01 Manual And Guidelines For Application Use Of Simplified WHO VA Tool 2016v1 5

01_Manual%20and%20%20guidelines%20for%20application%20and%20use%20of%20simplified%20WHO%20VA%20tool_2016v1-5

01_Manual%20and%20%20guidelines%20for%20application%20and%20use%20of%20simplified%20WHO%20VA%20tool_2016v1-5

User Manual:

Open the PDF directly: View PDF .
Page Count: 37

Download
Open PDF In Browser	View PDF

Verbal autopsy standards:
The 2016 WHO verbal autopsy instrument
V1.5

WHO Library Cataloguing-in-Publication Data
Verbal autopsy standards: the 2016 WHO verbal autopsy instrument.
© World Health Organization 2017

Some rights reserved. This work is available under the Creative Commons AttributionNonCommercial-ShareAlike 3.0 IGO license
(CC BY-NC -SA 3.0 IGO creativecommons.org/licenses/by-nc-sa/3.0/igo/).
Under the terms of this license, you may copy, redistribute the work for non-commercial
purposes, provided the work is appropriately cited, as indicated below. In any use of this work,
there should be no suggestion that WHO endorses any specific organization, products or services.
The use of the WHO logo is not permitted. You are not allowed to adapt the work. If you create a
translation of this work, you should add the following disclaimer along with the suggested
citation: “This translation was not created by the World Health Organization (WHO). WHO is not
responsible for the content or accuracy of this translation. The original English edition shall be
the binding and authentic edition”.
Any mediation relating to disputes arising under the license shall be conducted in accordance
with the mediation rules of the World Intellectual Property Organization..
Suggested citation. Manual for the training of interviewers on the use of the 2016 WHO VA
instrument, Geneva: World Health Organization; 2017; License: CC BY-NC-SA 3.0 IGO.
Sales, rights and licensing. To purchase WHO publications, see
http://apps.who.int/bookorders. To submit requests for commercial use and queries on rights
and licensing, see http://www.who.int/about/licensing.
Third-party materials. If you wish to reuse material from this work that is attributed to a third
party, such as tables, figures or images, it is your responsibility to determine whether permission
is needed for that reuse and to obtain permission from the copyright holder. The risk of claims
resulting from infringement of any third-party-owned component in the work rests solely with
the user.
General disclaimers. The designations employed and the presentation of the material in this
publication do not imply the expression of any opinion whatsoever on the part of WHO
concerning the legal status of any country, territory, city or area or of its authorities, or
concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps
represent approximate border lines for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that
they are endorsed or recommended by WHO in preference to others of a similar nature that are
not mentioned. Errors and omissions excepted, the names of proprietary products are
distinguished by initial capital letters.
All reasonable precautions have been taken by WHO to verify the information contained in this
publication. However, the published material is being distributed without warranty of any kind,
either expressed or implied. The responsibility for the interpretation and use of the material lies
with the reader. In no event shall WHO be liable for damages arising from its use.

The mark “CDC” is owned by the US Dept. of Health and Human Services and is used with
permission. Use of this logo is not an endorsement by HHS or CDC of any particular product,
service, or enterprise.

.

