Critical Time Intervention - Task-Shifting for Individuals with Psychosis in Latin America: A Multi-stakeholder Qualitative Analysis of Implementation Barriers and Facilitators

Abstract

This article presents the results of a qualitative study conducted to understand the barriers and facilitators in implementing a pilot trial of Critical Time Intervention-Task-Shifting—a time-limited, community-based, recovery-oriented intervention for individuals with psychosis—in Rio de Janeiro, Brazil, and Santiago, Chile. Data included 40 semi-structured interviews with service users, task-shifting providers, and administrators. Analysis proceeded in three iterative phases and combined inductive and deductive approaches. Coding frameworks for implementation factors, and whether or not they acted as barriers and facilitators, were developed and refined using many domains and constructs from the Consolidated Framework for Implementation Research. Barriers and facilitators were ultimately grouped into five domains: 1-Personal; 2-Interpersonal; 3-Intervention; 4-Mental Health System; and 5-Contextual. A rating system was also developed and applied, which enabled comparisons across stakeholders and study sites. Major facilitators included intervention characteristics such as the roles of the task-shifting providers and community-based care. Top barriers included mental health stigma and community conditions (violence). Nevertheless, the findings suggest that Critical Time Intervention-Task-Shifting is largely acceptable and feasible, and could contribute to efforts to strengthen community mental health systems of care for individuals with psychosis in Latin America, especially in advancing the task-shifting strategy and the recovery-oriented approach.

Since the Declaration of Caracas in 1990, many Latin American countries including Brazil and Chile have accelerated their efforts to reform specialty care for individuals with mental disorders (Pan American Health Organization/World Health Organization [PAHO/WHO], 1990). A major shift was moving away from large psychiatric hospitals towards a rights-based approach, and the promotion of community-based care—i.e., mental health care delivered in community settings (PAHO/WHO, 2005).

In Brazil, the Brazilian Psychiatric Reform, which began with the re-establishment of the democracy, the creation of the 1988 Constitution, and followed by the creation of the Unified Health System, the Sistema Único de Saúde, was not an isolated project, but part of a larger social movement that involved civil society, health professionals, user organizations, and policymakers (Delgado, 2011). In 2001, a national mental health law was enacted and an expansive network of community mental health centers (CMHCs) named Centros de Atenção Psicossocial (CAPS) were developed to offer alternatives to asylums (Onocko-Campos et al., 2017). A psychosocial approach, including community outreach, was immediately adopted by the CAPS although staff shortage chronically threatened its actual implementation. In this context, there was a broad and comprehensive national anti-asylum movement who fought for more dignified assistance and against the abandonment of individuals with psychosis in psychiatric asylums and gave way to the creation of the CAPS’ network (Vasconcelos & Desviat, 2017).

In the 1960s, Chile made significant progress towards community mental health care through different programs (including those led by Dr. Marconi, Dr. Weinstein and Dr. Cordero), but they were all dismantled in 1973 with the military coup d’etat against Salvador Allende. With the return of democracy in 1990, mental health reforms were discussed for several years and in 1993 a National Mental Health Plan was formulated, which promulgated the country towards steady progress in community care for people with mental health conditions, including a greater integration of mental health care with primary healthcare centers (PHCs) and the development of local CMHCs known as Centro de Salud Mental Comunitaria (Alvarado et al., 2012).

Both Brazil and Chile were able to significantly decrease the number of people committed to psychiatric hospitals and long-stay beds. Between 1990 and 2017 in Chile, an estimated 90% reduction in these beds occurred, while beds in general hospitals for individuals with mental health problems were doubled (Mundt et al., 2022). Similarly in Brazil, between 1996 and 2014 beds in psychiatric hospitals decreased by 65% (Ministério da Saúde do Brasil, 2015). Notwithstanding these advances, both countries have limited experience in implementing evidence-based psychosocial interventions for individuals with psychosis.

Critical Time Intervention—Task-Shifting

From 2014 to 2017, a pilot randomized controlled trial (RCT) of Critical Time Intervention—Task-Shifting (CTI-TS) was piloted in Rio de Janeiro, Brazil, and Santiago, Chile. The trial sought to investigate the effectiveness of the Critical Time Intervention (CTI) recovery-oriented approach—and the additional task-shifting strategy—in improving the quality of life of individuals with psychosis in these two Latin American contexts (Mascayano et al., 2019). The evidence-based CTI model offers time-limited (9 months) psychosocial support provided during a ‘critical’ period of transition (e.g., from hospital to community living), through three successive and continuous phases (Initiation, Try-Out, and Transfer of Care), and has been adapted for many types of critical transitions (see www.criticaltime.org).

In CTI-TS, instead of mental health specialists (i.e., psychiatrists, psychologists), ‘worker pairs’ of community mental health workers (CMHWs) and peer support workers (PSWs) were trained and supervised by mental health professionals to provide recovery-oriented community reintegration support for adult individuals with psychosis in the selected CMHCs. CMHWs were individuals committed to providing social support within community settings. PSWs were primarily selected for their similar background with users and were expected to share their own mental health recovery experiences with them (Mascayano et al., 2019).

The roles of the worker pairs were distinct and complementary. CMHWs worked more closely with userś families in engaging family members and improving family support while PSWs helped users improve their role in the family through mentorship, instilling hope and role modeling to users in a way that only a person with lived experience of mental illness could do. CMHWs strengthened user/family ties to CMHCs and PHCs while PSWs strengthened old and created new user ties to informal community supports (e.g., self-help groups, leisure activities). However, CTI-TS worker pairs in Chile functioned more independently from (though in communication with) the mental health professionals (e.g., psychiatrists, clinical psychologists) to carry out their specific tasks, while the task-shifting providers in Brazil worked more closely with mental health professionals from the CMHCs, who also functioned as supervisors.

