Introduction

A recovery-orientation is now the accepted and preferred method of community mental health service delivery in the United States (Farkas et al. 2005; Surgeon General’s Report 1999). Recovery principles emphasize the essential role of collaborative relationships between consumers and providers that support consumers’ goals and preferred futures (Green et al. 2008). Writings on recovery also describe the detrimental impact of directive or coercive relationships between mental health providers and adults living with mental illness (Davidson et al. 2005; Farkas et al. 2005). However, factors that shape the nature of professional relationships between consumers and providers working in a recovery model of service delivery are not well understood (Nath et al. 2012).

Previous research has demonstrated the importance of the therapeutic alliance between mental health professionals and their clients as a predictor of positive outcomes (McCabe and Priebe 2004; Priebe and Gruyters 1993), and is arguably among the most important factors in psychotherapy (Lambert and Barley 2002). A strong therapeutic working alliance is predicated on establishing a trusting relationship between the client and clinician where the clinician acknowledges the inherent power differential in the therapeutic context (Farkas et al. 2005). Scholars have noted, however, that a recovery-orientation may influence providers’ delivery of services within the therapeutic context. Davidson et al. (2005) warn of the dangers of increased social pressure on consumers to recover and the possibility that clinicians might encourage clients to take on new challenges prematurely in the name of “recovery.” Such social pressure on providers to insure the outcome of “recovery” for consumers may result in the use of directive practices that are more nuanced than overt forms of coercion. Provider directiveness can be conceptualized as the “degree to which practitioners try to influence clients to accept a solution or course of action preferred by the practitioner” (Healy 2008, p. 72). Qualitative findings suggest that providers intentionally engage in a number of directive practices with their clients (Healy 2008).

Provider directiveness is counter to the principles of mental health recovery that call for shared participation in decision making between consumer and provider (Davidson et al. 2005). Specifically, directiveness may occur when providers initially listen to consumers’ goals, but ultimately make decisions and direct consumers in ways that are inconsistent with consumers’ stated goals. Previous studies suggest that providers’ directiveness can result in increased client resistance to treatment (Beutler et al. 2001) and that less directiveness by therapists resulted in outcomes for reactant clients in treatment for alcoholism (Karno and Longabaugh 2005). For example, in a study of 51 consumers involved with mental health courts, Pratt et al. (2013) found that consumers’ perceptions of negative pressures (i.e., perceived injustices during court proceedings) from professionals statistically predicted their future involvement with the criminal justice system.

Workforce demands on providers in community mental health settings such as excessive amounts of paperwork and large client caseloads are well documented and contribute to high job turnover rates and professional burnout (Gitter 2006). However, some research suggests that work with consumers in a recovery-orientation can have positive professional and personal benefits for mental health providers themselves. In their community sample of 114 community mental health providers, Kraus and Stein (2013) found that providers’ perceptions of agency recovery-orientation were positively related to their job satisfaction after controlling for individual and setting characteristics such as age, level of education, and caseload size. Wilson and Crowe (2008) found that psychiatric nurses in their qualitative study described higher levels of job satisfaction when they felt they had strong working alliances with consumers. Stein and Craft (2007) found that mental health care providers generally reported a sense of personal growth as a result of their work with adults with serious mental illness. In a qualitative study of community psychiatrists, Carpenter-Song and Torrey (2015) found that cultivating strong relationships with consumers was identified by these clinicians as one essential element of their work.

If a recovery model of mental health care promotes collaborative relationships between providers and consumers, then both parties can potentially experience personal rewards and insights. However, previous research has not examined relationship factors such as therapeutic alliance and provider directiveness in the same research study, although there is evidence to suggest that both of these factors are important in relationships between providers and consumers for consumer outcomes (Carpenter-Song and Torrey 2015; Healy 2008; Karno and Longabaugh 2005; Pratt et al. 2013). Moreover, researchers have yet to identify specific factors in their relationships with consumers that may be associated with providers’ own feelings of job satisfaction and personal well-being. Studies that consider providers’ perceptions of their relationship with consumers within a larger context of community mental health workforce demands can have direct implications for the implementation and delivery of recovery-oriented community mental health services.

