Abstract
The present study assessed the potential of a health behavior model used to explain adherence to treatment for chronic illnesses, the Integrative Behavioral Prediction Model (IBPM), to better understand therapy engagement (e.g., low participation) for child therapy in community-based service settings. Qualitative interview methods were used to assess the fit of the IBPM to therapy engagement. Caregivers of children (n = 17) who had successfully completed therapy, were at risk of dropping out, or terminated prematurely at a community mental health clinic were interviewed. Clinic therapists and administrative staff were also interviewed (n = 8). From the perspective of caregivers, therapists, and administrative staff, most IBPM elements—cognitions, intentions, and environmental/contextual factors—appear to be relevant to therapy engagement. Other factors, such as personal and psychological barriers (e.g., poor fit with therapist), not found in the IBPM also may influence therapy engagement. It appears that the core elements of the IBPM may translate to child therapy, though future research is needed to evaluate the generalizability of the study findings. Thus, health behavior models (e.g., IBPM) may improve our understanding of factors contributing to poor therapy engagement for children receiving psychosocial therapy in community-based service settings.
An oft-cited challenge for mental health therapists who work with the families of children and adolescents (hereafter we use “children” to refer to children and adolescents unless distinctions need to be made) in community-based service settings is poor therapy engagement (Kazdin, Holland, & Crowley, 1997; Manfred-Gilham, Sales, & Koeske, 2002; McKay & Bannon, 2004). Therapy engagement refers to client participation in therapeutic activities to achieve therapeutic goals that are identified and agreed on by the client and therapist (Lequerica & Kortte, 2010; McKay & Bannon, 2004; Staudt, 2006). Up to half of the children and families who initiate services in community-based mental health clinics drop out prematurely (McKay & Bannon, 2004). Caregivers play a critical role in treatment attendance as they consent for children to participate in therapy and provide transportation (Shirk & Russell, 1998). For the children and families who attend therapy, low involvement in therapeutic activities can undermine the alliance and contribute to drop out (McLeod et al., 2014; Patterson & Chamberlain, 1994). Poor therapy engagement is thus a widespread problem in practice settings (McKay & Bannon, 2004).
Poor therapy engagement may reduce the effectiveness of interventions. Clients who drop out of therapy receive inadequate treatment and often are in the greatest need of mental health services (Kazdin & Mazurick, 1994; Pina, Silverman, Weems, Kurtines, & Goldman, 2003). For therapists, poor therapy engagement contributes to a waste of resources, the loss of health care dollars, and reduced productivity (Bosworth, Oddone, & Weinberger, 2006; Klein, Stone, Hicks, & Pritchard, 2003; Patterson & Chamberlain, 1994). Poor therapy engagement can also produce additional costs such as client suffering; therapist and/or client frustration, anger, and hopelessness; and reduction in the quality of life for both clients and their therapists (Bosworth et al., 2006).
To date, there have been few efforts to understand therapy engagement in child psychotherapy. In one innovative line of research, Kazdin and colleagues (1997) examined the extent to which four barriers—caregiver perception of stressors, the alliance, treatment relevance, treatment demandingness—influence attendance. Kazdin and colleagues found that more barriers were associated with worse attendance and outcomes (Kazdin et al., 1997). Termed the barriers-to-treatment model, this research has advanced understanding of how factors that might emerge during therapy can contribute to dropout. However, it is also important to identify factors present from the beginning of therapy (e.g., cognitions, motivation, client factors) that may contribute to drop out throughout therapy (Swift & Greenberg, 2014). The barriers-to-treatment model has limited application for this purpose (Nock & Kazdin, 2001).
A few conceptual models focus more narrowly on client involvement (Karver, Handelsman, Fields, & Bickman, 2005; King, Currie, & Petersen, 2014). These models center on client participation during in-session therapeutic activities and are conceptually closer to client involvement (Chu & Kendall, 2004, 2009; McLeod et al., 2014). In contrast, our definition of therapy engagement encompasses both treatment attendance and ongoing participation in therapeutic activities (McKay & Bannon, 2004). We use this broader definition because poor treatment attendance represents a serious problem in community mental health clinics, and regular attendance is a necessary prerequisite to client involvement. To improve the quality of care in these settings we believe that it is important to expand understanding of the factors that contribute to poor therapy engagement over the course of treatment.
To advance the field, a theoretical model that identifies risk factors for poor therapy engagement is needed. Health behavior theories that explain adherence to treatment for chronic illnesses (e.g., diabetes) may help to identify such factors. In the United States, chronic illnesses are a leading cause of death, disease, and disability (Heron, 2013). Health behavior theories converge to suggest that behavioral change occurs in two stages: (a) adopting new cognitions and intentions and (b) initiating behavior change. To manage a chronic illness, individuals must adhere to medical recommendations, make significant behavioral and/or lifestyle changes (e.g., diet), and maintain these changes over time. As psychotherapy requires individuals to make similar changes, lessons from health behavior science might inform efforts to understand and ultimately optimize therapy engagement. To date, research has identified an array of factors that influence adherence to treatment for chronic pediatric illnesses, including client, family, disease, regime and provider characteristics (Cheng & Walter, 2006). To our knowledge, health behavior theories have not yet been used to aid understanding of therapy engagement in child therapy.
