How Integrating Mental Healthcare into Primary Healthcare Changes Health-Seeking Behavior: A Realist Evaluation and Extension of PRIME in Nepal
Victor A. Bodiut
Master Thesis Medical Anthropology and Sociology (MAS), 15EC Word count (excl. references): 21.797
Student number: 11260653 University of Amsterdam
Supervisor: dr. Danny de Vries 2nd Reader: dr. Mark Jordans Local Supervisor: Nagendra Luitel
Local Translator: Paru Pandey
22-08-2019 Amsterdam, Netherlands
Table of Contents
1. Acknowledgments………...3
2. Citing and formatting conventions………..4
2.1 Abbreviation list………5
3. Abstract………7
4. Background………..9
5. The current research………...13
5.1 Health-seeking behavior: a conceptual direction………15
5.2 Realist evaluation: an analytical framework………...18
5.3 Research questions………...22
6. Materials, methods and data analysis……….23
7. Ethical considerations and reflexivity………27
8. Results and discussion………...30
8.1 Community preparedness and engagement in mental healthcare………30
8.2 Community judgement about mental ill health………49
8.3 Involvement of close-knit community in treatment……….60
9. Conclusions and future directions………..64
9.1 What now, that PRIME has ended?...65
9.2 The health-seeking process: a proposal………..67
10. Recommendations………..70
Acknowledgments
I would like to thank everyone who contributed to this work, both directly and indirectly. From my parents, family and friends who made the whole journey possible, to my academic mentors, supervisors and peers who constantly raised the bar of excellence, to the staff, health workers and people who welcomed me, guided me, and opened their world to me. This thesis is a data-driven story about people told by people and filtered through my eyes.
PS: Dear reader, this will be a rather intricate read, but it is my hope that I will be a voice of clarity guiding you through it. I aim to write in a clear, direct and personal way.
One of the special moments in Chitwan, Nepal together with health workers.
Citing and formatting conventions
All in-text citations and reference bibliography follow the American Anthropological Association (AAA) style.
Direct quotations from published work shorter than 2 sentences are marked by the double quotation marks “…” and kept within the paragraph format. Quotations longer than 2 sentences and quotations from respondents are marked by the double quotation marks “…” and indented to the right with 0.5".
Reference to terminology that is either used for specific naming, idioms, mottos, quote-in-quote, ironical, highly specialized, not used for its literal meaning, or has been used as such in literature is marked by the single quotations marks ‘…’ and kept within the paragraph format.
Reference to theoretical or analytical concepts is marked by italicizing the text at first appearance only.
Abbreviation list
APA – American Psychological Association AUD – Alcohol Use Disorder
AUDIT – Alcohol Use Disorder Identification Test CD – Communicable Disease
CDC - Centers for Disease Control and Prevention CHW – Community Health Workers
CIDT – Community Informant Detection Tool CMO – Context-Mechanism-Outcome
DALY – Disability-adjusted Life-year
DFID – Department for International Development DISC-12 – Discrimination and Stigma Scale FCHV – Female Community Health Volunteers FGD – Focus Group Discussion
GBD – Global Burden of Disease HBM – Health Belief Model HDI – Human Development Index HIC – High-income Countries HP – Health Post
IHME – Institute for Health Metrics and Evaluation IHME – Institute for Health Metrics and Evaluation IRB – Institutional Review Board
KII – Key Informant Interview
LMIC – Low- and Middle-income Countries MHCP - Mental Healthcare Plans
mhGAP – Mental Health Gap Action Program MNS – Mental, Neurological and Substance Use MoHP – Ministry of Health and Population NCD – Non-communicable Disease
NHRC – Nepal Health Research Council PHC – Primary Health Clinic
PHQ-9 – Patient’s Health Questionnaire
PRIME – Program for Improving Mental Healthcare PTSD – Post-traumatic Stress Disorder
RCT – Randomized Control Trial RPC – Research Program Consortium RQ – Research Question
SDG – Sustainable Development Goal SDI – Sociodemographic Index STD – Sexually Transmitted Disease sub-HP – Sub-Health Post
ToC – Theory of Change
TPB – Theory of Planned Behavior
TPO – Transcultural Psychosocial Organization TRA – Theory of Reasoned Action
UCT – University of Cape Town UvA – University of Amsterdam WHO – World Health Organization
Abstract
One way to reduce the gap in treatment of mental disorders in low-resource settings is to
incorporate mental healthcare into existing routine healthcare. The primary aim of this thesis was to assess how such implementation in a low-resource rural setting changes how people seek and reach mental treatment. Specifically, a mixed-method assessment following the realist evaluation strategy focusing on the health-seeking behavior concept at the community level was performed on the Program for Improving Mental Health Care (PRIME) 2011-2019 in Chitwan, Nepal. Findings show that implementing a primary mental healthcare system is feasible and efficient for improving health-seeking and treatment coverage for mental illness. The key factors were
increasing community preparedness and engagement through establishing a trust-based network of identification, referral, and professional treatment including both formal and non-formal care and reducing stigma and discrimination through raising awareness at the community and family level. Recommendations for further policy are proposed.
Keywords: global mental health; mental healthcare; primary healthcare; task-shifting;
How Integrating Mental Healthcare into Primary Healthcare Changes Health-Seeking Behavior: A Realist Evaluation and Extension of PRIME in Nepal
“I would probably not be alive if the PRIME project did not help me” – service user
recovered from 20+ years of alcohol addiction.
This was a happy case in the end, but the 20 years of severe addiction were very sad, and perhaps avoidable. My thesis is about avoiding those 20 years, about saving lives from the sharp claws of mental illness, about reaching people that would otherwise struggle alone with limited options of treatment, about their tough journeys in hope to get well even when lacking a clear idea of what is happening and what needs to be done, about implementing an infrastructure of care where it is most needed even when knowing that it will not solve everything, about finding solutions in low-resource settings together with the community in need, about learning from trial and error and doing it better next time, about hope. My thesis is about understanding how
integrating mental healthcare into an existing system of primary healthcare in rural Nepal narrows the treatment gap in mental care, influences mental health in the community, and changes what people go through to get better.
Before diving into the matter, I want to clarify my positionality towards treatment in mental health and throughout the work of the last months. During fieldwork in Nepal, I aimed to be as open and curious as possible, like an open book that lets itself to be written. Still, we all carry our baggage. My baggage is my previous training in psychology and neuroscience in the Netherlands, before turning to medical anthropology and sociology. Therefore, my position is that I acknowledge, embrace and cherish the diversity of treatment practices in mental illness across the globe and in rural Nepal implicitly, such as alternative, traditional, faith and religious healing, but I think everyone should have the chance to seek and receive professional modern psychological and psychiatric help. This should not be enforced by any means, should be adapted to the context and culture, integrated in the local structure of human resources as much as
possible, incorporated in the sufferer’s lifeworld and explanation of illness, but it should be one of the options on the sufferer’s table that the sufferer is well aware of. The reader should know that this is my position throughout the thesis, and that I will never, ever disregard any form of healing but rather aim to understand it.
