The Effects of Organizational Culture on Mental Health Service Engagement of Transition Age Youth HyunSoo Kim, PhD Elizabeth M. Tracy, PhD David E. Biegel, PhD Meeyoung O. Min, PhD Michelle R. Munson, PhD Abstract Nationwide, there is a growing concern in understanding mental health service engagement among transition age youth. The ecological perspective suggests that there are multiple barriers to service engagement which exist on varying levels of the ecosystem. Based on the socio-technical theory and organizational culture theory, this study examined the impact of organization-level characteristics on perceived service engagement and the moderating role of organizational culture on practitioner-level characteristics affecting youth service engagement. A cross-sectional survey research design was used to address the research questions. The data were collected from 279 practitioners from 27 mental health service organizations representing three major metropolitan areas in Ohio. Hierarchical linear modeling was used to address a nested structure. Findings revealed that location of organization, service setting, and organizational culture had significant effects on the continuation of services. In addition, the relationship between service coordination and resource knowledge and service engagement was moderated by organizational culture.

Address correspondence to HyunSoo Kim, PhD, Department of Social Welfare, Dongguk University-Gyeongju, 123 Dongdae-ro Gyeongju-si, Gyeongju, Gyeongbuk, Republic of Korea 780-714. Email: [email protected]. Elizabeth M. Tracy, PhD, Mandel School of Applied Social Sciences, Case Western Reserve University, Cleveland, OH, USA. David E. Biegel, PhD, Mandel School of Applied Social Sciences, Case Western Reserve University, Cleveland, OH, USA. Meeyoung O. Min, PhD, Mandel School of Applied Social Sciences, Case Western Reserve University, Cleveland, OH, USA. Michelle R. Munson, PhD, Silver School of Social Work, New York University, New York, NY, USA.

)

Journal of Behavioral Health Services & Research, 2014. 1–20. c 2014 National Council for Behavioral Health. DOI 10.1007/s11414-014-9406-y

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

Introduction Mental health and mental health service disengagement among transition age youth Prevalence rates of mental disorders among transition age youth (TAY), those between the ages of 18 and 30, are extremely elevated.1Among young adults, the lifetime prevalence rate of mood disorders, for example, has been found to be two times that of older age cohorts.2 Further, the onset of many mental disorders occurs during this time,3 a time when transition and uncertainty are common and stress is high.4 Also, continuity of mental disorders from childhood to adulthood is common.5 Beyond prevalence, research has documented the difficulties young adults with mental illnesses face as they transition to adulthood, such as poor education outcomes, underemployment, and increased risk of suicidality.6 These important young adult concerns are even more challenging if transition age youths with mental disorders do not engage in treatment for their mental health issues. Client disengagement is considered a significant obstacle affecting both the efficacy of mental health treatment and treatment outcomes.7,8 Mental health treatment is likely to be less effective if young adults receive less than full treatment doses. Further, client disengagement wastes staff time and mental health resources, denies access to others in need, and limits the number of people an agency or practice can serve.9 For these reasons, among others, disengagement in mental health services is critical to understand. Research has shown that for many TAY, the transition to adulthood leads to disconnection with mental health services.10,11 One study found that TAY are often making decisions about treatment more on their own.11 This can be problematic if during adolescence their access and engagement in service was facilitated and/or encouraged by their parents and/or guardians. A recent study found that some TAY perceive the support during adolescence as particularly helpful in their continued use of services.12 Specific rates of disengagement in mental health services among TAY are extremely difficult to estimate, since there are few studies specifically focused on this population. However, recent approximations of results from studies of children’s mental health engagement suggest that 40 to 60% of children who receive treatment terminate it prematurely.13,14 Olfson et al. (2009) utilized a national sample of those over age 18 and found estimates on premature treatment termination ranging from 22 to 32%.15 Other research found that attrition rates at initial intake appointments can range from 48 to 62% among those youth accepted for an evaluation.16 Across the population of children, youth, and young adults, high drop-out rates exist in various service settings and treatment stages, and client engagement is an essential yet challenging component in effective mental health service delivery.17 Research focused on addressing this persistent problem in new and innovative ways is needed.

Predictors of service disengagement: theoretical and empirical research Engagement has widely been addressed in the children’s literature,18 particularly focusing on individual- or family-based theoretical frameworks. This research indicates that barriers to engagement can be found within a client’s behavior, the beliefs and experiences of the family, lack of sensitivity by professionals, and agency procedures. More relevant to the present study, Munson et al. (2012) published a theoretical framework for understanding young adult mental health service use decisions.12 The study, which focuses on the service use experiences of 18- to 25-year-olds over the transition to adulthood, found a unique set of barriers and facilitators of service use among transitioning young adults, including stigmarelated concerns, emotional reactions, and the need for increased knowledge to make increasingly independent decisions regarding services. This study also found that macro-level factors, such as community factors, agency factors, and social relationships, play a significant role in service use decisions and engagement. The study suggests that research on engagement in mental health services needs to focus on the individual, social, and community levels and, in particular, examine how these different levels of influence impact engagement.

The Journal of Behavioral Health Services & Research

2014

There are numerous studies examining barriers to engagement from the client/family perspective, but comparatively few have considered the perspective of the service provider and the overarching organizational culture. The present study will examine barriers to engagement among transition-age youth from the perspective of the organization and provider. A shift of focus: the role of the organization and the provider in service engagement The present study focuses on organization and provider variables and therefore draws upon sociotechnical theory, theory of organizational culture, and resource dependence theory. Socio-technical theory integrates the social context and core technical processes of an organization in order to understand how each affects the other.19 The socio-technical theory suggests that organizational-level factors (e.g., type of organization, service setting, culture, and structure) can explain the variance in an organizations’ “core technology” (e.g., engagement strategies, treatment), which subsequently impacts overall service provision.20 The model of organizational effectiveness views organizations as creating a social context (e.g., organizational culture) within which the technical work of the organization is performed.21 This model assumes that the organization’s “core technology” (e.g., type of treatment, engagement strategies and skills) is embedded within a social context that is created by the organization.20 The organization’s successful performance depends as much on social processes in the organization as on technical processes. This is salient in mental health services research as it suggests that effective performance in mental health service provision depends on both social and technical processes, as well as the practitioners’ expectations, perceptions, and attitudes.20 According to the socio-technical theory, an organization’s social context can complement and enhance the adoption and successful implementation of new technologies (e.g., new treatment model or engagement strategies), present barriers to the adoption of new technologies, or truncate or adapt a technology in ways that reduce the technology’s effectiveness. In turn, the perceptions of an organizational social context may influence clinician performance such as service engagement. These aspects of service engagement among transition age youth have been largely ignored thus far in research. According to the theory of organizational culture, culture can affect work performance and organizational effectiveness by influencing the individual workers directly or indirectly.22 Organizational culture typically refers to basic assumptions, values and behavioral norms, and expectations found in an organization or its subunits.23 The theory of organizational culture suggests that the social norms, expectations (e.g., the extent to which practitioners are expected to be proficient in their work), meaning, perceptions (e.g., whether practitioners perceive a high level of personal engagement in their work with clients), and attitudes (e.g., practitioners’ commitment to the organization in which they work) are the keys to understanding individual behavior in organizations and organizational effectiveness.20 These factors can encourage or inhibit the adoption of the best practices, strengthen or weaken fidelity to established protocols, support or attenuate positive relationships between service providers and consumers, and increase or decrease the availability, responsiveness, and continuity of services provided by the organization. In addition, there are some organizational culture studies that explore whether or not organizational culture generates a moderating effect that affects the relationship between individual/team level and organizational level (i.e., cross-level interaction). For example, Glisson and James’s (2002) study revealed that more constructive cultures were associated with higher service quality (e.g., continuity, availability).24 This study also revealed that there are cross-level interaction relationships that link constructive culture to individual attitude and perceptions and behavior. Organizational success in resource dependence theory (RDT) is defined as organizations maximizing their power.25 According to RDT, practitioners lacking in essential resources for engaging TAY will seek to establish relationships with others in order to obtain needed resources.26 That is, to achieve success in the organization, it is necessary to facilitate the delivery of service by coordinating the contributions of multiple service providers and scheduling services so that they are provided without

