Implementation and scale-up of integrated depression care in South Africa: An observational implementation research protocol

Abstract

Background:

People with chronic physical conditions who have comorbid depression experience poorer health outcomes. This problem has received scant attention in low-and middle-income countries (LMICs). The aim of this study is to refine and promote the scale-up of an evidence-based task-sharing collaborative care model for depression co-morbid with chronic disease in primary health care settings, called the Mental health INTegration (MhINT) programme, in South Africa.

Methods:

Adopting a learning health system approach, this study uses an in-site, iterative observational implementation science design. The first stage comprises assessment of the original MhINT model under real-world conditions in an urban sub-district in KwaZulu-Natal, South Africa, in order to inform refinement of the model and its implementation strategies. The second stage comprises assessment of the refined model across urban, peri-urban and rural contexts. In both stages, population-level effects are assessed using the RE-AIM (Reach-Effectiveness-Adoption-Implementation-Maintenance) evaluation framework, with various sources of data including secondary data collection and a patient cohort study (N=550). The Consolidated Framework for Implementation Research (CFIR) is used to understand contextual determinants of implementation success involving quantitative and qualitative interviews (Stage 1, N=78; Stage 2, N=282).

Results:

The study results will provide refined intervention components and implementation strategies to enable scale-up of the MhINT model for depression in South Africa.

Next Steps:

Strengthening ongoing engagements with policy makers and managers; providing technical support for implementation; and capacity building of policy makers and managers in implementation science to promote wider dissemination and sustainment of the intervention.

Editor’s Note:

In partnership with Milton L. Wainberg, M.D., Psychiatric Services is publishing protocols to address the gap between global mental health research and treatment. These protocols present large-scale, global mental health implementation studies soon to begin or under way. Taking an implementation science approach, the protocols describe key design and analytic choices for delivery of evidence-based practices to improve global mental health care. This series represents the best of our current science, and we hope these articles inform and inspire.

Background

South Africa faces a substantial burden of multimorbid chronic diseases, including HIV, cardiovascular disease, diabetes, and common mental disorders (CMDs) (depression, anxiety) and substance use disorders (1–4). Depression is of particular concern. In addition to being the most prevalent individual disorder in South Africa (5); people with chronic physical diseases are more likely to suffer from depression than the general population (6, 7); depression co-morbid with chronic physical conditions compromises patient self-care and adherence to physical disease treatment regimens (8–10); and there are poorer health outcomes and increased mortality in this patient population (11, 12).

In the context of a 75% treatment gap for CMDs in South Africa (13), the country’s national mental health policy emphasizes the integration of mental healthcare, including depression care, into primary healthcare (PHC) via task-sharing (14). This is in line with health systems reforms towards horizontal integrated programming to respond to the multimorbid burden of disease in South Africa (14).

PHC-based task-sharing of pharmacological and psychosocial treatment models for common mental disorders have proven effective in LMICs (15, 16). Further, integrated collaborative stepped-care that includes task sharing has been found to be efficient and effective for depression care. Most of this evidence comes from controlled trials in high- and middle-income countries (17–19). There are few evidence-based models of collaborative stepped care for integrated depression care in real-world, lower-resource contexts (20–22). There is also a dearth of evidence on implementation strategies that are enabling of uptake of integrated primary depression care within routine primary care in LMICs.