Background
In a global trend where communicable diseases (CDs) such as tuberculosis and HIV AIDS scaled down and non-communicable diseases (NCDs) such as hypertension and diabetes scaled up, mental illness is on the top of the NCD list (Patel et al. 2018; Bloom, Cafiero, and Jané-Llopis 2012). Biomedically, mental (e.g., depression, bipolar affective disorders,
schizophrenia), neurological (e.g., epilepsy, migraine, multiple sclerosis) and substance use (e.g., alcohol, opioid, tobacco) together coined MNS disorders are the number one contributor to the NCD-related Global Burden of Disease (GBD; Mathers and Loncar 2006; Stein and Seedat 2007; Aguilar-Gaxiola et al. 2011). MNS disorders currently account for an estimated 14% of the GBD and more than 1 in 10 lost years of health globally (Mathers and Loncar 2006; Patel et al. 2018). WHO states that among the top 10 leading causes of disability, four are represented by mental illness and about half billion people suffer of some sort of mental-related problem (Cohen 2001; Patel et al. 2007). Moreover, the MNS-related disability-adjusted life-years (DALYs), a metric constructed to measure the number of years in an individual’s life lived with a health-altering disability has been on a constant rise globally, see Figure 1 (Patel et al. 2018).
Figure 1. “The rising burden of mental and substance use disorders, Alzheimer’s disease and other dementias, and suicide (self-harm) by SDI groups” from Patel et al. (2018). SDI refers to the sociodemographic development of a given region.
To make things worse, MNS disorders are a major risk factor for premature mortality and suicide globally, especially in the young population (Patel et al. 2018; Bloom, Cafiero, and Jané-Llopis 2012; Liu et al. 2017). Although suicide rates have decreased globally in the past 3 decades from about 15 to 10 per 100.000 people, they are on a steady trend in the young
population (Naghavi 2019; Langhinrichsen-Rohling, Friend, and Powell 2009). The Centers for Disease Control and Prevention (CDC) places suicide as the second leading cause of death within the youth cohort of 10 to 24 years (CDC, 2014). These high rates of suicide for this particular age group do not come by chance. Most symptoms of mental health problems become apparent in adolescence and young adulthood, and this cohort is highly vulnerable to social, emotional and neurocognitive disturbances lingering into adult life (Patel et al. 2007; Patel et al. 2018; Rickwood, Thomas, and Bradford 2012). This points towards a gateway of direction towards essential prevention and protection involving a well-balanced and nourishing childhood, healthy social integration, and accessible support and treatment very early-on.
Despite the prevalence, suffering and burden associated with mental illness, there is a disproportion in terms of governmental health budgets and policy coverage allocated for mental health globally that leads to large proportions of people not receiving care (WHO 2008; Patel et al. 2018). This becomes most relevant when comparing cross-national low-resource and high-resource settings, with a significant treatment gap in mental health between low- and middle-income countries (LMICs) and high-middle-income countries (HICs; World Health Organization 2008; Patel et al. 2018). To put things in perspective, about 80% of people affected by MNS disorders are from LMICs, and they benefit from about 10% of global resources in mental health (Jacob and Patel 2014). Whereas the treatment access and follow-up rates for priority MNS disorders such as depression and alcohol use disorder (AUD) are roughly around 50-70% in HICs, they are in the range of 0-20% in LMICs (Mitchell, Vaze, and Rao 2009; Mitchell et al. 2012; Rathod et al. 2018; De Silva et al. 2016). In other words, roughly 4 out of 5 people struggling with mental disorders in LMICs receive no care and treatment and the attention given to mental health in these contexts is low (Lund et al. 2012; Patel, et al. 2007). This treatment gap is even higher for severe neurological disorders such as epilepsy (WHO 2008). Unattended mental illness acts in a bi-directional comorbidity with other health and socio-economic factors, increasing the chances for poverty, unemployment, homelessness, poor physical and reproductive health, injuries, and
co-morbid substance abuse (Patel and Kleinman 2003; Prince et al. 2007; Lund et al. 2010; Folsom et al. 2005; Butterworth et al. 2012; Luitel et al. 2017).
In a scheme of inadequate mental healthcare infrastructure, limited assessment and treatment knowledge and options, scarcity of financial and human resources, the lack and
unequal distribution of mental health professionals is at the core of the treatment gap. In LMICs, the data shows an estimated < 1 psychiatrist, psychologist, psychiatric nurse and social worker per 100.000 inhabitants compared to > 10 in HICs, and a shortage of about 1.2 million mental health workers, see Figure 2 (WHO 2008; Kakuma et al. 2011; Saraceno et al. 2007; Kohn et al. 2004). Notably, there is a possibility that the data presented to date is under-representative of the real situation especially in low-resource settings, in part because of the high stigma associated with mental illness, the lack of general knowledge and awareness regarding differences in the manifestation of mental disorders, conditions that do not fit the biomedical categorization (e.g., symptomatology regarded as psychosomatic, witchcraft or spirit possession), and traditional healing practices.
Figure 2. “Human resources for mental health care in each income group of countries, per 100.000 population. (Source: Mental Health Atlas, WHO 2005)”, from (WHO 2008).
This gap in mental treatment has been the top priority for the Global Mental Health movement, “a coalition of individuals and institutions committed to collective actions that aim to
close the treatment gap for people living with mental disorders worldwide” (Patel et al. 2011:88; see also http://www.globalmentalhealth.org/). An innovative strategy that has received
international consensus and interest is that one way to narrow the treatment gap is to integrate mental healthcare into community and primary healthcare in low-resource settings (Patel et al. 2007; Patel et al. 2011). WHO defines primary healthcare as follows:
“Primary healthcare is essential healthcare based on methods and technology made universally accessible to individuals, families and communities through their active participation and at an affordable cost. Primary healthcare is an integral part of a country’s health system and, ideally, its main focus. It is the first level of contact for individuals, families and communities and enables healthcare to be delivered as close as possible to where people live and work. It is therefore the first element of the care continuum” (World Health Organisation 2006).
Integrating mental healthcare into primary healthcare involves training and supervising existing skilled but non-specialized healthcare personnel to screen, assess, treat and refer people struggling with mental health issues, a process of task-shifting and task-sharing (Cohen 2001). Most of the healthcare services related to mental health globally come from non-specialist health professionals within routine healthcare settings (Jacob and Patel 2014). Such task-shifting has the potential to provide deliverable and cost-efficient treatment, increase opportunities and accessibility for mental health services, fix the scarcity of trained personnel, and reduce stigma-related barriers, all governed by a more holistic approach to mental healthcare (Patel et al. 2007; Patel et al. 2007; World Health Organization 2008). This rather theoretically simple but
practically less simple idea is at the very core of the thesis and its potential fascinates me. Concurrently, with the goal of scaling up care for MNS disorders in routine healthcare settings globally, the WHO released the Mental Health Gap Action Program (mhGAP)
intervention guideline in 2008 (adding the mhGAP operations manual more recently; World Health Organization 2008; WHO 2018) and the Mental Health Action Plan 2013–2020 (Saxena, Funk, and Chisholm 2013). Last year, the Lancet Commission (Patel et al. 2018) focused on guidelines towards sustainable development goals (SDGs) in global mental health, guided by four principles: (i) mental health is a global public good highly relevant to all countries, because in the context of mental health all countries can be considered in development; (ii) mental health
takes the form of a continuum from mild short-term distress, to chronic, progressive, and severe conditions, highlighting the limitations of a traditional binary of the type ‘disorder vs. no
disorder’; (iii) individual mental health is a complex interaction of genetic, neurodevelopmental, psychological, social, and environmental processes (i.e., biopsychosocial), especially relevant during the high plasticity of early-life course; and (iv) mental health is a fundamental human right for all people, everywhere. Following these guidelines, intervention programs have been picked up at international, national and local levels in low-resources settings across the globe (e.g., Jordans et al. 2010; Honikman et al. 2012).