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

any delay that might adversely affect the service recipient’s condition. If socio-technical theory and organizational culture theory are related to organizational effectiveness, RDT is related to organizational efficiency. While organizational effectiveness is a measure of whether an organization can create acceptable outcomes, an external measure, organizational efficiency, or how well an organization does what it intends to do given its resources, is an internal measure. The RDT suggests that how well an organization maximizes power (e.g., service coordination, cooperation, teamwork) is the key to understanding organization’s success in performance, specifically successful service engagement. The present study The objective of this research was to expand the examination of engagement beyond individual and family level factors, focusing on practitioner and organization level factors, while also examining this perplexing problem among a particularly vulnerable and understudied population, transition age youth. Guided by socio-technical theory and organizational culture theory, this study explored two broad research questions: (1) how organization-level characteristics are related to TAY service engagement, while controlling for practitioner-level characteristics and (2) how organizational culture moderates the relationship between practitioner professional characteristics and the youth service engagement in mental health services (see Fig 1).

Methods Study design and participants The present study utilized a cross-sectional survey research design to explore the relationship between service providers and service engagement among transition age youth (TAY). Two hundred seventy practitioners from 27 mental health service organizations in OH were selected using convenience sampling method. As potential target organizations, there were 78 mental health organizations in which represent three metropolitan areas (Cleveland, Columbus, Cincinnati) in OH. Inclusion and exclusion

Figure 1 Hypothesis testing model

The Journal of Behavioral Health Services & Research

2014

criteria for the study sample are identified. County contracted urban community mental health service organizations that provided counseling, individual treatment programs, community psychiatric supportive treatment, community-based treatment foster care, and/or outpatient mental health clinics were contacted by letter or phone to recruit participants. Residential settings and inpatient hospital mental health organizations were excluded because many of the individuals admitted to the programs were discharged involuntarily after a certain amount of treatment, which would impact the measure of engagement. Practitioners who offered TAY mental health services and those who work directly with TAY were the proposed study participants. Professionals who had worked in the agency for more than 1 year were invited to complete the survey. This inclusion criterion was established to ensure that the participating staff were familiar with the organization’s culture. The broader categories included (clinical) social workers, mental health counselors, case managers, supervisors, therapists as well as many other professionals who had similar job titles. Executive directors or administrators were excluded because they are often removed from the provision of direct services. A total of 287 surveys were sent to providers, and a total of 258 surveys were returned, for an overall response rate of 89.8%. Five surveys were not included in the analysis because three of them were completed by non-eligible individuals (e.g., administrators), and two of them included no responses to any of the SES items or OCMH items on the survey The total sample of this study (including the sample from the pilot test −26 practitioners, 4 organizations), was 279 practitioners from 27 organizations. The average number of participants from an organization was 10.7 (SD=5.6), ranging from 3 to 23. Measures The primary questions and hypotheses that were addressed in the current study focused on (1) how organization-level characteristics are related to TAY service engagement, while controlling for practitioner-level characteristics and (2) how organizational culture moderates the relationship between practitioner professional characteristics and the youth service engagement in mental health services. The dependent variable was practitioners’ perceived service engagement of transition age youth. Independent variables were categorized by two levels (practitioner level and organizational level). Practitioner level variables (as a function of control variable) include (1) socio-demographic characteristics (race, gender, age, education, years in organization) and (2) professional characteristics (in-service training, resource knowledge, use of referrals, coordination, service provision, case load size, service barriers, teamwork). Organizational level variables include (1) organization characteristics (location, service setting) and (2) organizational culture. Organizational culture was viewed as independent variable as well as moderating variable. Variables are summarized as sets of dependent, independent, moderating, and control variables. All variables’ categories and operational definitions and internal consistency in this study are presented in Table 1. Service engagement The Service Engagement Scale (SES) is a 14-item measure consisting of statements assessing client engagement with services. Respondents rate each item on a four-point Likert scale: 0=not at all or rarely, 1=sometimes, 2=often, and 3=most of the time.27 Negatively worded questions were reverse scored. The total score ranges from 0 to 42, with higher scores indicating greater levels of engaging with services. Item content includes (1) availability (three items), which refers to the client being available for arranged appointments (e.g., “Clients avoids making appointments”), (2) collaboration (three items), which refers to the client actively participating in the management of illness (e.g., “Actively participates in planning treatment”), (3) help-seeking (four items), which refers to the client seeking help when needed (e.g., “Difficulty in asking for help”), and (4) treatment adherence (four items), which refers to the client’s attitude toward taking medication and coming to treatment (e.g., “Client refusing treatment”). In the present study, the total SES score

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

Table 1 Variables, operational definition, measures, and levels of measurement Dependent variable Service Engagement Scale (SES)a

Independent variables Practitioner level

Organizational level

Operational definition Practitioner’s rating on the level of client availability for treatment, collaboration, help seeking behaviors, and treatment adherence Operational definition

Socio-demographic characteristics Gender Sex as defined by practitioner Age Current age Race Ethnicity as defined by practitioner Education The highest level education achieved Years in organization The number of years the practitioner has worked in this agency Professional characteristics In-service training The total hours of in-service training within a month Use of referrals The degree of referrals made to other service systems Coordination Coordinating with other organizations Case load size The number of clients currently treating by practitioner Teamwork The frequency of individual and group staff meetings Resource knowledgea The number of service categories with which practitioner was familiar The number of different services that Service provisiona practitioner provided directly to youth The number of perceived gaps in the Service barriersa availability and continuity of services available to youths at the organization Organization characteristics Location The location of the population practitioner served Service setting The type of outpatient service setting Organizational culture The assumptions, beliefs, values and Organizational behavioral expectations shared Culture in Mental within organizations Health (OCMH)a Scale dimensions include professional support and trust (PST), professional values (PV), empowerment, avoidance, and control (EAC), and organization environment (OE)

Cronbach’s alpha α=0.73

Cronbach’s alpha

30-item “yes or no” checklist 21-item “yes or no” checklist 10-item “yes or no” checklist

PST α=0.92 PV α=0.90 EAC α=0.86 OE α=0.63

a

Indicated variables with standardized measurements or indices/checklists, Cronbach’s alpha was calculated for standardized measure only