The Integrative Behavioral Prediction Model (IBPM; Fishbein, 2000; Fishbein & Yzer, 2003; Yzer, 2012) could serve as a template for understanding therapy engagement. The IBPM is a synthesis of the major behavioral prediction theories in health behavior science—that is, health belief model (Rosenstock, 1974), Social cognitive theory (Bandura, 1977), theory of reasoned action (Fishbein & Ajzen, 1975), and the theory of planned behavior (Ajzen, 1991; Francis et al., 2004). In developing the IBPM, Fishbein (2000) concluded that three factors are sufficient to predict whether an individual will engage in a health behavior (e.g., adherence to a HIV medicine regimen): (a) an individual’s intention (e.g., motivation to follow medication schedule), (b) an individual’s skill set (e.g., ability to remember dosage schedule), and (c) the absence of environmental constraints (e.g., transportation to pharmacy). Primary determinants of intention include three core cognitions: (a) attitudes toward a behavior, (b) subjective norms about social pressure to perform a behavior, and (c) perceived behavioral control/self-efficacy, or the individual’s perception that he or she has control over and is confident in performing a behavior (Ajzen, 1991; Fishbein & Yzer, 2003; Francis et al., 2004). Each of these cognitions, in turn, influences intention (Fishbein & Yzer, 2003). For understanding therapy engagement in child therapy, the perspectives and cognitions of caregivers (parents, guardians, etc.) are of great importance to consider. As caregivers oftentimes consent for children to participate in therapy and provide transportation (Shirk & Russell, 1998), they represent the gate-keepers for the children receiving treatment.
Components of the IBPM have received empirical support and been used to inform the development of prevention programs (Fishbein, Hennessy, Kamb, & Project RESPECT Study Group, 2001; Fishbein & Yzer, 2003). Thus, the IBPM represents an ideal model to aid understanding of therapy engagement in child therapy. We have adapted the IBPM to focus on factors that might influence the extent to which a caregiver engages in child therapy. As depicted in Figure 1, it is hypothesized that therapy engagement is most likely to occur when a caregiver has a strong intention to participate, if the caregiver has the requisite skills and abilities to participate, and if no environmental factors interfere with participation (Fishbein, 2000; Fishbein et al., 2002). Furthermore, therapy engagement can be influenced through changes in intention (via cognitions), skills, and environmental factors. Importantly, this model implies that different interventions will be required for a caregiver who has formed an intention to engage in therapy, versus for a caregiver who has not. If the IBPM generalizes to child therapy, it could promote a more thorough and comprehensive understanding of therapy engagement.
Integrative Behavioral Prediction Model adapted for therapy engagement in child psychotherapy.
Although the IBPM may offer a framework that can help broaden understanding of factors that influence therapy engagement, the extent to which the model will generalize to child therapy is an open question. Research suggests that specific facets of the model may translate to caregiver perspectives around child therapy. For example, caregiver expectations about participation in child therapy have been found to predict treatment attendance (Nock & Kazdin, 2001). As another example, caregiver motivation to participate in therapy has also been found to predict treatment attendance (Nock & Photos, 2006). Whether health behavior models, such as the IBPM, apply to therapy engagement in child therapy is a novel question.
The purpose of this study is to assess the extent to which the IBPM components translate to child therapy. We utilized qualitative methods appropriate for translational research designed to apply an existing model to a new field (Brekke, Ell, & Palinkas, 2007; Glaser & Strauss, 1967). We used directed content analysis to examine the goodness of fit of the IBPM with child therapy and to make modifications (Hsieh & Shannon, 2005; Zhang & Wildemuth, 2009). Directed content analysis involves using an established theory or framework for initial coding (e.g., the IBPM) and allows themes to emerge from the data in subsequent coding. We gathered perspectives about factors related to therapy engagement from caregivers of children who had received mental health services at a community mental health center. The clinic is located in the United States and serves primarily low-income children and adolescents (aged 4–17 years). The clinic was selected for the study based on the formation of a research partnership built around shared interests; therapy engagement was a central area of concern for the clinic at the time of the study. For triangulation, we gathered perspectives from therapists and staff employed at the clinic.
Method
Participants
Participants were recruited from a community mental health clinic and included caregivers (n = 17) of children receiving mental health services at a community mental health clinic, their therapists (n = 6), and administrative staff (n = 2). Pseudonyms are used throughout the manuscript in lieu of real names (see Table 1). All clinic staff were approached and asked about participating in the study. Eligibility criteria for caregivers (n = 17) were that they spoke English and their children fell into one of three categories. These categories were (a) successful termination, therapy had ended and the therapist and caregiver agreed that therapy was complete (n = 5); (b) premature termination, therapy had ended (including due to caregivers/clients moving) but the therapist and/or caregiver did not agree that therapy had ended (n = 4); or (c) at risk of premature termination, the client had no-showed or canceled two or more consecutive sessions (n = 6).
Participant Pseudonyms and Demographics.
Note. AA = African American; C = Caucasian; U = unreported; NA = not applicable.
Caregivers and children
Fifteen interviews were conducted with 17 caregivers (2 males, 15 females; two caregivers were present at two interviews). Caregivers included biological mothers (n = 11), an adoptive mother (n = 1), maternal grandmothers (n = 3), a maternal grandfather (n = 1), and a mother’s boyfriend who was also a caregiver (n = 1). Three caregivers were Caucasian, 11 were African American, and three did not report race/ethnicity data.
