The details of the study development, protocol, design and background have been published elsewhere21. An exploratory non-controlled multi-site study was conducted in each of Bosnia and Herzegovina, Colombia and Uganda, with a target enrolment of at least 30 patients in each country.
The NIHR-funded Global Health Group on developing psycho-social interventions in low- and middle-income countries was established in 2017 to explore the application of resource-oriented interventions to improve community mental health care in people with severe mental illness23. The group was originally a collaboration between Queen Mary University of London in the UK and three original partner countries – Bosnia and Herzegovina, Colombia, and Uganda. The group was later expanded to include additional studies in Argentina, Peru and Pakistan. This group was specifically testing three resource-oriented interventions: Multi-family groups, Volunteer befriending, and DIALOG. During this work, the investigators identified and discussed the unmet need for psychosocial support in order to improve primary care delivery to support good mental and physical health. Of the three interventions being assessed, DIALOG + was identified as being the most likely to be implemented within primary care as it leverages the routine meetings that were already happening and this intervention was specifically designed to make routine care meetings therapeutically effective, but this had not been tested in primary care settings.
Clinicians received training and manuals describing the details of the intervention24. Enrolled participants were offered monthly DIALOG + sessions at routine clinic visits, over a period of approximately three months. Feasibility was determined by the extent to which the intervention was implemented as planned, and the how well the study recruited and retained participants over the study period. For assessing outcomes, key outcome criteria were assessed at baseline and post-intervention and compared.
The study used a consistent core protocol which ensured comparable implementation across the three country sites. The protocol also provided some flexibility to ensure suitability to the local contexts. This resulted in small differences in the inclusion criteria of patients.
Patients and procedures
Patients with long-term illnesses such as hypertension, diabetes, cardiovascular disease and chronic obstructive pulmonary disease were targeted for inclusion in the study, as these are often associated with mental distress and can impact on the patient’s quality of life25.
The following inclusion criteria were used:
-
Adult patients (16–65 in Bosnia and Herzegovina and Colombia; 18 and above in Uganda)
-
At least one long-term physical condition
-
Poor quality of life (< 5 on the Manchester Short Assessment of Quality of Life (MANSA), < 5.5 in Colombia where previous work with patients indicated that they score slightly higher on the MANSA26)
-
Capacity to provide informed consent
-
Command of the local language
-
Living within 20 kilometres of the clinic
-
Having attended the primary care clinic for at least six months
Given the exploratory and pragmatic nature of the study with the aim of including as many eligible primary care patients as possible, being unable or unwilling to provide informed consent was the only exclusion criterion.
The clinicians were required to be a qualified health professional working in the relevant clinic with no plans to leave their post within four months of recruitment, with no other exclusion criterion.
Participants and Consent
The aim was to recruit at least 30 participants in each country, in line with central limit theorem which states at 30 is the minimum number of participants required to determine meaningful parameter estimates for exploratory studies27. All patients and clinicians provided written informed consent prior to any data collection. The process for documenting and the procedures for conducting informed consent, including of patients with low literacy, were reviewed and approved by the local ethics committee. Capacity to consent was assessed at screening. The University of California Brief Assessment of Capacity to Consent (UBACC28) was used in Uganda, where it had been previously used. An adapted capacity to consent checklist based on the one published by the British Psychological Society was used in Bosnia and Herzegovina and Colombia for those participants for whom it was felt might be lacking capacity to consent29. Patients were not reimbursed for attendance at routine meetings but were reimbursed for travelling to research interviews and for the time they spent in these interviews, in line with amounts reviewed and approved by the ethics committees in each country.
Time periods
After enrolment, the patients received DIALOG + at their routine clinic appointments, approximately monthly. These sessions were delivered by their healthcare worker, using the DIALOG + application on a tablet computer. The intervention period was approximately three months (with some flexibility around COVID-19 restrictions which delayed some clinic appointments), and patients received up to three DIALOG + sessions.
Socio-demographic information, including clinical characteristics for the patients, was collected at baseline. Outcomes were measured at baseline and after the intervention period.
