What Would It Take? Stakeholders’ Views and Preferences for Implementing a Health Care Manager Program in Community Mental Health Clinics Under Health Care Reform

Abstract

Health care manager interventions can improve the physical health of people with serious mental illness (SMI). In this study, we used concepts from the theory of diffusion of innovations, the consolidated framework for implementation research and a taxonomy of implementation strategies to examine stakeholders’ recommendations for implementing a health care manager intervention in public mental health clinics serving Hispanics with SMI. A purposive sample of 20 stakeholders was recruited from mental health agencies, primary care clinics, and consumer advocacy organizations. We presented participants a vignette describing a health care manager intervention and used semistructured qualitative interviews to examine their views and recommendations for implementing this program. Interviews were recorded, professionally transcribed, and content analyzed. We found that a blend of implementation strategies that demonstrates local relative advantage, addresses cost concerns, and enhances compatibility to organizations and the client population is critical for moving health care manager interventions into practice.

Introduction

Chronic health conditions disproportionally impact people with serious mental illness (SMI; e.g., schizophrenia) compared with the general population, resulting in excess morbidity and premature mortality (Druss, Zhao, Von Esenwein, Morrato, & Marcus, 2011). Racial/ethnic minority status may contribute additional health risk. For instance, Hispanics with SMI have elevated rates of obesity, metabolic syndrome, and type 2 diabetes compared to non-Hispanic Whites with SMI (Hellerstein et al., 2007; Henderson et al., 2005; Kato, Currier, Gomez, Hall, & Gonzalez-Blanco, 2004). Lack of access and underutilization of primary care services and poor care coordination and quality of medical care are factors that exacerbate the health disparities faced by people with SMI (Druss, 2007).

Health care manager interventions can improve the access, coordination, and quality of medical care for people with SMI (Bartels et al., 2004; Druss et al., 2010; Kilbourne et al., 2008). For example, the Primary Care Access Referral and Evaluation program (PCARE) is an intervention that uses health care managers (e.g., registered nurses [RNs]) in community mental health clinics to increase patient activation around physical health issues and improve care coordination between mental health and primary care providers (Druss et al., 2010). In a randomized controlled trial using RNs to deliver PCARE at a mental health clinic serving predominantly African Americans, PCARE more than doubled the rate of receipt of preventive medical care and improved the quality of cardiometabolic care and mental-health–related quality of life among adults with SMI compared to usual care (Druss et al., 2010).

New Contribution

Despite these results, PCARE is not widely implemented in the community. The integration of medical and behavioral health services is a central element of medical homes and is at the forefront of health care reforms in the United States (Alakeson, Frank, & Katz, 2010), yet little is known about what could facilitate the translation of health care interventions, like PCARE, for people with SMI into community mental health settings. In this study, we address this implementation gap by examining stakeholders’ (e.g., mental health providers, consumer advocates) views and preferences for implementing PCARE in public outpatient mental health clinics serving predominantly Hispanics with SMI and using social workers instead of RNs to deliver this intervention.

Conceptual Framework

Multiple factors account for this implementation gap (e.g., fragmented billing systems, lack of staffing, staff resistance to change). Implementation theories, such as diffusion of innovation and the consolidated framework for implementation research (CFIR), indicate that the characteristics of the intervention (e.g., cost, compatibility), particularly stakeholders’ subjective evaluations of these characteristics, influence the implementation process (Damschroder et al., 2009; Rogers, 1995). In other words, “perceptions count” in implementation as they shape stakeholders’ evaluations and reactions toward an innovation, which in turn shapes how the innovation is adopted in a new setting (Rogers, 1995, p. 209).

The diffusion of innovations theory and CFIR specify that several intervention characteristics play a role in the implementation process (see Table 1). Examining perceptions of relative advantage reveals the value and importance that stakeholders place on different sources of evidence (e.g., published studies, guidelines, personal experiences) to shape not only their judgments about the practice innovation but also what they consider advantageous about the innovation. Inquiring about perceptions of cost can produce important insights about the financial approaches needed to sustain the innovation. Compatibility captures stakeholders’ views about the fit between the innovation, the setting, and the knowledge system (e.g., values, norms) of potential users of the innovation. In this study, we examined two aspects of stakeholders’ views about the compatibility of the intervention: to the organization and to its use with Hispanic patients, our population of interest. Complexity encapsulates stakeholders’ views about the scope, disruptiveness, burden, and the number of steps, procedures, and changes required to implement the innovation (Damschroder et al., 2009). Trialability provides potential users an opportunity to gain experience with the innovation, enables users to learn how it works, and serves as a way to reduce uncertainty about the innovation (Rogers, 1995). Lastly, social influence captures how stakeholders’ views of a practice innovation are shaped by the opinions of other individuals or organizations in the stakeholders’ social and professional networks.

Table 1.

Descriptions of Intervention Characteristics and Implementation Strategies.

Concepts	Description
Intervention  Characteristicsa
Relative advantage	The degree to which stakeholders perceive the practice innovation  as being better than the status quo.
Cost	The economic and human resources viewed as necessary for  adopting the innovation.
Compatibility	The degree to which an innovation is viewed as consistent with  the values, past experiences, and needs of potential  implementers of the innovation.
Complexity	The perceived difficulties and challenges for implementing the  innovation.
Trialability	The ability to pilot and/or experiment with the innovation on a  small basis before deciding to adopt the innovation on a larger  scale.
Social influence	The degree to which other individuals (e.g., opinion leaders) or  organizations in the stakeholders’ social network endorse or  support the implementation of the innovation.
Implementation  Strategiesb
Financing	Focuses on incentivizing the use of the clinical innovations and  providing resources for training, implementing, and sustaining the  innovation.
Restructuring	Focuses on shifting or altering staffing, professional roles and  responsibilities, physical structures, equipment, and data systems  to facilitate the implementation of a new practice innovation.
Education	Includes using multiple approaches (e.g., workshops, webinars)  to inform stakeholders about the practice innovation and  implementation efforts.
Planning	Consists of enabling stakeholders to gather data to identify and  assess potential needs and quality gaps, select strategies, build  buy-in, and develop relationships and leadership to support the  implementation process.
Quality   management	Focuses on using data systems and support networks to  continually evaluate and enhance the quality of care and the  fidelity of the practice innovation being implemented.
Attending to the   policy context	Focuses on using policy leverages—such as accrediting  bodies, licensing boards, and legal systems—to facilitate  implementation.