The Journal of Behavioral Health Services & Research

2014

was used because this measure adequately reflects the multiple dimensions of engagement with mental health services. The scale has high internal consistency, retest reliability, and good face validity and content validity, supported by clinician ratings of engagement.27 The SES has been used with samples of community-based mental health clinicians and their clients.27–34 In this study, skewness and kurtosis values were all in a range of+1 to −1, and there were no significant normality violations for any of the subscale. An examination of normal probability plots of residuals revealed no violations of normality, linearity, or homoscedasticity of residuals. The mean scores for the SES subscales and total scale scores were above the mid-point of the scale. Average scores for availability for treatment (M=5.8, SD=1.3), collaboration (M=5.7, SD=1.5), help seeking behaviors (M=6.4, SD=1.6), treatment adherence (M=7.5, SD=1.5), and total SES (M=25.4, SD=4.0) suggested, in general, practitioners of this study perceived above the mid-level of engagement among TAY clients in mental health services. Cronbach’s alpha statistics were computed to assess the internal consistency of the SES as a single dimension measure and for each of the subscales. Cronbach’s alpha was acceptable for the total SES (α=0.73). Item-total correlation demonstrated moderate to strong correlation to total scores with coefficients ranging from 0.23 to 0.45. Since this measure reflects the multiple dimensions of engagement with mental health services that are related (range of subscale correlations r=0.48∼0.66). Socio-demographic characteristics The socio-demographic variables included respondent’s gender, age, and race (White Caucasian; coded as 0 vs. non-White Caucasian; coded as 1). Practitioners were asked the number of years they had worked in the organization and the highest level of education achieved. Level of education was recoded into a dichotomous variable with “Less than masters (including community college, bachelors degree; coded as 0)” and “masters or above (including masters and doctorate degree; coded as 1).” In this study, participants were 81.4% female and 39.6 (SD=11.9) years old on average with a range of 23 to 73. The respondents were 79.6% Caucasian, and nearly two-thirds of the practitioners had a master’s degree (64.9%). Over half (51.5%) of practitioners had worked in the mental health field for more than 10 years. Mean years in the agency was 5.7 (SD=6.3) and ranged from 1 to 38 years. Almost three fifths (59.3%, N=16) of the organizations provided services to clients in urban areas (see Table 2). Professional characteristics In-service training was the total hours of in-service training they had within a month. Use of referrals was a four-point Likert scale item which assessed the level of referrals made to other service categories (e.g., “To what extend is providing youth with referrals to drug/alcohol, health, education, inpatient, and outpatient resources a part of your job?”). The average hours for the inservice training a month was 4.7 (SD=7.6) and ranged from 0–40 h. Coordination was a four-point Likert scale item which assessed the frequency of service coordination with other organizations (e.g., “When you work with a youth, how often do you coordinate with other agencies in providing services?”). Different from use of referrals, coordination of service included interagency treatment planning teams, targeted case management services, and individual family-based support services. Mean score of use of referrals and coordination was 1.9 (SD=0.9) and with a range of 0 to 3. Case load size is the total number of clients treated by the practitioner in the past 3 months. Average caseload size reported by participants in the past 3 months was 42 with a range of 1–112 cases. Almost half (49.5%) had less than 34 cases. Teamwork was measured with two items. The first item focused on frequency of individual meetings with supervisors or supervisees for receiving or providing supervision. The second item focused on frequency of group staff meetings. Each of these items was scored on a 4-point Likert scale that included daily (1), several times per week (2), weekly (3), and monthly (4). The responses to the items were summed to create a single score of frequency of teamwork ranging from 2 to 8, with

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

Table 2 Socio-demographic characteristics of practitioners Number of samples (%) Gender Female Male Age Race Caucasian African American Hispanic Asian Other Education Community college Bachelors Masters Doctorate Other Years in field Years in the organization

M (SD)

Min.

Max.

39.6 (11.9)

23

73

11.8 (9.1) 5.7 (6.3)

1 1

38 38

227 (81.4) 52 (18.6)

222 46 6 2 3

(79.6) (16.5) (2.2) (0.7) (1.1)

7 82 181 8 1

(2.6) (29.4) (64.9) (2.9) (0.4)

higher scores indicating more frequent teamwork with others in providing services. The average of teamwork score was 4 (SD=1.3) and with a range of 2 to 8. Resource knowledge was indicated by practitioner familiarity with resources in other service systems and the community (e.g., “The following list names a number of different categories of mental health, substance use, or other resources, please check any agencies or organizations that you are familiar with.”). Practitioners reported their familiarity with 30 service categories grouped into four major service domains: six for the health and education domain, seven for the inpatient behavioral health care domain, eight for the outpatient treatment domain, and nine for the “other” service domain.35 Practitioners indicated “yes” or “no” for each service category, and a total familiarity score (ranging from 0 to 30) was computed, with higher scores representing more practitioner knowledge of resource. In this study, on average, practitioners were familiar with 23 of the 30. Service provision was assessed through questions about direct provision for 21 possible mental health services (e.g., “In your service setting, which of the following services do you personally provide to clients?”).36,37 Practitioners reported whether they provided any of 21 types of services (e.g., assessment, individual treatment, etc.). A summed score of service provision (range from 0 to 21) was computed, and higher score represents more service provisions to clients. On average, practitioners provided eight different types of services on behalf of TAY, and ranged from 0 to 19. Service barriers were measured by perceived service barriers of the practitioners (e.g., “Concerning the availability and continuity of services available to youths at your agency, what are the gaps?”). This was a 10-item checklist of potential barriers/gaps (e.g., transition between services, coordinating services) which can affect the availability and continuity of services available to client at the organization.35 Practitioners checked whether the barrier existed at their agency. The measure yielded a

The Journal of Behavioral Health Services & Research

2014

summed score of perceived service barriers (ranging from 0 to 10); higher scores indicated more barriers in services. On average, practitioners perceived 3 barriers out of 10. Organization characteristics With regard to location, practitioners were asked to characterize their service area with the following question in the survey: “which best describes the location of the population you serve?” Response options included three categories dummy coded into two categories (as there were no practitioner responses to the rural option) for the purposes of this study: “suburban” (coded as 0) and “urban” (coded as 1). Practitioners further classified the outpatient service settings in which they worked: (1) private therapist, psychologist, psychiatrist, social worker or counselor, (2) mental health clinic or center, (3) partial day treatment program, or (4) in-home therapist, in-home counselor. Response options were dummy coded into two categories for the purposes of this study: “outpatient (1, 2, and 3: coded as 0)” and “outreach (4: coded as 1).” All practitioners in this study reported working in private non-profit organizations. Almost three fifths (59.3%, N=16) of the organizations in this study provided services to clients in urban areas. Twenty-one (77.8%) organizations were “outpatient” settings which included (1) private therapist, psychologist, psychiatrist, social worker or counselor (n=2, 7.4%), (2) mental health clinic or center (n=18, 66.7%), (3) partial day treatment program (n=1, 3.7%). Six (22.2%) organizations were “outreach” settings which provided in-home therapy and counseling services. Organizational culture The Organizational Culture in Mental Health (OCMH) scale measured aspects of the organizational culture.38 The OCMH was designed based on theoretical aspects of organizational dynamics in people-serving organizations. Compared to other organizational culture measures, the OCMH was beneficial for capturing cultural elements of mental health service organizations. This 60-item scale had been well validated and showed good reliability.38 In addition, a qualitative study39 of staff evaluations and perceptions of organizational culture in the adult outpatient mental health services center concurs on the presence of specific cultural components and provides construct validity for organizational elements identified by the OCMH. Respondents indicated on a six-point Likert scale from “Not at all” to “always.” Schiff’s study showed that the OCMH includes four substantial and two tentative subscales. However, in this study factor analysis of the data produced four substantial subscales with strong psychometric properties. Scale dimensions include professional support and trust (PST) (α=0.91), professional values (PV) (α=0.90), empowerment, avoidance, and control (EAC) (α=0.88), and climate: organization environment (OE) (α=0.69). Summed subscale scores were computed (possible range from 0 to 80 for PST and PV; 0 to 60 for EAC; 0 to 30 for OE), with higher scores representing constructive organizational culture and lower scores representing defensive organization culture in general. Specifically, the PST represents the nature of relationships and interactions among employees, supervisors, and managers.38 The scale measures an atmosphere of openness, responsiveness to feedback of both staff and clients, and responsiveness to individual needs (e.g., “Is there an atmosphere of trust, sincerity and mutual respect”). In mental health organizations with a high score on PST, freedom of expression is encouraged, and management is open to feedback from all levels of the organization.38(p.99) The PV focuses on professional attitudes and counseling values inherent in mental health professional practice (e.g., “Do staff show equal respect for all clients they serve”). This subscale reflects the core value of belief in the dignity and worth of people that have always been at the heart of counseling processes.38( p.100) The EAC dealt with issues of power and control that create dysfunction in the workplace (e.g., “Are there cliques that compete for power, preference and control”). Organizations scoring high on this factor are difficult places to work, but in the current study, the scores were reverse coded so that