A collaborative task sharing model for integrated primary depression care was developed through the PRogramme for Improving Mental Health CarE (PRIME) in South Africa (PRIME-SA) (23) and its efficacy and effectiveness evaluated through i) a repeat cross-sectional survey in clinics and a comparison group cohort study, with good outcomes at the facility and patient level (24); and ii) a pragmatic cluster randomized control trial using study-employed lay counsellors (25), which found that the model did not produce inferior outcomes to care as usual where psychological treatment was provided by mental health specialists (26). However, uptake of the model - in particular, identification and referral of chronic care patients by PHC providers with comorbid depression - was poor. Following the translational research continuum (27), this evidence-based model forms the initial Mental health Integration (MhINT) model evaluated by this observational implementation science protocol to understand how to promote implementation and broader dissemination of the model. Funding obtained from the Centres for Disease Control (CDC) (U2GGH001197) was used to provide technical support to the Department of Health (DoH) in the KwaZulu-Natal (KZN) province of South Africa to scale-up this model in one district as a test site. The intervention components of the initial MhINT model (Figure A1 available online) involve the provision of psychoeducation, screening and strengthened clinician assessment and diagnosis of depressive symptoms at the PHC facility level. The latter is achieved through the provision of enhanced mental health training in the use of Adult Primary Care (APC) (known internationally as the Practical Approach to Care Kit – PACK)(28). APC is a clinical decision-making tool providing integrated evidence-based guidelines for the diagnosis and treatment of chronic conditions. It is intended for use by all health care practitioners working at primary care level and has been rolled out nationally by the South African National Department of Health. Referral pathways are also strengthened. Depending on depressive symptom severity, patients are referred to: i) Existing non-specialist clinic based staff trained to provide psychosocial counselling using cognitive behavioural techniques and problem solving, both shown to have an evidence-base for being delivered within non-specialized health care settings (15); ii) Existing clinic-based PHC doctors with strengthened training to initiate psychotropic medication; and iii) Existing mental health specialists at district level. In order to optimize scale-up of this task-sharing model by the DoH, an initial set of implementation strategies was chosen based on practical experience and process evaluation conducted during the formative and outcome evaluation of the PRIME-SA model (23, 24). Table 1 describes the roles of various health systems actors in implementing this initial model, including the initial set of implementation strategies employed by MhINT. We have aligned these to the implementation strategies specified by the Expert Recommendations for Implementing Change (ERIC) (29).

Table 1: Overview of MhINT Model Provider Roles, Technical Support and Implementation Strategies.

Provider, Roles and MhINT technical support provided
Provider	Role		MhINT technical support including orientation workshops, training and mentorship 1
Mental health coordinator/district mental health task team.	Provide overall coordination, monitoring and evaluation.		District mental health task team supported through two day workshop and mentoring support to undertake a situational analysis that informs a district mental health care plan that is incorporated into district plans.
Psychologist.	Provides training, supervision.		Psychologists orientated and trained through a 4-day workshop in their roles of providing training, supervision and emotional support to PHC level within a task sharing approach.
Registered Psychological Counsellors/ Social.	Training, supervision, and supporting Lay Counsellors.		Orientation and training of registered counsellors or equivalent through a 4-day workshop to train and supervise PHC facility-based non-specialist counsellors.
PHC coordinators & Operational managers	Support mental health integration using Continuous Quality Improvement (CQI)		Two-day training workshop in CQI tools with CQI mentor provided by MhINT providing mentorship for PHC coordinators and facility managers in CQI to support PHC facilities in integration of depression care using CQI.
Facility managers.	Oversee implementation and integration.		Orientation to responsibilities of different role players in collaborative care model through a half-day workshop; Capacitated in CQI for monitoring implementation and data management.
PHC Staff Nurses/ Enrolled Nurses.	Conduct initial mental health screening on the PHC facility population.		As per Department of Health (DoH) guidelines. MhINT did not initially provide technical support
PHC Clinical Nurse Practitioners (CNPs).	Identifies CMDs, provides brief intervention, referral and re-assessment.		Existing facility trainers capacitated through a 3-day workshop to provide onsite sessions orientating CNPs to person-centred care and their role of case managers within the collaborative care model; equipping them with clinical communication skills for person-centered care, use of Adult Primary Care (APC) for treatment and referral of CMDs.
PHC doctors.	Initiate medication, monitor psychotropic medication.		Orientated to collaborative care model, APC and capacitated in mhGAP guidelines through a half-day workshop.
Lay counsellors/Enrolled nurses.	Provide evidence-based counselling (CMDs and adherence).		Oriented to collaborative care model; capacitated in manualised, depression counselling package using problem solving and cognitive behavioural techniques through a 5-day workshop. The training is followed by individual in-vivo supervision and monthly emotional support by the psychological counsellors/social workers
Outreach team leaders (OTL) (PHC Clinical Nurse Practitioners/Enrolled nurses).	Supervision of CHWs; Home visits of difficult cases.		As per DoH guidelines. MhINT did not initia provide technical support
Community Health Workers (CHWs).	Case identification, Psychoeducation; Tracing & linkage to care.		As per DoH guidelines. MhINT did not initia provide technical support
Implementation Strategies
MhINT Strategy	ERIC Strategy	Level	Purpose
Situational analysis	Conduct local needs assessment	System	Inform development of district mental health plan
Train-the-trainer model for building capacity	Use train-the-trainer strategies	Facility	Efficiently train primary care providers
Supportive supervision	Audit and feedback; provide clinical supervision	Provider	Mentor providers, monitor competency, and offer emotional support
Adult Primary Care (APC) decision support tool	Remind clinicians	Provider	Promote nurse-led identification and management of patients with depression and other common mental disorders
Continuous quality improvement	Develop and organize quality monitoring systems	System	Identify implementation bottlenecks and propose solutions through learning collective