The current research
One such intervention has been the research program consortium (RPC) Program for Improving Mental Health Care (PRIME; Lund et al. 2012), founded by UK’s Department for International Development (DFID). The global mental health network needed (and still needs) to understand how to go about implementing sustainable and deliverable mental healthcare
packages within existing primary care systems in diverse contexts with the prospect of up-scaling. Quite a lot is known about what works through the extensive use of randomized control trials (RCTs), but a lot less is known about how these interventions can be delivered in routine healthcare in low-resource settings (Lund et al. 2012). This was the primary aim of PRIME: to generate high-quality evidence that can inform policy and practice on the implementation and scaling-up of district-level mental healthcare plans (MHCPs) in low-resource settings (Lund et al., 2012). PRIME has been focusing on five high-priority MNS disorders (i.e., depression, maternal depression, AUD, psychosis and epilepsy) in five cross-country application sites (i.e., Ethiopia, India, South Africa, Uganda and Nepal) between 2011 and 2019. The program was primarily informed by the mhGAP and Mental Health Action Plan guidelines but contextualized and applied to each context (for more details on PRIME, also see http://www.prime.uct.ac.za/).
My focus is on the Nepalese context, where PRIME has been running in the Chitwan district in southern Nepal between 2011 and March 2019 through the non-governmental organization (NGO) Transcultural Psychosocial Organization (TPO) Nepal (see:
http://tponepal.org/) in partnership with the WHO and the Nepal Ministry of Health. PRIME in Nepal underwent three main phases: (i) an inception phase of formative qualitative research on the district site using situation analysis tools leading to MHCPs that were contextualized and
applicable for the particular setting of Chitwan; (ii) an implementation phase where the MHCPs were implemented across health facilities in the district and then evaluated; and (iii) the scaling-up and extension phase where several more facilities were included based on lessons from the first two phases.
With a population of about 26.5 million and a general life expectancy of 69 years at birth, Nepal is one of the poorest countries is South Asia, ranking 145th out of the 188 countries on the UN Human Development Index (HDI; De Silva et al. 2016). The treatment gap in mental health is still big in Nepal, where there is a lack of population-wide coverage of mental healthcare and most existing mental healthcare services are located in the big cities (e.g., Kathmandu, Pokhara; Luitel et al. 2017). Mental health policy in Nepal was developed in 1996 but never implemented, and out of the 75 districts in the country, only 7 had formal mental health services (Luitel et al. 2017). As Nepal finds itself in a fragile-state post-conflict setting (i.e., the Nepalese Civil War officially ended in 2006), most investigations of mental health focused on populations who participated in armed conflict in the post-traumatic stress disorder (PTSD) line of research (Luitel et al. 2017). As such, the attention given to the treatment gap in other mental disorders and the barriers to seeking and reaching care has been very low (Luitel et al. 2015; Luitel et al. 2017; Jordans et al. 2019). Luitel et al. (2017) reports scarce availability of decentralized mental health services, with only about 5% of people with AUD and less than 10% of people with depression receiving treatment from any type of health provider, and below 2% of patients seeking treatment at the primary healthcare level.
Among the major barriers to care reported in Luitel et al. (2017) are financial costs, stigma associated with being perceived as ‘weak’ or ‘crazy’ for having mental health problems and being too sick to seek help. Importantly, in this study these barriers did not seem to differ per demographics such as age, sex, education, caste, etc., which could imply that targeting the
district population as a whole is a feasible idea. Still, this needs further investigation. In another study, Kisa et al. (2016) found that systemic barriers to professional mental health services in Nepal included lack of awareness, economic burden and poverty, stigma and discrimination, myths and misconceptions regarding mental health problems, inaccessibility of services, unavailability of medicine, negative attitudes of health workers, lack of patient follow-up, inadequate mental health specialists, traditional beliefs, delays in familial decisions for seeking
care, and lack of social support. As these barriers are very intimate and specific from case to case, in-depth research at the individual-community level is therefore required to understand them from the perspective of the sufferer and his/her community, for a healthy integration of the MHCPs that does not focus solely on providing services but also takes the voice of the people into account (Shaikh 2008; WHO 2008).
Health-seeking behavior: a conceptual direction
This understanding is encapsulated by the concept of health-seeking behavior. The reasoning behind this concept is that there is a black box that needs to be targeted in-between having symptomatology of illness and the actual use of treatment. It is nowhere close to a 1-to-1 relationship between the two. There is an entire journey that the sufferer has to go through in engaging with health services that differs on an individual-level and is influenced by internal and external factors, and there is no certainty that the contact with the health provider will be made at all. The definitions of health-seeking behavior or related terminology are multiple and
decentralized, and in many cases not stated explicitly in studies (see Barker 2007; MacKian 2003). Example definitions of health-seeking behavior include “any action or inaction
undertaken by individuals who perceive themselves to have a health problem or to be ill for the purpose of finding an appropriate remedy” (Ward, Mertens, and Thomas 1997:21) or “a
sequence of remedial actions that individuals undertake to rectify perceived ill-health” (Ahmed et al. 2000). Rather, I would like to focus on the way the concept has been used in the literature, a literature that is vast and multidisciplinary. For the sake of simplicity, it can be divided into two main traditions: (i) the literature focusing on the illness response of the patient, or the ‘process’ of health-seeking; and (ii) the literature focusing on the utilization of a healthcare system, or the ‘end point’ of health-seeking (MacKian 2003).
