GLOBAL PSYCHIATRY

Objective: This paper describes the protocol of a process evaluation of a cluster randomised controlled trial designed to evaluate the effectiveness, cost-effectiveness and implementation of a digital mental health intervention, called DIALOG+, in five low- and middle-income countries in Southeast Europe (Bosnia and Herzegovina, Kosovo1, Montenegro, North Macedonia and Serbia). The objectives of the process evaluation are: a) to explore attributes of context that might impact on the implementation of the DIALOG+ intervention; b) to assess intervention fidelity and c) to explore patients’ and clinicians’ retrospective (i.e. experienced) acceptability of the intervention.
Materials and methods: This is a mixed-method process evaluation nested within the cluster randomised controlled trial. We adopted the guidance on process evaluations of complex interventions published by the United Kingdom Medical Research Council. Data collected during and after the trial, but prior to awareness of trial outcomes, include transcripts, questionnaire responses, routinely collected monitoring data and audio-recordings of intervention and control sessions. Data analysis is descriptive and involves triangulation methods to compare findings across countries, stakeholder groups (healthcare provider, patient) and data type (qualitative, quantitative).
Results: This work is part of a larger study entitled ‘Implementation of an effective and cost-effective intervention for patients with psychotic disorders in low and middle-income countries in Southeast Europe’ (IMPULSE). The study is funded by the European Union’s Horizon 2020 research and innovation programme. The IMPULSE trial recruited 81 clinicians and 458 patients. The clinician clusters were randomised to the intervention (six sessions of DIALOG+ over 12 months) or treatment-as-usual arm. Process data collection began in parallel with the trial, starting in April 2019. Data collection and analysis will be completed before the main trial findings are known. Process evaluation findings will be used to interpret the trial results including assessing the effect of context on outcomes.
Conclusion: This process evaluation will explore the context, intervention fidelity and acceptability to contextualise the trial results, help in optimising sustainability of the intervention and inform its future dissemination. The methods described here may also inform the development and implementation of other complex psychosocial interventions in low-resource settings.

Background: There are limited resources for improving mental health care across Europe, especially in Low-and-Middle- Income Countries (LMICs) in South-eastern Europe with fewer specialist staff and less funding. Scaling up psychosocial interventions that utilise available time and resources more effectively could improve care for people with psychosis in these settings. One intervention is DIALOG+, delivered via an app on a tablet computer: patients identify life areas to improve and clinicians use a solution-focussed process to help improve these areas. This intervention was piloted across mental healthcare systems in European LMICs, and focus groups were conducted to explore whether such interventions could use available resources effectively to improve care for psychosis in these settings.
Methods: Eleven focus groups were conducted with clinicians and patients with psychosis who used the intervention over three months during the pilot study, in Bosnia and Herzegovina, Kosovo United Nations Resolution, Montenegro, North Macedonia and Serbia. The Theoretical Domains Framework (TDF), which describes factors affecting engagement with healthcare interventions, structured topic guides and guided analysis. Codes from the data were mapped onto the TDF, analysed to identify barriers and facilitators, translated into English and checked for inter-rater reliability.
Results: 25 clinicians and 23 patients participated in focus groups. Clinicians’ barriers included limited time for sessions and difficulties working with acutely psychotic patients. Patients’ barriers were burden of greater concentration when engaging with DIALOG+ and feeling tense or disturbed during the sessions. Facilitators included motivation to use DIALOG+, positive opinions shared by others, perceived benefits for practice and improving clinician-patient conversations, relationships and care.
Conclusions: Barriers to using psychosocial interventions could be overcome even if resources cannot be increased. Despite limited time and other barriers to using DIALOG+, perceived benefits to practice and clinician-patient relationships suggest that psychosocial interventions can use available resources effectively to improve care for psychosis.