This study had a participatory design [23] and used mixed methods [24] to explore uses of an Internet-based intervention designed to support recovery-oriented practices in mental health care for people in need of long-term mental health care. The intervention was studied for a period of 6 months in two separate communities. Heeding calls for service user involvement in research [25,26], this study was conducted in collaboration with a service user consultant.

Norway has universal health care that is funded by the public as part of the through general and earmarked grants [27]. The municipalities are responsible for providing primary health care and social services, whereas the Regional Health Authorities provide specialist services (eg, acute wards, district psychiatric centers). “Communities” in this paper refer to both levels of care provided to residents of two municipalities in Norway. These communities differed in characteristics, thus providing an intended variation in context: a small rural community with approximately 5500 inhabitants within an area of 1493 km² versus a large community on the outskirts of the capital with approximately 52,000 inhabitants within an area of 100 km². Management in the two communities expressed commitments to policies promoting eHealth, user involvement, and collaborative practices. The largest community explicitly expressed commitments to recovery principles in policy and strategy documents [28].

The two communities became involved in the project through prior contacts with the principal investigator (DG). Multiple service entities at primary and specialist levels of care, as well as local service user organizations within the two communities, received written information and verbal presentations about the study. This information included the project’s overall aim of gaining insights into user needs and how e-recovery might facilitate or undermine service user involvement in treatment and collaboration with providers. Service users interested in participating in the study needed to fulfill the following criteria: they had to be older than 18 years; had to have received mental health services for at least 6 months prior to inclusion; and needed expectations of requiring services at least 6 months forward, electronic ID (see subsequently), and at least one provider willing to interact with them through ReConnect. As an exploratory study, efforts were made to recruit a wide range of participants in terms of gender, age, health issues, and types of ongoing support.

The study was approved by the Regional Committees for Medical and Health Research Ethics in Norway and the Privacy Protection Committees at the participating sites. Participants signed an online consent form before inclusion in the study.

Local steering committees were established in both communities and consisted of primary and secondary health representatives (both clinicians and authorities), information technology (IT) management, and service user representatives. Their mandate was to ensure access to necessary resources (eg, clinician time, IT support, localities for training), and that the project harmonized with ongoing activities. Two hours of group training and/or individual training were held within both communities for service users and providers initially and when requested during the study period.

After online registration and completion of the consent form, participants completed an online questionnaire containing questions about demographic characteristics, previous use of the Internet, and the following psychosocial measures:

Data on portal use were extracted from the ReConnect user log on the server. All log-ins and uses per module were recorded in the system log.

Descriptive statistics were used to analyze questionnaire and system log data using SPSS version 21 (SPSS Inc, Chicago, IL, USA). Data are presented as medians and range for continuous variables due to skewed distributions and as proportion and percentages for categorical data. Demographics from the two communities are aggregated together to protect anonymity.

Six focus group interviews were held, three with service users and three with providers, separately at each community after approximately 3 months of use. The focus group interviews were conducted by MS and LSE together. Questions were semistructured addressing current practices and expectations, experiences related to use, recovery processes, collaboration, and desired changes in the portal. A total of 12 cafés (six in each community) were moderated by LSE who facilitated discussions about use of ReConnect and about recovery processes. Both the focus group interviews and café sessions lasted approximately 1.5 hours and were audiorecorded. All focus group interviews were transcribed, whereas cafés were transcribed only when LSE’s notes indicated areas of interest to research analysis. Forum postings also served as data. Along with commenting on participant postings in the forum, LSE introduced topics for discussion relevant for recovery, such as hope, strengths, and citizenship. At times, authors DG, MS, and LSE discussed questions that LSE in turn posed to participants (eg, “Have any of you changed your way of collaborating with your provider through ReConnect and if so how?”).

Publicly available documents (eg, minutes of meetings) and personal communications regarding stakeholders’ actions following the study are referred to when reporting findings regarding the domain “organizational commitment” [35].

