Improving data improving health: Verbal Autopsy for health systems strengthening

The majority of deaths in Africa and Asia are never recorded. Verbal autopsy (VA) is a method to determine causes of death using an interview with witnesses of the death, data from which are interpreted to conclude medical causes. In this seminar, we will reflect on innovations in VA including automated and mobile VA, and examine contextualised approaches to understanding why people die outside facilities or without registration. We will also explore how information from VA can be made available as part of a health systems approach. The aim of the event is to consolidate methods that have the potential to bring together communities and policy makers with better data and improved capacity for evaluation. Join participants from the Information, Evidence and Research Unit at WHO, the Umeå WHO Collaborating Centre for VA, the Centre for Global Development at the University of Aberdeen, the Malaria Consortium, the Africa Health Research Institute, University College London Institute for Global Health and Lancet Global Health. Image credit © 2013 GMB Akash, Courtesy of Photoshare

---

PROGRAMME

0900 Coffee and Registration

0930 Welcome and opening remarks | Prof Ibrahim Abubakar UCL + Dr Zöe Mullan, Lancet Global Health

0945 Keynote: Monitoring mortality and causes of death in the SDG era | Dr Ties Boerma, WHO

1000 Keynote: Learning from mortality: pragmatic approaches to verbal autopsy for better health | Prof Peter Byass, Umeå Centre for Global Health Research

1020 Mobile VA, technological advances and ethical priorities | Dr Edward Fottrell, UCL

1040 Coffee

1100 Keynote: Social Autopsy, a method to examine barriers to health care, risky behaviours and missed opportunities for health intervention | Dr Karin Källander, Malaria Consortium

1120 Verbal Autopsy and Participatory Action research (VAPAR): People-centred and action-oriented methods in South Africa | Dr Lucia D’Ambruoso, University of Aberdeen

1140 Interfacing: making connections within and between states, through data dynamics, and with the international community | Dr Kobus Herbst, Africa Health Research Institute

1200 Panel discussion, Closing remarks

1230 Lunch

---

SPEAKER PROFILES

Dr Ties Boerma Director Department of Information, Evidence and Research World Health Organization

Dr. Boerma is the World Health Organization's Director of Information, Evidence and Research. He has over 30 years of experience working in global public health and research programmes, including 10 years at national and districts levels in Africa. He has worked for different United Nations organizations, bilateral donors, national governments and research institutions and has published extensively on AIDS, maternal and child health in epidemiological, demographic, and public health journals. A national of the Netherlands, he received his degree in medicine (MD) from the University of Groningen, and a PhD in medical demography from the University of Amsterdam.

Professor Peter Byass Director Umeå Centre for Global Health Research, Umeå University

Professor Byass has worked in Global Health for 30 years, particularly around Africa. Professor Byass likes to measure health in meaningful ways and expose some of the massive global inequities in health and disease.

Dr Lucia D'Ambruoso Lecturer in Global Health Centre for Global Development, University of Aberdeen

Dr D’Ambruoso is a social scientist interested in participatory methodologies and health surveillance for health system strengthening. Dr D’Ambruoso is lecturer in global health at the University of Aberdeen, a Global Affiliate of the Umeå Centre for Global Health Research and an honorary lecturer at the University of the Witwatersrand in South Africa.

Dr Edward Fottrell Senior Lecturer in Epidemiology and Global Health Institute for Global Health, University College London

Dr Fottrell is an epidemiologist with expertise in health measurement in resource-poor settings, community-based intervention research and population surveillance data processes. Since 2010 Ed has been the technical lead of experimental and quasi-experimental health and behaviour change interventions in Bangladesh. He has contributed to mortality and ‘verbal autopsy’ surveys in various settings, including Ethiopia, India and Somaliland. Recent research has focussed on digital data capture in VA, and the opportunities and ethical challenges this creates.

Dr Kobus Herbst Deputy Director Africa Health Research Institute

Dr Herbst is a public health consultant and data scientist with experience in managing projects in both the public health and health research sectors. At the Africa Health Research Institute, he was responsible for one of the earliest publications showing the decline in population mortality following the introduction of anti-retroviral treatment. He worked with Prof Peter Byass on the evaluation of the InterVA-4 automated verbal autopsy instrument and the curation of a large dataset from 21 INDEPTH sites on cause-specific mortality. Dr. Herbst leads the INDEPTH research data management programme and is the founder of the INDEPTH Data Repository, the first of its kind specializing in longitudinal population data from LMICs. (http://www.indepth-ishare.org).