As in the original CTI, CTI-TS had identical three phases. The Initiation phase involved intensive in vivo work and the elaboration of a recovery-oriented plan focused on a few selected areas (e.g., family, CMHCs, PHCs). The Try-out phase focused on lessening involvement and observing how the plan made in the prior phase worked in practice. Finally, Transfer of Care was devoted to ensuring the durability of what was previously achieved. The intervention was intended to complement— not replace—treatment as usual, while the role of CTI-TS workers was specifically designed to avoid becoming the primary source of care for the individuals (Mascayano et al., 2019).

CTI-TS was developed through a socioculturally sensitive adaptation process led primarily by local investigators (Dev et al., 2021; Mascayano et al., 2019). The CMHWs and PSWs were trained on CTI-TS in two phases, first with 54 h of didactic training and then with 9 months of on-the-job training, guided by a CTI-TS training manual (Conover & Restrepo-Toro, 2013). The core components and structures of the CTI-TS model are described in detail elsewhere (Conover & Restrepo-Toro, 2013), and the quantitative results of the pilot RCT effectiveness trial are presented in Mascayano et al. (2022). For the purpose of this paper, which aims to discuss the qualitative analyses of the implementation determinants, a summary of intervention characteristics across the two study sites is included in Table 1.

Table 1.

Intervention characteristics across CTI-TS Study Sites

Intervention characteristic	Rio de Janeiro, Brazil	Santiago, Chile
Participants receiving the intervention	25	30
CMHCs	3 CAPS	5 CSMCs
CMHWs	3 CMHWs	3 CMHWs
	Undergraduates from a training course for community health workers and having previous experience working with individuals with severe mental illness	Individuals with experience with and passion for working with vulnerable populations, familiarity with the areas surrounding the CMHCs, and community service experience
PSWs	3 PSWs	4 PSWs
	Selected from within the same CMHC where participants were recruited	Selected from CMHCs not participating in the study
Caseload	7–9 users per worker pair	8–12 users per worker pair

CMHCs community mental health centers, CAPS centro de atenção psicossocial, CSMCs centro de salud mental comunitaria, CMHWs community mental health workers, PSWs peer support workers

Evaluating Implementation Determinants of Critical Time Intervention-Task-Shifting

Upon conclusion of the CTI-TS trial, a series of semi-structured interviews with service users who received the intervention, and implementers [PSWs, CMHWs, administrators (i.e., clinical and managerial staff at CMHCs)] were conducted to evaluate the stakeholders’ perspectives on the barriers and facilitators experienced during the implementation of CTI-TS. This was done as part of the global Shared Research Project, which sought to identify common and distinct barriers and facilitators of different task-shifting mental health intervention models implemented by different Collaborative Hubs for International Research in Mental Health (National Institute of Mental Health, n.d.).

The Shared Research Project continued recent prior efforts in global mental health research to use implementation science (IS) to advance the field—specifically, by applying IS determinants frameworks to analyze and synthesize the knowledge base of the myriad factors that can impact the implementation outcomes of task-shifting mental health interventions. Having such an understanding can help researchers and implementers examine why certain intervention models arrive at different implementation outcomes in different contexts (Damschroder et al., 2009), and to better inform the selection of adaptations and implementation strategies for future implementation and scale up of task-shifting mental health intervention models in other settings.

In addition to an analysis across the different “hubs,” the Shared Research Project data from CTI-TS were also independently analyzed. The present study provides a comprehensive comparative analysis of the barriers and facilitators of CTI-TS, based on in-depth interviews with users, PSWs, CMHWs, and administrators from both Rio de Janeiro, Brazil, and Santiago, Chile. The research questions were: What are the implementation barriers and facilitators in the pilot CTI-TS trial, as perceived by the different stakeholders from both Rio de Janeiro, Brazil, and Santiago, Chile? What are the lessons learned regarding the acceptability and feasibility of task-shifting, recovery-oriented psychosocial interventions for individuals with psychosis in Latin America?

Methods

Settings

The CTI-TS qualitative Shared Research Project study was conducted between 2015 and 2016, in the CMHCs in Santiago, Chile, and Rio de Janeiro, Brazil that were participating in the pilot trial. The five participating CMHCs in Santiago served a population of approximately 1.2 million individuals, corresponding to about 18% of Santiago’s population (Regional Ministry of Health Secretariat of the Metropolitan Region, 2013). The percentage of people living below the poverty line in these coverage areas ranged between 21 and 42%; the average estimate for Chile is 21% (Chilean Ministry of Social Development, 2015). In Rio de Janeiro, the three participating CMHCs were primary adult mental health services for areas with approximately 2.4 million residents, representing 37% of Rio de Janeiro’s total population of 6.5 million (Regional Ministry of Health Secretariat of the Metropolitan Region, 2013). Rio is characterized by social and economic contrasts, with some neighborhoods (favelas) presenting high rates of poverty and violence (Perlman, 2006).

Participants

Data were collected from a total of n = 40 participants (see Table 2). Service users were randomly selected from among the 55 participants who received CTI-TS. All active PSWs and CMHWs were invited to participate in the interviews. The administrators, who were also mental health professionals (e.g., psychiatrists or psychologists), coordinated with CTI-TS on behalf of the CMHCs. All interviews were conducted by independent interviewers not associated with the intervention nor the randomized control trial. Users were interviewed by mental health professionals (e.g., clinical psychologists or social workers). Interviews were conducted in each site’s respective local language (Brazil: Portuguese; Chile: Spanish) and were transcribed verbatim and then translated to English by bilingual research assistants.

Table 2.

Sample Sizes of Study Participants

Stakeholder group	Rio de Janeiro, Brazil	Santiago, Chile	Total
Service Users	9	15	24
Peer Support Workers (PSWs)	3	3	6
Community Mental Health Workers (CMHWs)	3	2	5
Administrators	3	2	5
Total	18	22	40

Instrument Development

Semi-structured interview guides were created following a four-step process: First, we held meetings with the intervention development team, peer support experts, and several Latin American mental health workers to understand the CTI-TS model and training goals. Second, a team of masters-level research assistants conducted an extensive literature review on task-shifting initiatives and peer support work to identify major research areas of concern in the field. A list of themes was generated, and we composed one to three questions for each theme. Finally, we validated the draft instrument with key members of the local implementation team.