The caregivers’ children were clients receiving mental health services (n = 15; 5 males, 10 females). Children averaged 12.30 years of age (SD = 5.20 years, range = 4–17 years; 73.3% African American, 6.7% Caucasian, 20.0% biracial). At the time of the interviews, the average length of treatment was 24.80 weeks (SD = 22.30; range = 3–95). The children’s therapist-assigned Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM-IV-TR; American Psychiatric Association, 2000) diagnoses were acute stress disorder (n = 1), adjustment disorder with anxiety (n = 1), adjustment disorder with mixed disturbance of emotions and conduct (n = 5), adjustment disorder with disturbance of conduct (n = 2), attention-deficit/hyperactivity disorder combined type (n = 2), major depressive disorder, single episode, moderate (n = 1), mood disorder not otherwise specified (n = 1), posttraumatic stress disorder (n = 1), and sexual abuse of child (if focus of attention is on victim; n = 1). No other demographic data were gathered for caregivers or clients.
Therapists and administrative staff
Therapists (n = 6; 100% female; 100% social workers; 50.0% licensed) and administrative staff (n = 2) who had regular contact with caregivers (e.g., intakes, scheduling, providing child care) participated in the study. Other demographic data were not collected or are not reported here to prevent participant identification.
Participant Contact Procedures
For the study, project staff provided an overview of the study to the therapists and staff and then approached staff about participating. Six of the seven full-time therapists (85.7%) and two of the three staff (66.6%) participated. To recruit caregivers, therapists were asked to share information about the study with their clients’ caregivers. Interested caregivers were asked for permission for project staff to contact them about the study. A total of 75 caregivers provided permission. Of the 75 caregivers consenting to be contacted, 31 (41.3%) were not eligible due to their children being in active treatment. Eligible caregivers were then contacted by project staff and invited to participate in the study (n = 44 were eligible for participation based on eligibility criteria and were invited to participate). Of those eligible, 15 (34.1%) completed interviews. In the eligible group, 65.9% (n = 29) did not complete interviews due to being unable to be contacted (e.g., disconnected phone), declining to participate, or moving out of state.
Interview Procedures
Interviews were conducted between March 2010 and May 2011. Interviewers were four females (two Caucasian, one Latina, and one Asian American), and one male (Caucasian). All interviewers were doctoral students and faculty from a university psychology department; they had at least a bachelor’s degree (two PhDs, two MAs, and one BA) and all had prior interviewing experience. The interviewers/coders were trained by the two faculty members who had extensive interviewing and coding experience; one of the faculty members also had extensive experience conducting qualitative studies. Interviewers were trained on protection of human participants, conducting semi-structured interviews, and cultural considerations. The interview team met to review the interview procedures and to promote consistency across interviews. Interviews were digitally recorded, lasted approximately 60 min, and were conducted at the clinic or in caregivers’ homes. Interviewers asked participants open-ended questions from a script (available on request from the corresponding author). Trained research assistants (RAs) transcribed interviews. Transcripts were verified for accuracy by the interviewer and were imported into NVivo10 for qualitative analysis. Participating caregivers received US$20 gift certificates whereas a donation was given to the agency to compensate for the time of the therapists and clinic staff (US$60 per participant). All methods and procedures were approved by the university’s institutional review board.
Data Analysis
Qualitative coding occurred from June 2012 to September 2013 and was completed by a coding team of three doctoral students and two faculty members. The coding team also served as interviewers. The coders worked together to develop a codebook that included definitions for assigning codes, examples, and coding guidelines. The following Broad Codes were included in the initial codebook based on theoretical relevance within the IBPM, consistent with directed content analysis (e.g., Hsieh & Shannon, 2005): factors related to therapy motivation and engagement (cognitions), skill set, and environmental barriers and contextual factors. The basic unit of coded text (Zhang & Wildemuth, 2009) was one exchange between the interviewer and interviewee (one interviewer speaking turn and one participant speaking turn) to facilitate the interpretation of coded text by providing context of the conversation. Coding progressed through two phases. The first phase involved two coders double-coding transcripts on the Broad Codes generated by the IBPM. Each coder noted when a potential theme did not fit in the Broad Codes and noted potential sub-themes (both those within and beyond the IBPM) using coding memos. A third coder then reviewed the transcripts and resolved any disagreements. At the end of this phase, the coding team added a new Broad Code (personal/psychological barriers to engagement), refined the definitions of the existing Broad Codes, and generated a list of possible sub-themes falling within each Broad Code—both those that fit the IBPM (e.g., cognitions) and those that did not (e.g., attributions). The coding team finalized the sub-themes and produced and reviewed a final version of the codebook. In Phase 2, sub-themes were assigned in each Broad Code. Transcripts were randomly ordered and sub-theme codes were assigned by two coders. Decision rules and memos were used to increase consistency and differentiation across coders, and the final coding was reviewed for consistency by the first author and disagreements were resolved. Table 2 includes the final codes and number of interviews in which the themes were coded. Themes are described below by presenting summaries of information in the transcripts, illustrative paraphrases, and verbatim quotes.
Emergent Themes in Caregiver and Therapist/Staff Interviews.
Note. T/S = therapist/staff; CG = caregiver.
A total of 15 caregiver interviews were conducted with 17 caregivers (two caregivers were present in two interviews).