¹Details of the MhINT training and orientation workshops can be found in the training and orientation manuals on the Centre for Rural Health (CRH) website www.crh.ukzn.ac.za

We describe the Southern African Research Consortium for Mental health INTegration (SMhINT) evaluation protocol. Our aim is to observe real-world implementation of the initial MhINT model described above to identify opportunities for refinement and adaptation of this initial model to optimize widespread implementation and dissemination of the model in South Africa.

Methods

Learning Health Systems Approach

We are guided by a learning health systems approach, where policy makers, researchers, service providers, and patients work together as a collective to iteratively coproduce new knowledge and engage in shared decision-making in order to strengthen the health system and health outcomes (30). Critically, learning health systems strive to identify interventions and implementation strategies that work in routine contexts: generating new evidence through research when required, solving practical problems of service delivery and engaging in rigorous evaluation of program effectiveness to improve quality across the health system (30). This study builds on a sustained partnership between the investigators and the South African Departments of Health at the national, provincial, and district levels through the PRIME and MhINT programmes, collectively amounting to six years of collaboration.

Study Design

We will use an in-site, iterative, observational implementation science design (27) structured around the four phases of intervention scale-up set out by Barker et al. (31): i) An initial ‘set-up’ phase where the initial MhINT model was introduced into the system; ii) Early assessment and refinement of this initial model into a ‘scalable unit’ for inclusion in routine PHC services; iii) Assessment of the ‘scalable unit’ across a variety of contexts that would be encountered at scale; and iv) ‘Going to full scale’ where a larger number of sites adopt the refined MhINT model. The initial ‘set-up’ phase lasted from 2016–2018, involving the introduction of the MhINT model into the study site district. This protocol covers the second and third phases (renamed as Stages 1 and 2, respectively). Barker’s fourth phase is beyond the scope of this study.

We combine the Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM) framework (32) and Consolidated Framework for Implementation Science Research (CFIR) (33) with mixed methods data collection (34) illustrated in Figure 1. Within the context of our overall aim, the specific study objectives are described below:

Figure 1: Scale-up Evaluation Design.

Stage 1: Assessment of the Initial MhINT Model (urban setting)

Objective 1.1.

Assess the RE-AIM outcomes of the initial “set-up” MhINT model in one sub-district.

Objective 1.2.

Assess CFIR determinants of the dissemination, impact and sustainability of the initial “set-up” MhINT model across multiple domains.

Objective 1.3.

Engage in a participatory process with key stakeholders to recommendE refinements to strengthen the MhINT model and its implementation strategies into a “scalable unit”.

Stage 2: Assessment of the Scalable Unit Across Diverse Contexts (urban, peri-urban, rural settings)

Objective 2.1.

Assess the RE-AIM outcomes of the strengthened MhINT ‘scalable unit’ across diverse contexts (urban, peri-urban and rural)

Objective 2.2.

Assess CIFR determinants of the dissemination, impact and sustainability of the “scalable unit” MhINT model across diverse contexts (urban, peri-urban, rural) to inform further adaptations for going to “full-scale”.

Objective 2.3.

Engage in a participatory process with key stakeholders to recommend refinements to strengthen the MhINT model and its implementation strategies into a “full scale” model for wider scale-up.

RE-AIM outlines the critical elements of the population-level effects of health interventions (32). Definitions of the RE-AIM elements are provided in Table 2. Semi-structured interviews targeting CFIR domains will be used to explore the implementation component of RE-AIM. CFIR delineates constructs associated with implementation success across several domains, including i) intervention characteristics like goodness-of-fit; ii) inner setting or health system characteristics that may aid or abet integration, including organizational culture; iii) characteristics of the outer setting or external environment, such as policies and community needs; iv) characteristics of individuals, including patient and provider needs; and v) implementation processes, such as harnessing of support from key stakeholders to improve uptake (33, 35).

Table 2: Characteristics of Amajuba sub-districts: Newcastle, Emadlangeni and Dannhauser.