The interview guides across these four stakeholder groups were largely similar except for a few thematic areas. Interview guides for PSWs and CMHWs included specific questions around their relationship with each other, as well as their opinions on task-shifting (for example, social congruence with user and impact of sharing recovery story for PSWs, and CMHWs’ role in engaging families in users’ recovery, respectively). Lastly, the administrator’s interview guide incorporated their perspectives on users, PSWs and CMHWs, on the inter-relationships and dynamics among these stakeholders, and on the intervention itself. The number of questions on the interview guide across all stakeholders ranged between 20 and 45. See Supplemental Files 1 and 2 for a list of topics and questions included in stakeholder interviews.

Analysis

Data analysis was carried out in three phases described in detail below. Broadly, a codebook was iteratively developed in phases 1 and 2, using a combination of inductive and deductive coding, to arrive at a list of constructs that uniformly represent all stakeholder perspectives. In phase 3, a rating system was developed to identify the valence of each construct as a facilitator, barrier, or a neutral factor in the implementation of CTI-TS.

Phase 1: Development of Codebooks for Each Stakeholder Group

First, an open coding strategy was used with Santiago user interview transcripts to assign ‘codes’ that were categorized as either ‘barriers’, ‘facilitators’ or ‘neutral’ to intervention implementation and scale up. For example, “lack of family support” would be coded as a barrier; “lived experience of PSW enables empathetic connection” a facilitator, and a neutral code would be a factor that neither exerted a hindering nor beneficial effect, or whose effect was neutralized, such as “coping with neighborhood violence”. This process was also applied to user transcripts from Rio de Janeiro, with codes phrased similarly or combined if constructs were similar or related to those from Santiago users.

These initial user codebooks were then applied to PSW and CMHW interview transcripts, which were expanded upon with new emerging codes using the same inductive and iterative approach. Separate codebooks were developed for administrator interviews, as the content of the interviews were considerably different. At the end of this phase, there were eight codebooks, one per stakeholder group at each site.

Phase 2: Consolidation of Codebooks

In phase 2, we revised each stakeholder’s codebook to consolidate across sites, and then compared codebooks across stakeholders to facilitate inter-stakeholder comparisons. This entailed re-naming codes, and re-organizing or merging them for uniformity in domains, themes, and codes (i.e., constructs).

At this time, every codebook had three ‘frameworks’: barriers, facilitators, and neutral. In each framework, codes were categorized into five domains: (1) personal characteristics, (2) interpersonal characteristics, (3) intervention characteristics, (4) mental health systems and policies, and (5) contextual factors. These were created by combining domains from ecological systems theory (Bronfenbrenner, 2005) and the Consolidated Framework for Implementation Research [CFIR] (Damschroder et al., 2009, 2022). We ground our determinants framework with the ‘personal characteristics’ domain (which is the ‘individuals’ domain in the CFIR) as opposed centering on the ‘innovation’ [i.e., ‘intervention’] domain as did the CFIR. We then followed the ecological systems theory’s (Bronfenbrenner, 2005) framework of spheres of influence emanating from the micro- to meso- to macro-levels to add the following overlapping, larger domains. Domain 2, interpersonal characteristics, represents the meso-level characteristics that emerged as important determinants connecting the individuals to the contextual factors and settings. Of note, the CFIR does not clearly specify factors related to the interpersonal domain. In Domain 3, intervention characteristics, we combined the ‘innovation’ and ‘implementation process’ domains of the CFIR. Our Domain 4, mental health system and policies, is equivalent to the ‘inner setting’ domain in the CFIR, and Domain 5, contextual factors, the ‘outer setting’ domain.

The final process of consolidation involved merging codes across the barriers, facilitators, and neutral frameworks. We created a table where rows included the list of all codes (grouped into their corresponding domains and themes) identified in the facilitator, barrier, or neutral frameworks. Eight columns represented each of the stakeholder groups (4) in each site (*2).

Phase 3: Rating Barriers and Facilitators

Upon completion of the table of codes consolidated across barriers (−), facilitators (+), and neutral (=) frameworks, we rated each code (i.e., implementation factor) with 1, 2, or 3 signs to represent the significance or ‘weight’ attributed to that code, based on the number and degree of importance of the stakeholders’ responses. This rating system (see Supplementary file 3) was derived following group discussions among the analysis team (PTL, MA, TK, SD), and was meant to ensure consistency in assessing the importance of the factors. When a code was not discussed by a specific stakeholder group, it was left blank.

Results

Implementation factors manifested as facilitators, barriers, or neutral within the five domains are presented below. A detailed list of these factors indicating strength of impact is provided in Supplementary File 3. In the participant notation system, “S” or “R” represents Santiago and Rio, respectively; thus, CMHW R2 would be Rio-based CMHW #2.

Personal Characteristics

Lived Experience of Mental Illness and Recovery

All the stakeholders identified factors related to this theme as significantly impeding intervention engagement. Users frequently talked about how medication side effects and feelings of anxiety, hopelessness, and depression posed limits to their recovery. Other stakeholders also noted how symptoms hampered user participation (e.g. a user not motivated enough to get up from bed and meet with the PSWs and CMHWs).

While PSWs appreciated their own recovery orientation (e.g., insights about recovery process, not feeling defined by mental illness) and ability to take care of their mental health (e.g., boundaries around work, taking medication), some users and CMHWs commented on the PSWs’ challenges in their recovery process, including mental illness symptoms:

One of them [PSW] - he was very confused (…) I would fill out the form and (…) sometimes he had a totally different view of what happened (…) I would sit with him [and say] “try to remember” (…) I was there too (…) accompanying him (…) as if he was also a user (CMHW R3)

Knowledge, Attitudes, Beliefs, and Skills About Mental Illness, Treatment and Resources

These characteristics—endorsed by and about users, PSWs, and CMHWs—included explanatory models of mental illness and treatment, knowledge about community resources, capacity for empathetic connection, and interpersonal skills. Users in both sites indicated their own Knowledge, Attitudes, Beliefs, and Skills (KABS) about mental illness, treatment, and resources as strong facilitators, but they, and CMHWs across sites, also noted them as barriers or as neutral.