^aDescriptions of intervention characteristics are derived from the diffusion of innovation theory (Rogers, 1995) and the consolidated framework of implementation research (Damschroder et al., 2009).

^bDescriptions of implementation strategies are derived from the taxonomy developed by Powell and colleagues (2012).

Stakeholders’ evaluations of these intervention characteristics convey their values, preferences, and concerns about a new program and can be used to select and tailor strategies needed to implement the innovation into routine practice (Wensing, Bosch, & Grol, 2010). Implementation strategies are “systematic intervention process[es] to adopt and integrate evidence-based health innovations into usual care” (Powell et al., 2012, p. 124). In this study, we used a taxonomy developed by Powell et al. (2012) of six commonly used implementation strategies (see Table 1) in the health and mental health fields to categorize and organize the implementation themes that emerged from stakeholders’ discussions of PCARE and their suggestions for how to implement this program in routine practice.

In all, we used these six intervention characteristics and six implementation strategies as guiding constructs to: (a) describe stakeholders’ views of PCARE, (b) identify which intervention characteristics they value most, and (c) explore the link between these perceptions and known implementation strategies. Our goal is to formulate new insights into potential implementation strategies that could be used to facilitate the implementation of health care manager programs such as PCARE into public outpatient mental health clinics serving Hispanics with SMI.

Methods

Overview

This study is part of a larger multiphase project that is modifying PCARE to a new patient population (Hispanics) and provider group (social workers) and assessing the feasibility and acceptability of this adapted version of PCARE at a public outpatient mental health clinic in New York City (Cabassa, Druss, Wang & Lewis-Fernández, 2011). To inform program modifications and implementation, twenty semistructured qualitative interviews with stakeholders were conducted between September 2011 and May 2013 before the modified program was implemented at the study site. We began implementing the modified program at the study site in October 2013. At the time of this submission, we are collecting data to examine the feasibility and acceptability of this intervention. In this article, we present only the findings that emerged from these interviews regarding how to best implement this type of health care manager intervention in public mental health clinics. Program modifications that emerged from these interviews and other methodologies (e.g., patient focus groups, community advisory board input) are presented in another article.(Cabassa et al., 2014) Study procedures were approved by the institutional review boards at the New York State Psychiatric Institute and Columbia University.

PCARE Program

The PCARE Program is a 12-month intervention in which health care managers at a mental health clinic work individually with patients to facilitate and coordinate their primary care services and address the patient, provider, and system-level obstacles that prevent people with SMI from using primary care services (Druss et al., 2010). The core elements of PCARE are care coordination and patient activation. The health care manager serves as a bridge, coordinator, and advocate between patients and their primary care and mental health providers, and partners with all of them to ensure patients’ preventive primary care needs (e.g., immunizations, screenings) are properly identified, monitored, and managed. The health care manager addresses the fragmentation of care between the mental health and primary care sectors by making sure vital patient medical information (e.g., patients’ medications, health conditions) is shared across systems of care and providers. To enhance patient activation, the health care manager serves as an advocate and coach helping patients develop the knowledge and skills to actively engage and participate in their own health care and enhance their personal abilities to self-manage their health issues (Hibbard & Tusler, 2007). The health care manager uses motivational interviewing techniques and action plans to enhance health behavior change and patient activation. The ultimate goals of this health care manager intervention are to increase the receipt of preventive primary and cardiovascular care, patient activation, and health-related quality of life, and ultimately reduce clients’ risk of cardiovascular disease.

Sample

We used a purposive sampling approach to recruit individuals from four stakeholder groups: mental health providers, primary care providers, administrators, and consumer advocates. These groups were chosen because they are the potential implementers of health care managers’ programs in our community. Eligible participants were employed at least part-time at their agencies and currently involved in services aimed at improving the physical health of people with SMI. We used a combination of approaches to identify participants, including staff meeting presentations at mental health and primary care clinics, referrals from the Director of Research at the New York State Office of Mental Health, and nominations of colleagues from participants.

Mental health providers (e.g., psychiatrists, social workers) were recruited from our study site, a public outpatient mental health clinic located in Northern Manhattan in New York City that serves predominantly Spanish-speaking Hispanics of Dominican and Puerto Rican descent. Primary care providers (e.g., physicians, nurses) were recruited from community primary care clinics that provide medical services to patients from our study site. Administrators included individuals from multiple disciplines (e.g., psychiatry, social work) who were in leadership positions at mental health agencies that were affiliated to or received contracts from the New York State Office of Mental Health. Consumer advocates were individuals who self-identified as former consumers of mental health services and were employed either at mental health agencies or consumer advocacy organizations.

Qualitative Interviews

Semistructured qualitative interviews were conducted by the first author either in person at the participants’ work sites (n = 11) or via telephone (n = 9). We allowed telephone interviews in order to accommodate participants’ preferences and busy schedules. Interviews were audiotaped, professionally transcribed, and lasted approximately 60 min.