	Newcastle [urban]	Dannhauser [periurban]	Emadlangeni [rural]
Land area	1 855 km2	1 516 km2	3 539 km2
Population	363 236	105 341	36 869
Poverty rates	56.3%	78.6%	80.7%
Number of households	90 347	20 844	6 667
Income
No income	28%	83%	34%
Health resources
Hospitals	3	0	1
Community health centres	0	1	0
PHC facilities	14	10	2
WBPHCOTs	5	5	2 (1 per PHC)
Mobile points	12	36	79
Mental health specialists
Psychologists	2	0	0
Sessional Psychiatrist	1	0	0

Study Sites

The study site is the Amajuba District, in the northwest of KZN, where the MhINT model was implemented. It has three diverse sub-districts: urban (Newcastle); peri-urban (Dannhauser); and rural (eMadlangeni) (see Table 2). The assessment of the initial MhINT “set-up” model (Objectives 1.1 and 1.2) will be undertaken in 10 of the 14 PHC facilities in the urban Newcastle sub-district. Assessment of the refined MhINT “scalable unit” model (Objectives 2.1 and 2.2), is planned to take place across all three sub-districts; providing the opportunity to observe how the refined model and implementation strategies performs in the initial urban site (Newcastle), as well as whether adaptations are needed for peri-urban and rural contexts. All PHC facilities are serviced by full-time PHC nurses comprising Professional Nurses with a 4-year degree/diploma, Enrolled Nurses who have a two-year diploma, sessional PHC doctors, and HIV counsellors. Limited mental health specialists (n=3) are available at referral facilities.

This study received ethical approval from the University of KwaZulu-Natal Biomedical Research Ethics Committee, Reference BF190/17.

Study Procedures

Objectives 1.1 and 2.1: RE-AIM assessment:

In order to assess the RE-AIM outcomes, a number of data sources were used with accompanying sub-protocols for the collection of these data which are summarized in Table 3. We provide more details on the sub-protocols below.

Table 3:

Key Variables and Data Sources for Objectives 1.1, 1.2. 2.1 and 2.2

Key Variable	Source	Sub-Protocol	Frequency
Reach (Individual-level analyses of the proportion and characteristics of the target population that received the intervention along the cascade of care)
% of loss to follow-up (LTF) patients screened for CMDs at community level *	Clinic records	Secondary data	Quarterly
Characteristics of (LTF) chronic care patients screened and not screened at community level	Clinic records	Secondary data	Quarterly
% of chronic care patients screened for CMDs at facility level	District Health Information System (DHIS) data	Secondary data	Quarterly
Characteristics of chronic care patients screened and not screened at facility level	Patient cohorts	Cohort	Quarterly
% of positive chronic care patients screened who screen positive for depression	Cohort data	Cohort	Once-off
Characteristics of chronic care patients who screened positive and negative*	Cohort data	Cohort	Once-off
% of chronic care patients screening positive who are diagnosed and referred	Cohort data	Cohort	Once-off
Characteristics of chronic care patients screening positive who are diagnosed and referred	Cohort data	Cohort	Once-off
% of chronic care patients referred for counselling who receive at least one counselling session	Cohort data, project records	Cohort, Secondary data	Once-off; Quarterly
Characteristics of referred patients receiving one or more counselling sessions/not receiving any sessions	Cohort data; CFIR interviews with patients receiving one or more counselling sessions and those receiving no sessions.	Cohort; Qualitative data	Once-off
% of clinic population receiving mental health treatment initiation	DHIS data	Secondary	Monthly
Effectiveness (Real-world effectiveness on patient-level outcomes)
Depressive symptoms	Patient cohorts	3-month cohort data	Once-off
Disability	Patient cohorts	3-month cohort data	Once-off
Adherence to Prescribed Medications	Patient cohorts	3-month cohort data	Once-off
Perceived Stress	Patient cohorts	3-month cohort data	Once-off
Adoption (Organizational-level outcome referring to the proportion and characteristics of settings/service providers that adopt the intervention)
Facility level rate of morning talk	Facility records	Secondary data	Monthly
Characteristics of facilities with	Facility profiles; ORIC; MICA;	Secondary data;	Annually; Once-
greater/fewer morning talks on CMDs per month.	CFIR interviews with facility managers and counsellors/health promoters.	Quantitative data; Qualitative data	off.
Facility level rate of screening of chronic care patients for CMDs	DHIS data	Secondary data	Monthly
Characteristics of facilities	Facility profiles; ORIC; MICA;	Secondary data;	Annually; Once
achieving Department of Health screening targets (35% of headcount/less than 35% of headcount	CFIR interviews with facility managers and enrolled nurses	Quantitative data; Qualitative data	off
Facility level rate of diagnosis and referral of screen positive chronic care patients for depression	Cohort	Cohort data	Once-off
Characteristics of facilities with	Facility profiles; ORIC; MICA;	Secondary data;	Annually; Once
higher/lower rate of diagnosis and referral of screen positive chronic care patients for depression	CFIR interviews with facility managers and PHC nurses	Quantitative data; Qualitative data	off
Facility level rate of referred patients’ uptake of one or more counselling sessions	MhINT Project records	Secondary data	Quarterly
Characteristics of facilities with	Facility profiles; ORIC; MICA;	Secondary data;	Annually; Once
higher/lower referred patient uptake of one or more counselling sessions	CFIR interviews with facility managers, counsellors	Quantitative data; Qualitative data	off
Facility rate of mental health treatment initiation over time	DHIS data	Secondary	Monthly
% of providers who diagnosed one or more CMD patients	MhINT project records	Secondary data	Monthly
Characteristics of providers who refer/don’t refer	CFIR interviews with nurses who refer/don’t refer	Qualitative data	Once off
Implementation (Extent to which the intervention was implemented with consistency and fidelity along the cascade of care, adaptations made during the study, and cost)
Consistency of morning talks over time per facility	Facility records; CFIR interviews with facility managers and counsellors/health promoters;	Secondary data; Qualitative data	Quarterly; Once-off
Quality and consistency of screening over time per facility	Facility records; CFIR interviews with facility managers and enrolled nurses;	Secondary data; Qualitative data	Quarterly; Once-off
Quality & consistency of diagnosis & referrals per facility	Cohort; MhINT project records over time; CFIR interviews with facility managers and PHC nurses.	Cohort; Secondary data; Qualitative data	Quarterly;;;Once- off
Fidelity of counselling intervention	Fidelity checklists; CFIR interviews with clinic counsellors.	MhINT project data; Qualitative data	Once-off
Cost of intervention*	Costing analysis	Costing	Quarterly
Adaptations to intervention	Project CQI records; CFIR interviews with facility managers	Secondary data; Qualitative data	Quarterly; Once-off
Maintenance (Organizational-level institutionalization of the programme over time, as well as the individual-level sustainability of health outcomes)
Stability of effects of the intervention on patient level outcomes of effectiveness over time	Patient Cohort	3 and 9-month cohort data	Once-off
Characteristics of patients who	Patient Cohort; CFIR interviews	3 and 9-month	Once-off
had stability of effects over time	with patients who maintained	cohort data;
and those that didn’t	stability of effects over time and those that relapsed.	Qualitative data
Institutionalization of intervention	Audit of routine use of MhINT tools, processes and training materials at district level; CFIR interviews with district managers.	Secondary data; Qualitative data.	Once-off

^*Data only collected in step 3 when revised ‘scale-up unit’ is assessed

Sub-protocols

Secondary data on project implementation, including project records and routinely collected data, will be gathered from all participating clinics from the start to the end of the cohort studies in Stages 1 and 2. Routine service delivery data on screening and treatment rates will be extracted from the provincial District Health Information System (DHIS). Data that are only collected by facilities (e.g., tracking delivery of some implementation strategies, including morning talks) will be extracted by fieldworkers. MhINT project monitoring data will include counselling intervention fidelity using fidelity checklists and records from supervisory visits; referral rates will be extracted from referral forms; facility and staff characteristics gleaned from facility profiles; patient counselling uptake assessed from patient tracking forms; implementation of training packages and use of standardized operating procedures extracted from training records and continuous quality improvement (CQI) reports.

The cohort study will estimate the effects of nurse depression diagnosis, as well as referral for depression management, on participants’ depression outcomes, treatment adherence, stress, disability, income and employment.

Enrolment of eligible participants for the cohorts will be carried out by trained research workers, using PHQ-9, before consultation with a nurse and independently of possible diagnosis of depression by professional nurses using APC guidelines, and who will not know participants’ PHQ-9 scores. Whether each participant has been diagnosed as having depression by a nurse is determined by an exit interview conducted after the participant’s consultation with a nurse. Participants will be divided into three groups: patients diagnosed and referred for care by the nurse; patients diagnosed and not referred; and patients not diagnosed. Data will be collected at baseline at their first point of contact with the program, at 3 months, as well as≥ 9 months follow-up to ascertain stability of effects.

