Verbal Autopsy with Participatory Action Research (VA-PAR): Developing a people-centred health systems research methodology

Lead Research Organisation: University of Aberdeen
Department Name: Division of Applied Health Sciences

Abstract

Recent estimates suggest that two-thirds of the world's deaths pass unrecorded. This situation seriously limits the ability of health systems to respond to the needs of vulnerable and excluded populations. Developing methods to reliably understand why people die in populations with weak health systems is therefore an important strategy for saving lives.

Registering medical cause of death has long been considered essential for public health. Despite increasing globalisation, in many resource-poor countries, universal registration of vital events remains lacking and uncertain estimates provide an inadequate basis for policy and planning. In addition, the social inequalities and social contexts play an important role in shaping health for disadvantaged groups. Even with the data available, insufficient attention is paid to the root causes of mortality in resource poor settings.

Considering these factors, there is an urgent need to improve health information about and for marginalised populations to inform public health responses. The proposed research will develop an extension to an established method called Verbal Autopsy (VA). VA is used to measure the levels and causes of death in populations where large numbers die outside health facilities or without registration. The development will help local health systems to assess their own health situations, identify priorities and develop action plans for positive health change.

We will do this through three phases of work.
- Firstly, we will develop improved ways to classify causes of death by combing information on medical causes with data on the circumstances of deaths (seeking and using care at the time of death). In settings where health services are under-funded, weak and fragmented, these can often play a crucial role. The classifications will also be developed in consultation with local health planners to be of practical use.
- Secondly, we will develop an additional method to gain the views of local communities on long-standing health challenges. This will allow us to further understand how social, economic and health systems issues influence availability, accessibility, acceptability, and quality of care. The method may also help foster social inclusion in health.
- In the final phase, we will consult with higher levels of the health system about the method. This will help develop how we use the extended cause of death classifications combined with community knowledge. The aim here is to explore how the method could be used in an ongoing fashion to connect health surveillance to service organisation in an inclusive process. This approach encourages sustainable health gains.

Data, poverty and inequality exist in complex co-dependency: less data exist on the health of the poor than the rich, raising important questions about the relationship between material and data poverty. In settings where health systems are fragile and under-resourced, where new burdens of disease are rapidly emerging, and where large and diverse populations are excluded from access to health care, innovative approaches that connect the registration of vital events to health care systems in a people-centred approach are needed.

The approach employs a bottom up philosophy connecting with population data at source. In the longer term, it is envisaged that the method will contribute to more rigorous health data at population level in an inclusive process that can affect sustainable health gains through better data and improved capacity for evaluation.

The work will be conducted in an research centre in rural South Africa established for over 20 years. The extent of data available and the richness of experience in health research allows us to develop a method with partners who enact a broader commitment to registration of all individuals within a population.

Technical Summary

People-centred health systems (PCHS) is a recent progressive shift that has moved thinking beyond building-blocks models of health systems towards ones that centralise a human and relational nature. Despite the conceptual advance, empirical methods are lacking. The project seeks to develop methods for conducing and using Verbal Autopsy (VA) consistent with a PCHS approach by combining VA with Participatory Action Research (PAR) in a process connected to the health system at different levels. VA is a health surveillance technique that provides information on levels and causes of mortality in populations where deaths occur outside facilities and/or without registration. PAR is a process that aims to transform the roles of those participating from objects of research to active researchers and agents of change. It systematises local experience through collective analysis to generate valid forms of evidence on the relationships between health problems and their causes. Three phases of research are proposed. In Phase 1, we will conduct a secondary analysis of data gained through the application of the 2012 WHO VA standard in a Health and Demographic Surveillance Site (HDSS) in rural South Africa. Combing data on medical causes with new data on background characteristics of deaths, we will develop improved ways to classify causes in a method suitable for use at sub-district/district level. In Phase 2, local service users and providers will engage in a PAR process to review the results of Phase 1, set priorities for local services, and explore the potential for co-benefits related to empowerment and social inclusion. The final Phase 3 aims to consult at higher levels of the health system to consider how the method could be further applied and evaluated. The overall output is a practical and integrated methodology based on core standards that is contextually relevant and capable of affecting health gains by translating local priorities into actionable public health agendas.

Planned Impact

The ultimate beneficiaries of the proposed research are populations in which large numbers die outside health facilities and/or without proper registration. Estimates suggest that over two thirds of deaths worldwide pass without registration of cause, with over three quarters of these in Sub-Saharan Africa and South-East Asia. Fragilities in health systems are exacerbated by poor civil registration and vital statistics systems (CRVS): without an understanding of the situations of poor, marginalised and excluded groups, the problems they face can neither be identified nor addressed. In this context there is an urgent need to improve health information about marginalised populations.

