Understanding mental health system governance in India: perspectives of key stakeholders

Governance, the least studied health system component, comprises a system of rules and processes, and is a key determinant for effective decision making for health care planning. This study aims to identify institutional, legal and policy factors which are either barriers or facilitators for the implementation of integrated mental health in primary care in the India. Semi-structured interviews were conducted with 33 key informants at the district and national levels with policy makers, state level health care planners and district planners and managers in India. The data were analysed using thematic analysis using the qualitative software NViVO 10. Findings Participants stated that a conducive environment for mental health service delivery is necessary at the legislative, policy and planning levels, to facilitate integration of mental health into primary care. Amongst other factors, the need for active involvement of civil society and service user strengthening mental health information systems, and building the non-technical skills of the mental health workforce, were identified necessary to deliver adequate mental health services.


Introduction
Mental disorders pose a significant health burden in India 1, yet the treatment gap is large.
According to the National Mental Health Survey of India 2015-2016, the treatment gap for various mental disorders (including epilepsy, bipolar disorders, mood disorders, psychotic disorders, neurosis, alcohol use disorders and depressive disorders) ranges from 28% to 83%.
Even though effective treatments for mental disorders exist, many people in need are thus denied the potential for alleviation of their symptoms or recovery in India 2.
Efforts to reduce the treatment gap in India are underway, with integrated mental health care at the primary health care level central to India's mental health and broader health policy and plans. The national health policy in India was revised in 2017. Mental health care has been given specific objectives within this revised health policy. Directing public financing to increase the number of mental health specialists, creating a psychosocial support system in communities, and investing in digital technologies to improve access to mental health care have been envisioned as specific priority activities within the policy 3. India's first national mental health policies were published in 2014 4. The Mental Health Care Bill 2017 (GOI 5 and the National Mental Health Programme (NMHP) (MOHFW 6 envisage the delivery of mental health services by integrating these with primary care as the mechanism to close the treatment gap. As part of the NMHP, the District Mental Health Programme (DMHP) is a community mental health programme that aims to ensure mental health care for all by minimising the distress, disability and mortality arising from mental disorders.
India's mental health budget is a small proportion of the total health budget, with just 1.3% of the total health budget spent on the NMHP 7. Funds for mental health are primarily used for maintaining and running the national and state level mental health hospitals, leaving even less budget for primary care 8. Apart from tertiary care, a small proportion of these funds is also spent on district mental health programmes (DMHP); mostly for outpatient services.
The DMHP enables the diagnosis and treatment of people with mental disorders by primary care providers such as doctors, nurses and frontline health workers. Apart from providing drugs, treatment is also planned to involve psychological interventions delivered by nurses and counsellors. For the management of severe cases, or cases with treatment resistant conditions, an upward referral system is envisaged to facilitate specialised care either at district hospitals, mental hospitals or other speciality facilities (ICMR 9. However, current DMHP evaluations indicate its ineffective implementation . Strengthening the key health system pillars of service delivery, health workforce, health information systems, access to essential medicines, financing, leadership, and governance, are all necessary to make optimal use of these limited resources (WHO 10. Good governance is central to strengthening these health system pillars to create a system that enables integration. According to the World Health Organization 11, good governance ensures a "strategic policy environment where services are delivered effectively with better regulation and there exists coalition building and attention to system-design and accountability" (WHO 12. Therefore, poor governance can negatively affect mental health service delivery in primary care. The study forms part of the Emerald (Emerging mental health systems in low-and middle-income countries) research programme, aiming to generate evidence on scaling up delivery of mental health services in Lower and Middle-Income Countries and by doing so to strengthen mental health systems 13. The aim of this study is to understand the system-level governance barriers and facilitators faced by India in implementing mental health services at the primary care level, by using Sehore district of Madhya Pradesh as an illustrative example.