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

higher scores indicated fewer issues of power and control in the workplace. The OE focuses on aspects of the physical environment that ensure the workplace is comfortable and physically approachable (e.g., “Is there age-appropriate reading material for the waiting room”).38(p.101) In this study, the mean score of the subscale was above the mid-point of the possible range for the PST subscale (M=48.0, SD=14.8, range=0∼80). The mean score of PV was also above the mid-point of the possible range (M=58.5, SD=11.3, range=0∼80). Since the EAC was reverse coded, organizations scoring high on this subscale had constructive culture. The summed score of the responses to the EAC items was above the scale mid-point (M=41.7, SD=10.6, range=0∼60). The mean score of OE was also above the scale mid-point mean (M=19.5, SD=5.0, range=0∼30). Cronbach’s alpha for the PST (α=0.92), PV (α=0.90), and EAC (α=0.86) subscales were in the acceptable range. The OE yielded lower Cronbach’s alpha (α=0.63) suggesting that results should be interpreted with caution.

Analysis Due to the hierarchical structure of the data (level 1, practitioner; level 2, organization; individual practitioners nested within organizations), a multilevel modeling approach was used to examine individual practitioners’ perceptions of service engagement as they relate to both practitioner-level and organizational-level measures. The study data had a nested structure: practitioners within the same organization represented a cluster because they were exposed to similar organizational influences (e.g., organizational culture); they were more similar to one another than to practitioners in other organizations. Therefore, their responses to survey items were not independent, which would have been a violation of the independence of observations assumption in OLS regression.40 Hierarchical linear modeling (HLM) addresses this issue using HLM, version 7 (2010). In the current study, the following set of equation was tested: ðService engagementÞij ¼ β 0 j þ β 1 j ðgenderÞ þ β 2 j ðageÞ þ β 3 j ðyrs in organizationÞ þ β 4 j ðin‐service trainingÞ þ β 5 j ðuse of referralsÞ þ β 6 j ðcoordinationÞ þ β 7 j ðresource knowledgeÞ þ β 8 j ðservice provisionsÞþ β 9 j ðservice barriersÞ þ β 10 j ðraceÞ þ β 11 j ðeducationÞ þ β 12 j ðcase load sizeÞ þβ 13 j ð teamworkÞ þ r ij Level 2 model: β 0 j ¼ γ 00 þ γ 01 ðPSTÞ þ γ 02 ðPV Þ þ γ 03 ðEACÞ þ γ 04 ðOEÞ þ γ 05 ðlocationÞ þγ 06 ðsettingÞ þ u 0 j Level 1 model:

β 1 j ¼ γ 10 þ u 1 j β 2 j ¼ γ 20 þ u 2 j β 3 j ¼ γ 30 þ u 3 j β 4 j ¼ γ 40 þ u 4 j β 5 j ¼ γ 50 þ u 5 j β 6 j ¼ γ 60 þ u 6 j β 7 j ¼ γ 70 þ u 7 j β 8 j ¼ γ 80 þ u 8 j β 9 j ¼ γ 90 þ u 9 j β 10 j ¼ γ 100 þ u 10 j β 11 j ¼ γ 110 þ u 11 j β 12 j ¼ γ 120 þ u 12 j β 13 j ¼ γ 130 þ u 13 j

The Journal of Behavioral Health Services & Research

2014

(Service engagement)ij – The predicted service engagement of an individual practitioner (i) within a given organization ( j) β0j – Mean service engagement for organization j β1j∼β13j – Slopes for organization j γ00 – Level 2 intercept γ01∼γ06 – Level 2 slopes γ10∼γ130 – Means of the slopes across organizations rij =σ2 – Level 1 residual variance U0j =τ00 – Variance in the intercepts U1j∼U13j =τ11∼τ1313 – Variances in the slopes Level 1 reveals practitioner’s characteristics in relation to service engagement. Level 1 yields two outcomes (means and slopes) examined at level 2. Level 2 examination shows the extent organizations’ variations in organization characteristics and organizational culture that are associated with mean organizational service engagement and relationships observed at level 1, such as the relationships of practitioner characteristics to service engagement. According to Klein and Kozlowski (2000), when assessing shared unit constructs, researchers must demonstrate substantial within-group agreement before using the mean of unit members’ scores to represent the unit.41,42 Consensus within organizations was analyzed through the use of the rwg statistic for within-group consistency concerning the level of consensus that exists among organization members concerning organizational culture specific variables.24 The rwg was calculated by comparing an observed group variance to an expected random variance. It provided a measure of agreement for each group rather than an omnibus measure for the groups as a whole. The values of rwg vary between 0 and 1 with a high value indicating agreement among raters and a low value indicating a lack of agreement among raters. Generally, a rwg of 0.70 or higher is acceptable.41 In this study, the rwg indices were sufficiently large to justify aggregation. The range of average rwg scores (0.81∼0.96) indicated consensus within groups in the sample concerning the variables in question.43 Thus, this provided evidence that there was a high degree of agreement within each organization on OCMH subscales, and these variables could be aggregated to represent each organization.43 Consensus emerged concerning all four of the subscales that were included as level 2 variables in the HLM analyses. Specifically, with information provided from the multilevel model, the relationship between practitioner level predictors (demographic characteristics, resource knowledge, use of referrals, inservice training hours, coordination, teamwork, service provision, case load size, and service barriers) and their service engagement was controlled. After establishing that there was significant variance across organizations in the level 1 intercepts, the organization level hypothesis (type of organization, location, service setting, and organizational culture) was directly tested. Finally, after establishing that significant organization variance in the slopes was present in the multilevel model, it was then explored whether the variance in the slope across organizations is significantly related to the level 2 independent variable (e.g., organizational culture). This was a direct test for the crosslevel moderator. To assess the hypotheses, a sequence of models is required: the unconditional (one-way ANOVA), random-coefficient regression, intercept-as-outcomes, and slopes-as-outcomes models.44 Hypothesis testing model is presented in Figure 1. The first step in evaluating an HLM is equivalent to a one-way ANOVA test of dependent variables and yields variance component estimates and significance tests of the within- and between-group variance.41 An unconditional model did not include covariates at either the practitioner or organizational level. The unconditional model provided a baseline against which other models that included independent variables could be compared. In the current study, the model was performed to determine whether there were significant within- and betweenorganization differences in service engagement, which is the dependent variable in this study. The estimates for the unconditional model are shown in Table 3. The intraclass correlation

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

Table 3

t-ratio

−0.097 −1.647 1.745 −0.415 0.654 0.521 −0.208 −0.620 −0.187 1.870^ −3.605*** −3.175** −0.283

−0.004 −1.622 0.781 −0.263 0.024 0.018 −0.012 −0.178 −0.069 0.092 −0.017 −0.363 −0.079