The research is also intended to have a positive impact on local communities, policy-makers and health planners within national, provincial, district and sub-district health offices. Information on cause-specific disease profiles, their health systems and social determinants, and priorities for actionable public health agendas will have been rigorously verified at local community and systems levels. Co-constructed practical knowledge will be thoroughly embedded in, and thus relevant to, local health systems contexts. These outputs will be disseminated in local communities and to health authorities to investigate their validity and relevance, and to reflect on the method as a means of sustainable and self-administered evaluation and planning.

In addition, the research responds to the World Health Organization (WHO)'s current effort to develop CRVS systems in low and middle-income countries (LMICs). The proposed research will provide the piloted VA-PAR method promoting democratic knowledge production, capacity building and evidence-based advocacy. In this sense, the research is also intended to provide important information for WHO on the adaption of VA-based models as a medium term strategy towards universal vital registration. To enable these impacts, methodological and technical discussions on a) extended biosocial classifications for unregistered deaths and b) the explanatory potential of routine participatory adjuncts in health and demographic surveillance will be conducted with WHO. We will hold dissemination workshops with WHO in which the methods will be discussed in terms of the potential for sustainable and self-administered evaluation and planning.

The Development Grant work also aims to facilitate further application and evaluation of the method among wider audiences. We therefore propose to hold two further workshops in South Africa and the UK to engage with academic and non-academic audiences on further application and evaluation of the method. Firstly, with the International Network for the Demographic Evaluation of Populations and Their Health (INDEPTH), a southern-based, southern-led umbrella organisation of 50 independent health research centres that operate 48 HDSS sites in 20 LMICs (http://www.indepth-network.org). The meeting will consider how and for whose purposes health information is generated for sustainable action towards improved health outcomes. This workshop will also explore the potential to create further spaces for application and evaluation of the method. Secondly, in the UK through the Public Engagement in Research Unit (PERU) at the University of Aberdeen (http://www.abdn.ac.uk/engage/), we will conduct a national workshop on the relationships between material and data poverty, and on methods and opportunities for generating routine people-centred health information. Here we will also explore further application and evaluation of the method to enable the impacts envisaged.

Publications


10 25 50
D'Ambruoso L (2017) The case for verbal autopsy in health systems strengthening. in The Lancet. Global health
 
Description Disseration projects (n=2) linking postgraduate students from the University of Aberdeen to MRC/Wits Rural Public Health and Health Transitions Research Unit (Agincourt), South Africa
Geographic Reach Multiple continents/international 
Policy Influence Type Influenced training of practitioners or researchers
 
Description Launch if postgraduate taught course on Health Policy and Systems Research
Geographic Reach National 
Policy Influence Type Influenced training of practitioners or researchers
Impact In academic year 2015/16 the postgraduate taught course BU5568: Health Systems and Policy Research was launched at the University of Aberdeen. The course is led by labour economist Professor Ioannis Theodossiou and social scientist Dr Lucia D'Ambruoso as a cross-college collaboration. The course aims to take a global perspective on health policy and systems research (HPSR), providing students with knowledge and understanding of the key characteristics of health policy, health systems, and the contribution of research in this field. The course covers a wide variety of topics including health systems around the world, the effects of social inequalities on health inequalities and their repercussions for social policy, issues of public health policy, and how research can make credible contributions in this context. The principal aims are: • To introduce different health systems: what they are, how they are defined and financed and to encourage a critical and comparative perspective; • To explore health inequalities and health and social policy issues; • To provide an overview of HPSR. The course has also been registered with the Health Systems Global Training database. The course has been developed in response to work in the field of Health Policy and Systems Research, which has been through the Verbal Autopsy with Participatory Action Research (VAPAR) methodological development work (DFID/MRC/Wellcome Trust/ESRC Health Systems Research Initiative (MR/N005597/1) ).
URL http://www.abdn.ac.uk/msc/global-health
 
Description Placement project (n=1) linking University of Aberdeen postgraduate students to MRC/Wits Rural Public Health and Health Transitions Research Unit (Agincourt), South Africa
Geographic Reach Multiple continents/international 
Policy Influence Type Influenced training of practitioners or researchers
 
Description Making the Most of the Masters/Careers Service University of Aberdeen
Amount £500 (GBP)
Organisation University of Aberdeen 
Sector Academic/University
Country United Kingdom of Great Britain & Northern Ireland (UK)
Start 06/2016 
End 08/2016
 
