Lessons from the Emerald cross-country study

A key output of the Emerald study were lessons from scale-up attempts of evidence-based integrated mental health care packages in six LMICs. A key finding was that an add-on training approach, which has characterised much of the implementation of the WHO Mental Health Gap Action Programme (mhGAP), was insufficient to sustain integration, with the need for system strengthening interventions to enable integration emphasised (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). More specifically, this study reinforced the need for strengthening of the basic building blocks of the health care system for chronic care to provide an enabling and supportive system and environment for integrated mental health care. System reforms to support chronic care are necessary to respond to the global changing disease burden from predominantly acute conditions to chronic ones that often include comorbid mental health conditions. Based on a review of the available evidence, Thornicroft et al. suggest that the World Health Organization's Innovative Care for Chronic Care Framework (ICCCF) provides a useful framework to this end (Thornicroft et al., Reference Thornicroft, Ahuja, Barber, Chisholm, Collins, Docrat, Fairall, Lempp, Niaz, Ngo, Patel, Petersen, Prince, Semrau, Unutzer, Yueqin and Zhang2019). The ICCCF identifies three levels of the health system in a socioecological fashion: the policy and population level, meso or district/sub-district level, as well as the family and community level. At the policy and population level, policies and regulations that are health-promoting, for example, regulations relating to alcohol consumption and which may need to be implemented by other sectors, are important, as are governance issues, particularly of an inter-sectoral nature (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). At the meso-level, governance remains important, with frontline managers being responsible for translating and adapting programmes within organisational and systems governance architecture to customise and embed them within the system (Scott et al., Reference Scott, Schaay, Olckers, Nqana, Lehmann and Gilson2014). To this end, continuous quality improvement (CQI) was identified as an important vehicle for embedding implementation strategies in the Emerald study (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). In relation to human resources, person-centred care, which is subsumed under people-centred care, is central to care provision at the meso-level. It focuses on the totality of the person from a holistic biopsychosocial perspective; with clinical communication skills a helpful tool for understanding a person holistically (Stuart et al., Reference Stuart, Brown, Donner, McWhinney, Oater, Weston, Jordan and Tiopcco2000). This includes the identification of common mental disorders (CMDs) that may accompany patients' presenting physical complaints in PHC settings, as found in the Emerald study (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). Additionally, at the meso-level, clinical decision support tools are highlighted as helping to support providers with clinical expertise and skills to deliver the best possible care according to the latest evidence. In the Emerald study, integrated chronic care guidelines that included mental health were found helpful to embed task-sharing of mental health with primary care providers (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). Further, integrated clinical information systems that are central to capture information on multiple conditions, so as to optimise multimorbidity care as well as facilitate continuity of care necessary for long-term conditions are important; as is delivery system re-design to promote collaborative team-based care which has been found to be effective for PHC mental health-care integration in high-income countries (Archer et al., Reference Archer, Bower, Gilbody, Lovell, Richards, Gask, Dickens and Coventry2012). Being able to leverage existing established referral pathways and collaborations between levels of care and different sectors necessary for balanced care, as well as systems for dispensing chronic medication and information systems for tracing lost to care patients so as to reduce non-adherence and the chance of relapse, were all found helpful for facilitating uptake and embedding of the mental health integration strategies in the Emerald study (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). At the micro-level, achieving activated empowered and informed people who have control over their health and living with their condition over the long-term emerged as an important issue in the Emerald study and is central and in step with a people-centred approach. Micro-level interventions, however, require a robust community sub-system that includes community health workers, who interface with the formal health system, as well as households, community structures such as faith-based organisations, civic groups and local political structures, and other sectors providing services at a community level (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b). Establishing community supports such as peer-led support groups as well as mechanisms for tracing and tracking patient progress is emphasised. Promotion and prevention interventions are also key at this level, with people having control of their health and that of their communities ultimately being the hallmark of people-centredness. The Emerald study found existing community health worker outreach teams provided an important platform that could be leveraged for task sharing of psychosocial interventions and psychoeducation to improve demand for services as well as for tracing of non-adherent patients (Petersen et al., Reference Petersen, van Rensburg, Kigozi, Semrau, Hanlon, Fekadu, Gureje, Gurung, Jordans, Mntambo, Mugisha, Muke, Petrus, Shidhaye, Ssebunnya, Tekola, Upadhaya, Patel, Lund and Thornicroft2019b).