1.269 (20.025***) 8.413

68.322***

25.111

3.420** −1.531 2.483* −0.019 2.019^ −1.642

0.939 −0.475 0.077 −0.001 0.060 −0.117

2.789 (23.109) 8.142

−0.199 −1.468 1.378 −0.476 0.617 0.539 −0.181 −0.627 −0.110 2.037^ −3.015** −2.727* −0.305

48.589***

t-ratio

−0.009 −1.344 0.707 −0.283 0.024 0.020 −0.011 −0.188 −0.042 0.102 −0.017 −0.323 −0.085

24.803

Coefficient

6.996*** −4.071***

0.226 −0.032

2.827 (22.832) 7.957

2.762* −2.316* 2.634* −0.277 2.105* −1.823

−0.050 −1.620 1.425 −0.624 0.035 0.582 −0.887 −0.295 0.350 3.040** −2.551* −2.452* −0.519

49.673***

t-ratio

0.752 −0.709 0.079 −0.013 0.068 −0.120

−0.002 −1.570 0.732 −0.367 0.001 0.023 −0.041 −0.091 0.100 0.144 −0.014 −0.310 −0.139

24.972

Coefficient

Slopes-as-outcomes

1=male for gender; 1=non-white for race; 1=masters or above for education; 1=urban for location; 1=outreach setting for service setting; (−) indicates negative direction against anticipation ^pG0.1; *pG0.05; **pG0.01; ***pG0.001

83.709***

Coefficient

Intercept-as-outcomes

2014

0.989 (44.378**) 15.068

25.455

t-ratio

Random coefficient

Fixed effect Constant Level 1 Age Female Education Non-white Years in organization In-service training Resource knowledge Use of referrals Coordination Service provision Case load size Service barriers Teamwork Level 2 Location Service setting PST PV EAC OE Interaction Coordination×PV Resource knowledge×EAC Random effect Between variance (χ2) Within variance

Coefficient

Unconditional

HLM results of organizational culture on mental health service engagement

The Journal of Behavioral Health Services & Research

coefficient (ICC) or the ratio of between-group variance to total variance (τ00/(τ00 +σ2)) indicated that approximately 6% of the overall variance in service engagement lies between organizations. The variance component was statistically significant (χ2 =44.38, p=0.014), indicating that there was sufficient variability in the average levels of service engagement in different organizations to warrant further modeling. The significant fixed effect for the parameter estimate of grand mean service engagement indicates that the mean service engagement for all practitioners across all organizations was significantly different from zero.

Results Research question 1. How are organization-level characteristics related to TAY service engagement, while controlling for practitioner-level characteristics? At first, to control for the practitioner level variables in the study, the random-coefficient regression model was used. In this step, covariates at the practitioner level were added to the level 1. To see how organization-level characteristics are related to TAY service engagement at level 2, the practitioners’ demographic and professional attributes varied across the population of organizations at level 1 should be controlled. The values of t-ratio provided in Table 3 suggested that the grand mean intercept (pG0.001) and mean slopes of service barriers (p=0.004) and case load size (p=0.001) were significantly different from zero. This result suggested that, on average, practitioners who had higher perceptions of service barriers and greater case load size were associated with lower levels of perceived youth service engagement. Based on significant variance in the intercept terms across organizations assessed from the randomcoefficient regression model, to explore the main effect of organizational level predictors, the intercepts-as-outcomes model was performed. The purpose of this model is to see whether the betweenorganization variance was significantly related to organization characteristics and organizational culture. Namely, level 2 examined variations in mean organizational service engagement as a function of organizational characteristics through to be related organizational-level predictors (PST, PV, EAC, OE, location, service setting). In this model, six organizational-level predictors were added. Table 3 provides a summary of the results of the intercepts-as-outcomes model. As Table 3 indicates, the parameter of location was significant (p=0.003), implying that the practitioners who provided service to urban populations perceived higher levels of service engagement. The parameter of PST in Table 3 revealed that service engagement was significantly associated with PST (p=0.022). Even though the relationship between service engagement and EAC was not significant (p=0.057), it showed the anticipated direction. These findings indicated that in an organization where the level of professional support and trust was higher and issues of power and control in the workplace were fewer, practitioners perceived that their youth clients were more engaged in mental health service. In order to obtain the amount of intercept variance accounted for by organizational variables, the variance in the τ00 from the random-coefficient regression model (τ00 =1.27, the total between-organization variance in the intercept term across organizations) with the variance in the τ00 for the current model (τ00 =2.80, the residual variance in the intercept after accounting for level 2 variables) was compared. The R2 indicated that organizational level variables accounted for 21% of between-group variance in the intercepts. Research question 2. Does organizational culture moderate the relationship between practitioner professional characteristics and the youth service engagement in mental health services? The second aim of this study represented a cross-level moderation effect, in which organizational culture variables (e.g., PV, EAC) were hypothesized to moderate the relationship between

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

practitioner level variables (e.g., coordination, resource knowledge) and service engagement. Utilizing the slopes-as-outcomes model, an examination was made to see if the variance in level 1 slopes across organizations was significantly related to organizational culture. That is, organizational culture variables were hypothesized to affect the slopes (direction and/or strength) of the relationship between practitioner’s professional characteristic and service engagement. Among the possible interactions between the six professional characteristics and four cultural dimensions (total of 24 interaction terms), the pairs correlated with the given outcome at pG0.20 (6 interaction terms) were entered in the model. Then, using the series of slopes-as-outcomes model based on the hypotheses, a more parsimonious model was developed. That is, interaction terms that failed to reach levels of significance were trimmed from the model. Results presented in Table 3 indicate that two of the level 2 slopes (PV; EAC) were significant both at pG0.001 level, thereby supporting moderation hypotheses. This positive parameter estimate for PV (0.226) suggests that the slope relating coordination to service engagement became stronger (steeper). That is, in an organization which has higher level of professional values, the relationship between coordination with other organizations and service engagement became stronger. Interestingly, this negative parameter estimate for EAC (−0.032) suggests that the slope relating resource knowledge to service engagement became weaker (flatter). Specifically, fewer issues of power and control in the organization weakened the positive relationship between resource knowledge and levels of service engagement. The R2 indicated that 63% of variance in the relationship between coordination and service engagement is accounted for by PV, and 77% of variance in the relationship between resource knowledge and service engagement is accounted for by EAC. In this final model, additional level 1 and level 2 predictors were significantly related to service engagement. For the level 1 predictor, the practitioner who provided various types of service provision (p=0.005) perceived higher level of service engagement. For the level 2 predictors, in an organization where the issues of hierarchy and controls were fewer (p=0.048) and mental health clinic setting (p=0.031) (vs. outreach setting) practitioners perceived higher levels of TAY engagement in mental health service.