Title Extending Verbal Autopsy (VA) to Collect Information on Local Circumstances 
Description Verbal autopsy (VA) is a health surveillance technique used in low and middle-income countries to determine medical causes of death (CODs) for people who die outside hospitals and/or without registration. By virtue of the deaths it investigates, VA is also an opportunity to examine social exclusion from access to health systems. The aims were to develop a classification system for deaths investigated in VA cognisant of broader social and health systems determinants. A short set of questions on social and health systems circumstances at and around the time of death were developed and integrated into the WHO 2012 short form VA (SF-VA). Data was analysed from two census rounds 2012-13 in the Agincourt HDSS in rural South Africa where the SF-VA had been applied. InterVA-4 was used to develop cause-specific mortality fractions (CSMFs). Descriptive analyses were performed examining whether and how the new indicators varied by medical conditions and age/sex sub-groups. This work was conducted as part of the Verbal Autopsy with Participatory Action Research (VAPAR) methodological development work (DFID/MRC/Wellcome Trust/ESRC Health Systems Research Initiative (MR/N005597/1) ), Phase 1 
Type Of Material Model of mechanisms or symptoms - human 
Year Produced 2012 
Provided To Others? Yes  
Impact New ways to understand deaths investigated in VA: An analysis of 1,200 deaths investigated in Agincourt with VA containing new WHO indicators on social and health systems factors has been used to develop a novel classification system. The system complements data on medical causes and has identified multiple, reinforcing barriers to access including unaffordable costs, unavailable transport and under-utilisation of facility-based care. Supplementing VA with questions on background circumstances provides complementary information to CSMFs relevant for health planning. 
 
Title Participatory Action Research (PAR) to gain local knowledge on health and illness, including Photovoice 
Description The method aims to develop community partnerships to co-construct and act on new knowledge on health in rural settings. The method provides a means to engage with communities in a participatory process to review mortality data and extended COD classifications from Agincourt HDSS, to further amplify these data with local knowledge, to set priorities for local services, and explore the potential for benefits related to empowerment and social inclusion. The method has been developed as part of the Verbal Autopsy with Participatory Action Research (VAPAR) methodological development work (DFID/MRC/Wellcome Trust/ESRC Health Systems Research Initiative (MR/N005597/1) ), Phase 2. 
Type Of Material Model of mechanisms or symptoms - human 
Provided To Others? No  
Impact - PAR process with local communities: Using the new VA outputs in a deliberative process with village-based groups, new insights have been gained on specific mechanisms through which health risks are accumulated and access to services is limited. Issues identified include transactional sex, harmful traditional practices, alcohol/substance abuse and debt, compounded by denial of treatment, insensitive staff and poor confidentiality in clinics. Photovoice was also used to explore participatory photography - Partnerships: the process has been developed with Mpumalanga provincial health authority and local communities in Agincourt. The partnerships are evidenced in a paper, conference abstract and the current application co-authored by health authority representatives and researchers. The health authority also plans to initiate a community health forum responding to needs expressed in the PAR element. 
 
Description Collaboration with ASH Scotland (Taking Action on Smoking and Health) 
Organisation Action on Smoking and Health (Scotland ASH)
Country United Kingdom of Great Britain & Northern Ireland (UK) 
Sector Charity/Non Profit 
PI Contribution The Verbal Autopsy with Participatory Action Research (VAPAR) methodological development work (DFID/MRC/Wellcome Trust/ESRC Health Systems Research Initiative (MR/N005597/1) ) has underscored the need to do research that generates information that is relevant to policy/practitioners. An ongoing collaboration with ASH Scotland has been further fostered by this awareness. ASH Scotland do important work nationally and internationally to reduce harm caused by tobacco. Working with the Chief Executive ASH Scotland through an annual seminar generously provided to the University of Aberdeen has provided important insights on how to conduct research for policy and practice, and engage more broadly to maximise impacts. The collaboration with ASH Scotland has been prioritised this year to extend insights from the group to the VAPAR work.
Collaborator Contribution Annual seminar
Impact Annual seminar at the University of Aberdeen: Creating a tobacco free generation: how far have we come from and where to next? 16 February 2016 Sheila Duffy, CEO, ASH Scotland
Start Year 2014
 