Present Study

The present study examined the role of mental health care providers’ perceptions of their professional relationships with consumers in understanding their reports of agency recovery-oriented services and their own sense of job satisfaction and personal growth. Specifically, the present study investigated the relative contribution providers’ views of working alliance and directiveness in their relationships with consumers in accounting for variation in providers’ perceptions of recovery-orientation, job satisfaction, and personal growth as a result of working with consumers.

Based on theoretical principles of mental health recovery and findings from previous research, it was hypothesized that (1) providers’ perceptions of working alliance, and not provider directiveness, would be positively related to providers’ reports of recovery-orientation, job satisfaction, and personal growth. After accounting for the role of individual provider characteristics, perceived workforce demands and recovery-orientation, it was expected that (2) providers’ perceptions of their relationships with consumers would be the strongest predictors of providers’ reports of job satisfaction, and personal growth.

Method

Participant Characteristics and Procedure

The present sample consisted of 105 mental health professionals (80 women, 22 men, 3 did not say) employed at seven community mental health centers in Virginia who worked with adults diagnosed with a serious mental illness as categorized by the Diagnostic and Statistical Manual of Mental Disorders—Fourth Edition, Text Edition (DSM-IV-TR; APA 2000). Demographic characteristics of the sample are presented in Table 1. Mental health professionals were employed as: Counseling/Psychological staff, n = 68, which included masters level counselors/therapists, social workers, rehabilitation and occupational therapists, and staff psychologists; Medical staff, n = 12, which included nursing staff and psychiatrists; and Direct Services staff, n = 25, which included case managers. As seen in Table 1, a majority of participants in the sample (52 %) were in their early 40’s, married, Caucasian, and held at least a Masters degree (66 %). Participants reported providing mental health services for an average of 16 years (SD = 9.8). The majority of the present sample reported working with adults with a serious mental illness (64 %), substance abuse disorders (7.6 %), and co-occurring disorders (7.6 %).

Table 1 Sample demographic characteristics
Full size table

The institutional review board (IRB) at a large, public, Midwestern university approved the study for data collection. Community Mental Health Centers (CMHCs) were recruited via an agency recruitment text sent to a total of seventeen CMHCs in the Commonwealth of Virginia. Mental health providers (i.e., case managers, masters-level counselors, psychiatric nurses, social workers, occupational and rehabilitation therapists, psychologists, psychiatrists) received an invitation by email to voluntarily decide to participate in an online survey. Providers were offered the chance to enter a random raffle to win one of two $75.00 gift cards as incentive for study participation.

Measures

Working Alliance Inventory (WAI)

The WAI (Horvath and Greenberg 1989) is among the most widely used self-report measure to assess the working alliance between therapist and consumer (Hatcher and Gillaspy 2006). The measure has a total of 36-items with three 12-item subscales that reflect perceptions of task, bond, and goal alliance. The measure uses a seven-point scale (1 = Never to 7 = Always) and has been shown to have acceptable internal consistency and construct validity (Hatcher and Gillaspy 2006). An overall mean score was used in the present study with higher scores reflecting greater perceived working alliance with consumers. The internal consistency for the WAI in the present sample was .93.

The Recovery Self-Assessment-Revised (RSA-R)

The RSA-R (O’Connell et al. 2005) is a 36-item self-report measure that assesses perceptions or practices that are considered consistent with a recovery service orientation from the perspective of mental health providers. The RSA-R has five factors including: life goals, involvement, diversity of treatment options, choice, and individually-tailored services and has been shown to have acceptable internal consistency (O’Connell et al. 2005). Providers respond to items using a five-point Likert scale (1 = strongly disagree to 5 = strongly agree). An overall mean score indicating higher perceived agency recovery-oriented services was used in the present study and overall internal consistency of the measure in the present sample was .93.

Job Satisfaction Survey (JSS)

The JSS (Spector 1985) provides an overall job satisfaction score based on nine facets of employment (e.g., pay, promotion, supervision, fringe benefits, contingent rewards, etc.). Respondents were asked to rate nine statements (e.g., “I feel I am being paid a fair amount for the work I do”) using a six-point scale (0 = disagree very much to 6 = agree very much) with higher scores indicating greater job satisfaction. The measure has shown acceptable test re-test reliability, construct validity, and internal consistency (Spector 1985; Kraus and Stein 2013). The internal consistency coefficient of the JSS in the present study was .64.