A target sample size of 550 participants is set for the cohort study (200 not diagnosed, 150 diagnosed and not referred, and 200 diagnosed and referred). This sample is estimated to provide 86% power to detect a difference in mean PHQ-9 score between the diagnosed and referred and undiagnosed groups of 2.0 (5.3 vs 7.3) points, with 5% significance, assuming SD=5.1, intraclass correlation of 0.1, and 20% loss to follow-up (mean=7.3 and SD=5.1 are from 6- month follow-up in the PRIME trial). The sample size of 200 participants in the “diagnosed and referred” group and 150 in the “diagnosed but not referred group” will provide 81% power to detect a difference in mean PHQ-9 score between them of 2.0 (5.3 vs 7.3) units, with 5% significance, also assuming SD=5.1, intraclass correlation of 0.1, and 20% loss to follow-up. Numbers of patients recruited in each clinic may vary and depend on the numbers consecutively enrolled to meet the total sample sizes required. Inclusion criteria are: adults 18 years and above; time and ability to complete the full interview; willingness to provide informed consent and PHQ9≥9. Exclusion criteria are: PHQ9<9; inability to provide informed consent (e.g. presence of severe intellectual disability, currently experiencing an acute medical issue, or lack of private space for the interview). Fieldworkers will be trained to make these assessments. Patients with severe depression and end-of-life thoughts 7 days or more in the last 2 weeks will be accompanied by project staff to clinic professional nurses for care and instructed not to leave the patient unattended until the patient has been attended to.

Following provision of written informed consent procedures, participants will be interviewed in the local language (isiZulu) or English. Data will be collected using digital handheld devices and uploaded onto a secure server at the University of KwaZulu-Natal. Data quality assurance will be managed through daily quality checks of uploaded data and “in vivo” observations of interviews to ensure fidelity of questionnaire administration. The research team will actively contact all participants enrolled at baseline for follow-up (irrespective of whether they refused/discontinued recommended clinical care) using telephoning, home visits etc.

Socio-demographic variables, including questions on gender, age, language, ethnic group, educational and employment status, income and household composition will be recorded at baseline. Employment and income questions will be repeated at 3 and 9-month interviews to track whether the intervention impacts on change in economic status. Depression symptoms will be measured with the Patient Health Questionnaire (PHQ)-9, which is aligned with the Diagnostic and Statistical Manual 4 text revision (DSM-IV-TR) diagnostic criteria for major depressive disorder (36). It has been validated for diagnosis of depression with a cut off score of ≥9 in the chronic care population in South Africa (37). The PHQ-9 score will be the primary outcome of the cohort study. General disability will be measured with the 12-item World Health Organisation Disability Assessment Scale (WHODAS) 2 version that has been used across cultures, including in South Africa (38). Treatment adherence to prescribed medications will be measured with the Visual Analogue Scale (VAS) (39), previously used in South Africa (40) and other resource-limited settings, and is a cost effective alternative to measuring medication levels from biological samples (41, 42). The Perceived Stress Measure is a 10-item self-report measure that provides an assessment of the extent to which situations in a person’s life are appraised as stressful and has been used previously in South Africa (43).

The primary analyses will compare the mean PHQ-9 scores measured at 3 months, and the proportions of participants whose PHQ-9 scores decreased by at least 50% from baseline to 3 months, between the diagnosed but not referred group, the diagnosed and referred group, and the undiagnosed group, adjusted for propensity scores using the “t-effects” package in Stata version 16. The propensity scores will be the predicted probabilities of being diagnosed with depression and referred, respectively, at the baseline visit, from logistic regression models with baseline PHQ-9 scores, age, sex and socio-economic indicators as explanatory variables. These propensity scores will also be used to compare PHQ-9 scores, and all other outcomes at 3 as well as 9 months. The comparisons between groups will account for intraclass correlation of outcomes between clinics by using robust adjustment. We will also use multilevel mixed models to analyze pooled panel data from baseline, 3 and 9 months, with group, time and socio-demographic variables as covariates, and with patient and clinic as random effects. Finally, linear and logistic regression analyses will investigate additional predictors, moderators, and mediating factors of the outcomes (44).