Exemplar of optimizing health system reform policy windows: the case of South Africa

Recent calls for health reforms in LMICs towards horizontal chronic people-centred health systems (Agyepong et al., Reference Agyepong, Sewankambo, Binagwaho, Coll-Seck, Corrah, Ezeh, Fekadu, Kilonzo, Lamptey, Masiye, Mayosi, Mboup, Muyembe, Pate, Sidibe, Simons, Tlou, Gheorghe, Legido-Quigley, McManus, Ng, O'Leary, Enoch, Kassebaum and Piot2018; Thornicroft et al., Reference Thornicroft, Ahuja, Barber, Chisholm, Collins, Docrat, Fairall, Lempp, Niaz, Ngo, Patel, Petersen, Prince, Semrau, Unutzer, Yueqin and Zhang2019) provide a policy window to embed mental health as part of routine comprehensive primary health care so as to become part of reformed service delivery models from the outset rather than being seen as add-ons (Kingdon, Reference Kingdon2010). We demonstrate how such health reforms in South Africa (Pillay and Baron, Reference Pillay and Baron2011; Mahomed and Asmall, Reference Mahomed and Asmall2015) have provided opportunities to integrate mental health into the re-organisation of services from inception in the KwaZulu-Natal (KZN) province. An integrated collaborative care package for CMDs originally developed and tested through the Programme for Improving Mental Health Care (PRIME) research collaboration (Petersen et al., Reference Petersen, Bhana, Fairall, Selohilwe, Kathree, Baron, Rathod and Lund2019a) has been re-packaged for scale-up by the Mental health Integration Programme (MhINT) (https://www.crh.ukzn.ac.za). Implementing and scaling up of this package is being achieved through a sustained partnership between MhINT researchers and provincial and district Department of Health managers who are intimately involved in implementation strategy formulation along with providing essential bureaucratic facilitation for integrated care to become part of public sector governance and administration.

In line with the recommendation from Emerald, the most significant adaptation of the PRIME intervention through MhINT was the incorporation of CQI methodology as a mechanism for the health system to integrate mental health care into routine care. This was with the acknowledgement that whilst the collaborative care package for CMDs was developed for improving mental health care services in PHC, integration of mental health services overall is a system-wide endeavour. As such, MhINT utilised CQI not only as a vehicle for implementation but also for all the preparation activities at the different levels of the health system.

At the Provincial policy level, the MhINT programme provided strategic support necessary for translation of mental health care policies into implementable actions at the meso and community level. This included (i) supporting the integration of mental health strategic plans into the broader provincial annual performance plans in order to ensure allocation of finances to support the project of integration and facilitate person-centric approach to health service planning; (ii) reviewing and revising mental health data elements available in the clinical health information systems to facilitate improved monitoring of mental health services at primary health care; (iii) supporting the development of standardised operating procedures for mental health screening at PHC level; and (iv) putting in place a national and provincial steering committee to routinely review bottlenecks and innovations in implementation at the meso- and micro-levels (see Fig. 1).

Fig. 1. Integration of mental health into existing platforms at meso- and micro-levels in the KwaZulu-Natal province, South Africa.

At the meso-level, the MhINT programme has had to be responsive to the unique contextual factors that support or hinder the integration of mental health care using CQI strategies and tools to develop innovative solutions for the latter. To overcome the challenge of the integration of mental health into PHC being seen as additional work to existing responsibilities, MhINT aligned its activities with the Integrated Clinical Services Management (ICSM) model of managed care for chronic conditions. This has been done through strengthening the existing PHC process of clinical management and (i) locating mental health screening in the same space where all screening for vital signs are conducted. MhINT further strengthened the screening process by introducing a validated brief mental health (BMH) screening tool and collaboratively developing processes to facilitate referral of patients in need of emergency mental health care or further assessment at PHC; (ii) enhancing clinicians' training on existing clinical decision-making tools for managing chronic diseases [integrated chronic care guidelines called Adult Primary Care (APC)] (Fairall et al., Reference Fairall, Bateman, Cornick, Faris, Timmerman, Folb, Bachmann, Zwarenstein and Smith2015), so that these guides can be used to diagnose and manage CMDs; (iii) strengthening referral pathways by training identified PHC-based staff to provide a structured psycho-social counselling intervention to ensure a mental health service in the PHC for CMDs. Furthermore, MhINT has provided a mentorship and supervision model to different cadres at the different levels of the healthcare system to foster the spirit of collaboration and strengthen existing referral pathways. For instance, district psychologists have been equipped with skills to train and support mid-level mental health care workers in the system such as social workers, who in turn support and supervise the PHC-based staff who provide the psycho-social intervention at PHC level to patients with CMDs (which to date have included HIV counsellors and enrolled nurses). Additionally, PHC supervisors have been provided training and mentorship in CQI so that they are able to facilitate the development of change ideas in response to implementation bottlenecks and facilitate the embedding of the intervention.

At the micro-level, the MhINT programme has further supported the integration of mental health into routine outreach team activities by developing a training curriculum to enhance mental health literacy and developing an accessible psychoeducation tool which is also used to detect community members who require mental health services.

Through the CQI process, lessons from implementation at the micro- and meso-levels have informed the refinement of the KZN Mental Health Integration strategy resulting in several actions that promote the institutionalisation of the MhINT model. These include (i) inclusion of the MhINT psycho-social intervention training in a pre-service curriculum of auxiliary social workers in KZN who will be based at PHC facilities, so that there is a larger dedicated pool of healthcare workers that can provide psychosocial interventions for CMDs at PHC level; (ii) clinical communication skills have been integrated into the diagnosis and management of mental health problems in the APC guide from outset; (iii) mental health psychoeducation and detection of mental health problems at the community level has been included into the community health worker curriculum from the outset; (iv) the province has adopted the validated BMH screening tool to be used by enrolled nurses in the vital signs station across the province.