Acknowledgements
The production of this manual would not have been possible without the collaboration and
support of numerous organizations, institutions and individuals. The organizations
contributing to this work include World Health Organization (WHO), Bill & Melinda Gates
Foundation, Bloomberg Philanthropies Data for Health (D4H), Health Metrics Network
(HMN), INDEPTH Network, Institute for Health Metrics and Evaluation (IHME), All India
Institute of Medical Sciences, Australian National University, India National Institute of
Medical Statistics, International Centre for Diarrhoeal Disease Research (ICDDR), US Centres
for Disease Control and Prevention (CDC), Federal University of Minas Gerais, Ghana Health
Service, Ifakara Health Institute, London School of Hygiene & Tropical Medicine (LSHTM),
Johns Hopkins Bloomberg School of Public Health, National Institute of Health Research
Indonesia, National Institute of Medical Statistics of India; Norwegian Institute of Public
Health, Swiss Tropical and Public Health Institute, Ohio State University, Thailand Ministry of
Public Health, The University of Queensland, UCL Centre for International Health and
Development, Umeå University, United Nations Population Fund (UNFPA), Office of the
United Nations High Commissioner for Refugees (UNHCR), University of Alexandria,
University of the Witwatersrand, Uttar Pradesh Center for Maternal, Neonatal and Child
Health
For the 2014 and 2016 version of the VA instrument, acknowledgement is given to the WHO
working group on Verbal Autopsy, including Shams El Arifeen, International Centre for
Diarrhoeal Disease Research; Daniel Chandramohan, London School of Hygiene & Tropical
Medicine; Samuel Clark, University of Washington; Lalit Dandona, Public Health Foundation
of India; Abraham Flaxman, IHME; Bernardo Hernandez Prado, IHME; Robert Jakob, WHO;
Henry Kalter, Johns Hopkins Bloomberg School of Public Health; Soewarta Kosen, National
Institute of Health Research Indonesia; Jordana Leitao, Angola; Erin Nichols, CDC; Arvind
Pandey, National Institute of Medical Statistics, India; Chalapati Rao, Australian National
University; Ian Riley, Professor Emeritus, University of Queensland; Philip Setel, Vital
Strategies, D4H
This verbal autopsy instrument was produced and reviewed in collaboration with a WHO ledexpert group including Shams El Arifeen, International Centre for Diarrhoeal Disease Research;
Kanitta Bundhamcharoen, Thailand Ministry of Public Health; Peter Byass, Umeå University
Centre for Global Health; Daniel Chandramohan, London School of Hygiene & Tropical
Medicine; Chanpen Choprapawon, Health Policy and Strategic Bureau; Samuel Clark,
University of Washington; Don de Savigny, Swiss Tropical and Public Health Institute; Dr
Abraham Flaxman, IHME; Edward Fottrell, UCL Centre for International Health and
Development; Elizabeth França, Federal University of Minas Gerais; Frederik Frøen,
Norwegian Institute of Public Health; Gihan Gewaifel, University of Alexandria; Bernardo
Hernandez, IHME; Abraham Hodgson, Ghana Health Service; Sennen Hounton, UNFPA;
Kathleen Kahn, University of the Witwatersrand; Henry Kalter, Johns Hopkins Bloomberg
School of Public Health; Soewarta Kosen, National Institute of Health Research Indonesia;
Anand Krishnan, All India Institute of Medical Sciences; Vishwajeet Kumar, Uttar Pradesh
Center for Maternal, Neonatal and Child Health; Jordana Leitao, London School of Hygiene &
Tropical Medicine; Alan Lopez, The University of Queensland; Rafael Lozano, IHME; Honorati
Masanja, Ifakara Health Institute; Lene Mikkelsen, University of Queensland; Dean Yergens,