The former tradition is mostly based on socio-psychological models of predicting decision-making in health behavior such as the Theory of Reasoned Action (TRA; Fishbein, Ajzen, et al. 1975) on the importance of attitudes and subjective norms in behavior, the Theory of Planned Behavior (TPB; Ajzen 1991) on intention as the closest link to behavior, the Health Belief Model (HBM; Rosenstock 1990) on the role of perceptual factors and experiences in health behavior, or Andersen’s Behavioral Model of Health Services Use (Andersen 1995) on predisposing, enabling and need factors. The shortcoming of these models is that they generally
assume the sufferer to be a rational decision-maker even when sick, taking information in,
reviewing it and outputting behavioral intentions. They fall short in incorporating the spontaneity and unpredictability of decision-making, the way that people take decisions (Fazio 1990), the complex web of social contexts and determinants in which the sufferer is embedded in that will inevitably shape decision-making. It places too much emphasis on the individual rational actor and misses the thick description of real-life illness (MacKian 2003). As Ward et al. (1997:21) point out in their work on STDs (where many lessons on health-seeking behavior come from),
“It is erroneous to adopt simplistic models of health-seeking behavior based on the following rationalistic assumptions: that symptoms of disease or 'risky' behaviors are always identified and/or defined in health terms; that recognition of symptoms will necessarily or automatically result in health-seeking behavior; that health-seeking behavior will always take the form that scientific medicine thinks is most appropriate”. The latter tradition is mostly based on socio-anthropological work that aims to solve this problem by focusing on barriers and determinants that stand between the patient and the service, such as cultural (e.g., the role of women in a society), social (e.g., caste discrimination),
economic (e.g., costs of care), geographical (e.g., access to care) and organizational (e.g., perceived quality of services) aspects (MacKian 2003). This is very much in line with the PRIME-related work of Lund et al. (2018) which aimed to link the SDG guidelines with determinants of mental disorders focusing on demographic, economic, neighborhood,
environmental, and socio-cultural levels, see Figure 4. Rather than sticking to determinants and barriers however, the health-seeking behavior concept aims to map the care pathway that the sufferer goes through in seeking care while taking these determinants and barriers into account.
Although the focus has been on formal healthcare systems, it is acknowledged that this pathway can involve more informal, non-biomedical sources of treatment, such as self-help, trusted community members, kinship, traditional healers, folk healers, faith healers, untrained allopathic doctors, religious rituals, spiritual gurus, etc. (Nichter 1991; Glover 2009; Kleinman 1980; Kleinman and Gale 1982; Tipping and Segall 1995; Thirthalli et al. 2016). Moreover, it is also acknowledged that the pathway is not always a logical flow of steps dictated by the labelling of mental disorders, that the sufferer takes to reach treatment. For instance, Emmerson et al. (2004) suggested that the usage of mental health systems might be better described as a maze or
a ‘non-system’, rather than a clear pathway. Consistently, Burns and Tomita (2015) argued that in talking about care pathways, we have to realize that they are very often non-linear, recursive, complex, and can involve multiple treatments simultaneously.
Figure 4. “Social and cultural determinants of mental disorders and the Sustainable Development Goals: a conceptual framework”, from Lund et al. (2018).
My current understanding of the literature is that as the interplay of theoretical and empirical understanding has developed along the years, the concept of health-seeking behavior has somewhat united the two traditions, in the sense that psychological models tend to also acknowledge the influence of contextual factors on health-seeking and pathway descriptions tend to be interested in the reasoning and decision-making of its actors as well. This is also how I view and use the concept. It is clear to me that health-seeking behavior cannot be understood in isolation from the contextual variables around it, and that these aspects will not only influence the health-seeking steps per se but will also determine how the illness is understood, processed, incorporated, communicated and turned or not into action.
In sum, in order to understand health-seeking behavior in mental healthcare, I have to start by understanding the illness responses of people when it comes to experiencing mental illness and then make the connection with the utilization of mental healthcare. This calls for an emic investigation at the levels of the individual and his/her community. In line with recent acknowledgements, I use health-seeking behavior to mean the process that the sufferer goes through in seeking healing, however unpredictable, chaotic, diverse and complex, and not assuming a clear-cut sequence of logical steps. As such, I am also not assuming models of decision-making a-priori, but rather aim to learn the mechanisms of health-seeking behavior in the specific circumstances of Chitwan, Nepal and how they changed or failed to change with PRIME.
Realist evaluation: an analytical framework
Fortunately, this is exactly what the realist evaluation strategy proposes, a theory-driven technique introduced by Pawson and Tilley (1997; 2001) and used increasingly in the last years to evaluate complex health interventions in specific real-life contexts (Pawson et al. 2004; Pawson, Walshe, and Greenhalgh 2005; Pawson 2013). In a realist evaluation, the underlying expected mechanism and impact of the intervention are first made specific (i.e., how the
intervention should work in theory). Then, empirical evidence is analyzed and contrasted to this proposed theoretical framework, called the program theory, to support, contradict or refine it. The results are a back-and-forth process of increasing understanding between the theory and empirical data with a special focus to the context(s) (C) in which the intervention is applied, the generative mechanism(s) (M) by which it works in practice, and the actual outcome(s) (O) it produces (i.e., the CMO configurations).
Perhaps the most relevant realization at this point is that public health interventions do not act as independent actors, or agents of change, but are embedded in the environment they are placed in (i.e., organizational, social, economic, political and cultural; Rycroft-Malone et al. 2012; Wong et al. 2012). They imply a complex interaction of internal (e.g., training, treatment, supervision) and external factors (e.g., poverty level, stigmatization, policy on psychotropic medication) that needs to be integrated (Pawson 2013). This interaction is relevant at every level of the health intervention, from its design, testing and implementation to its running time, up-scaling, evaluation and refining. Standard experimental designs used to evaluate clinical
interventions such as RCTs fall short in capturing such design complexity (Mackenzie et al. 2010).
The theory-driven family of evaluation strategies is a collection of methods that aims at incorporating this complexity by emphasizing on how and why an intervention works, besides just observing pre-post outcome measurements (Coryn et al. 2011; Funnell and Rogers 2011). Actually, the backbone of PRIME was built using another strategy from this theory-driven family, the Theory of Change (ToC; De Silva et al. 2014; Breuer et al. 2015). Introduced by Weiss (1995), the ToC framework starts by stating the long-term desired outcomes (i.e., what the implementation wants to achieve) and works its way back through defining a sequence of
intermediate outcomes that are necessary to happen in order to reach that end goal (Vogel 2012). The causal processes between these outcomes, as well as the assumptions and contextual factors that might influence them are made explicit.
In PRIME, the ToC framework shown in Figure 5 (open in original format at
http://www.prime.uct.ac.za/toc) followed three levels of implementation, the health organization level (represented in blue, e.g., Ministry of Health and WHO partnerships), the health facility level (represented in light green, e.g., primary health clinics PHCs, health posts HPs) and the community level (represented in dark green, e.g., community health workers CHWs, female community health volunteers FCHVs, community members). Although the three levels were designed to work in interplay, my primary focus is the community level as the closest unit of analysis that can penetrate into the intimate world of health-seeking behavior.
A major advantage of ToC is that it allows for a process of constant fine-tuning through regular discussions of multiple stakeholders on the levels of implementation, coupled with evidence-based feedback from the implementation site, making it an ideal tool for planning, designing and evaluating public health interventions as they run (Breuer et al. 2015). In contrast, the realist evaluation has received its merits especially for its strength in retrospectively
reconstructing and verifying the program theory and its mechanisms in light of the outcomes and context of the intervention, in other words its post-hoc evaluative value. As PRIME has just ended in March 2019, I consider the realist evaluation of PRIME to be an ideal extension, a fresh and critical angle to look at what PRIME did and did not manage to do.
To elaborate, the realist strategy is embedded in realism, a philosophical framework of understanding the world that has become a theory-driven methodological paradigm within social sciences. Realism sits somewhere between positivism (i.e., there is an objective reality that can be directly observed through the scientific method) and constructivism (i.e., we cannot directly observe reality because the reality we study is created by the collective human mind). It posits that social reality cannot be measured directly, but through thorough and systematic
investigations of the causal mechanisms, contexts and outcomes of its events it can be
understood (Pawson and Tilley 1997). As such, the purpose of realism is to understand how and under what conditions things work (McEvoy and Richards 2003).