Study design, settings and participants
A qualitative research design was used, with key-informant interviews (KIIs) with policy makers, planners and managers working within the mental health system at the national and state levels in India, and in one district site (Sehore in Madhya Pradesh, India). Sehore district was selected as an example in the semi structured question guide as setting for interviewing programme managers, service providers and planners as it is one of the districts where mental health services are implemented into primary care through PRIME programme 14. A baseline situational analysis of mental health services in Sehore district has been described previously 15. While there are a number of different governance assessment frameworks in the literature, this study was informed by a hybrid of two governance frameworks. The first is that of Siddiqui and colleagues 16 who describe ten governance principles (see Table 1 for definitions of these principles). These principles are rule of law; strategic vision; participation and consensus orientation; transparency; responsiveness; equity; effectiveness and efficiency; accountability; intelligence and information; and ethics. The second framework is that of Mikkelsen-Lopez and colleagues 17 who incorporated an evaluation of governance as it pertains to the WHO's six health system building blocks.
Siddiqui's governance principles have been combined with Mikkelsen-Lopez and colleagues' health systems approach (see Table 1) to create a combined framework, which has previously been used to assess governance to identify and address barriers to health systems' performance 18. This framework recognises that Siddiqui's principles are inter-related. For instance, both the responsiveness and effectiveness principles are concerned with meeting the needs of the population while making the best use of resources. Similarly, the transparency and accountability principles are closely intertwined.
[Insert Table 1  Policy Group were also interviewed. District level respondents were selected from the community health centres where mental health services were integrated with primary care.
There is was a poor representation of stakeholders at district level because of a poor response to interview requests.

Data collection and tools
A common cross-country semi-structured interview guide was developed for the six low-and middle-income countries included in the Emerald study, which focused on identifying and addressing barriers to health system performance, based on the hybrid framework in Table 1.
This guide was subsequently adapted for the context of India, and details on the district and state level mental health plans were added. The main topics covered included domains for rule of law, strategic vision, participation and collaboration, responsiveness and effectiveness, equity and inclusiveness, ethics and oversight and accountability and transparency 19 . Three interviewers working as research associates in the project were trained in qualitative research methodologies and had prior experience in conducting qualitative research. Two separate guides were used for district and state/national level respondents. Data were collected in June-July 2015. The interviews were audio taped and field notes were taken after receiving participants' consent. Interviews were transcribed verbatim and the interviews, which were conducted in the local Hindi language (n=33), were subsequently translated into English.

Data analysis
The Framework Approach was used to guide the analysis of qualitative data 20, which includes stages of familiarisation with the data, indexing and charting, coding of the interviews using NVivo-10 software, and mapping and interpretation of results by two independent researchers 21,22. Data were analysed using both deductive and inductive approaches, using the health governance framework domains as 'a priori' themes, with sub-themes under each category subsequently derived inductively from the data.
The overarching nodes/themes were made up of the different elements of the governance framework, with sub-themes extracted from the respondents'' responses. A table of all the thick descriptive data coded in NVIVO-10, were then exported to Excel. Within Excel, the responses under each overarching node/theme were compared across different groups of stakeholders, to identify common sub-themes in participant responses by the lead researcher as well as providing information on the number r of participants who endorsed a sub-theme

Results
A total of 33 in-depth interviews with key informants were conducted at the national (n=19), state (n=6) and district (n=8) levels. Table 2 provides the number and type of respondents included in the study. The results are presented under the governance principles as set out below.
[Insert Table 2 about here]