The 2016 WHO verbal autopsy instrument

v 1.1

University of Calgary; Thorkild Tylleskär, University of Bergen; Jørn Ivar Klungsøyr, University
of Bergen; Carlos Navarro-Colorado, Centers for Disease Control and Prevention (CDC); Erin
Nichols, CDC; Sam Notzon, CDC; Arvind Pandey, National Institute of Medical Statistics of
India; National Institute of Health Research Indonesia; Mohammad Hafiz; Chalapati Rao, The
University of Queensland; Rasooly, Afghanistan Ministry of Public Health; Ian Riley, The
University of Queensland; Osman Sankoh, INDEPTH Network; Paul Spiegel , UNHCR; Carla
Abou-Zahr; Derege Kebede, WHO; William Soumbey Alley, WHO; Fatima Marinho, WHO;
Mohamed Ali, WHO; Enrique Loyola, WHO; Jyotsna Chikersal, WHO; Jun Gao, WHO; Robert
Jakob, WHO; Giuseppe Annunziata, WHO; Rajiv Bahl, WHO; Kidist Bartolomeus, WHO; Ties
Boerma, WHO; Bedirhan Ustun, WHO; Doris Chou, WHO; Lulu Muhe, WHO; Matthews
Mathai; Marc Amexo, HMN.
For technical support in the design and production of the XLS forms and review of the skip
logic, acknowledgement is given to Carolyn Gulas and Matt Berg, ONA; and Aurelio Di
Pasquale, Rajib Mitra, Vinit Mishra, Nicolas Maire, Swiss Tropical and Public Health Institute;
and Anuraj Shankar, Alisa Pedrana, Mandri Apriatni, Harvard School of Public Health.
For technical support in the design of the paper forms and related reviews, acknowledgement
is given to Do Yoon Kwon; Kyung Mok Ko; Jinsung Jung; Eunji Jo; Sangbin Han; Junseong
Kim; Youngseok Kim; Eunyoung Cho; Yoon Joo Cho; Inah Kim; Sehee Kim; Kwansoo Lee;
Young Jae Chun; Soojin Moon, Chalapati Rao, Matthew Kelly.

Table of Contents
1

Purpose and content .........................................................................................1

2

VA 2016 Implementation Kit .............................................................................3

3

Introduction to verbal autopsy..........................................................................3

4

The development of the 2016_WHO verbal autopsy instrument ......................7

5

Application and implementation of the 2016 WHO VA instrument ................10

6

Bibliography ....................................................................................................22

3.1
3.2
4.1
4.2
questions
5.1
5.2
5.3
5.3.1
5.3.2
5.3.3
5.4
5.5
5.6
5.7
5.8
5.8.1
5.8.2
5.8.3
5.8.4
5.9
5.10

Historical background ................................................................................................................................ 4
Uses and users of VA data......................................................................................................................... 5
2016 List of causes of death for VA ..................................................................................................... 7
List of indicators and their definitions, relevant age and sex groups and sample
8
Sections of the 2016 WHO VA instrument .................................................................................... 11
Technical description of the Table of Indicators (ODK XLS) ............................................... 12
Sample questionnaires............................................................................................................................ 14
Sample VA questionnaire 1: death of a child aged under four weeks ................................. 14
Sample VA questionnaire 2: death of a child aged four weeks to 11 years....................... 14
Sample VA questionnaire 3: death of a person aged 12 years and above ......................... 14
Guidelines on augmentation and local adaptation ................................................................... 14
Translation .................................................................................................................................................... 15
Vital registration ........................................................................................................................................ 16
Age categories of death ........................................................................................................................... 16
Infrastructure............................................................................................................................................... 16
Interviewers.................................................................................................................................................... 17
Data collection software, database, technology and staff ........................................................ 18
Methods for determining causes of death ........................................................................................ 19
Legal requirements, privacy, confidentiality, informed consent............................................ 20
Appropriate respondents and recall period ................................................................................ 20
Use of verbal autopsy-generated data ............................................................................................ 20

Appendix 1: 2016 cause of death list for verbal autopsy with corresponding ICD-10 codes
(identical with 2014)..........................................................................................1