To examine stakeholders’ views of PCARE, we presented each participant a vignette describing PCARE (available upon request). The vignette was read by the interviewer, and participants were given a copy of the vignette to read along and refer to during the interview. We then asked participants a series of open-ended questions and probes to examine their opinions of PCARE and their views of how to implement this program in their communities. Our interview guide (available upon request) was informed by constructs examining different intervention characteristics (e.g., relative advantage, compatibility) derived from diffusion of innovation theory (Rogers, 1995) and CFIR (Damschroder et al., 2009). Example of questions included the following: What advantages, if any, do you see PCARE having over existing services? (relative advantage); what similarities, if any, do you see between PCARE and the existing services offered at public outpatient mental health clinics? (compatibility); and how would you pay or bill for the services offered in PCARE? (cost). We also asked participants a series of questions about how they would implement a program like PCARE (e.g., Based on your experiences, what would facilitate the implementation of PCARE? What resources would you need? What procedures would you use to implement PCARE?).

To further examine participants’ views of PCARE, we presented participants with a list of 11 statements printed on separate index cards that represented the following intervention characteristics: relative advantage, cost, compatibility to the organization, compatibility to Hispanics, social influence, complexity, and trialability (statements are available upon request). Participants were asked to read each statement, pick the three that they deemed most important, and rank-order their choices from most important to least important in helping them decide to implement PCARE. We then asked participants to explain their choices and ranking. At the end of the interview, participants also completed a short survey to collect sociodemographic and work experience data. The interviewer also developed summaries after each interview describing participants’ responses and the insights generated from the interview. These summaries were reviewed and discussed during weekly team meetings to examine emerging themes and patterns in our data and determine whether we were achieving data saturation for each of our stakeholder groups. After 20 interviews, the team concluded that data saturation had been achieved for each stakeholder group. Since we were not learning anything new from these interviews.

Data Analysis

Frequencies and measures of dispersion (e.g., means, standard deviations) were used to describe sample characteristics and stakeholders’ ranking of intervention characteristics using SPSS version 21. We stratified stakeholders’ rankings of intervention characteristics by stakeholder type (e.g., mental health provider, administrator) to examine and report similarities and differences in these rankings by stakeholder groups.

We used a directed content analysis approach to inform our qualitative analysis (Hsieh & Shannon, 2005). This is a deductive analytical strategy that employs existing theories and constructs to inform and structure the coding process and allows for the validation, expansion, and refutation of frameworks and theories (Yang et al., 2014). We started with key concepts derived from the intervention characteristics (e.g., relative advantage, compatibility) specified in the diffusion of innovation theory and CFIR to develop and operationalize an initial coding scheme. The first two authors independently read each transcript and interview summaries, noting segments of text that corresponded to our initial codes, and drafted analytical memos describing the application of codes. We then met on a weekly basis for several months to present and discuss our application of codes, development of emerging codes not included in our initial coding scheme, identification of patterns and themes, and general interpretations of our data. Our discussions during these meetings and our meeting notes were then used to develop a final code book. Example of codes included compatibility of PCARE, relative advantage of PCARE, cost, and implementation strategies, among others.

Interview transcripts and interviewer summaries were entered into Altas.ti (Muhr, 2004), an analytical software used to organize and manage qualitative data coding and analysis. One of the initial readers (APG) then coded all qualitative data in Atlas.ti under the supervision of the principal investigator of the project (LJC). We generated queries and reports in Atlas.ti to further examine the text captured in the codes related to intervention characteristics and implementation strategies. We then conducted a second-level coding process in which we applied the taxonomy of implementation strategies derived from Powell et al. (2012) to the text captured in the implementation strategies code. This enabled us to examine and categorize how the implementation ideas discussed by stakeholders did or did not correspond with known implementation strategies. Finally, through review and discussion of our coding reports and queries, we identified discernable patterns in our codes and wrote analytical memos describing the emergence of themes from the data noting similarities and differences between stakeholder groups. We used the following strategies to ensure the trustworthiness of our analysis and guard against common biases inherent in directive content analysis (e.g., overemphasis on theory): generation of an audit trail documenting our analytical decisions, peer debriefing meetings, member-checking presentations to the project’s community advisory board, analysis of cases that did not fit our coding scheme, and generating new codes that emerged from our data that were not present in our guiding theories and frameworks (Padgett, 1998).

Results

Sample Characteristics

Our sample was composed of 20 stakeholders: five mental health providers, five primary care physicians, five administrators, and five consumer advocates. Sample characteristics are presented in Table 1. Fifty-five percent of our sample was female and half was either African American or Hispanic. Stakeholders came from different professions and had worked, on average, 10 years at their organization at the time of these interviews (see Table 2).

Table 2.

Sample Characteristics (N = 20).

Sample Characteristics	Mean (SD)	N	%
Age	47.9 (10.6)
Female		11	55
Ethnicity/race
Non-Hispanic White		9	45
Hispanic		7	35
African American		3	15
Other		1	5
Profession
Psychiatrist		5	25
Primary care physician		5	25
Social worker		3	15
Peer specialist/consumer advocate		5	25
Other (e.g., nurse, administrator)		2	10
Average years working at organization	10.1 (6.5)

Stakeholders’ Views of PCARE

All stakeholders had a positive view of PCARE, describing it as a useful, helpful, and necessary program that made sense to them because it addressed the physical health needs and gaps in medical services faced by people with SMI. The key program characteristics that they valued included assistance with care coordination, giving patients individual attention to address physical health issues, and having a health care manager who could regularly assess and monitor patients’ physical health problems and could serve as a communication bridge between primary care and mental health providers. Stakeholders also liked the holistic and patient-centered approach of PCARE, as expressed by the following primary care provider:

I like that it’s a holistic approach to the patient, and I like that it’s patient-centered. I like that there is monitoring throughout the process, but it doesn’t seem like anything’s being started without appropriate follow-up and monitoring, whether it’s lab conditions, or lab results, or a patient condition. I like that it’s tackling issues that I think are major issues in our population here, being physical activity, smoking, diet. I like that there’s a person that they can build a trusting relationship with; that there’s someone that they can identify with, even if the physician changes, or there’s a lot of change in their surroundings for whatever reason; maybe housing or different issues like that. They know that there’s a manager; they have someone that they can relate to, that knows their story.