Discussion Relationship between practitioner characteristics and service engagement In this study, to explore the organization level predictors and their association with perceived service engagement, the practitioner-level characteristics were controlled in the analysis. That means this research assumed that practitioner socio-demographic and professional characteristics are related to practitioners’ perceptions of service engagement. First, the results indicated that the amount of individual service provision was positively related to service engagement: on average, the practitioners provided eight different types of service actions. That is, TAY clients received multiple types of care. Having more service types involved in their care increases the likelihood that a greater number of service needs are addressed. This result reinforced the importance of practitioners’ multiple roles as gateways to care for TAY. Second, a greater case load size was associated with a lower level of service engagement. This finding is concordant with the existing research. Cunningham and Henggeler (1999) discussed that the size of the practitioner’s caseload or the burden of other priorities can affect how much time and effort a practitioner can commit to engaging a challenging client.45 As Staudt (2007) indicated, many settings do not provide optimal conditions for practitioners to successfully engage clients.46 For example, some mental health organizations may require practitioners to schedule seven or eight clients a day. It is not likely that practitioners will be predisposed to follow up with clients who miss appointments when “noshows” require much needed time to complete paper work and return phone calls. Third, a practitioner’s perception of a low number of service barriers was associated with a higher level of

The Journal of Behavioral Health Services & Research

2014

service engagement. This finding is consistent with Kazdin et al.’s (1997) evidence that therapists’ perceptions of barriers predicted client treatment continuation more strongly than did the client’s own self-report.13 In the current study, one of the most frequently reported service barriers perceived by practitioners was a service delay—“transition between services (41%).” In other words, organizations which helped TAY with transitional services may have a greater chance of continuity of care. For example, some organizations provided specific types of services which could promote transitions for TAY with mental health problems: case management, transition planning, follow-up on referrals, and long-term planning. Realistically, for the TAY population, service delay is frequent because the services they received from the child welfare system are no longer available. For instance, in the attrition study of Barrett et al. (2009), service delays of more than 1 day were reported by clients as a significant influence on their decision to leave treatment.9 Moreover, the longer a client has to wait for services, the more likely the patient is to withdraw, particularly if the wait is more than 1 week. These factors may prove especially influential for clients referred by hospitals or other service organizations. Therefore, anything that can be done to respond promptly to clients and reduce the length of time a person has to wait before beginning treatment is likely to increase the level of engagement. Relationship between organization characteristics and service engagement Based on the socio-technical theory, the first aim of this study was to explore the organization level predictors and their association with perceived service engagement. Under the framework, predictors for perceived service engagement were considered a predisposing organization’s structure characteristics and organization culture factors. First, as it was hypothesized, an organization serving an urban population had a higher level of service engagement than suburban. Sixteen organizations in this study provided service to TAY clients in urban areas. Eleven organizations were servicing a suburban population. A number of understandable accessibility issues related to organization location are common to most studies of service engagement.47Most frequently, transportation problems are cited, and distance travelled is another correlate. The studies’ assertion is that rural or suburban areas with less public transportation may produce barriers of service engagement. Interestingly, results in the study of Mitchell and Selmes (2007) indicated that service providers in urban areas were more likely to accept parent involvement in decision-making, so differences in perceptions of family involvement by community type had been anticipated.47 Barrett and colleagues (2009) suggest that “although the research about environmental influences on attrition is less conclusive than that on other domains, factors such as the accessibility of the clinic and the office environment are important considerations in seeking ways to increase client retention.”9In addition, research has shown that most mental health care professionals tend to be concentrated in urban areas and are less likely to be found in the most rural sections of the country.9 Therefore, consistent with the above studies, this study also supported the assertion that the problem of accessibility may be intensified depending on where youths live. TAY in suburban areas in need of mental health care may not be able to engage well in the services and locate appropriate services. Second, as HLM results indicated, the type of service setting was a significant predictor for TAY service engagement. Consistent with this prediction, mental health clinic settings had higher levels of service engagement than outreach service settings. In this study, four types of service settings were collapsed into two categories: “outpatient (77.8%)” and “outreach (22.2%).” Six organizations were in the categories of “outreach,” and those organizations provided both regular outpatient services and home-based intervention as well. Home-based intervention is an intensive type of mental health counseling designed to help stabilize difficult family situations so that youth with severe disorders can continue to live at home and families can remain intact.48 This finding is consistent with O’Mahen and Flynn’s (2008) study that women with depression indicated that they had greater confidence in treatments delivered in a professional office or clinic vs. at home.49

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

Although some studies have shown that the type of outreach intervention (e.g., Assertive-outreach Community Teams) is more successful at engaging clients,27,30 the study samples were all “highrisk” adult client populations (e.g., clients with severe mental disorders, schizophrenia). Another possible reason why the clinic setting had a positive relationship with engagement is that the programs designed to facilitate educational and vocational advancement and other supportive relationships besides interventions that improve functioning and symptoms are more attractive for the youth population. Also, it may be the a perfect place for sharing information regarding the system of care communities for supporting youth experiencing many challenges with transitioning from the child service system to the adult service system. Third, constructive organizational culture was related to a higher level of perceived service engagement. These results support a growing literature that finds positive relationships between constructive organizational culture and staff appraisals of agency functioning and programming.53–55 Specifically, in terms of professional support and trust (PST), freedom of expression and management openness to feedback from all levels of the organization were more likely to have a higher level of engagement. The professional support touches upon issues such as positive feedback for a job well done and public recognition of accomplishments. Trust provides the basis for honest and open feedback, positive response to criticism, and a two-way evaluation process.38 This openness extends to soliciting feedback from clients and responding in an acceptable manner to criticism and opposing views. Consistent with the prediction in this study, professional support and trust was an important predictor for TAY service engagement. In terms of empowerment, avoidance, and control (EAC), consistent with the hypothesis prediction, fewer issues of power and control in the organization were more likely to be associated with a higher level of engagement. EAC was concerned with issues of power and control such as hierarchical decision-making and readiness to make administrative decisions in a timely fashion. The informal power structure of the organization is reflected in the organization’s handling of conflict among staff and the avoidance of conflict. Organizations scoring low on this subscale are difficult places to work and are organizations that are more likely to have low levels of service engagement. The organization environment scale (OE) deals with aspects of the physical environment that ensures the workplace is comfortable and physically approachable.38 In the study of Barrett and colleagues (2009), many patients complained that the building was uninviting, with waiting rooms congested and uncomfortable, all patients (adults, children, psychotic patients, etc.) waited in a single room, and treatment rooms were small and poorly ventilated.9It was demonstrated in that study that physical environment changes can make a difference in retention. Although environmental factors may be important considerations in seeking ways to increase client engagement,9 the analysis of the influences of organizational environment (OE) on service engagement was not significant in this study; the hypothesis for OE subscale was not supported. Professional attitudes and values (PV) is about the belief in the dignity and worth of people. Ethical issues abound in this professional values scale. It was expected that a higher score on PV would predict a higher level of service engagement, but the hypothesis for PV subscale was not supported. One of the possible reasons why the PV has no main effect is that clients have a prejudice or mistrust of the relationship with professionals.12 Although PV had no main effect on service engagement, it had a moderator effect on the relationship between service coordination and service engagement. Cross-level moderation relationship on service engagement Based on organizational culture theory, the second aim of this study was to explore the moderating role of organizational culture on practitioner-level characteristics that affect youth service engagement. First, the result indicated that a higher score on professional values strengthened the positive relationship between service coordination with other organizations and the level of TAY service engagement. The core values such as the belief in the clients’ ability to change have always been at the