Instead of controlling for contextual variability and human factors as it is usually done in evaluations of clinical interventions, the realist strategy acknowledges that complex applications of healthcare programs can only work through the human factor. It is the reasoning, decision-making and interactions of its actors and participants that makes a health intervention active (Pawson et al., 2005). This is important because analyzing PRIME through a lens that would not allow for such complex daily human interactions would compromise the analysis from the start. How the community reasons, responds and interacts with the new infrastructure of mental healthcare is the core unit of analysis.
This change in the reasoning and actions of the actors is described by how realist
evaluation uses causality. In the classic understanding of causality also used in RCTs, the causal relation between intervention X and effect Y is established when a change in X causes a change in Y. In contrast, the realist tradition writes about the generative model of causality, where in order to understand the causal outcome relationship between X and Y one needs to understand the underlying mechanism between X and Y, as well as the context of their relationship. The generative model asks, ‘what actually generates the change that the health intervention aims at?’, ‘what happens in the reasoning and actions of the actors and participants that causes an
intervention to work or not?’, ‘what are the psychological and social drivers of change in this particular setting that causes the mechanism to work or not?’. The realist evaluation thus
proposes and tests hypotheses of several CMO configurations, and the question of ‘what works?’ turns into ‘what is it about this program that works, for whom, in which regards, and in what circumstances?’ (Greenhalgh et al. 2011; Pawson, Walshe, and Greenhalgh 2005;
Rycroft-Malone et al. 2012). This, to me, is a much more interesting and valuable understanding for the development of policy in public mental health. In (Pawson, Walshe, and Greenhalgh
2005:S1:21) words, “A realist review does not provide simple answers to complex questions. It will not tell policy-makers or managers whether something works or not, but will provide the policy and practice community with the kind of rich, detailed and highly practical understanding of complex social interventions which is likely to be of much more use to them when planning and implementing programs at a national, regional or local level”.
For clarification, it should be noted that I am not using the realist evaluation methodology mot-a-mot as described by Pawson and Tilley (1997) as I did not do a systematic literature review on a specific topic with the purpose of synthesizing the evidence towards further policy. This would be the standard description of its strategy. Rather, I am doing an evaluation of PRIME focusing on the change in health-seeking behavior using the realist strategy. Of course, as part of the process I am reviewing PRIME literature, but the main focus is the intervention per se, and the main source of evidence is internal PRIME data. The end purpose of increasing the understanding of mental healthcare interventions and propose solutions for review, up-scaling, application in other settings or policy change is the same here as in the original description of Pawson and Tilley (1997).
Research questions
Therefore, the general research question (RQ) is ‘how does the implementation of mental healthcare into primary healthcare in a low-resource setting influence health-seeking behavior at the community level?’. The specific RQ is ‘how did PRIME influence health-seeking behavior at the community level in Chitwan, Nepal?’. To tackle this RQ, I operationalized the analysis through the framework proposed in Figure 6, that situates and organizes the research both conceptually and empirically. On a case-by-case, the framework starts with identifying the intermediate outcome(s) proposed by the PRIME ToC at the community level and verifies to what extent it has been achieved (O). It then examines the underlying generative mechanism(s) of why and how the outcome was met or not (M). Critically, how the specific circumstances of context shaped this explanation is then analyzed (C). Lastly, the CMO configuration that results is compared and contrasted with the original ToC program theory. Across this process, the guiding question is ‘what does this mean for community-level health-seeking behavior?’. Before
discussing the research findings, let me elaborate on the methodological details of the investigation.
Figure 6. The operationalization framework of analysis following the realist evaluation CMO configuration for PRIME-related change in community-level health-seeking behavior.
Materials, methods and data analysis
The setting of PRIME was the Chitwan district, part of Province nr. 3 with Bharatpur as the district capital, located in the Central-South region of Nepal touching on the border with India at about a 7h drive from the capital Kathmandu, see Figure 7. The total population of the district is 575.058 inhabitants, the literacy rate is about 79% (12% higher than the national average), the rate of people living in rural areas is about 73%, and the existing health facilities are 2 hospitals, 4 PHCs, 5 HPs, and 41 sub-health posts (sub-HPs), with 2 specialized
psychiatrists in the whole district (Lund et al. 2012). However, Chitwan has the advantage that some mental health services and specialists are available at the district hospitals and medical colleges. The targeted population of PRIME in the current data is 10 out of the 38 municipalities of the district, consisting of 108.368 inhabitants on an area of 342 km2, the targeted health facilities are 9 HPs and one PHC, the primary health-providing staff are medical officers, health assistants and auxiliary health workers, and the main primary care services provided are
outpatient care, immunization, family planning, natal and pre- and post-natal care. PRIME reported a total of about 4500 visits of mental health complaints across the 8 years.
Figure 7. Map of Nepal showing the position of the Chitwan district highlighted in red (source researchgate.net).
The current dataset includes both qualitative and quantitative data that are analyzed in tandem, see Figure 8 for an overview. The realist evaluation methodology is not bound to any specific method, allowing for such mixed-method analysis (Greenhalgh et al. 2011; Pawson, Walshe, and Greenhalgh 2005). The qualitative data can be most logically split into two batches based on the time of collection, while the quantitative data is one batch. The first qualitative batch was collected in 2011 as part of the baseline core qualitative assessment of the population of Chitwan used in the planning of PRIME. It focuses on the same 3 levels (i.e., health
organization, health facility, and community) and probes on 5 main themes (i.e., demand and access, delivery, recovery, accountability, and stakeholder views on mental health and research uptake). The data consists of 33 key informant interviews (KIIs) and 9 focus group discussions (FGDs). The KII was defined as “an individual interview with knowledgeable people in certain fields, usually taking about 1-2 hours”. The FGD was defined as an “interview that will involve a group of people belonging to the same level to discuss on certain issues that will take about 1-2 hours”. The probing followed similar patterns across the batch, of the type “what is the mental health situation in your community?”, “how aware are the people of mental illness?”, “is the community able to identify and refer cases and if so who should do it?”, “what are the main
barriers to receiving care, do they differ for disadvantaged groups?”, “is the idea of task-shifting a good one for health personnel and how much training do they need”, etc. (PRIME Core Qualitative Assessment, 2011; De Silva et al. 2016). Before the probing, case vignettes describing symptomatology were presented for depression, AUD, psychosis and epilepsy to freshen everyone’s knowledge and make the discussion vivid.
Figure 8. Overview of current data.