Description Collaboration with Aberdeen City Council 
Organisation Aberdeen City Council
Country United Kingdom of Great Britain & Northern Ireland (UK) 
Sector Public 
PI Contribution Methods developed in the Verbal Autopsy with Participatory Action research (VAPAR) Development Award DFID/MRC/Wellcome Trust/ESRC Health Systems Research Initiative (MR/N005597/1), have been applied locally in a dissertation (Masters) project. Conducted in the Summer of 2015, this project engaged with community members in a deprived area of Aberdeen City to examine food poverty and insecurity in the context of the use of food banks using participatory methodologies developed in the VAPAR work. More broadly, the work reflects an awareness of the importance to link research projects with health and other authorities to ensure from the outset that research is designed to provide information that is of practical relevance, as part of a health systems and policy research approach.
Collaborator Contribution The Community Learning and Development Officer, Ms Emma McPherson helped gain access to the local community, which whom a series of facilitated discussions were held on the nature and extent of food insecurity.
Impact Student dissertation: Household food insecurity and food poverty in North East Scotland: a community-based participatory research study. A thesis presented for the degree of Master of Science (MSc) Global Health and Management, University of Aberdeen. July 2015. Student: Khyati Thaker, Bachelor of Physiotherapy, Rajiv Gandhi University of Health Sciences, India
Start Year 2015
 
Description Collaboration with Mpumalanga Department of Health Maternal Child Women and Youth Health (MCWYH) Directorate (Province) 
Organisation Mpumalanga Department of Health
Country South Africa, Republic of 
Sector Public 
PI Contribution The Verbal Autopsy with Participatory Action Research (VAPAR) methodological development work (DFID/MRC/Wellcome Trust/ESRC Health Systems Research Initiative (MR/N005597/1) ) has been developed with Mpumalanga provincial health authority and local communities in Agincourt. The partnerships are evidenced in a paper, conference abstract and the new application to Call 3 of the Health Systems Research Initiative co-authored by health authority representatives and researchers (D'Ambruoso et al, 2015; 2016a; 2016b). The health authority also plans to initiate a community health forum responding to needs expressed in the PAR element (Personal communication, 2015). D'Ambruoso L, Kahn K, Wagner R, Twine R, van der Merwe M. Moving from medical to health systems classifications of deaths. Manuscript submitted to BMC Glob Health Res Pol, 2015. D'Ambruoso L et al. Capturing community priorities. Fourth Glob Symp Health Sys Res; 2016a. D'Ambruoso L et al. Verbal Autopsy with Participatory Action Research (VAPAR): expanding the knowledge base through partnerships for action on health equity. Outline Application to DFID/ESRC/MRC/Wellcome Health Systems Research Initiative; 2016b. Personal Communication. S Masinga Mpumalanga Department of Health, 2015.
Collaborator Contribution Partners at the Department of Health have contributed through facilitated discussions, through co-authoured manuscripts for publication in international journals and through development of a funding proposal.
Impact PAPERS D'Ambruoso L, Kahn K, Wagner R, Twine R, van der Merwe M. Moving from medical to health systems classifications of deaths. Manuscript submitted to BMC Glob Health Res Pol, 2015. CONFERENCE ABSTRACTS D'Ambruoso L et al. Capturing community priorities. Fourth Glob Symp Health Sys Res; 2016a. FUNDING APPLICATIONS D'Ambruoso L et al. Verbal Autopsy with Participatory Action Research (VAPAR): expanding the knowledge base through partnerships for action on health equity. Outline Application to DFID/ESRC/MRC/Wellcome Health Systems Research Initiative; 2016b. EXPRESSED INTENTION BY PROVINCIAL DEPT. OF HEALTH TO INITIATE A COMMUNITY-EDUCATION FORUM AS A RESULT OF THE VAPAR DEVELOPMENT WORK Personal Communication. S Masinga Mpumalanga Department of Health, 2015.
Start Year 2015
 
Description Poverty as a human rights issue: Rights, social participation and ways of increasing children's voice in the debate Scottish Universities Insight Institute Friday, 26 June 2015 from 10:00 to 16:00 (BST) Aberdeen, United Kingdom 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Year(s) Of Engagement Activity 2015
 
Description Government Consultation - Scottish Government International Development discussion event - Glasgow University's Randolph Hall - Wednesday 9 March 2016 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Year(s) Of Engagement Activity 2016
URL https://www.eventbrite.co.uk/e/scottish-government-international-development-consultation-academia-e...
 
Description Lecture as part of the Queen Margaret University Grad Challenge Lecture Series at the Institute of Global Health and Development 25th February 2016 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Year(s) Of Engagement Activity 2016
URL http://www.qmu.ac.uk/iihd/whats%20on/whats_on.htm#DAmbruoso