Implementing a program like PCARE was also described as a needed program that could lessen some of the burdens and responsibilities that mental health providers face when treating patients with multiple physical and mental health conditions, as captured by the comments from the following program administrator:

I think it [having a health care manager] would be a relief … that it’s one less thing that we have to monitor that we can focus back again on mental health and know that someone is going to be making sure that they have that primary care appointment, that they follow up with it, that they’re taking the medications prescribed, that they understand what it is. Yeah. It’ll be a relief.

Despite these positive views, stakeholders had several concerns, mostly about how to integrate PCARE into routine care. A central concern, mostly discussed by mental health providers and administrators, had to do with the training, clinical experience, and supervision that a person would need to have to be an effective health care manager, particularly if providers without formal medical training (e.g., social workers) are assigned to this role. Another concern, mostly reported by mental health providers, had to do with the added burden and responsibilities this program could create for existing staff if this health care manager role was added to their normal duties. Mental health providers who voiced this concern worried that this could overwhelm existing staff and create resistance toward adopting PCARE if no new resources were allocated to support the new role. These concerns start to point toward implementation strategies discussed in the sections below.

Stakeholders’ Rankings and Discussion of Intervention Characteristics

Table 3 summarizes stakeholders’ rankings of intervention characteristics for the total sample and by stakeholder type. Relative advantage was the most frequently endorsed characteristic, especially by mental health and primary care providers and consumers advocates. As stated by a primary care provider, “evidence that it helps is the most important” and was considered an essential characteristic for supporting PCARE. Stakeholders valued a range of outcomes supporting the effectiveness of PCARE, including improvements in connecting patients to primary care, reducing risks for cardiovascular disease, helping patients make healthy lifestyle changes, and improving patients’ quality of life. Some administrators and mental health providers prioritized local evidence over evidence from published reports, as it demonstrated that the program addressed their patients’ needs and health issues and showed local positive impact, as exemplified by the comments from the following program administrator:

Table 3.

Stakeholders’ Ranking of Intervention Characteristics.

	Mental Health Provider (N = 5)		Primary Care Provider (N = 5)		Consumer Advocate (N = 5)		Administrator (N = 5)		Total (N = 20)
Intervention Characteristics	N	%	N	%	N	%	N	%	N	%
First Choice
Relative advantage	3	60	4	80	2	40	1	20	10	5O
Cost	0	0	0	0	1	20	3	60	4	20
Social influence	0	0	0	0	2	40	0	0	2	10
Compatibility to   organization	1	20	0	0	0	0	1	20	2	10
Compatibility to   Hispanics	1	20	1	20	0	0	0	0	2	10
Complexity	0	0	0	0	0	0	0	0	0	0
Trialability	0	0	0	0	0	0	0	0	0	0
Second Choice
Relative advantage	1	20	0	0	2	40	2	40	5	25
Cost	0	0	0	0	2	40	1	20	3	15
Social influence	0	0	0	0	0	0	0	0	0	0
Compatibility to   organization	1	20	2	60	0	0	0	0	3	15
Compatibility to   Hispanics	0	0	2	60	1	20	0	0	3	15
Complexity	2	40	1	20	0	0	1	20	4	20
Trialability	1	20	0	0	0	0	1	20	2	10
Third Choice
Relative advantage	0	0	1	20	0	0	0	0	1	5
Cost	3	60	2	40	2	40	0	0	7	35
Social influence	0	0	0	0	0	0	0	0	0	0
Compatibility to   organization	1	20	0	0	1	20	1	20	3	15
Compatibility to   Hispanics	0	0	0	0	0	0	1	20	1	5
Complexity	0	0	0	0	0	0	1	20	1	5
Trialability	1	20	2	40	2	40	2	40	7	35

So I think it has to be a wow factor of so-and-so who I never thought would lose weight, lost weight… . Oh my god they are exercising … I never thought they would… . It has to be a behavioral change I think that staff didn’t expect.

Relative advantage was followed by cost, selected mostly by administrators and consumer advocates as their top choice. These administrators and consumer advocates stressed the importance of making sure that the intervention is reimbursable through public or private insurance in order to generate revenues to offset intervention costs and support its implementation over time. As stated by an administrator, “if it’s not reimbursable, people are less likely to do it.” Cost savings was another critical issue for generating support from policy makers to implement PCARE on a larger scale, as discussed by this consumer advocate:

I know that the first thing that gets people’s attention here [referring to policy makers] is that this is going to save money and to a certain extent, we acknowledge short-term investments for long-term savings … people want to hear about how you’re going to save them money.

Following cost, stakeholders’ top choices revolved around issues of social influence, compatibility to the organization, and compatibility to Hispanic patients. Social influence was only endorsed by consumer advocates and it was among their top choices. The endorsement of trusted leaders and national organizations (e.g., Substance Abuse and Mental Health Services Administration) was considered a key priority for consumer advocates and their approval mattered in shaping their decision to support PCARE.

Compatibility to the organization was primarily endorsed by mental health providers, administrators, and primary care providers and centered on issues of fit between the intervention and the organization’s services, staff attitudes, responsibilities, and duties. These stakeholders mentioned that compatibility to the organization was closely tied to how the new program is introduced and perceived by the staff; they also mentioned that to enhance compatibility, organizations need help in planning, coordinating, and managing this change process in order to develop leadership support, increase staff buy-in, and avoid resistance. Compatibility to the organization was also discussed from a broader perspective, particularly among primary care providers and administrators who were engaged in service transformation efforts (e.g., development of medical homes) at their sites. PCARE was viewed, in the words of an administrator, as “fitting very nicely” with these efforts as it provided a tested model for integrating health care manager services in community mental health clinics.

Compatibility to Hispanics was mostly endorsed by mental health and primary care providers and consumer advocates and focused on making sure that the intervention fit patients’ culture, needs, and socioeconomic status and that it was delivered in the patients’ preferred language. Another important component for enhancing compatibility to Hispanics that was valued primarily by consumer advocates was that PCARE conveyed recovery-oriented and patient-centered values and principles when presented to the community.