The second batch was collected as part of my ethnographic research in Nepal, from March to May 2019 as PRIME was ending its course. It constitutes of 10+ informal discussions of 1-2 hours each, both in groups and one-to-one (i.e., equivalent of KIIs and FGDs
respectively), 20+ hours of observation, a community sensitization meeting, usual conversation and my daily diary. This ethnographic data was informal only in the sense of no formal ethical clearance because PRIME was ending and there was no time for institutional approval. My audience included service users and their kinship, community healers, community leaders, FHCVs, health workers, and health professionals. Most of the meetings took place in 3 HPs and 1 PHC, but I also visited other clinics, hospitals, pharmacies and household healing centers across the Chitwan district. Managing locations and contacts was facilitated by the TPO at both locations (Kathmandu and Bharatpur). About half of the interactions could be managed in
English, and the other half were translated live from English to Nepali and vice-versa by my local translator, health professional and friend, compensated financially for the effort.
The quantitative data is part of the survey evaluation during the implementation phase of PRIME. Specifically, the longitudinal following of a cohort of service users assessed before the MHCPs in 2014-2015 (baseline), after 3 months (midline), and after 12 months in 2015-2016 (endline) in order to track individual-level changes in mental health. From a total of N = 2139 participants screened, N = 578 were identified as cases of mental ill health and investigated further. This sample included depression, AUD, psychosis and epilepsy diagnoses. Accordingly, the sampling method was consecutive or systematic sampling of diagnosed patients from primary care clinics. The data collection method was interviewer-administered questionnaire at baseline, midline and endline in Nepali. Besides detailed demographics, it probed on clinical history, treatment details, medication and adherence, disability assessment and functional impairment, economic activity, inpatient and outpatient care, community-based psychosocial help, and stigma (DISC-12; Thornicroft et al. 2009; De Silva et al. 2016).
For detailed longitudinal assessment, at some points I only focused on the depression and AUD cohorts as they had the most representative sample sizes and included control groups. From the N = 578 who screened positive, 36% (N = 209) were diagnosed with depression and 40% (N = 232) with AUD. The treatment sample sizes (post-cleaning) were N = 137 for
depression and N = 175 for AUD. These are participants who received treatment within PRIME-related facilities. The control groups were patients with either diagnosed depression N = 72 or AUD N = 57 who did not follow PRIME treatment but the usual care existing in the region. Depression was assessed through the Patient’s Health Questionnaire PHQ-9 tool (Kroenke, Spitzer, and Williams 2001), while AUD through the Alcohol Use Disorders Identification Test AUDIT (Babor et al. 2001).
In terms of data analysis, it was governed by the analytical framework described in Figure 6 and further detailed in the sequential scheme showed in Figure 9. This applied for both qualitative and quantitative data. Within the qualitative data, I organized, coded, re-structured, analyzed, and summarized primarily using NVivo (Castleberry 2014). The quantitative data was analyzed by means of descriptive statistics, frequency statistics, group mean comparison and regression analysis using SPSS (IBM 2017), but JASP (JASP Team 2018) and other free
software were used for visualization purposes. Triangulation between the types of data was a very important safety check across the analysis.
Figure 9. Scheme of the sequential analysis during realist evaluation.
Ethical considerations and reflexivity
The Code of Human Research Ethics and respective declarations of the United Nations, World Health Organization, American Anthropology Association, but also the National Ethical Guidelines for Health Research in Nepal And Standard Operating Procedures (Nepal Health Research Council NHRC, 2011) were strictly followed. The PRIME project has been running in Nepal for almost a decade, therefore having had Institutional Review Board (IRB) ethical clearance. A data use agreement form has been signed between myself Victor Andrei Bodiut representing the University of Amsterdam as data receiver and the University of Cape Town
(UCT) as data provider in May 2019, with the description “Analysis of the PRIME cohort study data (Nepal Cohort Study) and baseline qualitative screening (Core Qualitative Questionnaire)”.
All data were anonymized and treated confidentially, communicated only internally between me, University of Amsterdam, PRIME and TPO Nepal. I pledged to sharing no personal data of any subject in my work with any external parties unless agreed otherwise with the
supervision, or in case of life-threatening situations (e.g., immediate suicidal intention). The data was always stored on my personal password-protected laptop and external hard-drive, in a password-protected folder. I followed the UvA protocol strictly, including the guidelines on using the data for writing a master thesis. In case of publication, the regulations of the publisher will be followed upon agreement with co-authors. For the data collected by me, verbal informed consent was a strict requirement, and I aimed to inform my respondents to the best of my
knowledge about my role and my intentions.
My objective has always been for the findings and insights of this work to have a beneficial practical impact towards everyone involved, from the patients, to the community, to PRIME, to the Nepalese public policy, to mental healthcare generally. I argue that science (and global mental health implicitly) has the responsibility of aiming for a real, applied, positive impact for its population of interest. This is also why the policy recommendation is included in here. I strongly resonate with the notion of “do good, not only avoid doing bad” in research.
My general positionality towards the topic was described at the very start of the thesis. Regarding my experience in the field specifically, it was difficult to wear the coat of the
objective observer. I cared for the people, I wanted to get involved as much as I could, I wanted to help and in some cases I did. I tried to be attentive when forming very strong relationships with people and avoid promises I could not keep knowing that I will be there for a limited time only. But when such relationships did happen, I did not enforce them to end, but cherished them. This is normal human behavior and I would like to think that the empathy lead to ‘richer’ data, rather than ‘biased’ data. In fact, besides the confidentiality and respect in collaboration, it was the empathy that made people feel at ease enough to go deeper into their experiences. The last thing I wanted was to spread insecurity and coldness that would close people down, the feeling that I am an intruder rushing through their routines for quick data. The infrastructure and
facilitating the human interactions. Moreover, I considered the subgroup of people struggling with mental ill health to be a vulnerable group that required extra carefulness. Minimizing any risks of participation for this group was critical.
Then, my positionality was rather tricky to grasp at first for the people I collaborated with, especially the staff of the health facilities. They first assumed that I am part of the PRIME team, then they learned I come from the University of Amsterdam as an anthropologist, but I am also trained in psychology, and I am actually from Romania, and I planned to play tennis in Kathmandu, etc. All this information might have been confusing, but this sparked their curiosity more than anything. The multiple role stance was not easy to handle from my side either. At times I felt as part of the NGO team together with people close to my age, working together. At other times I felt like the lonely wolf sleeping in hostels and going through places and meeting new people. At other times I felt part of the health facilities, spending time with the health workers and getting to know their stories, how their kids went abroad for schooling and the sort. Add to this the knowledge in the back of my head that my background will shape the way I understand this new world and my internal urge to help as much as I can, and you will have a feeling of the fieldwork experience. I thus aimed to stay aware, focused, reflexive, organized, keep prejudgments and categorizations to a minimum, keep a strong diary and communicate with the people around me. At times, I just went with the flow.
Then, the fact that I had the translator always with me and she has been working in PRIME for years might have influenced the light in which the people colored PRIME in general. Perhaps more favorable because of her presence there. Perhaps not. I cannot know for sure. Also, the fact that she was there might on the one hand make people feel comfortable to talk more freely, or on the other hand make them more aware of the local perception. The translation per se certainly had an influence both on the interactions and on the data itself. There were times when I felt that much more has been said in Nepali than was translated in English, not intentionally, but lost in translation or reformulated in a way that was manageable for the translator. I aimed to clarify important terminology whenever possible and always asked if there was anything that was skipped when translating. Anyways, without my translator and the TPO to guide me, the whole fieldwork would not have been even remotely as fruitful.