The Journal of Behavioral Health Services & Research

2014

heart of the counseling process, and they are fundamental in human interactions that involve people changing behaviors. Practitioners’ values clearly influence the kinds of relationships they have with clients, colleagues, and members in other service categories.50 Practitioners make choices about the people with whom they want to work and collaborate. Practitioners’ values also influence their decisions about the intervention methods they will use in their work with clients and co-workers. For example, some practitioners choose to devote their careers to clients they perceive as victims among TAY with mental illness, such as victims of domestic violence and child abuse. The practitioners may prefer to use service coordination such as individual family-based supportive services in their collaboration with other organizations, believing that these are the most effective services for improving the situation. Therefore, in the organization which has a higher score on professional values, service coordination is more likely to influence the level of service engagement. Originally it was suggested that practitioners who perceived less control of the hierarchy in the organization had a stronger relationship between resource knowledge and service engagement, but the hypothesis was not supported. Although the hypothesis was not supported, interestingly, there was statistically significant relationship indicating that resource knowledge is a better predictor of service engagement in the organization with more hierarchical control. According to Schiff (2009), organizations scoring high on EAC are difficult places to work because of issues in hierarchy and control.38 Senior administrators are in command, and staff are given little input in the decisionmaking process and things are slow to change. The organization can be bureaucratic and respond slowly to changing client needs and the setting within the mental health system. Communication across various sections can be poor especially horizontal communication. In this type of organization, a practitioner’s resource knowledge and familiarity with other organizations which provide necessary service for TAY clients may be more effective for engaging the client. This is an important finding because TAY in the mental health system have multiple behavioral health and environmental problems requiring services from multiple service sectors.35 For example, while multiple services may not promptly be available for a client at the organization due to the dysfunctions in the organization, more resource knowledge of other service sectors may increase the likelihood that a greater number of service needs are addressed, which is the ultimate goal of the continuity of service. Limitations All research has limitations, and this study has limitations to consider when interpreting the results. First, the measurement of service engagement among providers is in its infancy, and this study is one of the first to use the Service Engagement Scale (SES). Previous use of the SES indicates that it is a reasonably good measure of a provider’s perception of client service engagement.27–34 Because the premise of this study was that service engagement was a complex area, with multiple dimensions of service engagement, using the total score was reasonable for the purpose of this study. In addition, service engagement only measured the perceptions of providers. Without considering the discrepancies of engagement scores between clients and providers, this presents some challenges for construct validity. Another issue is that because the perception of engagement was retrospectively assessed (e.g., “within the past 3 months”), the results may be subject to recall bias. Additional research supporting its validity and reliability is needed. Finally, all measures have been completed by the same respondents. It is widely assumed that common method bias inflates relationships between variables measured by self-reports.51,52 Therefore, results should still be interpreted cautiously until it can be tested on a wider range of populations.

Implications for Behavioral Health The findings of this study suggest implications for behavioral health in mental health service systems. It is important to consider why some practitioners/organizations are more successful than

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

others in engaging transition age youth in mental health services. To date, this question has not been examined among this population. Given the present findings, constructive organizational culture as a potential predictor of service engagement may be a good, though underestimated, predictor for building higher levels of engagement between practitioners and their clients. It may be valuable to consider strategies that target the PST, PV, and EAC of organizational culture with a focus on service provision by the teams/units within organizations. The activities geared toward sharing knowledge and opinions may impact perceptions of influence among staff in mental health agencies, and these staff may benefit from team building exercises. The present findings from the moderation analysis add to this literature by showing that constructive organizational cultures encourage the service coordination efforts that result in TAY’s higher level of service engagement. Also, it suggests that even though an organization has dysfunction, practitioners’ resource knowledge can improve service engagement. That is, the practitioners could adjust their core technologies (e.g., resource knowledge, service coordination) to their unique culture and discover effective ways to provide their services. For practitioners, these findings contribute to a better understanding of the role played by organizational culture in the effectiveness of mental health systems and the impact of organizational culture on service engagement. Further, alternative explanations are possible, such as organizations with more openness and sharing of knowledge may attract more competent and similar-minded professionals who are better able to engage clients, which has been found before.53 Although provider organizations cannot easily change their attributes, they can improve their organizational social context (e.g., organizational culture) with various types of training or strategies to facilitate both the clinician and the youths’ engagement in services. In order to improve organizational culture, interventions must target the specific factors in the organizational context of mental health agencies that tend to deleteriously affect practitioner work attitudes and the associated level of service engagement. Optimal interventions will reduce the dysfunction in organizations, by clarifying roles and maximizing professional support, while simultaneously improving the personal accomplishment and work environment. On the other side of the spectrum, the organizations could adjust their core technologies (e.g., engagement strategies) to their unique structure, culture, or climate. There has been evidence that core technologies in human service organizations are soft, malleable, and more often than not adopted to fit their organizations’ existing social contexts.20 That is, the fit between the organization’s culture and engagement strategies is achieved by adapting or reinventing the engagement strategies rather than by changing the organizational culture to support the core technology.56 Therefore, administrators may need to consider interventions designed and tested to improve engagement and decrease attrition that are congruent with their organization.

Acknowledgments This project was funded by the Ohio Department of Mental Health, under grant number ODMH 10.1260. Conflict of Interest There are no conflicts of interest declared.

References 1. Kessler RC, Birnbaum H, Bromet E, et al. Age differences in major depression: results from the national comorbidity survey replication (NCS-R). Psychological Medicine 2009; 40: 225–237. 2. Kessler RC, McGonagle KA, Zhao S, et al. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: Results from the National Comorbidity Survey. Archives of General Psychiatry 1994; 51(1): 8–19. 3. Newman DL, Moffitt TE, Caspi A, et al. Psychiatric disorder in a birth cohort young adults: Prevalence, comorbidity, clinical significance, and new case incidence from ages 11–21. Journal of Consulting and Clinical Psychology 1996; 64: 552–562.