Rule of Law
India's revised Mental Health Care Bill was approved by Parliament in March 2017. The bill was at a draft stage at the time of interviews and was later approved. However, levels of awareness of mental health laws amongst citizens, as reported by state and district authorities, were low. Within our respondents, awareness was least amongst medical officers.
The Mental Health Care Bill was generally regarded as a necessary reform. Participants speculated, however, that the implementation of the new act would be challenging due to a lack of ownership at various administrative and executive levels. As one respondent stated: "It is difficult you see… There is a draft new mental health bill online, it is extensive and then having a structure to enforce them and implement is a separate task altogether. They don't own it, no one does". (Consultant, national level).
Respondents from caregiver organisations pointed out the importance of collective ownership for the mental health problem, which can be revealed by involving various civil society organisations. Many respondents highlighted that in the formulation of the new bill, service users, civil society, public health professionals, psychiatrists and psychologists were consulted. However, service user and care giver organisations commented that their level of involvement in the implementation of the bill was marginal. Respondents from service user and care giver organisations indicated that they would have liked more active involvement in all spheres of the Bill, including its preparation and implementation. Respondents indicated that implementation of the Bill to be facilitated, government needs to ensure that the necessary budget is available.

Strategic vision
Strategic vision, as conceptualised in the health governance framework employed here, relates to a broad and long-term perspective on overall health and human development. The revised mental health policy envisions universal health coverage for all its citizens by incorporating both a preventive and promotive healthcare agenda in its developmental policies 3. There was a general consensus by participants that there is increased scope for the development of mental health action plans at the national or district level in the next few years, Decentralisation was seen as pivotal to primary health care 23. In the health sector in India, decentralisation is exercised through Programme Implementation Plans (PIPs) where PIPs are consolidated at the state and/or district level in consultation with partners at the district and/or block levels, respectively. Respondents highlighted that there is considerable tension between the state, district and block levels, primarily due to power dynamics. The issue of distribution of power is said to be problematic and has affected the level of participation of peripheral level staff in the process of decentralisation in health.
State level respondents felt that mental health was not a regular agenda item on district planning committees, at which state PIPs are translated into strategies and budgets are allocated to achieve the required health outcomes. Most states do not include mental health in the PIPs. Participants did, however, suggest that a shift towards non-communicable diseases (NCDs) and approaches that target multiple chronic conditions, including mental health services, may help facilitate the integration of mental health services.
However, there was one example of decentralisation in mental health in Madhya Pradesh that was mentioned stated by a few respondents, which is the "Mental Health Action Plan 365".
The Mental Health Action Plan 365 is a state level initiative to align the state level mental health programme with the new mental health policy released in 2014. District level officials, service users and service providers were consulted by the National Health Mission in formulating this plan to address community needs.
Responsiveness and effectiveness of the health system Effectiveness and efficiency in delivering mental health services are essential for achieving better health outcomes. Under this heading, system level facilitators and barriers related to the system building blocks were explored, such as health workforce, financing, medicines and technologies and services delivery.

Human resources
Human resources mainly comprise two groups; service providers and health managers, including other support staff. Respondents highlighted that the number of service providers is profoundly low. Mental health providers at the state level in Madhya Pradesh stated that contractual job policies and poor opportunities for professional growth have contributed to this scarcity:

Medicines and technologies
Mobilising psychotropic drugs from tertiary care to primary care is key to integrating mental health at the primary level. While the essential drug list (EDL) approved by the Ministry of Health and the respective state directorates includes 18 psychotropic drugs for district and sub-district hospitals, many respondents in this study highlighted that none of these are available in primary health care settings. District officials expressed concerns regarding a vicious cycle of poor procurement of psychotropic drugs and poor demand for mental health services, since drugs will only be provided when there is a demonstrated need for them in the clinics. This vicious cycle was also mentioned by a state level planner: Respondents noted that the managers at the district store, under the control of the Chief Medical and Health Office, are mainly responsible for supplying medicines to primary care providers in primary and community health centres. The district store is also involved whenever there is a need for local purchasing of these medicines. All officials responsible for the procurement and supply of essential drugs to primary care facilities were reported to be cooperative, according to participants. However, medicines that are not included in the EDL or are less in demand often face delays in procurement due to the lengthy administrative process, such as gaining approvals from the local purchase committee, putting through new purchase orders and conducting quality checks.
Supervisors of mental health programmes at the state level believed that continuous supervision and hand-holding is required for facilitating drug procurement in any national programme.
"It is not possible to have buffer supplies; we do not have a proper pharmacist also at sub-district level. This helper who works as a pharmacist has to do some reporting also. So, it is difficult in that sense… But yes, if you follow up actively their people who A further sub-theme related to this governance element related to the lack of inter-sectoral and intra-sectoral collaboration suggesting a weaker collaboration within the health system but also across various sectors including social sector, education and justice. Within the health system, the respondents described the system as characterised by fragmentation between the private and public sectors, and between modern and traditional care.
They also described ambiguous partitions between prevention, primary, secondary and tertiary levels of care as well as between the state and national levels.