Both types of CTI-TS providers spoke positively about their own capacity for this type of work but were more reserved in discussing those of their coworkers. For example, PSWs typically underscored their own empathetic connection and interpersonal skills that positively influenced users’ experience during the intervention:

I increased my empathy, (…) my ability to listen, (…) to understand (…) I told the users I worked with… ‘You have a possibility that I did not have. I did not have a guide to lead me.’ (…) I was on my own, understand? (…) I had to face the problem alone (…) I like this very much and… I had no problem doing this. (PSW S2)

However, some users, especially those in Santiago, and CMHWs expressed some competency- and professionalism- related concerns about PSWs which included their low self-esteem, lack of reliability and punctuality, being inappropriately dressed, skipping supervision and discomfort to share recovery story.

No stakeholder negatively commented on professionalism-related characteristics about CMHWs, though occasionally Santiago users raised questions about some CMHWs’ skills (e.g., perceived paternalistic attitude, difficulty in conveying empathy/recovery orientation, and providing contradictory information). Similar to PSWs, CMHWs emphasized their own enthusiasm for work, prior relevant experience, interpersonal skills, job competency and knowledge of social and community resources. CMHW S3 said: “It [the intervention] respected the users, respected the space, the family; I listened (…) to everything they said. So, I always had these things on my mind. (…[to]…) respect them as people.”

Attitudes and Beliefs About Engaging in the Intervention

All stakeholder groups held mostly positive attitudes about the intervention, and there were broad agreements by the other stakeholders, with one exception. Although users themselves rated positively their understanding of the intervention (e.g., differentiating between social and medical aspects) and investment in it, some CMHWs and PSWs expressed disappointment in users’ level of commitment (e.g., limited awareness about CTI-TS personal benefits, canceling meetings for other priorities).

The worker pairs’ own intervention-related attitudes appeared especially strong facilitators, and included job motivation and satisfaction. Attributions and reasons amongst PSWs included advanced recovery, ability to interact with and help other users, and the positive impact on their family. A PSW from Santiago expressed his satisfaction with being involved in the intervention:

They [users and health professionals] valued me (…) I can tell that from their gestures (…) They have confidence in me (…) They are interested in what I have to say. This isn’t to show off, but I think I did a good job. (PSW S2)

CMHWs noted their own personal factors (e.g., desire to be challenged), the satisfactory work environment, and their work’s positive impact on community and self: “I have always participated in (…) different types of organizations, schools, cultural centers (…) always had something that motivates me to be with people. I think that’s why it’s also been easier to work on the issue of mental health” (CMHW S1). Administrators, too, were positive in their overall attitude towards CTI-TS, calling it “effective,” and “feasible.”

Users’ Personality Characteristics, Life Events, and Socioeconomic Characteristics

Participants mentioned user personality characteristics and life events either as neutral factors (e.g., personal values on family and home, change in living situation) or barriers (e.g. providing unreliable information, difficulties socializing) for intervention implementation. In terms of socioeconomic context, the lack of financial resources (e.g., money for transportation, access to mobile phones) was highlighted by multiple stakeholder groups from Rio. Stakeholders across sites mentioned how the intervention engagement and activities were affected by where the user lived; for example, a safe neighborhood versus an inaccessible favela, and how well the users and CTI-TS providers knew about safer routes. Homelessness also affected CTI-TS engagement: “There was a woman who lived on the streets; sometimes she showed up, sometimes she did not,” explained Administrator R1.

Interpersonal Characteristics

Users’ Interactions and Relationships

Users’ relationships with the task-shifting providers emerged as important and highly positive for CTI-TS. Users described the attachment, personal connection and trust they felt in this egalitarian relationship, given the often-similar backgrounds of the CMHWs and PSWs to the users. Administrator S2 shared an example of a meaningful relationship:

[PSW S2] became someone close in [a user’s] life.. […] He created networks. In a way they were profoundly different, right? This [user] went around with a mohawk and with chains… He was… wild. And [PSW S2] was a man […] ‘in a shirt,’ let’s say. So, it was weird to see them together, right? …All the same, they were friends in some way… It moved me to see… that they [understood each other…]

Relationships between users and their families were also critical but more fraught. Support—or lack thereof—from users’ families, including parents, spouses, siblings, children, and cousins, influenced implementation and perceived effectiveness of CTI-TS. For example, PSW S3 perceived that family understanding and support during CTI-TS facilitated early observable changes in the form of initiative-taking for one user with a developmental delay. However, he also noted: “Other families were frustrated with the problems that were happening …they didn’t want to understand, at times,” and described how a brother of a less motivated user threatened the user: “if you don’t behave, you’re going [to the hospital].”

When present, the nature of family support could be emotional (e.g., being patient about change), instrumental (e.g., transporting the user to recovery-oriented activities or providing financial assistance), or motivational (e.g., urging users to “not miss” treatment). Detrimental family effects on the user appeared to relate to a lack of involvement, negative interactions around treatment, conflict, or stressful home environment, and family misunderstanding of mental illness. User S3 shared she was apprehensive the home-based intervention would be ill-received by her family whom she described as “kind of behind in their thinking… [they think] someone ‘suffers depression because they’re stupid.’” In Rio, it was found that dependence on family members for attending or being connected to treatment (e.g., through the phone or visiting the CMHC) could impede intervention engagement.

Interestingly, users’ relationships with friends or the broader community seemed to have little effect on the intervention participation. Neighbors and friends were rarely mentioned, and when they were, the implication seemed neutral (e.g., having no communication with neighbors or not helping in the intervention despite their relationship being good). Finally, some users mentioned that their trusting relationship with non-CTI-TS healthcare staff was facilitative for CTI-TS, as the staff could help create positive expectations of the intervention. Users also noted their relationships were restricted at times due to the less time non-CTI-TS professionals had available.