Case Manager Personal Growth Scale (CMPG)

The CMPG (Stein and Craft 2007) is a 16-item self-report measure that assesses case mangers’ sense of personal growth as a result of their work with consumers. Developed from the stress-related growth literature, CMPG items reflect aspects of personal, relationship, and spiritual growth as a result of working with consumers (Stein and Craft 2007). The measure uses a five-point Likert scale (1 = strongly disagree to 5 = strongly agree) and an overall mean score was used in the present study. The measure has been shown to have acceptable psychometric properties (Stein and Craft 2007). In the present study, the internal consistency coefficient of the CMPG scale was .91.

Marlowe–Crowne Social Desirability Scale—Short Form (MCSD)

The MCSD (Strahan and Gerbasi 1972) is a 10-item self-report measure that assesses how likely a person is to respond in a socially acceptable manner or in a manner that would be viewed favorably by others. Responses are scored as either true or false (1 = true and 0 = false). The measure has a high internal consistency and has been found to highly correlate with standard scale of social desirability (Fisher and Fick 1993). The alpha coefficient for the measure in the present study was .67.

Provider Directiveness Scale

The Provider Directiveness Scale (see Table 2) is a 9-item self-report measure developed for the present study to examine providers’ endorsement of engaging in specific behavioral actions that direct consumers in ways that may be inconsistent with consumers’ views and preferences. Participants are asked to respond to statements about their level of directedness with consumers using a 5-point Likert scale (1 = Strongly disagree to 5 = Strongly agree). In consultation with several mental health professionals, a pool of items was developed to reflect provider interactions with consumers that included directiveness regarding: client’s social relationships and finances; treatment; medication and housing. Items were then rated by two independent judges who selected the best representations of provider directiveness (kappa coefficient r = .84, p < .001 for agreement between judges), which resulted in the final 9-item scale. (See Table 2 for scale items). In terms of construct validity, the total scores on the Provider Directiveness Scale were negatively correlated with total scores on the WAI (Horvath and Greenberg 1989), a measure of therapeutic working alliance (r = −.30, p < .01). Scores on the Provider Directiveness Scale were not significantly correlated with scores on a measure of social desirability (MCSD). In the present study an overall mean score was used, with higher scores indicating higher levels of provider directiveness. The internal consistency of the Provider Directiveness Scale in the present study was .67.

Table 2 Provider directiveness scale items
Full size table

Demographic Questionnaire

Demographic data such as age, gender, ethnicity, education level, and marital status were collected on all participants. Detailed employment history information and perceived job demand characteristics that included primary service population, current caseload size, length of job tenure, hours spent per week in direct contact with consumers, and income were also obtained for the present sample.

Results

Preliminary Analysis

Descriptive statistics for measures used in the present study are found in Table 3. Two separate multivariate analysis of variance (MANOVA) analyses were conducted to test for differences on study dependent variables (agency recovery-orientation, job satisfaction, and personal growth) as a function of providers’ professional roles (counseling staff, medical staff, direct services staff) and geographical location of agency of employment (Metropolitan Statistical Area; Non-Metropolitan Statistical Area). Results indicated no significant differences in providers scores on dependent variables as a function of professional role, F(6,105) = .60; p = ns; Wilk’s λ = .97, \(\eta_{p}^{2}\) = .02, or location of agency, F(3,105) = 2.40; p = ns; Wilk’s λ = .93, \(\eta_{p}^{2}\) = .07. Provider role and location of agency were not included in subsequent analyses.

Table 3 Descriptive statistics for present study measures
Full size table

Individual Provider Characteristics, Reports of Agency Recovery-Orientation and Provider Well-Being

A series of one-way analysis of variance (ANOVA) analyses were conducted to examine whether scores on the study measures of recovery-orientation, job satisfaction, personal growth, working alliance, and mental health provider directiveness differed as a function of individual characteristics (i.e., gender, age) and perceived workforce demands (i.e., current caseload size, work with consumers with SMI). Significant differences in overall scores on study measures were found as a function of gender. Women (M = 5.6; SD = .48) were more likely than men (M = 5.3; SD = .60) to report higher perceptions of personal growth, F(1,101) = 4.49; p < .05, and more likely than men to report higher scores on working alliance, F(1, 101) = 4.95; p < .05. Significant differences in overall scores on study measures were found as a function of age such that participants between the ages of 55 and 64 were more likely than those participants between the ages of 25 and 34 to report higher scores of agency recovery-orientation, F(4,101) = 2.75; p < .05. Participants who reported working with consumers who had a serious mental illness (M = 3.1; SD = .67) as opposed to other types of mental health difficulties (M = 4.3; SD = .60) reported lower scores on agency recovery-orientation, F(1, 104) = 7.41; p < .01.