Equity and inclusiveness
Mental disorders affect people of all demographic groups, but those vulnerable groups who are already socially or economically marginalised are particularly at risk, leading to further exclusion and in some cases contributors in violations of their human rights. Respondents in this study noted that geographical inequities, mental health stigma, and the use of traditional treatments hinder equitable provision of mental health services to some of the most at-risk groups.
Respondents indicated that there was inequity in relation to the implementation of the District Mental Health Programme across states in India. Furthermore, a respondent pointed out that due to diversity in Indian states, there was also inequity in the reach of services to vulnerable populations and differing barriers to generating awareness of services among groups of low socio-economic status.
"I think that like in India there are so many differences, you can't have a same approach for, let's say, a district in Tamil

Ethics and oversight
Ethical considerations in mental health treatment and in conducting research were both explored in this section. Regarding the ethos around treatment, national level respondents noted that the regulations around coerced psychiatric in-patient care were poor. It should be noted that these responses were gathered before the new mental health care act was approved, however. The new Act includes advance directives, which allow a person with mental illness to demand in advance how he/she wishes to be cared for when they lack capacity 5.
Another ethical concern brought up by many respondents was the irrational prescription of psychotropic medicines. Both the content of medicines and the completeness of prescription sheets were pointed to as reasons for irrational prescriptions.
Safeguards for conducting research, as set by research bodies such as the Indian Council of Medical Research 9, were considered to be comprehensive. However, policy group members were against the stringent guidelines for conducting clinical trials for newer antipsychotics.
According to them, these guidelines have prevented the implementation of new clinical trials, forcing researchers and service providers to rely on evidence originating from other countries.

Information systems
This principle focuses on monitoring and evaluation, and its importance in planning and evaluating performance. Data availability on mental health was reported to have improved substantially over the last five years. In particular, respondents noted an improvement in the collection of data related to the number of reported suicides. However, the data on mental health is often derived from large scale surveys and not from the routine health management information systems, as stated by a few participants. According to one national level Routine information systems in states like Madhya Pradesh contain indicators on major and minor disorders, as part of monitoring for the DMHP. However, this system of classifying mental disorders into major (relates to severe mental disorders like schizophrenia) and minor (relates to common mental disorders such as depression) is outdated and needs revision, according to state level respondents. Respondents also underlined other issues pertaining to the quality of the data for mental health collected through routine systems. Some said that data collected through routine information systems are not a true depiction of the burden, since severe mental disorders such as psychosis tend to be included and common mental disorders such as anxiety and depression are left out. Others reflected on the lack of comprehensive information in these routine systems. Absent or incomplete information, such as on medical history and prior treatment, further impedes the use of data collected by routine systems.
"What is happening is most cumulative work on one to one bases in an OPD