Trustworthiness of Data
Procedures were undertaken to enhance the validity and trustworthiness of the data (Barbour, 2001; Creswell & Miller, 2000; Drisko, 2005; Meyrick, 2006; Zhang & Wildemuth, 2009). We aimed for “transparency and systematicity” (Meyrick, 2006, p. 803) through all procedures. We demonstrate transparency in the description of the methods (e.g., recruiting participants, information about the researchers), data collection, and data analysis plan. Interviews were transcribed by RAs and then verified for accuracy by each interviewer and an RA (who had not completed the original transcription). The transcripts were coded multiple times, by multiple coders, across two coding phases. Meetings, journaling, and memos were used to communicate about coding disagreements and to refine coding frames (Barbour, 2001). We provide frequencies for the themes that were coded in Table 2. Clear coding definitions and procedures were provided to coders (Zhang & Wildemuth, 2009). Finally, we provide participant quotes to illustrate codes, add depth to the code descriptions, and allow readers to form their own interpretations (Drisko, 2005).
Results
Sixteen themes emerged that focused on factors that may influence therapy motivation and engagement (see Table 2). These themes are categorized and discussed in the following sections: Factors Related to Therapy Motivation and Engagement (Caregiver Cognitions), Personal/Psychological Barriers to Engagement, Skill Set, and Environmental Factors and Contextual Factors.
Factors Related to Therapy Motivation and Engagement (Caregiver Cognitions)
Within the domain of Caregiver Cognitions, the themes that emerged can be further categorized into four domains: Attributions, Attitudes/Expectations, Subjective Norms, and Perceived Behavioral Control.
Attributions about the origin of presenting problems
Attributions are how individuals understand the causes of other individual’s behavior—that is, their belief about causes (Malle, 1999). Here, attributions of how participants understood the causes of child symptoms (internal and external) were present in many interviews.
Internal attributions
About half (n = 8) of the caregivers shared that the problems leading them to seek therapy were due to child behavior and problems with the children themselves. Caregivers expressed distress with challenging child behaviors, concern about their children, and a desire to get help (e.g., parenting strategies). For instance, Ms. Khan shared about her daughter,
She’s defiant . . . she verbally abusive, she physically abusive . . . And I just feel that somebody that don’t know her is gonna hurt her. Similarly, Ms. Owens noted, I needed the help for her . . . she get upset so quick for no reason.
Therapists/staff (n = 7) also noted that caregivers (and sometimes children) make internal attributions (e.g., that presenting problems were “within the child”) and that this may make it difficult to discuss family or parent-focused components of treatment with caregivers. For instance, Ms. Pritchard stated,
. . . it seems to me that a lot of the guardians think that the problem is with the kid and it’s not. Or it’s, you know only partially with the kid and they’re really resistant to participating in the treatment themselves, because they just want their kid to be fixed. . . .
External attributions
Few (n = 3) of caregivers mentioned their own parenting skills, caregiver separation/divorce, or trauma/stress as factors that may cause or maintain the child’s presenting problems. For instance, Ms. Charter shared, “I was . . . in and out of their life for ten years . . . when I finally got them back it was a big adjustment for them to have a parent tell them what they can do and cannot do.” Similarly, Ms. Eaddy shared her reflection on the impact of her parenting skills on her daughter’s behavior: “I was always, she would get attention for the bad behavior and she wasn’t getting that much attention for the good behavior, so I realized . . . ”
In contrast, several of the therapists/staff (n = 4) mentioned “bad parenting” and “family systems issues” as causing the child’s difficulties. Therapists noted that caregivers who made external attributions are likely to engage in therapy. For instance, Ms. Vance (therapist) shared, “I think that they’re [caregivers who do not see the problem as solely the child’s problem] a little bit more receptive.” Also, to illustrate, Ms. Taylor shared,
I find it hard to give help on parenting things like that. I don’t find most of my families to be very receptive to it . . . they kind of just want, they want to place a problem on a child and they want it to be fixed. These are big generalizations I’m making, but just I think a lot of the issues are probably caused by the parenting. . . .
Attitudes/expectations about therapy
Attitudes are an individual’s overall evaluation of a particular behavior—here, therapy engagement, or participating in or attending treatment—and expectations reflect a person’s belief that something will happen or will occur. Therapy outcome expectations and role/involvement expectations were present in the interviews.
Therapy outcome expectations
Therapy outcome expectations were coded when participants shared their thoughts about what outcomes caregivers expected from therapy. Caregivers expressed therapy outcome expectations such as a desire to better understand their children through therapy and to identify causes of emotional/behavioral problems. For example, Ms. Hurst said she hoped therapy would “get to the bottom of whatever the underlying issues are.” Ms. Eaddy shared, “I wasn’t expecting miracles but I expected to at least you know get them some help.” Caregivers also wanted to see improvements in their own parenting skills, and their children’s behavior and communication with parents and other adults. To illustrate, Ms. Owens expressed, “I wanted somebody that was able to teach me and help me with my kids.” Mr. Ferris stated, “We wanted her to have more self-esteem, open up more with her problems.”