Lastly, the characteristics of complexity and trialability were less prominent in this sample: at best, they represented stakeholders’ second and third choices, particularly for mental health providers, administrators, and primary care providers. The discussions of complexity focused on how to address the concerns that were mentioned in the previous section regarding the added burden and responsibilities PCARE could create for existing staff if they take on this new role. The discussions about trialability focused on how trying out the program in a small scale with a few patients could be very beneficial to an organization as it would help staff familiarize themselves with the program and would provide valuable information about how to adjust the program and the existing services before implementing the program on a larger scale.

Implementation Strategies and Themes

Stakeholders indicated that a blend of implementation strategies was needed to move PCARE into routine practice. In Table 4, we present the implementation strategies identified by stakeholders organized by type of strategy, summary of key themes per strategy, stakeholder group endorsing each strategy, and their link to intervention characteristics. All six implementation strategies specified in the taxonomy developed by Powell et al. (2012) were present in our data. We also added one strategy, cultural adaptations, as it was discussed prominently in our interviews.

Table 4.

Implementation Strategies.

Implementation Strategy		Key Implementation Themes Identified by Stakeholders	Stakeholder Group Endorsing Strategy	Intervention Characteristic Addressed by Strategy
Financial	•	Adjust or develop billing codes to match services included in PCARE	Administrators,  mental health  providers, primary  care providers, and  consumer advocates	Cost, complexity
	•	Split intervention tasks among existing staff to maximize human resources
	•	Use start-up funds to support initial implementation
Restructuring	•	Reduce staff burden by hiring new staff or splitting intervention tasks among existing staff	Administrators, mental  health providers,  and primary care  providers	Compatibility to  organization,  complexity
	•	Tackle staff resistance to change via clarification of responsibilities and involving staff in the implementation process with leadership support
Education	•	Make training materials and curriculums available and accessible	Administrators and  mental health  providers	Complexity,  trialability
	•	Use interactive and experiential clinical training formats to learn PCARE
	•	Include training to develop skills in team-based care
Planning	•	Develop planning process that involves stakeholders at multiple levels to build buy-in, improve trust, and address resistant to change	Administrators and  mental health  providers	Compatibility to  organization
	•	Involve top leadership and management to facilitate changes in organizational policies and procedures
	•	Involve organizations bringing the new program into the planning process to develop trust and tailor implementation approach to the realities of the organization
Quality  management	•	Use of continuous measurement systems to manage program and monitor process and outcome measures.	Administrators,  mental health  providers, primary  care providers, and  consumer advocates	Relative  advantage
	•	Generate local evidence of PCARE’s impact across multiple outcomes to support program implementation
Attending to  the policy  context	•	Anticipate unintended consequences of implementing PCARE	Administrators and  mental health  providers	Compatibility to  organization,  cost
	•	Think strategically and act proactively to position organization in the changing policy context
Cultural  adaptations	•	Attention to the sociocultural and economic characteristics of the patient population being served is critical for making PCARE compatible with patients’ needs, preferences, and culture	Mental health  providers, primary  care providers, and  consumer advocates	Compatibility to  Hispanics
	•	PCARE needs to convey recovery-oriented and patient-centered values and principles
	•	PCARE needs to be delivered in the patients’ preferred language and pay attention to langue in patient navigation

Financial strategies were discussed by all stakeholders, but it was a prominent theme for administrators and consumer advocates. These strategies focus on having financial plans to help organizations adjust or develop new billing codes that match PCARE services in order to generate revenues to support its implementation. Another approach, linked to the restructuring strategies discussed below, focused on splitting PCARE tasks among existing staff members based on their expertise to reduce staff burden and maximize existing human resources, as suggested by the following mental health provider:

So it’s almost like dividing functions according to your discipline. So the nurses do this and the social workers do that, right. Because you are taking something away from what they are doing, that frees them up to do this. So it’s a shifting of responsibility because it’s not really adding more of a burden.

Several administrators also discussed the importance of having some start-up funds to cover the initial investments in time, resources, and personnel associated with program implementation. The financial strategies discussed addressed cost and complexity issues.

Restructuring strategies, discussed mostly by administrators, mental health providers, and primary care providers, focused on reducing staff burden and resistance to change using a variety of approaches. One suggestion was to hire new staff dedicated to the health care manager role, as summarized by this primary care provider:

Having one person full-time at one site makes it a lot better for the chances that they are going to gain trust of all, for the patients and providers… . Because to really make changes in the severely mentally ill for weight loss, smoking cessation, blood pressure, I mean, that requires intense contact, follow-up, and education.

Although appealing, many stakeholders, especially administrators and mental health providers, questioned the feasibility of this solution since public mental health clinics have limited resources to add new staff. An alternative approach to help reduce staff burden was to split tasks among existing staff as described above. This approach was supported by some stakeholders, but raised serious doubts among others—primarily mental health providers—who worried that shifting tasks could create, in the words of a mental health provider, “lots of upheaval” among the staff as some would view this change as incompatible with their job titles and responsibilities. Stakeholders—particularly administrators—also cautioned that the organizational and staffing changes required to implement PCARE in public mental health clinics could confront serious challenges from staff members with deep-rooted skepticism toward organizational changes that introduce new ways of doing things, as described by this administrator:

So this speaks to like when you’re introducing something new, the staff, they call it the state mentality. That there is some staff that they’re not going to do anything more than what they’re doing and when you introduce ideas of, that they should be part of, a new way of doing something, they challenge it… . Their defense has become, “we don’t need to learn any new things because we provide very good services to our patients; we’re very good to them.”… And “don’t come tell me that you know better than me how to provide service to these patients.”

To address these concerns, administrators and mental health providers suggested the need to include staff members in the planning process and for organization leaders to voice their support for the introduction of PCARE in order to reduce staff resistance to change. Restructuring strategies discussed addressed compatibility to organization and complexity issues.