The Journal of Behavioral Health Services & Research

2014

4. Arnett JJ. Emerging adulthood: A theory of adulthood from the late teens through the twenties. American Psychologist 2000; 55(5): 469– 480. 5. Kim-Cohen J, Caspi A, Moffitt TE, et al. Prior juvenile diagnoses in adults with mental disorder: Developmental follow-back of a prospective-longitudinal cohort. Archives of General Psychiatry 2003; 60: 709–717. 6. Gralinski-Baker JH, Hauser ST, Billings RL, Allen JP. Risks along the road to adulthood: Challenges faced by youth with serious mental disorders. In: DW Osgood, EM Foster, C Flanagan, GR Ruth (Eds). On your own without a net: The transition to adulthood for vulnerable populations. Chicago, IL: University of Chicago Press, 2005. 7. Baydar MN, Reid J, Webster-Stratto C. The role of mental health factors and program engagement in the effectiveness of a preventive parenting program for head start mothers. Child Development 2003; 74(5): 1433–1453. 8. Meyers RJ, Miller WR, Smith JE, et al. A randomized trial of two methods for engaging treatment-refusing drug users through concerned significant others. Journal of Consulting and Clinical Psychology 2002; 70: 1182–1185. 9. Barrett MS, Chua W, Crits-Christoph P, et al. Early withdrawal from mental health treatment: Implications for psychotherapy practice. Psychotherapy: Theory, Research, Practice, Training 2009; 46(2): 247–267. 10. Munson MR, Jaccard J, Smalling S, et al. Static, dynamic, integrated, and contextualized: A framework for understanding mental health service utilization among young adults. Social Science & Medicine 2012; 75(8): 1441–1449. 11. McMillen JC, Ramesh R. Pediatric to adult mental health service use of young people leaving the foster care system. Journal of Adolescent Health 2009; 44(1): 7–13. 12. Munson MR, Scott LD Jr, Kim H. Making the transition from care with care. Encyclopedia of Adolescence 2011. 13. Kazdin AE, Holland L, Crowley M, et al. Barriers to treatment participation scale: Evaluation and validation in the context of child outpatient treatment. Journal of Child Psychology & Psychiatry & Allied Disciplines 1997; 38(8): 1051–1062. 14. U.S. Department of Health and Human Services. Mental health: A report of the Surgeon General. Rockville, MD: Public Health Service, Office of the Surgeon General, 1999. 15. Olfson M, Mojtabai R, Sampson NA, et al. Dropout from outpatient mental health care in the United States. Psychiatric Services 2009; 60(7): 898–907. 16. Harrison ME, McKay MM, Bannon Jr WM. Inner-city child mental health service use: The real question is why youth and families do not use services. Community Mental Health Journal 2004; 40(2): 119–131. 17. Thompson S, Bender K, Lantry J, et al. Treatment engagement: Building therapeutic alliance in home-based treatment with adolescents and their families. Contemporary Family Therapy: An International Journal 2007; 29(1/2): 39–55. 18. McKay MM, McCadam K, Gonzales J. Addressing the barriers to mental health services for inner city children and their caretakers. Community Mental Health Journal 1996; 32: 353–361. 19. Porras JI, Robertson PJ. Organizational development: Theory, practice, and research. In: MD Dunnette, LM Hough (Eds). Handbook of industrial and organizational psychology. Palo Alto: Consulting Psychologists Press, 1992, pp. 719–822. 20. Glisson C, Landsverk J, Schoenwald S, et al. Assessing the Organizational Social Context (OSC) of mental health services: Implications for research and practice. Administration and Policy in Mental Health and Mental Health Services Research 2008; 35(1–2): 98–113. 21. Burns T, Stalker GM. The management of innovation. London: Tavistock Publications, 1961. 22. Aarons GA, Sawitzky AC. Organizational culture and climate and mental health provider attitudes toward evidence-based practice. Psychological Services 2006; 3(1): 61–72. 23. Rousseau DM. Assessing organizational culture: The case for multiple methods. In: B Schneider (Ed). Organizational climate and culture. San Francisco: Jossey-Bass, 1990, pp. 153–192. 24. Glisson C, James LR. The cross-level effects of culture and climate in human service teams. Journal of Organizational Behavior 2002; 23(6): 767–794. 25. Pfeffer J. Power in Organizations. Marshfield, MA: Pitman, 1981. 26. Pfeffer J. Developing resource dependence theory: how theory is affected by its environment. In: KG Smith, MA Hitt (Eds). Great Minds in Management: The Process of Theory Development, Oxford University Press, New York, 2005. 27. Tait L, Birchwood M, Trower P. A new scale (SES) to measure engagement with community mental health services. Journal of Mental Health 2002; 11(2): 191–198. 28. Tait L, Birchwood M, Trower P. Adapting to the challenge of psychosis: personal resilience and the use of sealing-over (avoidant) coping strategies. The British Journal of Psychiatry 2004; 185(5): 410–415. 29. Mulder CL, Koopmans G, Hengeveld M. Lack of motivation for treatment in emergency psychiatry patients. Social Psychiatry & Psychiatric Epidemiology 2005; 40(6): 484–488. 30. Davidson G, Campbell J. An examination of the use of coercion by assertive outreach and community mental health teams in Northern Ireland. British Journal of Social Work 2007; 37(3): 537–555. 31. Lecomte T, Spidel A, Leclerc C, et al. Predictors and profiles of treatment non-adherence and engagement in services problems in early psychosis. Schizophrenia Research 2008; 102(1–3): 295–302. 32. Fisher H, Theodore K, Power P, et al. Routine evaluation in first episode psychosis services: feasibility and results from the MiData project. Social Psychiatry & Psychiatric Epidemiology 2008; 43(12): 960–967. 33. Ruchlewska A, Mulder CL, Smulders R, et al. The effects of crisis plans for patients with psychotic and bipolar disorders: a randomised controlled trial. BMC Psychiatry 2009; 9: 41–48. 34. Staring ABP, Mulder CL, Duivenvoorden HJ, et al. Fewer symptoms vs. more side-effects in schizophrenia? Opposing pathways between antipsychotic medication compliance and quality of life. Schizophrenia Research 2009; 113(1): 27–33. 35. Bunger AC, Stiffman AR, Foster KA, et al. Child welfare workers’ connectivity to resources and youth’s receipt of services. Advances in Social Work 2009; 10(1): 19–38. 36. Stiffman AR, Hadley-Ives E, Dore P, et al. Youth’s access to mental health services: The role of providers’ training, resource connectivity, and assessment of need. Mental Health Services Research 2000; 2(3): 141–154. 37. Stiffman AR, Striley C, Horvath VE, et al. Organizational context and provider perception as determinants of mental health service use. Journal of Behavioral Health Services & Research 2001; 28(2): 1–17.

Organizational Culture on Mental Health Service Engagement

H. KIM et al.

38. Schiff JW. A measure of organizational culture in mental health clinics. Best Practice in Mental Health: An International Journal 2009; 5(2): 89–111. 39. Kerber J. Staff evaluations and perceptions of organizational culture: Implications for performance improvement and mental healthcare service evaluation. Unpublished doctoral dissertation, Iowa State University, 1997. 40. Goldstein H. Multilevel Statistical Models. 2nd ed. New York: John Wiley, 1995. 41. Hofmann DA, Griffin M, Gavin M. The application of hierarchical linear modeling to organizational research. In: KJ Klein, SWJ Kozlowski (Eds). Multilevel theory, research, and methods in organizations. San Francisco: Jossey-Bass, 2000, pp. 467–511. 42. Klein KJ, Kozlowski SWJ. From micro to meso: Critical steps in conceptualizing and conducting multilevel research. Organizational Research Methods 2000; 3(3): 211–236. 43. James LR, Demaree RG, Wolf G. rwg: An assessment of within-group interrater agreement. Journal of Applied Psychology 1993; 78(2): 306–309. 44. Raudenbush SW, Bryk AS.Hierarchical Linear Models: Applications and data analysis methods. 2nd ed. Newbury Park, CA: Sage, 2002. 45. Cunningham PB, Henggeler SW. Engaging multiproblem families in treatment: Lessons learned throughout the development of multisystemic therapy. Family Process 1999; 38(3): 265. 46. Staudt M. Treatment engagement with caregivers of at-risk children: Gaps in research and conceptualization. Journal of Child & Family Studies 2007; 16(2): 183–196. 47. Mitchell AJ, Selmes T. Why don’t patients attend their appointments? Maintaining engagement with psychiatric services. Advances in Psychiatric Treatment 2007; 12: 423–434. 48. Henggeler SW, Rowland MD, Randall J, et al. Home-based multisystemic therapy as an alternative to the hospitalization of youths in psychiatric crisis: Clinical outcomes. Journal of the American Academy of Child & Adolescent Psychiatry 1999; 38(11): 1331–1339. 49. O’Mahen HA, Flynn HA. Preferences and perceived barriers to treatment for depression during the perinatal period. Journal of Women’s Health 2008; 17(8): 1301–1309. 50. Reamer FG. Social Work Values and Ethics, 3rd ed., New York: Columbia University Press, 2006. 51. Lindell MK, Whitney DJ. Accounting for common method variance in cross sectional research designs. Journal of Applied Psychology 2001; 86: 114–121. 52. Podsakoff P, MacKenzie S, Lee J, et al. Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology 2003; 88: 879–903. 53. Glisson C, Green P. The effects of organizational culture and climate on the access to mental health care in child welfare and juvenile justice systems. Administration and Policy in Mental Health and Mental Health Services Research 2006; 33(4): 433–448. 54. Hemmelgarn AL, Glisson C, Dukes D. Emergency room culture and the emotional support component of family-centered care. Children's Health Care 2001; 30(2): 93–110. 55. Schein HE. Defining Organizational Culture. In: JM Shafritz, JS Ott (Eds). Classics of Organization Theory. Belmont, CA: Wadsworth, 1992, pp. 490-502. 56. Glisson C. Structure and technology in human service organizations. In: Y Hasenfeld (Ed). Human services as complex organizations. Beverly Hills, CA: Sage Publications, 1992.

The Journal of Behavioral Health Services & Research

2014