Although six caregivers showed an understanding that therapy “takes time” and effort from child clients and caregivers, all therapists/staff (n = 8) said that caregivers expected “quick fixes.” Therapists also said that this expectation could affect therapy engagement and retention. Ms. Pritchard shared, “The parents, the ones that are more likely to drop out early or you know, before they reached all the treatment goals are the ones that seem to want it to be done fast.” Therapists/staff also acknowledged that caregivers often do not know what to expect in therapy. Ms. Stevens shared, “they [caregivers] don’t know what to expect . . . the ones who haven’t had therapy before . . . it’s really a learning process of what therapy is and what we do here.”
Role/involvement expectations
Role/involvement expectations were coded when the participants discussed their views around the role that caregivers played (or should play) in the therapeutic interaction. Most caregivers (n = 12) expressed a desire to be more involved in therapy and to learn parenting skills. For example, Ms. Madigan said, “I expected more parenting [skills] . . . I didn’t really grasp the concept of the play therapy.” Ms. Dahl said, “I was expecting . . . how to deal with the situations that I was confronted with and I got what I expected.” Caregivers noted that their involvement was related to their therapists’ responsiveness and availability and their working relationship. Finally, some caregivers did not feel it was their role to be involved. For instance, Ms. Burns said that when the therapist would invite her to the therapy room she would reply, “No, this is her time. I’m not the one in therapy.”
Therapists and staff (n = 6) had mixed views about caregiver involvement with some indicating that caregivers did not want to be involved and others noting that caregivers did want to take an active role in therapy. For instance, Ms. Richardson shared, “[Caregivers] aren’t interested in going over the treatment plan or are not interested in, just participating.” In contrast, Ms. Stevens said, “So many [caregivers], I would say almost all of them [are involved and consistent] . . . if they’re interested in being here if it’s something that they enjoy doing.”
Subjective norms about therapy
Subjective norms are a person’s own estimate of the social pressure or support to perform or not perform a particular behavior—here, engaging in therapy. Twelve caregivers (no therapists/staff) expressed the belief that therapy was supported by important others in their lives and that support from family was often provided because of the child’s “need” for therapy. For example, Ms. Dahl shared, “[Everybody] was very supportive because they knew that I needed it and I needed it for the boys.” Some caregivers brought their children to therapy because they were told by another party to do so (e.g., court, school). Caregivers reported being open to the idea of getting help for themselves, showing positive norms about therapy. Four caregivers received unsupportive messages about attending therapy and thus were hesitant to share therapy information with their family. “Nobody needs to know [about the therapy],” shared Mr. Ferris. Caregivers explained that their family might have a misunderstanding of therapy or a preference for family privacy. Ms. Jacobs said, “They [my family] didn’t like it [taking my daughter to therapy] . . . cause they didn’t want to talk to nobody else.” Two therapists/staff expressed similar views.
Perceived behavioral control (self-efficacy) to engage in therapy
Perceived behavioral control (self-efficacy) is the extent to which an individual feels able to perform a behavior—here, to engage in therapy (participate in or attend treatment). Six caregivers and one therapist/staff discussed caregiver self-efficacy in engaging in various aspects of therapy. For example, Ms. Eaddy shared,
I mean it took a few weeks . . . before the youngest one . . . totally turned around and I noticed the difference in her that I realized, after talking with her therapist, that a lot of it was because of the things that I . . . had been doing that she had suggest that I do . . . it worked.
However, when caregivers were asked about whether they were able to do what was asked of them and to what extent they felt self-efficacy in engaging, most replied with short responses (e.g., “yeah”), and rarely elaborated. Six caregivers noted that self-efficacy to engage in therapy was either not present, inconsistent, or not applicable. Some caregivers discussed feeling that they were unable to implement the skills asked of them at all or that it was hard for them to confidently do this consistently. For instance, Ms. Dahl shared, “Yes to some extent [I had confidence in the skills I was taught] . . . [when] it did not work is when he got defiant.” Two caregiver interviews reflected that these caregivers were never asked to implement skills or engage in therapy (not applicable). Ms. Owens shared, “I wasn’t asked to do nothing with them, just make sure they come to their appointment.”
Personal/Psychological Barriers to Engagement
There were a number of individual-level (personal/psychological) factors that emerged as possible factors that could decrease motivation to engage in therapy. These contrast with environmental barriers, which refer to non-psychological factors that affected caregiver ability to attend and engage in therapy (e.g., logistical barriers).
Caregiver-related barriers
Caregivers (n = 11) discussed being disengaged due to not understanding what was happening in therapy, not liking the therapy process, or not seeing improvements. For instance, Ms. Ibsen stated, “I was not motivated at all. To me it was boring . . . What’s the point of even coming in because there’s no change there’s no improvement. . . .” Another barrier discussed by both caregivers and therapists was caregiver mental health difficulties (or substance abuse), stress, challenging life circumstances, or other competing responsibilities. Ms. Nesbitt said the following about her own participation:
[the therapist suggested] that I pick at least one particular day during the month and to spend time alone with my children . . . I didn’t follow through because I had other interruptions and distractions . . . it was actually a good idea… but I feel bad because I didn’t do what I was supposed to do.