Task-Shifting Providers’ Interactions and Relationships

The worker pair’s working relationship and rapport with each other was highly significant and endorsed by both stakeholder groups. PSWs appreciated the support they received from the CMHWs; PSW S1 shared that a CMHW was like “a salvation…Because apart from the work, [she is] a person that has a life, a family, problems…I had a very good work relationship with her.” Notably, CMHWs and PSWs across sites (though majorly in Santiago) reported some difficulties with a partner, for instance, issues with their partner’s disorganized nature, rigidity in thought or ‘bossiness.’ In Santiago each CTI-TS provider seemingly worked better with the one that was more like them, e.g., relaxed vs. structured. They attributed these issues to character differences or less time spent together.

The CMHWs’ and PSWs’ rapport with the user’s family, and family’s acceptance of the worker pair, were perceived as facilitating the intervention. CMHW R1 shared how “for a while, as soon as the intervention was over, his [the user’s] wife would call” updating him on the user’s progress and asking for help. This reflected both the strong alliance formed to aid user recovery, and the fluid boundaries that characterized these relationships. However, this rapport can take a long time to form, as some families show resistance to being helped by someone without a professional degree.

Among the interpersonal characteristics centered on the worker pair, PSWs’ own family support and enthusiasm was salient in their recovery and work with CTI-TS, but the same was not noted for CMHWs. Workers also sometimes noted less acceptance from non-CTI-TS healthcare staff, with PSWs in Santiago feeling stigmatized due to their user identity and CMHWs in Rio noting the CMHC staff took time to warm up to them.

Intervention Characteristics

Task-Shifting Providers

Shared experiences and similar backgrounds between users and PSWs, which enabled an equal (i.e. non-hierarchical) relationship between them, were seen as a significant facilitator in defining the roles of PSWs and CMHWs. PSW R2 described, “Most of them [the participants] even felt better, knowing they had a person who has been [a] patient giving them support.” He also noted his liminal position as a peer allayed any fear of the hospital setting and seeing individuals with psychosis, as someone who knew how the center could help.

However, there were more nuances and even challenges with the CTI-TS providers’ new roles. Users had difficulties understanding and distinguishing them from the staff at the CMHC.

In Santiago, users’ perceived social and professional hierarchies (e.g. between CTI-TS providers and doctors, or distinguishing two CMHWs’ education levels based on language) appeared as neutral or negative indicators. Additionally, some users and Santiago CMHWs expressed concerns about the emotional load on the CTI-TS providers.

The nature and types of support provided were perceived as a very strong facilitator, especially among users and worker pairs. The types of support mentioned included motivational, emotional, instrumental, and informational support. User S6 explained how the TS providers improved his self-esteem and motivated him to feel comfortable about his sexuality. CMHW R2 explained why he thought his intervention work made a difference:

I saw (User ‘x’)’s mother accept his illness, you know? I talked so much to her, “Oh, (mother name), there are some diseases that develop. That’s not the devil, you know?” She began to accept her son’s illness. (User ‘y’), we got him out of the… (…) we talked to him, he was able to leave the shelter and now lives in his own home…

Users viewed the personalized attention received from the CTI-TS worker pair as particularly helpful. However, user S13 expressed the need for more practical support in addressing the users’ socio-economic difficulties, such as housing, employment or finances.

Availability and scheduling, including appointment coordination, physical location of meeting spaces, and flexibility/availability of the worker pair and other professionals, was discussed significantly with mixed indications. Many participants viewed that having the worker pair increased the availability of psychosocial services for individuals with psychosis in these contexts, and it is generally easier to schedule appointments with them than with mental health professionals. However, users and TS providers also noted that they encountered significant challenges with respect to coordinating and scheduling appointments.

Most participants also viewed the assigned caseload as a facilitator, in that it was manageable, although some PSWs across sites and CMHWs in Rio felt that they worked with too many users. Relatedly, having the same TS provider throughout the intervention enabled the continuity of the relationship between the users and the TS providers, though the change in provider did not seem to have a negative effect when it did occur.

Compensation and payment, meanwhile, was viewed as a barrier by CTI-TS providers and administrators due to associated challenges like delayed or insufficient payment, having less benefits than a regular job, and resulting lack of motivation to meet project demands. Some PSWs expressed appreciation for money provided for transport or denied being affected by delayed compensation.

Community-based Care

The community-based nature of activities was positively regarded by stakeholders. Users appreciated the convenience, comfort and flexible boundaries associated with meeting at home. Rio administrators noted the intervention facilitated the CMHC’s services and reach into the community, inspiring CAPs to seek out territorial agents who could work in the community (Administrator R3) and enabling a homeless user to connect to her son (Administrator R1). “I think we [CAPS] got to pay attention to this, she had a child, the birth history, we paid attention to that [as a result of the TS providers efforts].”

A related advantage noted by CMHWs, PSWs and even users was the insight the worker pair gleaned into users’ social life, family dynamics, risk status and motivation. CMHWs noted its potential to include family members in the intervention. Conversely, CTI-TS providers had difficulties navigating boundaries and user expectations within the home and community setting (PSW S1).

A separate theme around the intervention taking place at the CMHC emerged. In Rio, the intervention activities taking place mainly at CAPS, partly due to fears of community violence, received mixed reactions. CMHWs had the most to say on this theme, with some noting positively that other users and their families expressed interest in being part of CTI-TS, and others commenting on the time it took to foster buy-in and cooperation from CAPs staff and understand the CMHC workings.

Intervention Complexity, Structure, Processes

Favorable comparisons of CTI-TS to existing services (i.e., relative advantage of the intervention) were highlighted by all stakeholder groups. These facilitators included: positive outcomes for users, the complementary nature of work between MH professionals and the CTI-TS providers, the providers engaging family members, and the international and university affiliation of CTI-TS creating benefits for users and staff at CMHC. However, these characteristics were also perceived as neutral by many participants. For instance, some participants initially doubted the intervention’s efficacy, and some viewed that there was insufficient family involvement, though these characteristics were not perceived to affect intervention success.