Results and discussion
Based on the intermediate outcomes specified by the PRIME ToC at the community level (see again the darker green boxes in Figure 5), I identified 3 main themes that PRIME aimed to act upon, as follows: ‘Community members are adequately trained (FCHV healers, community health workers, mother groups, teachers)’ + ‘Community members are capable of Mental Health identification and Referral’ + “Functioning community support & intervention system in place’ → community preparedness and engagement in mental healthcare; ‘Community is aware and sensitized about Mental Health’ + ‘Decrease in stigma and discrimination’ → community judgment about mental ill health; ‘Family members are involved in care process’ + the assumption (in pink-coral color) that a ‘Functioning peer support group is in place’ → involvement of close-knit community in treatment. I will now treat each theme in a separate chapter.
Community preparedness and engagement in mental healthcare
Outcome(s)
Within the PRIME ToC, the ‘Community members are adequately trained (FCHV healers, community health workers, mother groups, teachers)’ outcome was indicated by the number of trained human resources and frequency of training, the ‘Community members are capable of Mental Health identification and Referral’ outcome was indicated by the number of people referred to health facilities and the number of cases positively diagnosed in the PHCs, while the ‘Functioning community support & intervention system in place’ outcome was
indicated by the number of trained community members and source of patients in health centers. The main intervention identified at this level is the training courses provided by PRIME together with the assumption that the community is participating in transporting the patients to the health facilities. I identified several levels of action targeted by PRIME that are responsible for the changes in health-seeking behavior at this level, as follows.
Training
Existing staff of the PHCs and HPs, health workers, nurses, midwives, together with key community volunteers and members were selected and trained by either psychiatrists, clinical psychologists or psychosocial counsellors and provided with diagnostic flowcharts for
depression, AUD, psychosis, epilepsy, maternal depression, and sometimes anxiety and bipolar disorders.
“A number of people were trained in mental health assessment, as prescribers and non-prescribers. There was a 9-day training program by the TPO Nepal and the PRIME project for prescribers, and a 5-day training for the non-prescribers. The psychiatric and clinical psychologist personnel trained us, psychiatrists trained prescribers, and
counselors trained the non-prescribers. For instance, the Healthy Activity Program (HAP) health model from India was used in training for severe depression.” – HP Staff
This was a large-scale plan, with MHCPs being implemented in 10 health facilities at first, then scaled up to 36 in the scaling-up phase, coordinating training for about 200-250 non-psychiatric personnel. The division between a prescriber and a non-prescriber was done based on the qualification or education level of the trainee.
“The nursing staff is the non-prescriber and the difference to the prescriber is the education degree. The nursing staff has no authority to prescribe the medicine, only the health workers. That’s why the nursing staff are usually the counselors.” – Health Worker Compared to the situation before PRIME where mental healthcare was basically
inexistent at the level of primary health facilities, the staff has gained confidence in dealing with mental health patients not only symptomatically, but systematically based on classificatory diagnostics.
“Before PRIME we didn’t know about mental health and disorders, and if patients came for treatment then we used general health techniques or sometimes refer them to the district hospital, but here we had no treatment or classification of the cases. And
sometimes we would treat the patients symptomatically, like if someone comes and says ‘I have a headache or body pain’ we would just treat the symptom. But we did not know how to deal with mental cases. Now we are like specialists.” – Health Worker.
Another primary health staff remembers:
“One guy frequently came to the health facility and talked and cried and shared his family problems but at that time I was not trained, and I didn’t know how to deal with
these cases and the types of problems he was facing. So, I suggested what I thought it was best for him, like ‘don’t do this, don’t do that’. A couple of days later, he committed suicide. I still remember that. If I was trained before then I would have managed it.” Still, as the period of 9-day training for prescribing psychotropic medicine and 5-day training for counselling and psychosocial support is very short, I was very curious of how
confident the trainees were in their task-shifting and task-sharing abilities. Importantly, everyone mentioned the importance of a solid supervision mechanism on a weekly or monthly basis and periodical refresher trainings besides the initial training per se, which by itself would not have been enough. Even so, there remain some struggles, but the general feeling is positive.
“We are satisfied with this training and duration, we can help depression through this training, but if we would have more advanced level training then we can also apply it to more difficult cases (…) most of the psychosis cases we feel difficulty with, because they are not stable and don’t reach the medicine, they stay locked inside their house.” – Health Worker
Identification and referral
This issue of reaching the sufferer when the sufferer does not reach the health post brought up the identification and referral system aimed at by PRIME. Implemented through FCHVs, mother’s group members, and sensitized key community members, one of the main tools has been the Community Informant Detection Tool (CIDT; Jordans et al. 2015; Jordans et al. 2017) with the scope of increasing community detection of people with mental health problems and refer them to the health facilities. The CIDT contains symptom-based case vignettes of the different disorders contextualized to rural Nepal that trained community
members who are familiar with the community can use to identify people in need of mental care, see Figure 10 for an example used in depression. When people are identified as potentially having symptoms of mental ill health, they are offered the option of visiting a trained health professional in the health facility. Notably, PRIME specifically acknowledged the access to care problem as a separate entity from the availability of mental health services, which proved to be one of its strengths.
Figure 10. Example of contextualized CIDT format used for depression in PRIME, translated from the original Nepali.
“We had a network of female community volunteers, who went to the homes of the patients and brought them to the health post. They would go to homes and discuss with the patients like ‘do you have any problem, what is your problem?’ but indirectly, always indirectly. Not like ‘you have a problem’, but more like ‘I’m here to talk to you about the health post’, and the conversation goes around it. They also have leaflets with
descriptions and pictures that they will spread around saying ‘If you have or see this type of behavior you can go to the health post’. Nowadays, teachers, local leaders, they also know about mental health and refer people.” – Health Worker
And this is how a referral system came into place, a referral system that increased in complexity as PRIME went on. This referral system is at the core of increasing health-seeking behavior, scaling up the flow of patients opting for treatment at the health facilities.
“The female health volunteers are the direct link to the community, to the people. They know people from all over the area, they know what people do, their activities. They know what their problems are, what types of symptoms there have, they know
everything. They are our referral mechanism. We have 9 female health workers here (in that village region).” – Nursing Staff
“When we started to implement the program, around that time, around 10-15 patients started to come each month. Now the number has decreased, because most of the people already received service and many are cured.” – Health Worker
This flow of patients was consistent with an increase in the use of local health facilities as a choice for seeking care across health problems (both physical and mental). When asked about where they usually go for health consultations, 43% of the participants in the cohort study reported local health centers, 29% private offices, 18% the health provider’s home and 10% hospital outpatient facilities at baseline. After 1 year of PRIME, there was shift towards seeking care at the local health centers more, while private homes decreased in popularity. The numbers were now 60% for the local health centers, 25% for private offices, 8% for the health provider’s home and 6% for hospital outpatient facilities. Thus, the local health facilities became the choice of health-seeking behavior for the majority of people in the community across complaints, pointing to the large importance and potential of routine healthcare systems.
In particular cases, this referral system manifests itself also between different health facilities, from health facilities to hospitals and vice-versa.