A costing analysis will be performed in Stage 2 of the study for each sub-district context. We will estimate i) the incremental cost per outcome (change in PHQ-9 score, adherence, etc.); ii) cost to scale-up; and iii) perform a budget impact analysis (BIA) from the programmatic/payer perspective (i.e. DoH). We will estimate economic costs in this analysis, including actual financial outlays and costs of donated time and volunteer time (See online supplementary Table A1).

We will estimate the incremental costs of implementing the intervention compared to the current standard for the sub-district contexts. Intervention costs will include costs associated with start-up, personnel, transport, communication, consumables, and overhead costs. Cost data will be collected from the study budget, clinic expense reports, published information on labor costs, and staff interviews. To assess staff time spent on the intervention, we will conduct time and motion studies in all sub-districts while the intervention is running at full capacity. Together the micro-costing data, time and motion studies, and clinical outcomes will be used to estimate the average cost of providing services.

The costing results will be reported separately for each sub-district. For all key inputs and outputs, we will follow standard guidelines of the Second Panel of Cost-Effectiveness in Health and Medicine (45). We will conduct sensitivity analyses around key cost inputs to account for the uncertainty in our results. For budget impact analysis (BIA), we will consider direct program costs in the different sub-district contexts. Direct medical costs will be measured to ensure that DoH costs reflect the opportunity cost of the resources used in delivering services. Further, the “top-down” approach of expense report collection will be compared with the “bottom-up” micro-costing approach to triangulate and refine our cost estimates.

Objectives 1.2 and 2.2: CFIR interviews

In both stages, completion of data collection of the RE-AIM framework over a 12-month period will be followed by qualitative and quantitative interviews targeting CFIR domains to understand the implementation component of RE-AIM at both stages. Structured quantitative interviews will comprise: i) The Organizational Readiness for Implementing Change (ORIC) which is a 12-item measure based on Weiner’s theory of organizational readiness for change measuring change commitment and change efficacy (46); and ii) The Mental Illness: Clinicians’ Attitudes (MICA) scale which is a 16-item mental health-related stigma scale (47). Interview guides will be translated into isiZulu, and back-translated to ensure accuracy. All facility managers and service providers from the facilities across the sub-districts will be requested to complete the quantitative interviews. Semi-structured qualitative interviews will be conducted with facility-level managers, district managers, service providers, and patients (Stage 1, N=78; Stage 2, N=282) by trained interviewers (see online supplementary Table A2 for breakdown of sample sizes). With the exception of the counsellors (all of whom will be interviewed), participants will be sampled purposively according to the following strata. Facility-level managers according to high- and low-performing clinics; individual professional nurses according to high- and low-diagnoses rates; and patients according to high and low-uptake rates of counselling sessions. Interviews will be audiotaped and conducted in isiZulu or English, depending on participant preference following informed consent procedures. Audiotaped interviews will be translated where necessary from isiZulu into English and transcribed, with back-translation checks by multilingual members of the research team. Any discrepancies will be reconsidered in consultation with a third multilingual researcher (48). Transcripts will be analyzed using framework analysis typically used in health policy research. A thematic framework will first be developed using the CFIR domains. Transcript content will then be coded using the thematic framework whilst allowing for inductive themes to emerge. Cases will be grouped by the aforementioned participant strata to identify multi-level barriers and enabling factors (49).

Objectives 1.3 and 2.3: Participatory concept mapping

Purposively selected key stakeholders will be engaged in a participatory concept mapping exercise to reflect on the results of objectives 1.1/1.2 and 2.1/2.2. Collaborative learning sessions – specifically, process mapping and nominal group techniques of developing priorities – will be used to brainstorm potential strategies to overcome identified barriers and optimize facilitating factors. Identified strategies will be sorted and rated according to perceived importance and feasibility on the part of the participants. Approximately 22 stakeholders will be involved in workshops at each stage, grouped according to their designation of managers, providers and patients (see online supplementary Table A3). We will include a national representative responsible for community mental health services; all members of the provincial mental health directorate; key district managers responsible for primary health care and mental health; operational managers from all the participating facilities; a spread of high- and low-referring nurses purposively selected; as well OTLs/CHWs purposively selected on the basis of high and low number of referrals from the community level. Patients will also be purposively selected to provide a spread of those with high and low uptake of the counselling intervention. Informed by this process, the initial MhINT model will be refined in Stage 1 to develop a MhINT ‘scalable unit’ that will then be assessed across the diverse sub-districts in Stage 2, with sub-group analysis for each site using the methods outlined above using the RE-AIM framework and CFIR interviews. The same process will be repeated in Stage 2 to further refine the model for broader scale-up.