contact. There is not [a] health system approach in that you have [a] family folder
where you record the history of the family and ensure that they do receive the care as [and] when the time comes." (NCD consultant, National level) However, the new mental health plan was reported to have a revitalized mental health information systems component, seen as a facilitating factor towards an integrated information system for mental health.
"For mental health, there is hardly any reporting. Although you will be surprised but in the plan [it] has a separate budget for mental health, health management information system HMIS. [The} district part has been approved, cabinet has still to approve it. Once it is approved we can have a database, a HMIS for mental health. We will have to implement it" (MOHFW consultant, National level) The possibility of integrating mental health data with other priority programmes such as the maternal and child health and non-communicable disease programmes was also reported by respondents. A respondent explains integration as a feasible approach to improving data quality on mental health. Similarly, it was reported that transparency exists in other programmes but not in the field of mental health. Some measures such as establishing review boards to enhance accountability are starting to be rolled out for mental health, however 5.
National level respondents felt that the existence of an integrated finance division within MOHFW ensures transparent budgeting. It also acts as a watchdog in the Ministry of Health for finance-related activities. Again, monitoring to ensure the transparency of decisions around mental health services was perceived to be poor, although the Right to Information (RTI) is present which is designed to make information transparent to the public. This is primarily related to the low priority of mental health within public health, as highlighted by many respondents. The gap in the knowledge and awareness of respondents regarding mental ill health as suggested in the findings of this study needs to be addressed by increasing advocacy efforts to increase the public health priority of mental health in India. Recently involvement of policy makers and planners through long term long term engagement and mentoring to develop sustainable relationships, through awareness workshops and ongoing dialogue was recommended in a study in India 28. This can gradually enable buy-in for mental health at state and at national level in India.

Discussion
Third, in relation to the human resources element, this study indicated the need to strengthen the diverse cadres of the health workforce delivering mental health services, specially so of professionals with public health perspectives and implementation skills, needed to utilise the underspending in the Indian states. As recent study assessed the incorporation and assessment of implementation outcomes to improve implementation of mental health services at primary care in India. The implementation strategies such as active facilitation of primary care facility staff by an external support to improve quality of care is suggested 29 On the other hand in the field of public mental health, in various projects in India frontline workers in PRIME 14 and community health workers in Vidharbha Stress and Health Programme (VISHRAM) 33 have been successfully delivering psychosocial interventions and overall programme support. However, a critical mass of technical specialists such as psychologists and psychiatrists and public health specialists are still required to prevent staff burn out within this task sharing approach 34. Fourth, as for the finance element of the effectiveness principle, even though less than one percent of the national health care budget is allocated to mental health in countries such as India and China 8, in bigger states like Madhya Pradesh more than half of the budget allocated to mental health for 51 districts went unspent in 2015 35, which is consistent with our findings in the financing domain. One of the reasons for under-spending of the health budget that this study revealed under the principle of responsiveness and effectiveness is demand generation.
The gap between the community's needs and the availability of services in the community is huge. In other studies, barriers to demand included low levels of mental health awareness, stigma, costs, attitudes and beliefs around the causation of mental disorders 8. Generating awareness on where to access mental health services, and training frontline workers, can assist in generating demand from the community 36,37. Once we are able to address underspent budget by generating demand by engaging new community based human resources, these new resources can be involved in delivering psychosocial interventions, as recommended by community projects such as VISHRAM 33. However, to move towards a more integrated approach, it is important these services are not only co-located but integrated 38. Especially considering treatment of mental health problems overlaps with other chronic care conditions and share pathways of determinants. 39. Therefore, a more biopsychosocial approach throughout the primary care 40 Re-emphasising rational prescriptions of anti-psychotic drugs within the medical curriculum and encouraging doctors to rely on evidence-based practice coupled with their clinical judgement has been advised to avoid unethical prescriptions in the field of mental health in India 46. However other studies in India suggest that these issues are not restricted to mental health, better regulation and rethinking in the medical education system is needed in both public and private sectors, across all areas of health 47. These recommendations stretch beyond the health sector to promote and protect the mental health of the population. To safeguard the rights of the patient, ethical considerations on coercive medication and irrational prescriptions need careful monitoring 48 5 Eighth, with regard to intelligence and information, this study found a dearth of evidence on basic mental health indicators such as prevalence and service coverage, which represents an important systems-level challenge in integrating mental health with general health services.
Similar to other literature 49, the need to integrate mental health indicators into routine information systems was emphasised. Literature also suggests that it is difficult to get data on basic services delivered in primary and community care settings 50. Our study found that the poor quality of routine data hinders its use. Other studies have attributed inaccurate and incomplete reporting to low staff motivation levels 51. Such inadequate information systems translate into weak evidence, leaving decision making for future planning unaccountable. For planning mental health service delivery, a range of service indicators have been proposed by projects such as MHaPP 52 in four African countries at the district and provincial levels. Similar work on strengthening routine information systems is now being undertaken in various LMICs 13 including India 53. Information systems within governance are the least studied system domains but are also leverage points in a health system affecting accountability of health programmes 12.
However, the quality of routine data from health information systems as a measure of population needs is dubious 54. Information on service use is instrumental in designing, implementing and improving mental health programmes. A progressively integrated system can save time, inform clinical decisions and contribute to quality of implementation. However, data from mental health information system can only provide data on service use and not suffice population needs. Since the time of the study, the National Mental Health Survey 2015-2016 31 have partially addressed this need for survey data which will help to inform evidence based service planning in future.
Overall the barriers and the recommendations although generated from examples from the context of Sehore district, they are likely to be applicable more widely and can foster an enabling environment needed for implementing mental health care at primary care settings.
There has been an increasing research on how to reduce the evidence and practice gap. This study has suggested some factors and potential strategies to be considered at the design stage of mental health implementation at primary care to reduce the evidence to practice gap Key Messages: 1. Build public health perspective and improve implementation skills within the health services to utilise the underspending of the mental health budget at state level in India.
2. Effectively utilising mental health budget by improving demand for services at the community level but also by introducing case managers at primary care level trained in both mental and physical conditions enabling a biopsychosocial approach to health care in India.
3. Strengthening mental health information systems to improve service use, to generate data on population needs and hence accountability in the Indian public health systems. 4. Enhancing inter-sectoral collaborative arrangements so as to directly and indirectly involve participation of service users and civil society organisations, although challenging, is to be emphasised.