Therapists/staff (n = 7) described that a lack of caregiver desire to be involved (in sessions and in terms of providing information) was a barrier to therapy engagement. Ms. Quaid described that “. . . if a mom had a bad day before she got here, she ain’t wanna come back and talk to the therapist.” Ms. Taylor shared her perspective about caregivers: “A lot of them don’t really want to be involved . . . [sitting in the waiting room], it’s like a time for them to sit and read.” Therapists/staff said that caregivers sometimes do not share all information about their child or family with therapists and that this hampers engagement and progress in therapy. They identified that caregiver personality/mood characteristics were also barriers, including being “hostile” (Ms. Richardson); “overly defensive,” “disrespectful,” “attacking,” and “combative” (Ms. Stevens); having a “sense of entitlement” and being “defensive” (Ms. Taylor); and “guarded,” “rejecting,” “devaluing,” “flat,” and “lack empathy for the child” (Ms. Vance). Therapists/staff also found caregiver mistrust of the clinic or therapists as a barrier. For instance, Ms. Taylor shared,
. . . it’s worst since our building’s so nice. It’s like we seem like these fancy people . . . we come off as snobs . . . I think there’s some trust issues and some cultural barriers and, . . . I feel like the relationships are so fragile and they can be broken so easily.
Perceived child barriers
Whereas children were not interviewed directly, caregivers (n = 11) and therapist/staff (n = 7) discussed perceived child-related barriers to caregivers and child clients engaging in therapy. Caregivers expressed that sometimes child clients exerted self-determination around treatment attendance for various reasons ranging from not feeling comfortable sharing personal information to feeling like therapy was not helpful to them. Ms. Abel stated, “I still try to encourage them you know, but . . . they feel like they don’t want to go. It’s like church you know, they go when they want to. . . .” Caregivers identified child clients’ temperament/mood characteristics as barriers to engagement including: “[having] this mean streak in him” (Ms. Dahl); “terrible [attitude],” “stubborn,” and “spoiled” (Ms. Gale); “absent minded” (Ms. Lombardy); and “being lazy” (Ms. Nesbitt). Therapists/staff also found child characteristics to be barriers, including being “angry” and “shut down and over controlled” (Ms. Taylor); “shut down,” “guarded,” “super anxious,” and “rejecting of you and devaluing” (Ms. Vance); and “devaluing and resistant” (Ms. Williams).
Both caregivers and therapists/staff shared that perceived child motivation to engage in therapy was hampered due to the influence of external factors on the child. These included reasons such as the therapy being mandated or negative peer pressure. One mother described her son’s lack of motivation was due to her own inability to come with him to sessions (“if momma’s not going to come then why waste my time . . . ” shared Ms. Nesbitt).
Therapist-related barriers
Undesirable therapist characteristics, personality, or behaviors were also brought up as barriers in seven caregiver interviews and four therapist/staff interviews. Caregivers perceived that therapist personality characteristics affected engagement. Ms. Burns described that “her therapist used to irk me, like get on my nerves . . . you have to be careful when you pair these kids up. Their personalities of everybody does not click.” The importance of rapport was noted in therapist/staff interviews; for instance, Ms. Stevens said,
Yeah I think a lot of it is based . . . rapport. I mean if the parent trusts you than they’ll . . . build a relationship with you to, to you know and be invested, otherwise they’re completely checked out. . . .
Other undesirable behaviors of therapists included therapists using interventions that caregivers perceive to be ineffective (e.g., play therapy) or not communicating with caregivers about the content of therapy sessions and the rationale for the interventions used. For instance, Ms. Ibsen said, “I didn’t really understand . . . What’s the object of my child playing with toys? . . . I think things could have been explained to me a whole lot more.” Similarly, Ms. Hurst stated, “I’d like to know . . . more about what the therapists are doing with the children. . . .”
Therapists/staff and caregivers both brought up therapist inexperience as a barrier to caregiver/client engagement. For instance, Ms. Richardson shared about her own inexperience: “I don’t know how to do that with that population and I don’t want to damage them you know. I can’t gauge, well am I pushing too far?” From the caregiver perspective, Ms. Madigan shared, “I know you have a degree but . . . she just only could go a certain point because of her, you know she didn’t have any children. . . .” Some therapists were aware of the role of their own personality/behavior on engagement. For instance, Ms. Vance said, “I tend to be not aggressive but . . . I really put things out there . . . asking them to change too much . . . too quickly can make them uncomfortable and not come back. . . .” Similarly, Ms. Taylor described having difficulties in building rapport and trust with teenage clients: “They’re harder for me to build rapport with . . . it’s a distrust of you know of an authority. I just seem like another teacher or another adult.”
Skill Set
Skill deficits related to the performance of behaviors learned in treatment that may interfere with therapy engagement. For example, if a caregiver is asked to expose his or her anxious child to a feared stimulus, but does not have the skills to regulate his or her own anxiety caused by exposing the anxious child, then the caregiver is less likely to complete the assignment. Issues with skill set impacting therapy engagement were present in five caregiver interviews and five therapist/staff interviews. However, the coded text for Skill Set generally included very short participant responses.
Caregivers cited that they needed time to “get used to” (Ms. Eaddy) doing the skills asked of them. Caregivers shared that implementing strategies asked of them was challenging, and that sometimes their frustration or own strong affect was a barrier to implementing skills learned. For instance, Ms. Owens said, “[The therapist] was tryin’ to help [me] at the time with how to deal with my kids . . . without goin’ off and havin’ tantrums myself. . . .” Caregivers described frustration with the child as a major hurdle to successfully implementing the skills. As an illustration, Ms. Ibsen described her experience of trying to implement skills taught by her child’s therapist: “. . . sometimes I get frustrated so I just like give up.” Therapists/staff discussed that child clients sometimes did not have the skills to independently implement skills learned in session (e.g., relaxation) or to do homework assigned at home. To illustrate, Ms. Williams stated, “. . . some kids are just also so wound so tightly that they . . . can’t loosen up to do an act . . . relaxation—it’s too intimidating . . . or threatening.”