Timing and duration of the intervention received mixed feedback. Participants praised the timing of the intervention—i.e., engaging with users at a critical time of transition— but many users and TS providers had a difficult time transitioning through intervention phases and after CTI-TS ended. Most CTI-TS providers felt that the intervention and pre-decided phases were too brief, and personalization was needed; PSW S1 described how while some users did not seem to need the intervention where visits were ongoing, others started regressing with the withdrawal of visits:

After they were able to find someone to help them, they were abandoned again (…) when I see them in the field it’s so different to (…) statistics data (…) It was (…) us who were at the frontline in each situation. (…) I think that the time…(should) depend on each user.

In terms of intervention manualization and protocols, the training manual, intervention procedures and administrative work were viewed mostly as facilitators or neutral factors. CMHWs and Santiago PSWs expressed that the protocol was less rigid than current psychiatric treatments, and the confidentiality promoted facilitates user engagement. Some TS providers appreciated that the administrative forms were culturally adapted, which helped organize the work and reporting requirements. Santiago users appreciated the concluding ceremony. Nonetheless, several participants cited the intervention procedures and requirements as too rigid and expressed difficulties in properly terminating the intervention.

With regards to team processes, communication channels and structures among intervention providers and CMHC coordinating staff, as well as team interactions, were rated as strong facilitators by CTI-TS providers and administrators. The Santiago CMHWs noted they got help from CMHC staff with respect to matters like timely appointments, user crisis support and modification of treatment plans. PSWs especially appreciated the various avenues of communicating and getting support from other team members. For example, PSW S3 described how he and CMHW S1 were troubled by a user who appeared dishonest about ongoing substance use, and how they discussed concerns with each other and the CTI-TS supervisors. “Bottomline: we are a team,” explained PSW S2.

Administrators noted challenges in integrating with other CMHC staff, and that it took time; Administrator R2 shared, “Many times, I would have to keep reminding the staff constantly, ‘Guys, remember that there is research going on at CAPS,’” and how the CAPS staff did not attend the CTI-TS meetings though they were invited.

Training and Supervision

The training and supervision model received positive feedback. The CTI-TS workers noted they could roleplay and gained helpful knowledge and skills, such as empathetic listening, motivational support, risk evaluation, sharing of their own experiences (for peers), and knowledge about resources and mental health. CMHWs also noted it served to build the foundational relationship with PSWs and understand what to expect from users. In addition, peers reported that the training also helped them personally self-reflect and feel empowered outside work. Some acknowledged, however, that in actual practice experiences could differ from what was discussed in training.

Supervision was also perceived as a strong facilitator. It enabled further development of skills introduced in training and provided spaces for open dialogue with an experienced professional. However, PSW S1 noted she felt constrained by the limitations placed on her (to not diagnose, to work as a PSW instead of CMHW) and that there was not enough warmth from the team or focus on self-care: “Because we would arrive (…) on Mondays, and the papers were there and the board. [There was] not even a ‘good morning’— ‘How are you, how was your weekend? How is your family, how is your daughter?’” She suggested that reiki or de-stressing sessions would have been helpful for the TS Providers engaged in this work.

Mental Health System Characteristics

Prominent implementation factors related to the mental health system can be divided into two levels: those at the CMHC/PHC level, and those at the policy and federal level.

Community Mental Health Center/Primary Healthcare level

Characteristics of CMHCs were largely appointed as significant barriers or neutral according to most stakeholder groups, although administrators were more positive on the facilitating aspects of CMHCs. Pre-existing community outreach activities at CMHCs, when present even at nascent stages, facilitated greater acceptance for the intervention. Limited resources and difficulties working with persons at CAPs (such as physical space for privacy or opportunity to discuss cases with workers) were specially highlighted in Rio, where the intervention was more closely connected to the CMHCs. CMHWs across sites pointed to the “broken chain” (CMHW S1) of communication, and poor coordination and continuity of care between local CMHCs and other agencies (e.g. PHC services, church, and NGOs) which, though partially alleviated by CTI-TS, also hindered the intervention; it resulted in limited referrals between agencies and missed opportunities for redressal of issues like the informal resource interfering with user engagement with the CMHC (CMHW R1: “it took a long time to find out that (the user) sometimes missed meetings because she crocheted at the church or because she was doing astigmatism spiritual treatment there”).

Administrators shed light on aspects of involvement of local CMHC/PHC in the intervention. While Santiago administrators mentioned CMHC involvement helped CTI-TS access users who needed support, they agreed with Rio administrators on the restrictions of limited time, finances and overburdened personnel at the PHCs. These acted as intervention barriers, though some relief was provided by CTI-TS. Further, the PHCs had no dedicated mental health team that could facilitate CTI-TS. Administrator S1 noted that “the characteristics of hospitals or centers are dependent on how they are shaped in their organization chart… the social environment where they are located, the public policies that are implemented.” She explained that these contextualized characteristics and processes would need to be accounted for in an intervention like CTI-TS. For instance, problems like crime that were applicable to her hospital’s locality may require special considerations.

Policy and Federal Level

Social benefits (e.g., bus pass, pensions) and general aspects like the intervention aligning well with the ongoing psychiatric reform in the region were highlighted as facilitators, while difficulties in accessing or availing benefits (because they were time consuming or required computers) appeared as impediments. Both sites acknowledged that the lack of financial investment in public health limited the intervention. Administrators spoke of the unstable economic and political climate resulting in cuts in funding and shortage of resources, both at community care and hospital level. Users and PSWs too endorsed this problem and called for greater investment: “to give health the importance it deserves, you need to spend more money on health” (PSW S2).

Contextual Characteristics

Mental Health Stigma

Demonstrated primarily against the users and the PSWs, stigma was reported as a major barrier by nearly all stakeholder groups. The most common types of stigmas were internal, family, and community. A few users did not want to disclose their problems as they did not trust those around them, sometimes due to associated negative experiences, or because they anticipated pity. For example, User S5 associated his depression and difficulty working with an instance of being offered alms. Similarly, some users’ families held negative stereotypes of mental health difficulties and were resistant to home visits by CTI-TS providers due to anticipated stigma from neighbors. CMHWs across sites and administrators in Santiago noted that anti-stigma education and public initiatives were successful in reducing it over time; CMHW R1 spoke about how a user’s husband requested them to take their daughter, who was “very reluctant to accept that her mother was mentally disturbed” to the family clinic, which discussed prejudice and discrimination with family members. “The daughter came to us at CAPS, and there she was able to create a working bond in both CAPS and the community.”