Education strategies, discussed mostly by administrators, mental health providers, and primary care providers, focused on the type of training and skills staff needed to take on the role of health care managers. Several educational ideas were suggested, such as program manuals and materials that are easily accessible and available via websites, a free training curriculum for program staff, and ensuring the curriculum was adapted for staff with no formal medical training (e.g., social workers). With these types of training opportunities available, stakeholders thought that nonmedical mental health providers, such as social workers, could take on this role given their skills in system navigation, case management, and familiarity with patients, but stressed that they still would need appropriate supervision by a medical provider (e.g., a RN). In terms of training formats, several stakeholders wanted more hands-on experiential approaches that include face-to-face contact with trainers and on-site supervision, as described by an administrator: “training should move beyond webinars and not force staff to stare at a computer but be more involved; more face-to-face contact and more hands-on supervision.” Several primary care providers also indicated that staff members taking on the role of health care managers would benefit from developing skills in team-based care in order to “hone their skills in things like role definition, negotiation, closing loops, following-up, and checking back,” which were viewed as central to the health care manager role. Education strategies discussed addressed issues of complexity and trialability.

Planning strategies, discussed primarily by administrators and mental health providers, focused on building buy-in and trust and addressing staff resistance to change. These stakeholders viewed planning as an important step that could help prepare organizations for the implementation process by tackling staff doubts about the new program. As stated by an administrator, planning must be collaborative in nature in order to increase acceptance, as “listening to people as opposed to imposing something without hearing what they have to say … goes a long way in getting people to accept it.” Involving top management was seen as critical because implementing PCARE may require policy and procedure changes to integrate the program into normal operations. As expressed by a mental health provider, “without that buy-in, then everything will not fall into place.” Stakeholders also commented that planning needed to include the players and/or organizations, bringing in the new program to avoid common pitfalls in the implementation process and tailoring the implementation approach to the realities of the new setting, as described by the following mental health provider:

I think a big pitfall is when outside agencies come in and say: we’ve got the model and if you just do what we say … you’ll be fine. As opposed to coming in and saying: There is a model, if there is a way to get the model to jive with the way you guys do things, then it will work. And part of what we can offer you is an opportunity to help you think through how this model might integrate. They [referring to the organization implementing the new program] are going to be much more receptive… . I think that’s real work. Because the model is going to be implemented differently in every clinic.

The planning strategies discussed addressed compatibility to organization issues.

Quality management strategies, discussed by all stakeholder groups, focused on the importance of being able to track and monitor program outcomes in order to facilitate its management and demonstrate its impact. Several ideas were proposed to capture and track program outcomes, including designing a patient registry for the program and having patient outcome data easily available and accessible in order to review them in staff meetings. The goal was to develop an infrastructure that could be easily used by administrators and staff to track and monitor client outcomes, facilitate care coordination and care management decisions, and ensure that the program was having its intended effects. The ability to track these outcomes in real time was seen as critical for generating local evidence of the program’s effectiveness and for building local support and enthusiasm for sustaining the implementation of the program over time. Quality management strategies discussed addressed relative advantage issues.

Strategies for attending to the policy context were discussed less prominently in our interviews compared with the other implementation strategies but were mentioned by several administrators and mental health providers. These strategies focused on two main issues. First, implementing PCARE could create unintended negative consequences for the organization as it opens the door for uncovering physical health issues that the staff at a mental health clinic may not be equipped to handle, thus increasing referrals to medical services (e.g., the emergency room [ER]). As described by a mental health provider, this can put staff in a difficult quandary as they are faced with ethical and professional duties to act when a medical issue is identified, but they are also cognizant of the negative impact this could have on their organization, which could “get cited” by state agencies and funders and incur higher costs by referring people to emergency services:

More is falling into your lap and being responsible for … but it’s just like somebody wants their blood pressure taken and I know they’re not taking their medication … if their blood pressure is high, who’s responsible? What are you going to do about it? Then what do we do? So is it sometimes, we’re better not knowing? It’s like I saw a TV show about lawyers. It’s like I don’t want to know where my client is because then I’m breaking the law. So sometimes you’re better off not knowing. But if we do find it, we send them to the ER or Urgent Care. And so then are we calling Emergency Management Services and the patient can’t afford the ambulance? Who’s going to pay for us to take the patient if we are going to be taking them there to the Urgent Care or the ER? Who’s going to pay for that? Or are we going to get cited more for the patients going to the ER more often?

This quandary highlights how complex policy issues associated with costs and quality of care metrics (e.g., use of the ER) that are used to monitor programs are present in providers’ decisions and need to be carefully considered when introducing a new program such as PCARE. To address these unintended consequences, mental health providers suggested including administrators and funders in the planning process so that these issues are discussed and taken into consideration.

The second issue discussed by administrators and mental health providers focused on how important it is for organization leaders to think strategically about policy changes in their communities in order to identify opportunities and resources that could facilitate the implementation of PCARE. For instance, an administrator talked about how her organization was paying close attention to the development of medical homes in her community and actively forming partnerships with medical organizations involved in competing for medical home contracts in order to connect her clients to these new services and opportunities. In other words, she was proactively positioning her organization to take advantage of this new service transformation. The attending to the policy context strategy discussed addressed compatibility to the organization and cost issues.

Cultural adaptation strategies, discussed primarily by consumer advocates and mental health and primary care providers, focused on three main issues. First, stakeholders talked about the importance of making sure that PCARE services were respectful of and compatible with the culture and economic realities of Hispanic patients. Several stakeholders worried that if treatment recommendation and goals were not culturally informed and grounded in the economic resources of the patient population, patients may feel shamed and disengage from care, as discussed by the following consumer advocate:

So folks [referring to Hispanic patients] are made to sort of feel bad and shame for failing to comply with … what the doctors would prescribe as ideal behavior… . There’s a lack of cultural understanding when you tell certain folks to do yoga and go for hikes and if that doesn’t really match up with a lot of people’s sensibilities.