Environmental Barriers and Contextual Factors
Environmental barriers included non-psychological factors that affected caregiver/client ability to attend and engage in therapy such as physical, logistical, fiscal, or social barriers. Contextual factors of the mental health clinic (e.g., organizational climate) can also affect therapy engagement. In this domain, there were no novel environmental barriers or contextual factors that emerged that have not been previously documented in the literature. Consistent with prior literature (e.g., Kazdin et al., 1997), the following environmental barriers and contextual factors sub-codes were identified: agency environment (physical attributes of the clinic facility affecting engagement), organizational vibe (characteristics demonstrating the climate and culture of the agency), agency logistics, work/family schedule, and/or transportation. Agency environment was discussed by caregivers (n = 9) and in two therapist/staff interviews. Caregivers elaborated on the specific clinic characteristics including details concerning waiting room (e.g., too noisy), clinic room environment (e.g., child friendly), and availability of tangible resources (e.g., art supplies). Caregivers (n = 10) and therapists/staff (n = 2) expanded on the specific characteristics of the organizational vibe that affect engagement, including having a respectful, professional, and flexible atmosphere, and having special events involving clients and families (e.g., holiday parties). For instance, Ms. Dahl shared, “Respect, there was a lot of respect showed to us.” Agency logistics (e.g., flexibility of scheduling, lack of follow-up, case management concerns, insurance reimbursement issues) were brought up frequently by caregivers (n = 15) and therapists/staff (n = 7). For instance, Ms. Taylor shared: . . . yeah the little administrative things just are relationship killers . . . people’s appointments getting deleted . . . parents can tell when you’re too busy for them. . . .” Finally, work/family schedule and/or transportation (e.g., school hours, work schedules, life stressors, transportation issues) emerged frequently among caregivers (n = 15) and therapists/staff (n = 7) as influencing engagement.
Discussion
Poor therapy engagement affects the quality of care in community mental health care settings. However, the factors that contribute to poor therapy engagement in child psychotherapy provided in community-based service settings are not fully understood. This study used qualitative interview methods to assess the potential of the IBPM, a health behavior model, to promote a better understanding of factors influencing therapy engagement, with a particular emphasis on the role of the caregiver. An important strength of this study is that we were able to gather perspectives from caregivers who were at risk of “dropping out” of therapy or who had terminated therapy, a population that is difficult to access. Overall, our findings suggest that most elements of the IBPM may be helpful in understanding engagement in child therapy. Core elements of the IBPM—caregiver cognitions, intentions, and environmental/contextual factors (see Figure 2)—appeared to fit child therapy, though some IBPM elements may not translate as readily (i.e., skill set). In addition, a few child, caregiver, and therapist factors not found in the IBPM emerged as potential barriers to therapy engagement (i.e., personality characteristics, therapist experience, etc.). Overall, these findings indicate that health behavior models, such as the IBPM, might help expand our understanding of therapy engagement in child therapy.
IBPM adapted for therapy engagement in child psychotherapy including emergent themes.
Consistent with the IBPM, our findings suggest that caregiver cognitions (attitudes/expectations, subjective norms, perceived behavioral control) may play a role in therapy engagement in child therapy. Two types of attitudes/expectations emerged in the interviews: caregiver treatment expectations and caregiver attributions about the reason for the child’s presenting problems. The finding that caregiver cognitions could play a role in therapy engagement suggests that a central element of the IBPM may translate to child therapy, as cognitive factors (e.g., attitudes) play a pivotal role in determining health behaviors (Fishbein, 2000), and in this application, therapy engagement. We now consider both types of attitudes/expectations and their potential relation with therapy engagement in child therapy.
Our findings suggest that caregiver treatment expectations may play a role in therapy engagement. Therapists/staff reported that they believe caregivers who are more likely to drop out of therapy also expect outcomes to happen rapidly and do not expect to be involved in the therapy. Conversely, a number of caregivers reported that they wanted to be more involved in therapy, and recognized that therapy would take time (contrasting with therapists/staff perceptions that caregivers want rapid outcomes). Some caregivers shared that they believed it was “not their place” to be involved in therapy, and that therapy was a time for the child to receive help. These findings suggest that a potential mismatch may exist between caregivers and therapists/staff in their expectations for how therapy produces change—that is, outcome expectations—and the role the caregiver will play in therapy—that is, role expectations (Arnkoff, Glass, & Shapiro, 2002; Dew & Bickman, 2005; Nock & Kazdin, 2001). If present, such mismatches could contribute to problems with therapy engagement, especially if they are not discussed. Previous research supports this viewpoint as caregivers with more accurate treatment expectations are more likely to attend treatment (McCabe, 2002; Nock & Kazdin, 2001).
Our findings also suggest that caregiver attributions about the cause(s) of their child’s presenting problems may also contribute to therapy engagement. Attributions are not included in the IBPM, but do represent a belief that may influence health behaviors (Bandura, 1977), and there is research indicating that caregiver attributions may affect treatment attendance (Morrissey-Kane & Prinz, 1999). One way caregiver attributions might play a role in therapy engagement is through willingness to participate in therapy. For example, caregivers who believe that the cause of a problem lies within their child may resist participating in therapeutic activities targeting caregiver behavior. Perhaps caregivers who believe they can change their child’s behavior form stronger intentions to engage in treatment.