All transcripts from the focus groups, cafés, and forum were read through by authors MS, LSE, and DG, who added codes corresponding to the first two of Braun and Clarke’s six stages thematic analysis approach [36]. For this paper, DG searched for codes and quotes that illustrated Le Boutillier et al’s [35] four practice domains: (1) personally defined recovery, (2) promoting citizenship, (3) working relationships, and (4) organizational commitment. Because our aim was not to present themes based on an analysis of the total dataset (a focus of subsequent publications), but rather use the data to illustrate the practice domains, we present 3 to 5 illustrative quotes per practice domain. MS and LSE, who were most familiar with the informants and contexts from which quotes were extracted, reviewed the quotes in terms of credibility in illustrating the practice domains.

The median number of log-ins was 17 (range 1-151) (Table 3). Median number of messages sent was 2 (range 0-43). Modules not used by most participants could be frequently used and valued by one or two participants. This was particularly the case for crisis plan, network map, the medication list, and the diary. Some reported that having the options was valued, even though they had not used them yet.

In the following, the experiences reported in focus groups, forum postings, and cafés by service users and providers are presented relative to Le Boutillier et al’s [35] four practice domains that were derived from a qualitative analysis of 30 international recovery-oriented practice guidance documents. These domains are summarized in Textbox 1. Although “practices” typically refers to actions taken by providers, we included the actions taken autonomously by service users through their use of ReConnect.

This descriptive and exploratory study sought to illuminate the question: how is an e-recovery tool used as an adjunct to ongoing community mental health practices and what role can it play in shifting practices toward recovery-oriented care?

The service users who used the portal became more involved in activities reflecting the first two of Le Boutillier et al’s [35] practice domains—personal recovery processes and citizenship—regardless of the practices of their provider. This was observable for the approximately 10 active forum posters and 17 café participants who also reported benefits similar to those reported in studies of online [37] and offline [38] peer support. Combining online and offline peer support with toolbox resources was an empowering common frame of reference for service users. Service users valued working more concretely on their personal life domains and goals, and in having a common vocabulary in discussing their experiences with peers. The opportunity to do so represented itself a shift toward recovery-oriented practices. Not only were service users offered a choice in terms of how they received mental health services, they could also choose to participate in defining their personal recovery processes and participate in community-promoting arenas. The service user who reported obsessive use of the forum, which was resolved together with his/her provider, was the only negative health-related experience reported among services users.

The positive role that the e-recovery portal played as a service separate from traditional services was highly dependent on the role played by the service user consultant (LSE) who moderated the forum and cafés. Although knowledge of optimal models for peer-run interventions is still evolving [12], communities who seek to promote recovery through similar portals will need to invest in similar types of expertise and role models for hope. Our experience suggests that the success of this role is closely linked with the acknowledgment of experiential knowledge as an asset within the community, in-depth familiarity of the principles of recovery, and the availability of discussion partners in health care when difficulties or dilemmas arise (LSE’s experiences will be elaborated on elsewhere). When sufficiently supported, such consultants with “lived experience” can contribute to mobilizing resources among service users and communities in ways that also can be valuable for improving the quality of health care services [12].

Use of the portal to augment treatment and its role relative to working relationships (the third practice domain) was less obvious. Dyad diversity, along with the nondirective way in which ReConnect was introduced to dyads (“use it as you see fit”), was reflected in highly diverse uses of the various portal resources. A total of 30% of providers never initiated or responded to messages, a source of frustration for service users. After 6 months, both service users and providers reported they were still discovering resources in the portal and adapting uses to their needs and preferences. This may partly be due to the shift in locus of control in that service users’ could now control the content of their own story and had a lowered threshold for linking documentation (eg, personal goals) to requests for follow-up. Both parties in working relationships can experience transitions of control as challenging [17,39], which likely adds to the time it takes to adjust.

Even if some dyads did not use the available resources in the portal to engage service users, the mere existence of the portal, and the dyads’ agreements to use it, inserted the topic of control into service users’ reports of their experiences in working with their provider. Some service users reported becoming empowered to make or request changes in the treatment they received, and that providers responded positively to these requests. However, such examples probably reflect good working relationships prior to use of ReConnect. Poor working relationships did not appear to improve through use of ReConnect, but rather were more clearly exposed as such. To explore how ReConnect can more systematically support working relationships in future studies, we have incorporated a short feedback-informed treatment measure to help dyads attend to the quality of their working relationship [40].