Of note, some aspects of mental health stigma were also noted as neutral by stakeholders, such as cognitive separation of self from others with the same diagnosis, non-disclosure as a coping strategy, and absence of or indifference towards stigma from neighbors or community.

Users also reported structural types of stigmas, such as those among professionals such as primary healthcare doctors. This often resulted in internalization of stigma where users rejected or feared the PSWs. Moreover, the PSWs felt discriminated against due to their user identity. A strong example was when a CTI-TS team member rejected the PSW’s request to take on a different role, which she experienced as being underestimated and restricted.

Social and Community Conditions

Though physical safety was commented upon in Santiago, it was far more salient in Rio, where community violence was noted as a major barrier by all stakeholders and interfered with CTI-TS providers’ home visits. The worker pair was also often viewed as “outsiders” in the favelas, which could impede them from accompanying users to activities. Interestingly, stakeholders in both the settings identified ways through which violence-related barriers were neutralized. For example, the worker pair timed their home visits to avoid any anticipated community violence, and encouraged users and families to go to events that were in favelas ‘closed’ to themselves—“sometimes the patient or the family would say ‘Ah! I know where it is,’” CMHW R3 explained, “So we suggested it, but without putting ourselves at risk.”

Users also cited religious faith as a coping strategy for violence.

Relatedly, users in both settings reported religion and spirituality to be a moderate to high facilitator for engagement and recovery. Users’ religious beliefs and religious communities provided a sense of comfort to them during their participation in the intervention. The CTI-TS providers did not seem to use faith the same way, but they feel community spaces could be facilitative: PSW R3 shared how involvement in a church body that discussed concerns and promoted self-acceptance was helpful to his work and understanding.

Socioeconomic factors were a major contextual barrier for the intervention. When families or users had financial struggles, they were observed to be less interested in the intervention or be compelled to choose between engaging in CTI-TS and earning money. Similarly, political and economic instability was described to affect availability of funds and resources for mental health (Administrator B1).

Discussion

This study represents an effort to comprehensively identify and characterize the barriers and facilitators reported by multiple stakeholders at two sites (Santiago, Chile, and Rio de Janeiro, Brazil) that implemented CTI-TS, a task-shifting recovery-oriented intervention aimed at promoting community reintegration for individuals with psychosis. We organized barriers and facilitators into five domains spanning from personal to contextual levels and developed a rating system to indicate the extent to which each factor positively and/or negatively impacted the implementation of CTI-TS according to each stakeholder group—namely, users, the task-shifting providers (i.e., PSWs, CMHWs), and administrators.

The findings suggest that the implementation of task-shifting, recovery-oriented, community-based psychosocial interventions such as CTI-TS for individuals with psychosis in Chile and Brazil is largely acceptable and feasible. In particular, CTI-TS was appreciated across stakeholders the most for the flexible and personalized psychosocial support provided by the task-shifting providers (CMHWs and PSWs) and its community-centric delivery of care. The worker pair, with distinct yet complementary roles, served as a necessary ‘bridge’ between highly specialized services at CMHCs and the psychosocial support in community settings, and could form a valued personal connection with the users. PSWs provided role modeling through their shared lived experience with users, while CMHWs facilitated connections to the CMHCs and community resources. The worker pair in turn appreciated the training and supervision they received, and the intervention was strengthened by partnerships and team processes amongst and between the CTI-TS providers, CTI-TS supervising team and the CMHC administrator/liaison.

However, several major challenges in implementation were identified. At the individual level, questions were raised on PSWs’ competencies and professionalism by both their CMHW partners and the users they were meant to help. Some users had difficulty understanding the worker pair roles, and Santiago users’ perceptions of PSW roles were heavily shaped by existing norms regarding social and professional hierarchies. At the interpersonal level, a lack of user family support, family conflict or misunderstanding of mental illness interfered with CTI-TS engagement. CTI-TS providers found it difficult to coordinate meetings with the users and families, and, along with users, expressed concerns about the limited duration of the intervention. At the mental health system level, Rio stakeholders noted that it took a long time to adjust with CAPS (and their limited resources) and to gain support from the staff. Most stakeholders also expressed a need for higher resource allocation to the program, including higher compensation for the CTI-TS providers. Regarding the contextual characteristics, the community conditions such as violent or unfriendly neighborhoods had to be navigated; they affected both user attendance of the intervention at the CMHC and worker pair home visits, especially in Rio. Furthermore, mental health stigma affected users’ levels of social support from family and community, as well as family and user willingness to have worker pairs visit the home.

Implications for Implementation Research

The systematic analysis of the adaptation and implementation of an evidenced-based psychosocial intervention, to the best of our knowledge, represents a novel effort in the region and may encourage similar future initiatives. Multiple stakeholder perspectives around the feasibility and acceptability of CTI-TS across both the sites offer several important lessons for future mental health implementation research in Latin America.

First, our findings highlight the importance of modifying intervention characteristics to fit the local context and meet the needs of the community within which it is being implemented (Baumann et al., 2017; Movsisyan et al., 2019). For instance, with community-level violence being noted as a major factor by users, the worker pair, and administrators alike, some of the rather spontaneous adaptations included in CTI-TS (e.g., worker pair timing their home visits) neutralized the potential negative effect. Another important aspect to consider is the central role that families play in caring for and supporting individuals with psychosis (Singh & Javed, 2020). In many LMICs, including in many Latin American countries, up to 90% of people with psychosis living in the community are cared for by their families (Rathod et al, 2017). Thus, family based strategies such as psychoeducation and engagement of family members in the treatment and care for individuals with psychosis could improve the success of the intervention (Morillo et al, 2022).