Results

Given that health systems strengthening requires adaptability, collaborative mechanisms, and routine data gathering (50), the adoption of a learning health systems approach reflected in the two stage design will facilitate the iterative examination of findings of this observational research design in collaboration with patients, providers, health managers, and researchers. Shared decisions on how to address bottlenecks to uptake of the initial “set-up” MhINT model identified in stage one, will lead to the co-production of novel and localized interventions and strategies for optimizing implementation and scale-up (the “scalable unit”), that will be tested across three diverse contexts (urban, peri-urban and rural), in stage two. Assessment of the strengthened model in these three different contexts with varying strengths and limitations will allow us to identify distinct implementation strategies needed to strengthen integration in urban, semi-urban and rural contexts. The findings should be helpful for strengthening the MhINT model for going to “full-scale”, with interventions and implementation strategies to assist with further scale-up across these typical diverse contexts by the South African government. Internationally, the findings will build on emerging lessons from previous research that suggests, among other things, the need for a whole-system strengthening approach to integrating mental health care in primary health care (51). It will also provide knowledge on strategies to optimize scale-up of collaborative care for integrated depression care comorbid with chronic physical conditions in other scarce-resource contexts.

Discussion and next steps

There is increasing recognition of i) the need for systems change to enable the integration of mental health, inclusive of depression, into primary health care (52); ii) the primacy of open and flexible models that allow for cross-context adaptations (53); and iii) the paucity of studies testing optimal implementation strategies to promote scale-up of integrated depression care in routine primary care in LMICs (22). Against this backdrop, while the results of this study protocol are still to be seen, the design used may be helpful for other implementation research collaborations that aim to promote uptake of integrated depression care into routine primary care services. Noteworthy is the iterative learning health systems approach - important for ensuring the uptake and sustainability of interventions and implementation strategies (54).

This approach, does, however, denote less control over the research process compared to controlled implementation research designs, as well as requiring greater investment of researchers’ time and resources in negotiations and capacity building activities (55). Further, in light of a higher probability of sustainability of mental health integration in the presence of systems reforms that are enabling of chronic care (51), investment of time and resources into such systems strengthening initiatives may also be necessary to create this enabling context. Next steps for the researchers engaged in the SMhINT study include sustaining ongoing engagements with policy makers, planners and managers; providing technical support for health systems strengthening to facilitate implementation; and building capacity of policy makers and managers in implementation science to promote wider dissemination and sustainment of the intervention.

Highlights:

This study protocol uses an in-site, iterative, observational implementation research design within a learning health system approach to assess the real-world implementation of an evidence-based collaborative model for integrated primary depression care.

Potential refinement of each component of the model at multiple levels (patients, providers, system) is evaluated simultaneously, in contrast to evaluating the contribution of a fixed implementation strategy on a set of predetermined outcomes.

The learning health systems approach promotes the alignment and response of the model to the priorities and practicalities of implementation across diverse, real-world settings, which should optimize implementation and wider dissemination.

Box 1: Key Challenges, Opportunities and Design Solutions.

Key Challenges

• Prior research suggested poor uptake into routine services of the initial MhINT evidence-based collaborative care package for integrated depression care being evaluated by SMhINT.

• In particular, identification and referral of comorbid depression in primary care patients was poor - this being the first step in the treatment cascade.

• This highlighted the need for implementation science to understand how to improve uptake and embedding of the package in routine care.

Key Advantages

• Heightened public health priority of integrated depression treatment in patients with multimorbid chronic diseases in South Africa given evidence of poorer health outcomes in chronic patients with comorbid depression.

• A policy window opportunity for integration afforded by health systems reforms towards integrated horizontal programming in South Africa.

• Funding for technical support for the scale-up of the initial MhINT package in one district – providing the opportunity to iteratively evaluate how to refine the package and implementation strategies for broader scale-up.

Design Solutions

• The adoption of a learning health system approach and iterative two-stage research design across varying contexts will enable the research team to be responsive to context and Department of Health needs.

• The strong collaborative relationship with the Department of Health will enable key learnings to be translated into policy changes necessary for institutionalization and broader scale-up.

• Locating the SMhINT study within broader efforts to support the Department of Health in health systems reforms towards horizontal programming will enable embedding of integrated depression care within these reforms from the outset.