Caseload size and work with consumers with SMI (yes/no) were included in subsequent statistical analysis to understand the extent to which providers’ reports of these workforce demands were related to the study dependent variables. Length of job tenure was significantly correlated with age of participants (r = .80, p < .001). Participants’ scores on social desirability were not significantly correlated with their scores on perceived recovery-orientation, job satisfaction, and personal growth, suggesting that providers were not responding to study measures based on social desirability.

Providers’ Views of the Client Relationship, Recovery-Orientation and Job Satisfaction

A series of three hierarchical multiple regression analyses were conducted to examine the relative contribution of participants’ individual characteristics, perceived workforce demands, and self-reported provider–consumer relationship variables in accounting for variation in scores of perceived recovery-orientation, job satisfaction, and personal growth. In the first analysis, perceived agency recovery-orientation was used as the criterion variable, and individual characteristics (providers’ age and gender) were entered into a first block, perceived workforce demands (case load size, work with consumers with SMI) were entered into a second block and providers’ views of the provider–consumer relationship variables (working alliance and provider directiveness) were entered as a third block in the regression equation.

As seen in Table 4, when using perceived agency recovery-orientation as the criterion variable, the overall regression model was significant, F(6, 97) = 4.73, p < .001, and explained 24 % of the total variance in perceived agency recovery-orientation. Age and working with consumers with SMI were significant in the regression model and together accounted for 14 % of the variance in agency recovery-orientation scores. When added to the regression model, working alliance and provider directiveness accounted for an additional 10 % of the variance in agency recovery-orientation scores above that of individual characteristics and perceived workforce demands. Findings suggest that regardless of their age and their work with consumers with SMI, providers who report higher levels of working alliance and higher levels of provider directiveness with consumers also reported higher levels of agency recovery-orientation.

Table 4 Hierarchical regression analysis
Full size table

When using scores on job satisfaction and personal growth as criterion variables in two separate hierarchical regression analyses, individual characteristics (age and gender) were entered into a first block, and perceived job demands (caseload size, work with consumers with SMI) were entered into a second block in the regression equation. Scores on agency recovery-orientation were entered into the third block of the equation to control for variance in well-being scores attributable to providers’ perceptions of working in a recovery paradigm. Providers’ perceptions of their relationship with consumers (perceived working alliance and provider directiveness) were entered as a fourth block in the regression equation.

When using self-reported job satisfaction as the criterion measure, the overall regression model was significant, F(7, 97) = 2.40, p < .05, and explained 16 % of the total variance. Individual provider characteristics and perceived workforce demands were not significantly related to providers’ reports of job satisfaction in the regression model. Providers’ scores on agency recovery-orientation were significant in the regression model and accounted for 8 % of the variance in job satisfaction scores. Working alliance accounted for an additional 6 % of the variance in providers’ reports of job satisfaction. Findings suggest that those providers with higher perceptions of agency recovery-orientation and a perceived higher working alliance also reported higher levels of job satisfaction.

When using perceived personal growth due to work with consumers as the criterion measure, the overall regression model was significant, F(7, 96) = 6.50, p < .00, and explained 34 % of the total variance. Results suggest that providers’ age was significantly positively related to personal growth scores, but scores on workforce demands were not significant in the regression model. Providers’ scores on agency recovery-orientation were significant and accounted for 6.8 % of the variance in personal growth scores. Working alliance and provider directiveness scores accounted for an additional 17 % of the variance in providers’ reports of personal growth. Results suggest that regardless of providers’ age and perceived recovery-orientation, providers who reported higher levels of working alliance and greater use of directive practices also reported higher levels of personal growth in their work with consumers.

