants indicated that the point system for allocating some input into daily success may have discouraging effects on client development if they are too austere in the event of the occasional shortcoming. We encourage programs that implement behavioral point systems to review how point systems are associated with facilitating client change and make accommodations that may promote practical expressions of this intervention that promote client success.

Our findings indicated that participants in a PHP demonstrated positive, desired change over time. For individuals whose symptom severity and level of functioning have created an overall low sense of well-being, PHP programs may provide the structure and support that are required to promote adjustment and resilience. Although reporting some contentions with the PHP structure, participants expressed a general acknowledgment that the routine and structure intrinsic within a PHP program was an impetus for change. Adolescence is an unpredictable time for many teenagers, which can be made tumultuous by the complications associated with mental health concerns; therefore, it is reasonable to suppose that the predictability of scheduling, personnel disposition, rules/consequences, and location of a PHP may be inherently therapeutic for some clients. Furthermore, PHP treatment may provide the appropriate combination of medical, mental health, and case management services within the least restrictive environment that some authors have suggested is needed to maximize treatment gains with adolescents (Garfield et al., 2010; James et al., 2006).

Suggestions for Practice

Based upon our findings, several suggestions for practice have emerged related to relational connection, group dynamics, and counselor competence. Foremost, establishing quality-working relationships is a preeminent condition for development and may supersede treatment modality. Given that many inpatient and PHPs are based on the medical model of treating client symptoms rather than promoting the development of the whole individual, we encourage counselors to conceptualize this relational core as a primary goal of the healing process. Furthermore, we encourage counselors to be aware of the group dynamics that promote supported vulnerability and accountability within the change process. While peer-to-peer groups tended to be more transformational for participants in our study, group size, splitting up, and open enrollment were viewed as barriers to change. Counselors are also encouraged to implement multiple modalities and formats for group therapies. Our participants also communicated a strong liking for the varied nature of group modalities that may stimulate insight and affective processing through a number of affective, behavioral, and cognitive structures. Finally, counselors are encouraged to balance the amount of exposure of clients to counseling interns. Although our participants found practicum and internship students to be a great support and helpful, the overt competence and skill of the master’s level clinicians were acknowledged as a catalyst for growth during treatment.

Limitations and Directions for Future Research

The findings of this preliminary evaluation of a PHP program have encouraging implications for the implementation of therapeutic milieus to treat adolescent onset psychological disorders, some caveats are noted. Although a mixed methodology sequential explanatory design with a single group of participants can provide some estimation of treatment effect over time, we cannot submit conclusive attributions of causal relationships related to treatment effectiveness. Future between-group investigations comparing PHP treatment to clients receiving alternative treatment modalities would support such inferences. In the event that a comparison group is unattainable, PHP programs are encouraged to implement single-case research designs that promote causal inferences based on the use of the participant as their own control group and yield a metric related to treatment effect size. Either of these evaluation strategies would allow PHP program coordinators to establish evidentiary support for their practices, which is consistent with the standards for empirically validated practices in clinical work (Chambless et al., 1996, 1998). The characteristic structure of PHP service delivery lends itself to strong internal validity related to treatment effect from admission to discharge; however, the present design was limited for identifying the degree that treatment gains were sustained following discharge. Future researchers are encouraged to collect longitudinal data during the aftercare programming phase of treatment or some other reasonable interval. Quantitative evaluation of these data would support inferences about the robustness of treatment magnitude; qualitative inquiries may promote an understanding of how the PHP experience has been contextualized by participants and what aspect of treatment has remained meaningful over time. Within these studies, researchers are encouraged to include boys within their focus group interviews to assure that equitable voice is given to their PHP experience. Also, larger sample sizes may promote the use of predictive analytics against clinical criterion that can be used by PHP administrators to identify the optimal length of treatment, which client characteristics (e.g., gender, history of self-harm, and history of substance abuse) are associated with greater treatment gains. Finally, we encourage future researchers to identify and implement assessments that are intended solely for use with children and/or adolescents. Although the SCL-90-R has a long-standing history of use in clinical research and provided a benchmark for comparison, the lengthy format and lack of updated clinical norms were suspect during this evaluation.

Conclusion

The prevalence of adolescent mental health diagnoses has conspicuous consequences for an individual’s adult well-being and, if untreated, individuals can experience a number of developmental impediments. This study has provided some preliminary support for PHPs as a potentially effective strategy for treating adolescents within an intensive therapeutic milieu. Although our sample was small and not compared against and an alternative treatment, we contend that the design was characterized by internal validity related to treatment, use of formal assessments, and demographic consistency. We encourage scientist-practitioners to implement more rigorous research designs and hope that corroboratory results will provide an impetus for the endorsement of PHP practices as an established evidence-supported intervention. In particular, more studies are needed that include larger samples, accurate reporting of data and demographics, judicious use of psychometric assessments, and evaluation of manualized treatment interventions with underrepresented populations.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders-5 (5th ed.). Washington, DC: Author.

Biegel

G. M.

Shapiro

S. L.

Brown

K. W.

Schubert

C. M.

(2009). Mindfulness-based stress reduction for the treatment of adolescent psychiatric outpatients: A randomized clinical trial. Journal of Consulting and Clinical Psychology, 77, 855–866. doi:10.1037/a0016241

Breslau

Lane

Sampson

Kessler

R. C.

(2008). Mental disorders and subsequent educational attainment in a US national sample. Journal of Psychiatric Research, 42, 708–716. doi:10.1016/j.jpsychires.2008.01.016

Chambless

D. L.

Baker

M. J.

Baucom

D. H.

Beutler

L. E.

Calhoun

K. S.

Crits-Christoph

… Woody

S. R.

(1998). Update on empirically validated therapies, II. The Clinical Psychologist, 51, 3–16.