vii

The 2016 WHO verbal autopsy instrument

1 Purpose and content
The 2016 version of the WHO verbal autopsy instrument is suitable for routine use. Based on
the 2012 instrument, experiences with the interim 2014 version, and in collaboration with the
authors of the different analytical software for assigning the cause of death, questions have been
added or edited to facilitate the use of the publicly available analytical software (InterVA,
SmartVA previously known as Tariff – simplified PHMRC, and InsilicoVA). Also, the skip
pattern has been edited based on the information obtained from field testing and cognitive
reviews of the 2012 instrument, and the 2014 interim version.
All materials (questionnaires, table of indicators with skip patterns, crosswalks, and this
manual) are available separately for download at
http://www.who.int/healthinfo/statistics/verbalautopsystandards.
The questions allow for responses with a simple yes or no answer, multiple choice, or a duration
in some instances. This approach makes the instrument usable with analytical software that
assigns causes of death. Some very few free text fields are included to allow adding information
that may be used in reviews but they are not used by the analytical software.
The instrument is designed for all age groups, including maternal and perinatal deaths, and also
deaths caused by injuries. A set of paper forms by age group shows the design of the instrument.
It is however recommended to use electronic data collection methods, based on the electronic
format of the published instrument.
Sets of questions address information relevant to vital registration and information relevant to
assessment of the cause of death and the context.
The 2016 instrument is based on the 2012 version of the WHO verbal autopsy instrument that
had been designed to become suitable for routine use. Compared to the 2007 instrument,
numbers of conditions and questions had been reduced, based on evidence from the field and
expert reviews. This document contains some references to the development of the 2012
instrument, because its development is the basis for this new version of the WHO instrument.
This manual informs users on how to use the 2016 WHO verbal autopsy (VA) instrument. The
components of this manual include:
•
•
•
•
•
•
•
•
•

Background on VA;
The full matrix of questions, definitions and related skip patterns;
Instructions on how to use the matrix of questions;
ODK compatible form for all ages, including skip patterns and calculated fields
Paper forms for illustration and data collection - where necessary
Information about available analytical software for assigning cause of death, including
crosswalks for publicly available analytical software: InterVA5, Smart VA (IHME)
and InsilicoVA (Washington University);
Criteria for setting up a data collection infrastructure and the related databases;
Instructions on how to adapt questionnaires for local use;
General cause of death certification and coding guidelines for applying the
International Statistical Classification of Diseases and Related Health Problems, tenth
revision (ICD-10)1 to VA; and
1

Purpose and content

•
•

A simplified cause of death list for VA with corresponding ICD-10 codes.
Crosswalks for InterVA4 and SmartVA are available for download separately

This manual and its resources are the products of the first one-year effort by an expert group
led by the World Health Organization (WHO), consisting of researchers, data users, and
government agencies, for the 2012 VA instrument, and an additional year of work of the WHO
Working Group on Verbal Autopsy (subgroup of the WHO Reference Group for Health
Statistics). The 2016 WHO VA instrument is intended to allow for simple and inexpensive
identification of causes of death in places where no other routine system is in place and will
serve the needs of countries’ civil registration and vital statistics (CRVS) systems.
Independently, this instrument can also be used in research and disease specific programmes.
All materials are easily and widely accessible on the WHO web site, in print, and will be
incorporated into diverse resource kits, intended for strengthening national vital statistics
systems. Additional language versions will be made available through similar channels.
Experience from the field and publications on the most widely used and validated VA
instruments and procedures (WHO VA standards, InterVA and Population Health Metrics
Research Consortium -PHMRC VA instrument)1-3 were systematically reviewed and also
assessed against experience in using analytical software for cause of death assignment (InterVA
and SmartVA). The utility of each VA question was discussed with VA users. Experiences
from field testing and cognitive reviews of the 2012 instrument, and the results of a
simplification of the PHMRC Tariff method also contributed to the development of 2016 WHO
VA instrument. These reviews and assessments have resulted in a simplified instrument with
a reduced number of questions and causes of death, compared to 2007. However, the number
of questions has slightly increased compared to the 2012 instrument, because some questions
were added, and some complex questions were split into two thus making sure they ask about
only one indicator at a time. Furthermore, the 2016 WHO VA instrument also facilitates the
use of publicly available analytical software for assigning the cause of death, including InterVA
and SmartVA.
The systematic application of the 2016 WHO VA instrument will facilitate the application of
VA in routine surveillance of vital events and introduce more consistency and crosscomparability of VA-derived mortality data. The correspondence table (Appendix 1) allows
for easy conversion to and from ICD-10.
The application of the 2016 Instrument in routine use and research with its standardized
international set of questions will facilitate the compilation of larger databases that finally would
provide the evidence for stepwise improvement of VA questionnaires internationally, and
become a basis for continuous development of analytical methods.