Discussion

Using a sample of 105 community mental health care providers, present study results highlight the importance of providers’ perceptions of their relationships with consumers in understanding their views of agency recovery-orientation and sense of well-being. No significant differences on scores of recovery, job satisfaction, or personal growth were found as a function of providers’ professional role or agency of employment. Overall, participants who reported working directly with consumers with serious mental illness reported greater perceptions of agency recovery-orientation than those providers who did not. Regardless of individual provider characteristics and perceived workforce demands, providers’ reports of higher levels of recovery-orientation were generally associated with their perceptions of higher levels of working alliance and higher levels of provider directiveness with consumers. Higher levels of working alliance with consumers were significantly related to providers’ reports of job satisfaction, while greater levels of working alliance and provider directiveness significantly predicted providers’ sense of personal growth as a result of their work with consumers.

Theory and research on mental health recovery emphasize the importance of a strong, collaborative alliance between providers and consumers (Green et al. 2008). Provider directiveness is thought to be inconsistent with recovery-oriented service goals that support consumer choice, autonomy, and self-determination (Davidson et al. 2005). Previous studies suggest that consumers who reported a stronger alliance with providers also reported better employment and housing outcomes (Priebe and Gruyters 1993). Prior research has also found that higher levels of provider directiveness were related to poorer outcomes for clients (Karno and Longabaugh 2005). The present study contributes to existing literature by simultaneously assessing providers’ reports of working alliance and directiveness with consumers within a single study. Contrary to our initial hypotheses based on previous studies, mental health care providers’ reports of higher levels of both working alliance and provider directiveness were related to their perceptions of higher levels of recovery-orientation and personal growth. Unlike previous studies of provider directiveness which relied on ratings of clinicians’ directiveness made by independent raters observing client sessions (Karno and Longabaugh 2005), the present research developed a self-report measure of provider directiveness based on specific behavioral actions that direct consumers in ways that may be inconsistent with consumers’ views and preferences.

There are a number of possible arguments in favor of using mild forms of directiveness in healthcare settings when it protects central values and interests of the patient (Sjöstrand and Helgesson 2008). Research suggests that clinicians may limit consumers’ autonomy under the auspices of ethical decision making and beneficence and not overt coercion (Sjöstrand and Helgesson 2008). In the present study, provider directiveness involves decisions and actions of providers that may serve to systematically reduce consumers’ self-determination.

Directiveness, as articulated in the present study is a more nuanced phenomenon than overt forms of coercion such as involuntary commitment, which is typically justified in the name of client safety and treatment adherence (Sjöstrand and Helgesson 2008).

In understanding present findings, it may be that providers see working alliance and directiveness within a larger set of helping behaviors available to them in their work with consumers. Providers may be aware that using their working alliance to direct consumers’ treatment in ways inconsistent with consumers’ wishes may limit consumers’ choice and autonomy (Sjöstrand and Helgesson 2008). In fact, the self-report measure of provider directiveness used in the present study specifically assessed providers’ reports of directiveness that were inconsistent with consumers’ wishes or goals. However, providers may consider directiveness within a strong working relationship with consumers as part of being actively engaged with consumers in their role as competent mental health professionals. For example, in a qualitative study of ten British psychiatric nurses, Lützén (1998) found that nurses reported using their values, knowledge, and experience to know when to limit patients’ autonomy, specifically in the context of “subtle coercion” (p. 103). This possibility is not altogether surprising, given that patients traditionally assume that providers are the best judges of which treatments would be effective, given their training and experience (Wheat 2009). Thus, providers may perceive that they are competent and actively engaged with consumers when providers direct treatment in ways that they believe are beneficial, regardless of consumers’ preferences or goals.

The present study replicates and extends findings from previous research on providers’ views of recovery and individual well-being (Kraus and Stein 2013; Wilson and Crowe 2008). For example, Kraus and Stein (2013) found that case managers who reported higher levels of recovery-orientation where they worked also reported lower levels of professional burnout and higher levels of job satisfaction. In the present study, higher levels of recovery-orientation were generally related to higher levels of job satisfaction in a sample of providers who occupied a variety of professional roles. Moreover, present findings suggest that providers’ reports of a stronger working alliance with consumers further contributed to providers’ reports of greater job satisfaction, after considering the contribution of perceived recovery-orientation. If replicated, present findings suggest that positive working alliance with consumers within a setting viewed as embodying recovery principles may be very important to professional well-being for mental health care providers.