Chambless

D. L.

Sanderson

W. C.

Shoham

Bennett Johnson

Pope

K. S.

Crits-Christoph

… McCurry

(1996). An update on empirically validated therapies. The Clinical Psychologist, 49, 5–18.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

Coll

K. M.

Juhnke

G. A.

Thorbo

Haas

Robinson

M. S.

(2008). Family disengagement of youth offenders: Implications for counselors. The Family Journal, 16, 359–363.

Coll

K. M.

Thorbo

Haas

(2006). Outcome evaluation of adolescent offender psychosocial development: A comparative study. Journal of Humanistic Counseling, Education, and Development, 45, 208–220.

10.

Derogatis

L. R.

(1994). Symptom checklist-90-Revised administration, scoring, and procedures manual. Minneapolis, MN: National Computer Systems, Inc.

11.

Faul

Erdfelder

Lang

A. G.

Buchner

(2007). G*Power 3: A flexible statistical power analysis for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175–191.

12.

Garfield

R. L.

Love

J. R.

Donohue

J. M.

(2010). Health reform and the scope of benefits for mental health and substance use disorder services. Psychiatric Services, 61, 1082–1086.

13.

Gunderson

J. G.

(1978) Defining the therapeutic process in psychiatric milieus. Psychiatry, 41, 327–335.

14.

Grossman

J. M.

Liang

(2008). Discrimination distress among Chinese American adolescents. Journal of Youth and Adolescence, 1, 1–11.

15.

Groves

Backer

van de Bosch

Miller

(2012). Dialectical behaviour therapy with adolescents. Child and Adolescent Mental Health, 17, 65–75. doi:10.1111/j.1475-3588.2011.00611.x

16.

Horvath

A. O.

Del Re

A. C.

Fluckiger

Symonds

(2011). Alliance and individual psychotherapy. Psychotherapy, 48, 9–16. doi:10.1037/a0022186

17.

James

Leslie

Hurlburt

Slymen

Landsverk

Davis

… Zhang

(2006). Children in out-of-home care: Entry into intensive or restrictive mental health and residential care placements. Journal of Emotional and Behavioral Disorders, 14, 196–208.

18.

Kessler

R. C.

Berglund

Demler

Jin

Walters

E. E.

(2005). Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry, 62, 617–627.

19.

Lenz

A. S.

Holman

R. L.

Dominguez

D. L.

(2010). Encouraging connections: Integrating expressive art and drama into therapeutic social skills training with adolescents. Journal of Creativity in Mental Health, 5, 142–157. doi:10.1080/15401383.2010.485086

20.

Lenz

A. S.

Speciale

Aquilar

J. V.

(2012). Relational-cultural therapy intervention with incarcerated adolescents: A single-case effectiveness design. Counseling Outcome Research and Evaluation, 3, 17–29.

21.

Liang

Tracy

Glen

Burns

Ting

(2007). The relational health indices: Confirming factor structure for use with men. The Australian Community Psychologist, 19, 35–52.

22.

Liang

Tracy

A. J.

Kenny

M. E.

Brogan

Gatha

(2010). The relational health indices for youth: An examination of reliability and validity aspects. Measurement and Evaluation in Counseling and Development, 42, 255–274. doi:101.177/0748175609354596

23.

Liang

Tracy

Taylor

C. A.

Williams

L. M.

Jordan

J. V.

(2002). The relational health indices: A study of women’s relationships. Psychology of Women Quarterly, 26, 25–35. doi:10.1111/1471-6402.00040

24.

Lincoln

Guba

(1985). Naturalistic inquiry. New York, NY: Sage.

25.

Minuchin

, (1974). Families and family therapy. Cambridge, MA: Harvard University Press.

26.

Moustakas

C. E.

(1994). Phenomenological research methods. Thousand Oaks, CA: Sage.

27.

Patton

M. Q.

(2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

28.

Piaget

(1963). The psychology of intelligence. New York, NY: Routledge.

29.

Schwarz

S. W.

(2009). Adolescent mental health in the United States. New York, NY: National Center for Children in Poverty.

30.

Self-Brown

S. R.

LeBlanc

Kelley

M. L.

Hanson

Laslie

Wingate

(2006). Effects of community violence exposure and parental mental health on the internalizing problems of urban adolescents. Violence and Victims, 2, 183–198. doi:10.1891/088667006780644253

31.

Shirk

S. R.

Karver

M. S.

Brown

(2011). The alliance in child and adolescent psychotherapy. Psychotherapy, 48, 17–24. doi:10.1037/a0022181

32.

Stramm

(1985). The hospital as a “Holding Environment.” International Journal of Therapeutic Communities, 6, 19–39.

33.

Utay

(2005). Improving social skills: A training presentation to parents. Education, 126, 251–258.

34.

Winnicott

D.W.

(1965). The maturational process and the facilitating environment: Studies in the theory of emotional development. New York, NY: International University Press.

35.

Zahn-Waxler

Shirtliff

E. A.

Marceau

(2008). Disorders of childhood and adolescence: Gender and psychopathology. Annual Review of Clinical Psychology, 4, 275–303.