Moreover, attention to economic obstacles to engaging in medical care were seen as important for making sure the intervention was compatible to Hispanics. As described by the following consumer advocate, addressing this material problem could go a long way in helping patients engage in medical care:

[What] I find so frustrating in mental health is that people come with real material problems and all we have to offer is therapy or nonmaterial intervention. And you know, that’s invalidating, it’s crazy. So the ability yeah to offer people help materially early on, I think will do a lot.

Second, consumer advocates talked about the importance of paying attention to how PCARE is marketed and presented to the community. From their perspective, the language and values that are used to describe PCARE are critical for communicating the intervention’s intent and philosophy. Consumer advocates liked that PCARE incorporated several recovery-oriented and person-centered principles into the intervention (e.g., coaching, coordination, shared-decision making), but cautioned that the use of the terminology associated with a health care “manager” could put people off as it implies that patients’ need to be managed and that health decisions are being driven by the provider without patients’ input or involvement. Several consumer advocates recommended using other terms to convey the partnership and collaborative approach driving the health care manager role in PCARE, such as a “health coach,” “health supporter,” and “health partner.”

Third, consumer advocates, primary care physicians, and mental health providers stressed the importance of making sure that the program is delivered in the patients’ preferred language in order to eliminate language barriers and provide patients the opportunities to address medical issues in their native language. Language issues were also discussed beyond the walls of the mental health clinic in the context of helping Hispanic patients navigate the health care systems as they come into contact with different personnel (e.g., receptionists) who may not speak Spanish. Several stakeholders mentioned the importance of addressing these language issues and suggested that either the health care manager or another staff member (e.g., peer specialist) accompany patients to their visit to address language barriers. Cultural adaptation strategies discussed addressed compatibility to Hispanics issues.

Discussion

In this study, we used concepts derived from two implementation theories (diffusion of innovation and CFIR) and a taxonomy of implementation strategies to examine stakeholders’ views of an existing health care manager intervention and formulate insights into implementation strategies that could be used to facilitate the use of this intervention in public outpatient mental health clinics serving Hispanics with SMI. We found that stakeholders from four different groups viewed the intervention positively as it addressed pressing health care needs in the patient population. The following intervention characteristics were central in shaping stakeholders’ opinions: relative advantage, cost, compatibility to the organization, and compatibility to Hispanics. Stakeholders’ concerns about this health care manager intervention centered on how best to integrate the complex intervention into routine practice, and their discussions on how to implement it indicated that a blend of implementation strategies (e.g., financial, quality management, restructuring, cultural adaptation) was needed to move this intervention into practice. Below we discuss the implications of our findings for practice and research.

Implications for Practice

A central intervention characteristic valued by our participants to support the uptake of this new practice innovation was the importance of showing that the innovation had relative advantage over existing services, particularly if this advantage was confirmed by local evidence of the intervention’s impact in addressing clients’ health care needs and reducing local gaps in the quality of care. This finding suggests that clinic administrators and mental health providers involved in integrating health and mental health services for people with SMI may benefit from using quality management strategies, particularly those that focus on continuous measurement, monitoring, and quality improvement of the program’s impact (Grol, Bosch, Hulscher, Eccles, & Wensing, 2007). Quality management strategies entail that administrators understand the principles and methods for quality improvement and develop or enhance their health information systems and infrastructure (e.g., patient registries, electronic medical records) and their collection of quality-of-care indicators to track health care processes and outcomes over time. Existing national initiatives, such as the Substance Abuse and Mental Health Services Administration and the Health Resources and Services Administration Center for Integrated Solutions (www.integration.samhsa.gov) and the National Quality Forum (www.qualityforum.org), that support the use of health information technologies and standardized quality-of-care indicators focusing on the physical health of people with SMI can facilitate the use of quality management strategies for moving complex health care manager interventions into practice.

Cost was another intervention characteristic discussed by our stakeholders, including how to pay for the new intervention and service transformation changes (e.g., initial start-up costs) and how to maximize existing human resources to facilitate the integration of the new program in routine practice. These cost concerns mirror the financial issues (e.g., reimbursement policies) known to impact the implementation and sustainability of integrated medical and behavioral health services to improve the health care of people with SMI (Druss & Mauer, 2010; Manderscheid & Kathol, 2014). Our findings indicate that financial strategies must play a prominent role in moving health care manager interventions into practice. Stakeholders suggested specific financial strategies to pay for these services, such as developing plans to adjust or change billing codes to reimburse for the new services (e.g., care manager visits), which will entail changes to insurance reimbursement policies at the local and state levels; accessing start-up funds from local foundations or government agencies to support the initial costs of program implementation; and utilizing existing staff and human resources to deliver the new services, which would require new training and skills development for this staff.

These financial strategies identified in our study aligned with several financial policies stipulated in the Patient Protection and Affordable Care Act and other national health care reforms initiatives. For example, the development of medical homes in some states allows them to be housed in community mental health centers, thus supporting the delivery and payment of health care management services in mental health settings (Alakeson et al., 2010). Moreover, the establishment of accountable care organizations is shifting fee-for-services policies to bundled payment schemes that could include coverage for health care manager services and link these payments to meeting quality of care standards (Druss & Mauer, 2010). These financial reforms could help pay for interventions such as PCARE in order to facilitate their implementation in public mental health clinics.