Present study results add to existing literature that examines the perceived impact of professional relationships with consumers on the personal lives of mental health providers. For example, Linley and Joseph (2007) found that clinicians’ sense of well-being was related to their reports of a strong working alliance with their clients. Stein and Craft (2007) found that case managers reported that they had made positive changes in their personal lives as a result of working with consumers with serious mental illness. In the present study, participants’ reports of stronger working alliance and greater directiveness with consumers were significantly related to their reports of higher levels of personal growth. In fact, the combination of working alliance and provider directiveness accounted for an additional 17 % of variance in providers’ reports of personal growth after controlling for individual characteristics, perceived job demands, and perceived agency recovery-orientation.

Although speculative, present results suggests that providers’ sense of active engagement with consumers is related to their own sense of personal growth. It is possible that providers feel a sense of competence in their role as helpers when they see themselves as creating a strong working alliance and using more directive interventions in their work with adults coping with mental illness. Providers’ sense of active engagement and competence in the helping role may be factors related to their own sense of personal growth as a result of working with consumers.

Study Limitations and Future Directions for Research and Practice

Present findings are limited in a number of respects. The present study used a relatively small, non-random sample of community mental health care providers in Virginia. Participants in the research were predominantly middle-aged, Caucasian women who generally reported job tenure of over 15 years. Providers in the sample occupied a number of professional roles in community mental health care, but the overall generalizability of study findings is not clear. It may be that the present sample reflects mental health professionals who are more established in their jobs and are particularly committed to working in community mental health. It is unclear how representative the present sample is to community mental health professionals found in Virginia or nationally. The cross-sectional nature of the study also did not allow for examination of the causal direction of associations between the independent and dependent variables. Present findings need to be replicated in cross-sectional and longitudinal research using larger and more diverse samples of mental health professionals. Similar to previous research (Gehrs and Goering 1994), the present study asked providers to think about their relationships with consumers in general, rather than focus on interactions with specific consumers. It is likely that providers’ vary in their levels of working alliance and directiveness based on the individual characteristics and circumstances of specific consumers. The present study offers no information about the role of consumers’ characteristics in shaping providers’ levels of working alliance or directiveness.

Present findings are provocative and provide interesting directions for future research. Results suggest that providers may not view directive practices that are inconsistent with consumers’ wishes as contradictory to a recovery-orientation when provider directiveness is accompanied with a strong working alliance with consumers. Additional research is needed to replicate and extend findings by examining providers’ rationales for directive practices in their relationship with consumers. Providers’ definitions of professional competence and active engagement with consumers within a recovery-oriented system may be important topics for future research on mental health recovery. Moreover, present findings suggest that future studies that include both providers and consumers in examining provider–consumer relationships may be critical for advancement of recovery-oriented services (Arnow et al. 2013). Researchers currently have little understanding of the level of congruence of perception between consumers and providers about basic features of their professional relationships or the possible role of providers’ well-being in facilitating recovery outcomes for consumers.

Findings from the present research also have several important implications for clinical practice. Recently, scholars have called for greater education and training of helping professionals to enable them to develop recovery-oriented skills in working with adults with mental illness (Anthony and Ashcroft 2010). For example, Anthony and Ashcroft (2010) describe the need for providers to develop “higher order helping skills” that are not geared to care-taking, stabilizing, or controlling consumers’ behaviors. If replicated, present findings support the need for greater dialogue among practitioners about the definition and use of directive practices within the context of recovery. Education, training, and interventions for providers that incorporate feedback from consumers about providers’ levels of directiveness may be useful in refining effective recovery-oriented skills for mental health practitioners.

Study results also suggest providers’ sense of personal well-being as a helping professional may be an additional factor related to their use of directiveness with consumers. Simply stated, providers generally feel good personally when they can tell consumers what they should be doing within the context of strong working relationships with consumers. Some level of provider directiveness may be shaped by symptoms expressed by consumers and providers’ concerns for consumers’ safety (Montesano et al. 2014). However, it is also possible that positive feelings about being a helper and ideas about professional competence and active engagement with consumers contribute to providers’ feelings of personal well-being. Provider education and training in recovery-oriented practices may need to be accompanied by a larger shift in professional culture regarding what it means to be a “professional helper.” The present study underscores the critical importance of ongoing research and continued dialogue about provider–consumer relationships associated with well-being for adults coping with mental illness and for mental health care providers themselves.