Second, community-based interventions that are time-limited may face challenges to their feasibility of implementation and sustainability in Latin America and other regions with a dearth of established systems of mental health care. As suggested by Valle (2020), implementation of new care models in Latin America may encounter resistance from healthcare providers as they lack the appropriate time, infrastructure, and financial means to integrate the intervention activities into their existing responsibilities. Accordingly, implementation and sustainability of community-based interventions in the region requires focused preparations and dedicated personnel and institutional resources (Aceituno et al., 2021; Saraceno et al., 2007), such as having different providers delivering the intervention activities like CTI-TS did.

Along the same lines, implementation research needs to understand and address barriers to such integration with community and mental health care systems. A focus on building readiness among existing local healthcare staff and organizations will be imperative to the successful implementation of similar recovery-oriented evidence-based interventions (Thibaudeau et al., 2021). In our study, participants in Rio, especially CMHWs, expressed that it took time to get buy-in from CAPS staff and to truly understand the characteristics and benefits of CTI-TS. Additionally, perhaps a better integration of CTI-TS within the community resources may have helped those who were prone to regressing after support from the worker pair ended.

Third, although the benefits of peer support are well documented in the literature (Chinman et al., 2014; Stastny, 2012; Watson, 2019), a central challenge for future task-shifting interventions is how to better define and support the role of peer providers. Despite each country having a relatively strong history of community mental health care compared to other countries in the region, both Chile and Brazil still lack experience with incorporating peers into mental health services (Stastny, 2012). In CTI-TS, PSWs’ role modeling for new service users was generally well received, but not unanimously and to different degrees across stakeholders and sites. PSWs were perceived by other stakeholders as having adequately delivered their tasks, and the PSWs themselves were proud and satisfied with their own work. However, some users from Chile held stigmatizing attitudes towards PSWs, perceiving that PSWs’ lived experience of mental health recovery could also be a liability (Agrest et al., 2019; Le et al., 2021). Additionally, CMHWs from both sites expressed mixed emotions about their coworkers’ competency.

These attitudes about PSW roles have often been cited as major barriers in recent reviews about peer support in mental health services (Davidson et al., 2012; Ibrahim et al., 2020; Mutschler et al., 2022). Several effective strategies for the implementation of peer services have been proposed, including having a clear job description and role clarification, having a senior administrator serving as a “champion” to systematically address implementation barriers as they arise, and providing supervision for peer staff to continuously strengthen their job skills and performance (Davidson et al., 2012). In particular, the importance of clearly articulating and communicating the benefits and roles of PSWs, especially among the healthcare staff, cannot be overstated. Only peer providers—because of their lived experiences—can offer users hope, positive self-disclosure, and role modeling, all of which are critical components for recovery (Davidson et al., 2012). These unique contributions from peer providers can be highlighted and discussed—for instance, within training and supervision sessions, to raise awareness of and acceptance of peer support work from other mental health providers. In addition, the introduction of PSWs into mental health care should be implemented along with strategies that target the healthcare system and organizational policies that support the PSWs’ work ((Gates & Akabas, 2007; Ibrahim et al., 2020; Mutschler et al., 2022).

Fourth, embedded practices and values based on the biomedical approach still pose significant challenges to the implementation of future recovery-oriented interventions. Recovery includes principles such as centering on the persons’ meaning and purpose in life, promoting hope and connectedness, adopting an equal status during therapeutic relationships, and developing skills and abilities while understanding limitations (SAMHSA, 2012); these require a fundamental paradigm shift. In both study sites—and across Latin America—symptom management remains the main treatment priority, and service users and family members’ deference to specialized mental health professionals prevail (Vasconcelos & Desviat, 2017). In Brazil, although the recovery approach was very much in alignment with the operating principles of the CAPS, budgetary constraints limited CAPS work within facilities and not widely implemented in the community (Evans et al., 2012; Onocko-Campos, 2019). In Chile, recovery represented a novel concept, one that offered a more holistic and realistic understanding of users’ needs. However, without intersectoral work, and investment in strengthening community resources such as employment support, mental health stigma reduction campaigns, and new training programs for mental health workers, progress for community integration of individuals with psychosis will remain stagnant. Accordingly, factors at the mental health system and contextual levels should be prioritized (O’Connell et al. 2021).

Limitations

The findings reported from our study should be considered in light of some limitations. First, a higher proportion of stakeholders were interviewed from Santiago (users: 15 of 30; 50%) as opposed to Rio (users: 9 of 25; 36%) due to logistical reasons. This may have led to a larger and broader representation of perspectives from Santiago. Second, it is likely that participant selection for interviews may have been based more on who was available and could be reached for the interview, rather than the diverse range of experiences with CTI-TS. Third, as noted by some participants, some of these interviews were conducted several months after the intervention ended; hence, the ability of the participants to recall specific details of experiences with CTI-TS may be subject to transience. Nonetheless participants comfortably shared their overall, long-term impressions. Of note, adolescents with psychosis were not targeted for the intervention despite the importance of addressing their needs and difficulties accessing services and receiving appropriate support in the community. Finally, some interviewees may have been positively biased in the impressions they shared as the interviewers were from local mental health teams. This is especially likely in the case of PSWs, whose liminal positions in the mental health system may have contributed to their notably more positive appraisals and fewer expressed complaints as compared to their CMHW partners.

Conclusion

In this paper, we described the process of developing, consolidating, rating, and comparing and contrasting the myriad of factors that can act as barriers and facilitators to the implementation of CTI-TS in Rio de Janeiro, Brazil, and Santiago, Chile, according to multiple stakeholders’ perspectives. The findings suggest that such a community-based, recovery-oriented task-shifting intervention, although still a relative novelty in Latin America, can contribute to efforts to expand community-based mental health services for individuals with psychosis in the region. As Patel and colleagues (2015) have asserted: “no evidence-based, multisectoral, holistic, and rights-based approach is too sophisticated for LMICs” (p. xi). However, these efforts require substantial additional investments at multiple levels of decision making.

Data Availability

The data that support the findings of this study are available from the corresponding author, PTDL, upon reasonable request.