The largest community whose leadership had committed to recovery principles (ie, fourth practice domain) was also prepared to address the infrastructure barriers to implementing ReConnect as a permanent service. This may reflect greater financial and political resources compared to the smaller community, who did not make implementation of ReConnect a priority. More importantly, however, the largest community viewed ReConnect as a means for more effectively reaching quantified policy goals for user involvement in individual action plans [28]. This type of match between organizational values and the characteristics of the eHealth tool is an important success criterion for eHealth implementations [41]. Once an organizational commitment is in place, portals such as ReConnect can facilitate more rapid shifts in practices toward recovery, in addition to more rapid dissemination of new knowledge within communities.

We are not able to offer plausible explanations for the lack of men despite considerable efforts to recruit them, an issue which future service design studies need to address. Our opportunistic selection of quotes from superficially coded data to illustrate the four practice domains is not a balanced reflection of the experiences of participants. Thorough inductive analyses of participant experiences relative to collaboration and personal recovery are forthcoming. Nevertheless, we argue that the approach in this paper is justified in light of our aim of exploring the role such portals might play in shifting practices.

This study complements reviews of technically supported self-management interventions in general [6,42-44], as well as more specific recovery-oriented self-management interventions [45,46]. Our own scoping review of e-recovery found 20 studies of six recovery-oriented portals in five countries [47]. These studies have promising, but as yet no definitive findings related to enhanced shared decision making [48], strengths and resilient self-care strategies [49], social connectedness and empowerment [7,50], and patient-centered care [51,52], to mention a few. This study is one of the few to use a participatory approach with an exploratory design using mixed methods and, to our knowledge, the first to discuss e-recovery findings more systematically in light of a recovery framework. Several of the components in our portal are similar to other solutions with promising findings (ie, access to health records [44], shared decision making [42], and peer support [45,46]). Combining multicomponents into a single portal, as we have done, increases the challenge of sorting out active ingredients. At the same time, our study of how such a multicomponent intervention is used and influences ongoing practices helps pave the way for implementation of subsequent, more evidence-supported interventions in communities.

Based on this exploratory study, the following hypotheses can be proposed for future studies:

Personal recovery: people who have Internet-based tools that help them articulate what is important to them, coupled with providers who help operationalize “what is important” into concrete goals, are more likely to become actively engaged in their recovery processes than those without such tools and support.

Citzenship: e-recovery portals that combine Internet-based peer support with local in-real-life peer support are more likely to lead to community engagement than those who have access exclusively to one or the other.

Working relationships: working relationships via e-recovery are more likely to be effective if coupled with low-threshold feedback mechanisms that monitor the quality of such relationships than those without.

Organizational commitment: organizations with commitments to recovery principles are more likely to invest in and benefit from e-recovery portals than those without such commitments.

Although policy-pushes toward recovery and eHealth are so far largely based on values and resource constraints, e-recovery is unlikely to survive without evidence of its efficacy in helping people live fulfilling lives. Progress toward efficacy trials will need to build on more in-depth understandings of how digital resources interplay with recovery processes and for which service users, dyads/teams, and community contexts. Future research would benefit from recovery researchers joining forces with computer scientists in sorting out key recovery-oriented factors that can be co-created, boosted, tested in larger controlled trials, and implemented through digital innovations.

The 24/7 availability of peer support and support for articulating personal goals in recovery processes represented itself a shift toward recovery-oriented practices within the participating communities. It is nevertheless the two practice domains, working relationships, and organizational commitment that are key to the more fundamental role that e-recovery portals can play in shifting practices toward recovery. Given organizational goals of monitoring service user involvement in care and the quality of working relationships, e-recovery portals can play a role in helping practices become more responsive to needs and aspirations as defined by service users.