2

2016 WHO verbal autopsy instrument

v 1.1

2 VA 2016 Implementation Kit
This manual provides an overview of structure, content and scientific background to the
WHO VA 2016 Instrument.
It is accompanied by
• A set of sample questionnaires - in case some setting wants to use paper
questionnaires. The design can be adapted, but the sequence and skip patterns need to
be preserved
• VA Field Interviewer Manual - with instructions for every question and some more
general information
• Manual for Training of Interviewers - to conduct a full training for interviewers
• Manual for the Training of Master Trainers and Supervisors
• Training curricula and template agendas for training of interviewers, master
trainers and interviewers

3 Introduction to verbal autopsy
Reliable data on the levels and causes of mortality are cornerstones for building a solid evidence
base for health policy, planning, monitoring and evaluation.
The main objective of VA is to describe the causes of death at the community level or population
level where civil registration and death certification systems are weak and where most people
die at home without having had contact with the health system.
In settings where the majority of deaths occur at home and where civil registration systems do
not function, there is little chance that deaths occurring away from health facilities will be
recorded and the cause of death certified. As a partial solution to this problem, VA has become
a primary source of information about causes of death in populations lacking vital registration
and medical certification. VA has become an essential public health tool for obtaining a
reasonable direct estimation of the cause structure of mortality at population level, although it
may not be an accurate method for attributing causes of death at the individual level.
Verbal autopsy is a method used to ascertain the cause of a death based on an interview with
next of kin or other caregivers. The interview is done using a standardized questionnaire that
elicits information on signs, symptoms, medical history and circumstances preceding death.
The cause of death, or the sequence of causes that led to death, are assigned based on the data
collected using the VA questionnaire and any other available information. Rules and
guidelines, algorithms or computer programs, may assist in interpreting the information
collected using the VA questionnaire to determine the cause of death11.
A standard VA instrument comprises a VA questionnaire, a list of causes of death or mortality
classification system, and sets of diagnostic criteria (either expert or data derived algorithms)
for assigning causes of death. The VA process consists of several steps, and many factors can
influence the cause specific mortality fractions estimated through this process.4

3

Introduction to verbal autopsy

3.1 Historical background
In Europe, before the 19th century when modern systems of death registration were
implemented, designated death searchers visited the households of deceased people to assess
the nature of deaths. The need for lay reporting of causes of death remained in low and middle
income countries where there was a lack of medical capacity to produce death certificates for
the population. As an alternative, in the 1950s and 60s in Asia and Africa, systematic interviews
by physicians were used to determine causes of death. Workers at the Narangwal project in
India labelled this new technique “verbal autopsy”5,6.
The interest of WHO in VA (formerly “lay reporting”) of health data was first demonstrated in
a publication by Dr. Yves Biraud in 1956. During the 1970s, WHO encouraged the use of lay
reporting of health information by people with no medical training, leading to development in
1975 of lay reporting forms (WHO 1978). Since the late 1970s and early 80s when the
Reproductive Age Mortality Studies (RAMOS), Matlab (Bangladesh) and Niakhar (Senegal)
questionnaires first emerged, several other questionnaires have been developed for use in
research settings and in national or large-scale regional surveys.5
The past two decades have seen a proliferation of interest, research and development in all
aspects of the VA process, including VA data-collection systems, VA questionnaires’ content
and format, cause of death assignment process, coding and tabulation of causes of death, and
validation of VA instruments.
In 2007, needs and demands for standardization led to the development and publication of the
WHO_2007 VA standard tools, which many researchers have adopted.7 The standards tools
included:8
•
•
•

Verbal autopsy questionnaires for three age groups (under four weeks; four weeks to
14 years; and 15 years and above);
Cause of death certification and coding resources consistent with the International
Classification of Diseases and Related Health Problems, tenth revision (ICD-10); and
A cause-of-death list for VA mapped according to the ICD-10.