Limitations of the study
One of the main limitations of our study is that the interviews focussed on assessing governance principles for integrated mental health service delivery with reference to one district, where mental health services are being integrated with general health services. Due to this limited focus and the small sample size, generalising these findings to other contexts within India should be done with caution. Second, some of the nuances around the human rights issues relating to persons with physical and mental disabilities such as the Convention on the Rights of Persons with Disabilities 48 to which India is a signatory, was not included in the topic guides and did not come out of the interviews and hence were not discussed. Third there was relatively limited representation of stakeholders at district level, despite 7 district stakeholders being approached and invited to participate in the study only 2 agreed. Reasons for refusal included time constraints and little to no involvement in the planning and delivery of mental health services in the district. Only two medical officers where mental health services were integrated into primary care through PRIME programme agreed to participate. Inclusion of more doctors and primary care workers could have represented the challenges they face in implementing mental health plans.
Furthermore, these findings are based on the perceptions of policy makers, planners and implementers, who may withhold information based on their organisation's policies. Our researchers faced difficulty in contacting respondents, who were service providers and planners and as such were extremely busy. Hence three of the interviews, although they provided rich information, were ended prematurely.

Conclusions
With a governance lens, this study draws on the experiences of stakeholders involved in delivering or planning mental health services. Governance sits as a backbone to addressing challenges in other health system blocks, focusing on health system inputs (human resources, financing, medicine and supplies), processes (interventions, plans and policies) to result in better health system outcomes (measured through routine information systems). Overall, the study has identified measures to mitigate the mental health disease burden in India by understanding governance principles affecting mental health service delivery using the specific case study of Sehore district in Madhya Pradesh. Two fundamental elements were identified in this study. First, poor implementation of plans and policies needs to be tackled through measures such as broadly strengthening a health system platform at primary care level to deliver mental health intervention, specifically strengthening teams of specialists and implementers/managers, and strengthening and integrating routine monitoring systems. Second, apart from these system level reorganization strategies, measures to generate awareness and involve civil society in service planning, legal and advocacy measures are also required to generate demand.