Dr Karin Källander Senior Research Advisor Malaria Consortium

Dr Källander leads research on maternal, neonatal and child health, including mobile health (mHealth). She is a specialist in health systems research, child health epidemiology and control of pneumonia, malaria and other infectious diseases. Dr Källander has over 14 years’ experience as researcher, lecturer, programme coordinator, and consultant in both development and emergency settings, including seven years living and working in a low-income country. She has a deep understanding of the health sector and working with governments and policy makers in Africa to achieve desired outcomes.

---

SESSION SUMMARIES

Keynote: Monitoring mortality and causes of death in the SDG era | Dr T Boerma, WHO

Cause-specific mortality is the foundation of public health and mortality targets are prominent in the Sustainable Development Goals 2016-2030. Data derived from death registration systems with medical certification and coding using the International Classification of Diseases 10th revision are considered the standard. Coverage and quality of such systems however is low in over 100 countries and an estimated two-thirds of deaths go unrecorded globally. To address these gaps death registration systems, as part of civil registration and vital statistics (CRVS) systems, need to be strengthened and new ways to obtain cause of death information are being developed. This includes wider application of verbal and social autopsy methods in the context of for instance improving CRVS systems, which has considerable potential to improve the available information on cause of death patterns in high mortality countries.

Keynote: Learning from mortality: pragmatic approaches to verbal autopsy for better health | Prof P Byass, UCGHR

In much of today’s world, individual deaths are unlikely to be registered – yet cause of death registration is the obvious entry point to understanding population health problems and needs. New approaches to verbal autopsy make this possible in principle – but need widely implementing.

Keynote: Social Autopsy, a method to examine barriers to health care, risky behaviours and missed opportunities for health intervention | Dr K Kallander, Malaria Consortium

Effective implementation of child survival interventions depends on improved understanding of cultural, social, and health system factors affecting utilisation of health care. Standardised instruments now exist for collecting and interpreting information on common barriers to health care, risky behaviours, and missed opportunities for health interventions in deceased children under 5 years old and pregnancy related deaths. One such tool is the Social Autopsy method. In this session, Dr Källander will describe the methodology, development, and summarise results of how this method can be used to help researchers, implementers and district stakeholders to understand where health care delays take place, which interventions can be most effective to prevent deaths, and which harmful behaviours most urgently need to be addressed through supervision or behaviour change communication.

Verbal Autopsy and Participatory Action research (VAPAR): People-centred and action-oriented methods in South Africa | Dr L D'Ambruoso, University of Aberdeen

This presentation will report on the development of a people-centred health systems research method called VAPAR (Verbal Autopsy and Participatory Action Research). VAPAR is a method with three main elements. First, VA is extended to describe mortality in terms of social and health systems circumstances, in order to complement information on medical causes. Second, the extended VA data are further verified and amplified with local knowledge gained via a Participatory Action Research (PAR) process in which communities organise evidence for action. Thirdly, and throughout, the method is embedded in the health system at different to ensure practical relevance.

Mobile VA, technological advances and ethical priorities | Dr E Fottrell, UCL

Using mobile phones to combine simplified VA data collection instruments with automated methods to assign cause of death creates an opportunity to greatly increase the coverage and timeliness of cause-specific mortality data. Drawing on recent field experiences, this talk will describe advances in mobile phone implementation of VA and highlight key operational and ethical challenges, as well as opportunities for public health action, that are emerging.

Interfacing: making connections within and between states, through data dynamics, and with the international community | Dr K Herbst, Africa Centre for Population Health

Verbal autopsy methodology is increasingly moving from the confines of health and demographic surveillance systems (HDSS) to CRVS systems. HDSSs are increasingly confronted by the need to not just observe local populations but to provide an intervention framework to respond to those observations. I will draw on my experience over 16 years in managing a HDSS site in an area heavily affected by the HIV pandemic, in conjunction with my experience in promoting and implementing data sharing in the INDEPTH Network of more than 50 HDSSs to reflect on the interfaces between HDSS and local health services, and HDSSs as data producers in a national and international context with data users as it pertains to verbal autopsy derived cause specific mortality data. The presentation will reflect on the relevance of these interfaces in increasing the relevance of VA at national and local level.