The WHO_2007 VA questionnaire has been modified in several projects to accommodate the
local needs. In 2011, evidence from use of the WHO instrument and related VA instruments
was reviewed to formulate the 2012_WHO_VA tools. Besides elimination of unreported causes
and focus on useful questions, it was designed to facilitate VA use in routine vital registration
systems to improve national cause-specific mortality data.
Over the past years, efforts have been made to develop and implement software programs for
automated interpretation of VA data to generate computer-based diagnosis of causes of death.
Currently, the two most commonly used programs are the InterVA method developed by the
Umea University (Sweden) and the Tariff method (SmartVA) developed by the Institute of
Health Metrics and Evaluation (USA). However, these two programs are based on slightly
different versions of VA questionnaires. So far, it is not clear which of these two methods
perform better and whether they complement each other, despite several comparative studies912
. In order to facilitate the application of the two commonly used and publicly available
automated methods for interpreting VA and to allow comparison the causes of death data
determined by these methods comparative analysis of these commonly used automated

4

2016 WHO verbal autopsy instrument

v 1.1

methods of VA interpretation, the WHO VA instrument was updated in 2016 to include all
input variables required for the optimum performance of these two software programs.
Experience from the field in using the 2012_WHO VA tools and the interim 2014_WHO VA
tools, and cognitive testing 13 provided further inputs in the refinement of the 2016_WHO VA
instrument.
This version of the WHO instrument is recommended for field implementation in conjunction
with vital registration to enable use of automated VA interpretation programs as well as by
physician coders of verbal autopsies.

3.2 Uses and users of VA data
VA is used in three main ways. First, it has been primarily used as a research tool in the context
of longitudinal population studies, intervention research or epidemiological studies. Second, it
has become a source of cause of death statistics to meet the demand for population-level causespecific mortality data to be used in policy, planning, priority setting and benchmarking. Third,
VA data are gaining acceptance as a source of cause of death statistics to be used for monitoring
progress and evaluating what works and what does not. Because vital registration coverage has
not significantly improved in developing countries, VA methods have been mainly applied in
the following data collection systems: clinical trials and large-scale epidemiological studies;
demographic surveillance systems; national sample surveillance systems; and household
surveys.6
Over the past decade, due to the growing demand for robust estimates of vital events and
determinants of health, the primary objective of health and demographic surveillance system
(HDSS) sites evolved to encompass: 1) the production of population-based health information
to support evidence-based health policies and 2) the monitoring and evaluation of health
interventions in settings where routine health information and vital registration systems are
incomplete. Most HDSS sites collect VA data and provide population level cause specific
mortality data. However, HDSS sites may not provide representative data for national estimates
of cause-specific mortality. Application of VA in large cross sectional surveys or in sample vital
registration systems have been used to obtain national and sub-national level mortality
estimates.,
Potential users of data generated using VA include communities, health care planners and
managers, researchers, global decision-makers and donors. While there is a degree of overlap,
these users have different perspectives on the uses of mortality data, which have an impact on
the desirable characteristics of VA instruments. Researchers, epidemiologists and global-level
decision-makers want VA data to inform burden of disease estimation and program evaluation,
implying that cause of death estimates must meet high accuracy standards and be comparable
over time and across countries.7 National and sub-national decision-makers and health system
managers require cause of death data for planning, budgeting and resource allocation and for
monitoring and reporting to donors, implying that VA data needs to be actionable and program
relevant.
There have been a few instances where VA has been administered on a large scale as an explicit
part of the development of national statistics.8 Users of VA have identified the need for simpler
data collection instruments coupled with convenient IT-based solutions (e.g. mobile phones or
hand-held devices). These large scale users of VA have a perspective different from that of

5

Introduction to verbal autopsy

researchers, giving priority to the VA instrument’s simplicity, feasibility and adaptability to
local contexts, cost-effectiveness and program relevance.7 A simplified VA instrument coupled
with automated methods to ascertain causes of death can be a stepping-stone to increase the
coverage of operational and representative civil and vital registration systems.