Compatibility to the organization was another intervention characteristic valued by participants. It involved clarifying how the intervention fits with existing services, staff attitudes, and responsibilities, and how changes associated with the introduction of the new intervention are managed within the organization. Organizational context can influence the implementation, quality, and outcomes of evidence-based practices (Aarons, Horowitz, Dlugosz, & Ehrhart, 2012). Our findings indicate that stakeholders, particularly those directing or delivering services, are keenly aware of these influences and suggested a series of organizational-level practices that are linked to restructuring and planning strategies. For instance, including frontline staff and program managers in the planning and implementation process can engender staff buy-in and help address their resistance to change by using their valuable input in the implementation process. Organization leaders should also pay close attention to staff burden when introducing a new practice and customize implementation efforts to address this issue by empowering existing staff with the training, resources, and incentives to take on the new practice in order to reduce resistance to change. In all, these findings suggest that existing organizational-level interventions (Aarons, Green, et al., 2012; Chamberlain et al., 2008; Glisson & Schoenwald, 2005) that use planned, systematic, and customized approaches to identify and address organizational-level barriers, involve organizational leaders, and build collaborations among stakeholders affected by the new practice innovation should facilitate the implementation of complex interventions (Damschroder et al., 2009).

Compatibility to Hispanics was another intervention characteristic valued by our participants and centered on clarifying how PCARE services fit patients’ culture, needs, socioeconomic status, and preferred language and how they align with recovery-oriented and patient-centered values. The cultural appropriateness of PCARE was an important consideration in stakeholders’ evaluation of this intervention. This suggests that implementers may need guidance and support to decide whether cultural adaptations are needed when moving an existing intervention into a new patient population and setting and, if necessary, to conduct these adaptations. Implementing practice innovations requires a balancing act between retaining the active ingredients of the intervention while customizing it to the realities of a particular setting and population. No simple approach currently exists for achieving this critical balance in routine practice, but using existing cultural adaptation models that provide guidelines, steps, and methods for conducting adaptations can help implementers determine what needs to be adapted and how to conduct these adaptations without compromising the core components of an intervention (Cabassa & Baumann, 2013).

Other intervention characteristics that can inform the implementation of health care manager intervention were mentioned by our participants but had less prominence in our sample. Social influence was important for consumer advocates and illustrates how the endorsement of trusted organizations shaped their support of a practice innovation. Discussions about the trialability and complexity of PCARE centered on efforts to learn and gain experience with the intervention and were linked to educational strategies. We found that stakeholders—particularly those administering and delivering these services—wanted access to materials and training curricula to learn the intervention and prefer training formats that are interactive, include supervision, and focus on developing skills in team-based care. These findings suggest that for health care manager interventions, training curricula must move beyond didactic formats (e.g., webinars, one-time workshops) to more experiential approaches that include demonstrations, role plays, ongoing supervision and feedback, and group-based learning. Enhancing skills in team-based care is also critical since the health care manager is a bridge between the patient and his or her mental health and primary care providers, defining and assigning tasks, establishing communication channels, and coordinating care. Lastly, attention to the policy context was mentioned by a few administrators and focused on how to address the unintended consequences of a program like PCARE, given its expansion of services within mental health clinics, and how to leverage existing health care reforms (e.g., creation of medical homes) to support the implementation of these types of interventions in mental health clinics.

Implications for Research

Our findings support the view that introducing complex interventions such as PCARE that require organizational transformations is an “uncertainty-reduction process” in which the benefits, value, cost, compatibility, and complexity of the innovation are questioned by potential implementers; efforts to facilitate implementation must address these uncertainties (Rogers, 1995, p. 216). Our study illustrates how the application of qualitative methods, particularly stakeholders’ interviews before an innovation is implemented, can be an important step in the implementation process. It helps uncover key uncertainties held by stakeholders as they make sense of the proposed innovation, which can inform the selection of implementation strategies needed to reduce these uncertainties.

Consistent with CFIR and the diffusion of innovation theory, we found that health care manager interventions require organizational changes at multiple levels, thus efforts to pitch and facilitate their implementation in public mental health clinics needs a combination of implementation approaches that move beyond offering training to staff on these interventions. This combination of strategies must help organizations demonstrate local relative advantage for the new intervention to build buy-in and support, address cost concerns to address the financial needs to sustain these new services, and enhance compatibility to organizations and the client population of interest in order to facilitate its integration into routine practice. In other words, blended implementation strategies that package discrete implementation approaches (e.g., quality improvement, financial strategies) that target multiple levels of the organization must accompany interventions such as PCARE in order to support their implementation in the public mental health system. More studies are needed to develop these blended implementation strategies and test their impact across different implementation outcomes (e.g., uptake, sustainability, fidelity). For example, a clustered randomized trial recently funded by the National Institute of Mental Health (R01MH102325) is examining the effectiveness of learning collaboratives—a blended implementation strategy that combines collective problem solving, planning, networking, and quality improvement methods (Schouten, Hulscher, van Everdingen, Huijsman, & Grol, 2008)—to implement an integrated health promotion program in 48 mental health organizations. Studies of this nature can contribute important knowledge that can inform mental health policies and services to support the organizational transformations needed to scale up the integration of health interventions in the mental health system to address the persistent health disparities afflicting people with SMI.

Several study limitations should be noted. Our purposive sample was drawn from one urban area in a large Eastern city in the United States; stakeholders from other communities may have different views of PCARE. Stakeholders’ perceptions of an innovation are not static and can change as they gain experience with the innovation (Rogers, 1995). Our cross-sectional design prevented us from capturing changes in stakeholders’ perceptions over time. Given our focus on examining stakeholders’ views and preferences for the implementation of PCARE in outpatient mental health clinics serving Hispanics, our finding may not generalize to other patient populations and settings.

Conclusion

Despite these limitations, our study indicates that attention to the fit between the characteristics of an intervention and the experiences, knowledge, and value system of potential implementers is critical for informing the implementation process as it can help identify potential barriers and point toward specific implementation strategies needed to address these obstacles. For health care manager interventions that require shifts in services, staff responsibilities, scope of practice and financing, we found that a blend of implementation strategies that helps to demonstrate local relative advantage, addresses cost concerns, and enhances compatibility to organizations and the client population of interest is critical for their implementation. Health care reform efforts focused on integrating medical and behavioral health services for people with SMI in public mental health clinics should invest in the development and use of blended implementation strategies to facilitate the implementation of health care manager interventions that can help improve the health and health care of people with SMI.