Our findings suggest that some environmental barriers may be related to therapy engagement. The IBPM highlights that environmental barriers can negatively affect therapy engagement, so this aspect of the model may translate to child therapy. This is consistent with previous research that highlights the role of environmental barriers in poor treatment attendance (Soderlund, Epstein, Quinn, & Cumblad, 1995; Tolan & McKay, 1996). Indeed caregivers’ and therapists’/staff responses reported that stressors and environmental barriers, such as work/family schedule and access to transportation, were contributing factors that may decrease therapy engagement. Although these factors may not be amenable to direct influence on the therapist’s part (e.g., transportation access issues), these are important factors for therapists to be aware of, and possibly may help caregivers and children problem-solve around these barriers.
Interestingly, our results suggest that contextual factors may be important for therapy engagement. A number of caregivers noted that the agency environment as well as the ease (or difficulty) of scheduling appointments influenced therapy engagement. This is consistent with research suggesting that organizational climate and culture predict service quality in child welfare and juvenile justice services (Glisson & Hemmelgarn, 1998; Glisson & James, 2002). Together, such findings suggest that contextual factors may affect the quality of mental health services in community settings. Agencies and organizations providing outpatient psychosocial therapy to children may find it helpful to be attuned to their clinic’s own stimulus value for families and potential benefits and challenges related to the clinic context.
Skill set, a component of the IBPM, was not well supported as a theme in our analyses. According to the IBPM, caregivers who do not possess the skills to carry out tasks related to therapeutic activities may be less likely to engage in treatment. However, in this study, very few comments relevant to skill set were made. It is possible that this component was not supported due to the nature of the questions asked. For example, the fact that the child clients were not interviewed may have decreased our ability to validate this component as caregivers were not consistently involved in therapy. However, it is also possible that skill set is of secondary importance in child therapy and does not represent an important contributor to engagement.
Personal and psychological barriers to engagement emerged as a new theme in our study. Personal characteristics of both caregivers and perceived characteristics of child clients have importance in terms of understanding engagement and being attuned to these factors may help to improve engagement. For instance, a therapist who notes that a caregiver is struggling with depression may provide a referral for that parent for mental health services. Other factors that emerged included being motivated by therapist responsiveness and availability (e.g., to fit sessions with the caregiver’s schedules, to talk to the caregiver outside of therapy session) and by a trusting and supportive relationship with the therapist. Conversely, the lack of these factors acted as barriers to engagement. The importance of a strong therapeutic relationship to promoting positive outcomes, including treatment attendance, has empirical support in child therapy (Karver, Handelsman, Fields, & Bickman, 2006; McLeod, 2011). Perhaps not surprisingly then, caregivers see certain therapist characteristics, such as trustworthiness and flexibility, as important to therapy engagement.
Although this study provides valuable information about engagement in child therapy, a few study limitations bear mentioning. First, child clients were not interviewed, so an important viewpoint is missing. As a result, a number of factors either under child control or related to child motivation that may contribute to therapy engagement may have been missed (e.g., the influence of peers on engagement) in the model. Second, the caregiver sample was recruited from a low-income, urban population served in a specific clinic, so applicability of this model to other populations and service settings needs to be assessed. For instance, it is possible that more affluent populations may experience fewer transportation barriers and other environmental barriers. In addition, we excluded non-English-speaking families. Also, there were race differences between the therapists and many of the clients. It may be beneficial in future studies to specifically examine the impact of language/cultural factors, as well as race/ethnicity differences between therapist and client, on therapy engagement. Third, though we sampled broadly across the ages served in the clinic (ages 4–17 years), it is possible that there may be differences between engagement perspectives between caregivers of younger and older children due to developmental differences. It will be important for future research to investigate if therapy engagement for caregivers differs across children and adolescents of different ages. Fourth, it is important to carefully consider sample representativeness as only 34% of eligible caregivers participated. Recruiting individuals for research from a pool of individuals who are not presently engaged in therapy is challenging. Although we were particularly interested in recruiting caregivers whose child had terminated prematurely, we were only able to recruit a limited number (four) of these participants. These caregivers were difficult to recruit due to not responding to our efforts to contact them. For these reasons, we believe that it is important for future research to investigate the generalizability of our findings. In all, we believe that this study provides a useful starting point for examining factors influencing engagement in a community-based service setting, but future work (quantitative and qualitative) is needed to replicate our findings and further explore the role of these factors across different settings and participants.
In conclusion, this study provides a novel application of a health behavior theory (the IBPM) to child psychosocial therapy. Our findings suggest that the IBPM may aid understanding of the factors that contribute to therapy engagement in child therapy, and that more broadly, health behavior theory may be useful when applied to therapy engagement. More research is, however, needed to identify and define the role of these factors in therapy engagement, determine if these findings generalize to other settings, and to specify the best ways for therapists to incorporate this information into their practice.
Footnotes
Acknowledgements
The authors acknowledge Edward G. Canada and Vivian Mann for their support and input on the project.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Preparation of this article was supported in part by a grant from the Virginia Commonwealth University Council for Community Engagement (principal investigators: McLeod, Corona).