6

2016 WHO verbal autopsy instrument

v 1.1

4 The development of the 2016_WHO verbal autopsy
instrument
The simplified instrument (2016_WHO VA instrument) comprises a short list of causes of
death of public health importance that can be ascertained from a limited number of questions
suitable for use in VA interviews and amenable to automated assignment of cause of death using
analytical software. The design of the VA questionnaire allows for addition of a narrative
section where so desired and addition of locally relevant questions and diagnoses. The
2016_WHO VA instrument has been informed by field testing and cognitive reviews13 of the
2012_WHO instrument, and the results of a simplification of the PHMRC Tariff method.
The WHO standard VA instrument published in 2007 has been successfully applied in many
research settings since then. In recent years, there has been growing interest in strengthening
countries’ CRVS systems, which has led to the demand for a more simplified and practical VA
instrument that is associated with IT applications for data collection and analysis.
To produce a simplified 2012 VA instrument, WHO carried out a systematic review of the use
of VA and led an expert group of researchers, data users and other stakeholders, in
collaboration with HMN, the University of Queensland (UQ) and the INDEPTH Network.
Based on the compiled experience and evidence from the most widely-used and validated VA
procedures (WHO VA standards, InterVA and PHMRC VA instrument),1-3,14 consensus was
reached on a simplified VA instrument for routine use as part of civil registration and vital
statistics systems in settings where many deaths are not medically certified.

4.1 2016 List of causes of death for VA
VA cannot ascertain all causes of death and, as many validation studies have shown, VA does
not perform equally well for all causes that it can ascertain. Taking these limitations into
account, the 2007_list of causes of death for VA was revised to develop the 2016 list that includes
all causes of death that could be ascertained with reasonable accuracy from a well-administered
VA interview (Appendix 1). In developing the 2016 WHO VA instrument, a review was
undertaken to compile evidence for the revision and simplification of the 2007 WHO VA
standard cause of death list. The review included VA research studies that used either physician
certified VA (PCVA) or automated analytical software to assign the cause of death15. In
addition to research studies, a review of the materials and inputs from VA experts provided
evidence on the feasibility and relevance of causes of death that can be reliably ascertained by
VA.
The revision and simplification of the 2007 standard WHO VA cause of death list was based
on:
•
•

The frequency of a given cause of death being reported in VA; the importance and
relevance of a given cause of death to global mortality levels1;
The cause of death can be addressed by public health interventions; and

1

In the simplification of the list of diagnoses, attention has been given to the Global Burden of Disease (GBD) groupings. In view of the ongoing
edits of the GBD, the correspondence table does not include GBD references. The mentioned ICD-10 codes may serve to distribute cases to the
relevant GBD groups.

7

The development of the 2016 WHO verbal autopsy instrument

•

The feasibility of the cause of death being ascertained through VA.

The list of causes of death resulting from the above process is presented in Appendix 1.
Using a minimum set of causes of death facilitates the merging and comparison of data from
VA on an international scale. The mapping of the list of causes of death using ICD-10 codes
shown in Appendix 1 allow comparison of mortality data determined by VA with ICD-10 coded
causes of death data ascertained using the international certificate of causes of death.

4.2 List of indicators and their definitions, relevant age and sex groups
and sample questions
The 2016_WHO VA questionnaire cause of death related indicators for all age groups in one
sheet, but only subsets are used for the different age groups. The list of indicators is subdivided
into 4 sections and 118 subgroups (Appendix 2). Within sections and subgroups of the
instrument, skip patterns are driven by the age and sex of the deceased and whether it, was a
maternal or perinatal death. The indicators were selected by reviewing the 2007_WHO VA
questionnaire; identifying unused and uninformative indicators from the field experience of
VA experts; and with input from cause of death assignment results using analytical software
(InterVA and Tariff), and cognitive assessment of the WHO 2012 Instrument. Field testing
with the interim 2014 Version and a final expert review and consolidation of the results led to
the formulation of the present list of indicators.