“Even when the medicine supply was late sometimes in some health facilities, in this period we referred the patients to another health facility to get continuous support and continuation of medicine”. – Head of one PHC
“Sometimes for the complicated cases we first take support from the psychiatrist by
Moreover, what was really impressive to find was that the referral system managed to incorporate other non-formal healing sources in the community, primarily traditional healers called ‘Dhami / Jhakris’.
“PRIME provided psychosocial and mental health-related training to so many Dhami Jhakris, and now there is a referral system between them and us. They do their treatment in their own way but once they cannot solve it they refer it here.” – Health Worker Treatment and treatment options
Once the patient is referred, s/he will most likely reach the health facility. The patient reaches thus another phase of the health-seeking behavior process. As mental healthcare at the local level has been a new endeavor for the community, it is important to recognize what exactly happens when a patient walks in the health facility, what are the treatment options available and how the process of treatment develops in time.
“So, an intake discussion is generally 5 to 10 minutes for a general medical problem. In case the patient shows mental health issues we take longer, usually 10-15 minutes, maybe half an hour if necessary. We ask what the problem is, we counsel also. Sometimes we have one mental case every day, but other times there is none.” – Head of one HP While describing a case of AUD, one health worker mentions:
“When a new case of alcohol abuse comes, I first identify if the patient used alcohol or not in that moment. So, I identify if the patient is stable or not stable. If s/he is stable, then we sit together and we talk first informally, in our culture you first have to talk informally like ‘hello, namaste, how are you?’ and that’s how we always would start a conversation here. Then I introduce myself to the patient and ask how the patient came to the facility – ‘are you personally interested to change your behavior, or your family forced you to come?’ (notice the key role of the family). After, I start to explain the advantages and disadvantages of using alcohol, we also use a brochure, I explain the brochure and try to educate them. And then the patients start to realize ‘oh, I have no advantages of alcohol, but many disadvantages’ and they start to share by themselves. I then ask about the economical aspect, because if a patient drinks for 20 years s/he will calculate the money lost on alcohol. They themselves realize then that they lost so many
things to alcohol. I make them realize and then I ask, ‘do you want to change your alcohol behavior or not’. The purpose is committing to treatment and change, so I can provide the necessary skills.”
Further, the following observation snippet shows an example of a follow-up treatment session, for a better understanding of the treatment process in the health facilities.
Figure 11. Observing a treatment session with a patient suffering from depression while recovering from psychosis. From left to right: patient, counselor, psychiatrist, counselor. Physicality was distorted on purpose for anonymity.
Usually, the reports indicated that patients start to share in-depth experiences about mental health symptoms after some sessions and not right from the start, which seemed to transfer from the cultural aspect of always starting conversations with friendly talk. The whole discussion is confidential unless there is immediate danger. The most common age groups are around 25 and 45. Following the intake session, the patient will come first weekly, then
bi-The treatment is ongoing at the moment of observation.
The patient is male, in his 40s, looks of Tibetan descent, the eyes are quite swollen.
He uses the hands a lot when talking, the hands are slightly shaking. He rubs his eyes quite often. The psychiatrist and psychological counselor listen carefully and are comforting. The psychiatrist watches his non-verbal behavior. He notes in the register from time to time, like checking symptomatology and progress.
The patient keeps a note diary.
He keeps talking and seems to share a lot of what is going on.
The medication is explained and most probably will be changed in dosage.
The tone of the psychiatrist and psychological counselor start to raise up, still friendly and comforting, but rather more decisive, pushing him for treatment continuation.
His mood looks low, but he nods and is receptive, ready to try and follow the treatment, listens carefully, seems to acknowledge the expertise of the health workers.
weekly, then monthly with the specification that patients can return whenever they feel bad. Medicine will be prescribed first for a few days, then prolonged when necessary, and will be procured with prescription from the pharmacies in the region. Moreover, primary healthcare is free of charge in Nepal (governmental policy), including the mental healthcare services
implemented through PRIME, which is a very big advantage in rural areas. Adherence and other challenges
If this seems to be all working smoothly and flawlessly, it is not the case. There have been challenges, unforeseen obstacles, drop-outs, relapses, discontinued medicine intake, and remote suicides. Service users sometimes find it hard to maintain treatment and come regularly, especially in cases of addiction. These challenges generally varied from case to case, from health worker to health worker, and from facility to facility.
“There are some challenges we remember. One case for instance, a female came to receive service in the health facility, the cause of her depression is that her husband suffers from HIV AIDS. She received counselling service but initially she did not share the husband’s problems. We asked what the main causes of her depression were, and she said, ‘I don’t know the cause, I have no problem, etc.’. But after 2-3 sessions she started to share her husband’s situation. Some cases already received support from other
hospitals or private clinics, and then again they came here for service, these are hard cases usually.” – Counselor
“There was an epilepsy patient who used to take a lot of cannabis and we provided treatment, he recovered quite well and then after some time he started again with the cannabis and symptoms started again. We have some cases where the people want to discontinue medicine, and then we need consultation if that is a good step or not, we don’t always know.” – HP Staff.
“In my experience, among the regular cases in this facility, the hardest to manage are the AUD cases because many patients go home and start drinking again, they relapse often. Then, we feel bad as service providers because the service users start again. The service users will not come back for treatment because they will feel guilty of relapsing after
heaving regular treatment and they will be scared to face eyes with us, they will feel guilty and ashamed.” – Counselor
Nevertheless, the general adherence level is high, I would say unexpectedly high. Patients reported good adherence both at baseline and endline across mental health disorders, in the range of 80-90%. Patients reported wanting to return to the health facility in most of cases. It should be noted here that the setting of formal survey administered by either a health professional or TPO staff might have acted as a pressuring factor for the respondents, even when confidential. However, the ethnographic reports seem to confirm high levels of adherence across PHCs.
“The total registered mental health cases here is about 90 and among them 4 or 5 cases have dropped out” (that is 5%). – Head of one PHC
“Most of the cases have good treatment adherence, and come regularly, and use more and more time per session also”. – Nursing Staff
“Initially, when he started the medication, it was difficult to manage the drinking. The urge to drink and shaking happened often. He thought he couldn’t manage it, so he came again to the health facility immediately. This was a very good decision. He then visited again, the health workers counseled me and worked on skills to manage the urge and sleeping problems. Then, he felt better, supported more”. – Counselor about Service User
Generative mechanism(s)
Several mechanisms have been identified as responsible for the change in health-seeking behavior and the general success of the mental healthcare system in the community. I interpret these mechanisms in light of their generative power in changing the reasoning, decision-making and actions of people, and not in their straightforward description. As such, the community detection campaigns made the switch from ‘I have no problem, or I do not know what I am suffering from’ to ‘I might have these symptoms that fit with the description of X’. The
increased availability and referral of services made the switch from ‘There is nowhere I can go’ to ‘I’ve heard about the local health post offering mental health services’. The affordability of costs made the switch from ‘Even if I want to go, I cannot pay for it’ to ‘The services are free or very cheap so I could give it a try’. The identification and increase in diagnosed patients made the switch from ‘I am the only one suffering of this’ to ‘Apparently others share my symptoms