COVID adjustments
At the time of planning for this study, it was not clear what the impact of ongoing and potentially changing pandemic restrictions would be21. It was anticipated, for example, that initially all sessions in Colombia would have been delivered remotely. As it happened, it was only necessary to do this for the three-month data follow-up sessions. In Bosnia and Herzegovina, recruitment of patients was delayed due to movement restrictions during partial lockdowns. In addition, recruited clinicians had increased workloads at certain times due to pandemics, so that some intervals between routine meetings were longer than the planned one-month period. There was no impact on activities due to COVID in Uganda.
Measures
The feasibility of the intervention was assessed as patient and clinician recruitment and attrition (with reasons for refusal and loss to follow-up) as well as the frequency and duration of sessions.
Three core feasibility criteria were pre-defined:
-
at least 75% of the anticipated 30 participants per country recruited, and,
-
at least 75% of approached and eligible participants enrolled in the study, and,
-
At least 75% retention in the study.
Completion of at least two thirds of the planned sessions was included as an additional feasibility criterion.
Socio-demographic information was captured on a questionnaire.
Outcome measures were:
-
Subjective quality of life was rated on the Manchester Short Assessment of Quality of Life (MANSA)30.
-
Symptoms of depression were assessed on the Patient Health Questionnaire (PHQ-8), a brief scale with good sensitivity and specificity based on the DSM-IV diagnostic criteria for depressive disorders 31.
-
Symptoms of anxiety were rated on the Generalised Anxiety Disorder Assessment (GAD-7) which has been used in primary care settings showing good validity and reliability32.
-
The objective social situation of patients was assessed using the Objective Social Outcomes Index (SIX)33.
Data Analysis
Quantitative data were analysed using IBM SPSS Statistics for Windows, Version 28.0.0.0 (190). For age, years old and range was reported. For socio-demographic and clinical characteristics, frequency counts and percentages were prepared. For quantitative outcomes, scores were calculated at baseline and for follow-up post intervention, and repeated measures t-tests were conducted with a p value of < 0.05 to assess changes in the outcomes. Effect sizes were calculated for all outcomes using Cohen’s effect size.
Settings
The study was conducted in sites in three partner countries, with different arrangement and management of primary care settings, as outlined in greater detail elsewhere21. The mental health measures in this study were not available to the treating clinicians and could not play any role in referrals for additional mental health care. No instances of patients being referred for mental health services was recorded in the study.
Bosnia and Herzegovina
The site in Bosnia and Herzegovina was in the Federation of Bosnia and Herzegovina, where health care provision is delivered through ten cantons each with their own health ministry with the Federal ministry of health providing a guidance role34,35. Within this system, there is a focus on continuity of care. Primary health care services include primary mental health care in the Community Mental Health Centres and family medicine. Outpatient clinics are situated in local communities in order to make sure that primary health care is accessible and available to all citizens. Patients were recruited from the Public Institution Health Care Centre of Sarajevo. The Public Institution of Health Care is the largest institution in the country offering primary health care36. The offices of the primary health care centre were used for delivery of the intervention, patient assessments, interviews and focus groups as well as training as they provided confidentiality and privacy.
Colombia
The study in Colombia was conducted in Bogotá with patients and clinicians recruited from the Javesalud Institución Prestadora de Servicios de Salud. Initially the plan was to recruit from both their Santa Beatriz and Toberin clinics, but the Toberin clinic shifted to treating only COVID patients and so all participants were recruited from Santa Beatriz. The Colombian healthcare system model places an emphasis on the role of primary care within the national public health strategy. Within the plans in place, the first route to access is provided through primary care, which aims to ensure timely access to medical care. Comorbidity is also managed through the primary care system as there is a shortage of specialists37. Another function is to identify the specific care required according to 16 different health risk groups, which includes mental health, violence-related and drug abuse related healthcare issues. However, there is limited capacity in primary care to manage long-term conditions in an integrated way, and expanding this capacity remains a priority.
Uganda
In Uganda primary care is delivered by both public and private sectors, with the public sector accounting for around 60% of care delivered38. National referral hospitals are the final points of the health services, but the specific referral pathway is often ignored due to a lack of clear gatekeeper control39. In practice this means that communities living near the referral hospitals use these as primary health care services, despite the established referral pathways. The study in Uganda was conducted in the outpatient clinics of Mityana district hospital and Masaka regional referral hospital. These outpatient clinics are for long-term physical illnesses, and are run by doctors, clinical officers and senior nurses.