Importantly, the structures and relationships fostered by MhINT have also been leveraged to strengthen the health system for tuberculosis (TB) patients who are particularly vulnerable to CMDs. Towards this end, a work package of the National Institute for Health Research (NIHR) Global Health Research Unit on Health System Strengthening in Sub-Saharan Africa (ASSET) (NIHR, 2019) for strengthening people-centred TB-care is presently being layered into the health system in the same district site as the MhINT programme. Taking an embedded participatory approach, this work package aims to develop person-centred TB care that can provide services and support to people with TB according to their individual, family and community needs, ultimately to reduce a TB mortality rate that is almost double that of the national average. By adapting critical TB programme dimensions such as screening and diagnosis to be less judgemental and stigmatising, while also considering the complexities of multimorbidities (especially with depression and HIV), this ASSET work package is generating bottom-up solutions to strengthen the district health system for improved people-centred care of people burdened by mental–physical multimorbidities. In a practical sense, this requires constant iterative cycles of planning and implementation, driven by a collaborative team consisting of researchers and government and community stakeholders, as well as programmatic flexibility to allow for action-oriented intervention development, testing and evaluation.

Implications for the global mental health scale-up agenda

These lessons from the ground have implications for the way in which we, as the global mental health community, approach scaling up of increased access to mental health care in scarce-resource contexts.

In the first instance, they sound a cautionary note to the call for global mental health financing similar to that invested in eradicating HIV and other epidemics (Kleinman et al., Reference Kleinman, Estrin, Usmani, Chisholm, Marquez, Evans and Saxena2016; Vigo et al., Reference Vigo, Patel, Becker, Bloom, Yip, Raviola, Saxena and Kleinman2019). Such global health initiatives in the past – especially in African countries – have operated vertically, bypassed country systems, influenced policy to suit multilateral agendas and have not sufficiently harmonised with other local initiatives (Mwisongo and Nabyonga-Orem, Reference Mwisongo and Nabyonga-Orem2016); resulting in untended consequences on health systems in the recipient countries. For example, HIV donor funding has been found to ‘crowd out’ other health issues – especially in contexts where there are limited human resources for health and where providers' attention becomes focused solely on the targets and indicators of the vertical donor funded priority programme (Grepin, Reference Grepin2012). This has often been to the detriment of other programmes and sustainability. This approach in the global mental health sphere may thus unintentionally entrench the perception of mental health as a vertical programme, and work against the goal of comprehensive integrated care and sustainability (Yasamy et al., Reference Yasamy, Maulik, Tomlinson, Lund, Van Ommeren and Saxena2011). In light of the increasing global burden of multimorbid chronic conditions, we suggest an expansion of the global mental health scale-up funding agenda to ensure that mental health is firmly part of funding streams for health systems strengthening to adequately prevent and manage multiple co-existing chronic conditions. By investing both in systems as enabling environments as well as in horizontal programming, incremental changes are more likely to be sustainable and ultimately, cost-effective.

Second, the inclusion of dialogical action principles such as the adoption of co-learning and co-design between researchers and communities are important to foster acceptability and sustainability. This requires a recasting of partnerships and collaborations between research institutes, organisations and funding bodies situated in high-income geographies, and researchers, policy-makers and implementers at a local level in LMICs, from expert-driven to emphasising co-learning, support and respect (Horner, Reference Horner2019); and balancing universalisation and standardisation on the one hand, with geographic and cultural variation on the other (Horner, Reference Horner2019). This may require increased time commitments, resources and investment in local control; with the flipside being that researchers and funders have less control over the ultimate outcomes of the intervention (Beran et al., Reference Beran, Lazo-Porras, Cardenas, Chappuis, Damasceno, Jha, Madede, Lachat, Perez Leon, Aya Pastrana, Pesantes, Singh, Sharma, Somerville, Suggs and Miranda2018).

Lastly, global policy shifts, as well as emerging policy shifts in LMICs towards strengthening of comprehensive integrated primary health care systems that are responsive to multimorbid chronic conditions, provide a policy window for the global mental health community to ensure that mental health is firmly integrated within these multimorbidity policy shifts and accompanying implementation strategies from the outset of such reforms. This should assist to address the low priority afforded to mental health in LMICs historically, alleviate the problem of it being seen as an add-on to existing services, improve health outcomes, reduce costs as well as potentially reduce institutional stigma (Thornicroft et al., Reference Thornicroft, Ahuja, Barber, Chisholm, Collins, Docrat, Fairall, Lempp, Niaz, Ngo, Patel, Petersen, Prince, Semrau, Unutzer, Yueqin and Zhang2019). It is also aligned with the recent call for a Lancet editorial calling for the global mental health community to ‘transform global health into a movement with mental health’ (Horton, Reference Horton2019).