8

2016 WHO verbal autopsy instrument

v 1.1

The table shows the distribution of questions by age group and brad section of the
questionnaire.
The levels refer to the skip questions. The ‘entry level’ questions will always have to be
asked.
Questions for CRVS are a set recommended by the UN Statistical Division. They may
not be necessary in all settings.
The questions asking for content of a medical certificate of cause of death may provide
helpful information in certain settings and in case such a certificate has been issued.
Segment/
depth
Personal
Entry level
Level 2
CRVS
Entry level
Level 2
CoD
Entry level
Level 2
Level 3
Context
Entry level
Level 2
Level 3
DeathCert
Entry level
Level 2
Level 3

Neonate

Child

19
15
4
13
2
11
122
38
65
19
23
7
10
6
12
1
1
10

21
17
4
18
2
16
161
86
72
3
23
12
10
1
12
1
1
10

Adult (incl.
maternal)
20
16
4
18
3
15
184
66
96
22
19
10
8
1
12
1
1
10

Grand Total
Entry level

189
63

235
118

253
96

The full set of indicators that were considered and the rationale for inclusion or exclusion of
each indicator in the 2016 VA instrument is available for download at
www.who.int/healthinfo/statistics/verbalautopsystandards. The list of indicators is further
described in Section 4.2.
The instrument allows for the addition of indicators of topical interest such as risk factors where
necessary. However, such augmentation of the 2016 instrument is not encouraged (see Section
4.4 “Guidelines on augmentation, and local adaptation”), because the 2016_instrument is
comprehensive and if correct cause of death certification and coding procedures are used, it
should be possible to generate comparable data over time across populations. Any addition of
indicators should be done in consultation with WHO in order to ensure operability of the
analytical software for assigning cause of death and comparability of results. See Section 4.4 for
further guidance.

9

Application and implementation of the 2016 WHO VA instrument

5 Application and implementation of the 2016 WHO VA
instrument
This section describes the application and implementation of the 2016WHO VA instrument.
The 2016 WHO VA instrument and supporting documentation for implementation include
the following components:
•

2016 Cause of Death List with ICD Codes (Appendix 1, described in Section 3.1,
identical with the 2014 version)

•

Table of Indicators (Appendix 2, described in Section 4.2): For each indicator, describes
the variable ID and the data type, defines threshold values to categorize numeric values,
defines skip patterns, and includes notes for translators and interviewers.

•

2016 WHO VA instrument
-

Excel- and XML-files and references to data collection platforms (posted on
WHO website at: www.who.int/healthinfo/statistics/verbalautopsystandards)

-

3 Sample (Paper) Questionnaires (Appendices 3-5)

•

Tools and guidance for remote data collection and storage in a database (described in
Section 4.8.2; will be posted on WHO website at:
www.who.int/healthinfo/statistics/verbalautopsystandards)

•

Instrument translation tables (described in Section 4.5, available for download at
www.who.int/healthinfo/statistics/verbalautopsystandards)

•

2016 WHO VA instrument training manuals (are posted on WHO website at:
www.who.int/healthinfo/statistics/verbalautopsystandards)

•

Analytical tools for cause of death assignment are available at:
InterVA : http://www.interva-4.net/
Tariff : http://www.healthdata.org/verbal-autopsy/tools

Details for application and implementation described in this section include: a description of
the structure and indicators included in the 2016 WHO VA instrument; guidelines for local
adaptation and translation of the instrument; guidelines for data collection, management, and
storage; and guidelines for cause of death assignment and use of VA data.

10

2016 WHO verbal autopsy instrument

v 1.1

5.1 Sections of the 2016 WHO VA instrument
The 2016 VA instrument contains both common sections and specific sections appropriate to
both the age and sex of the deceased.
NOTE: Age, sex, information about the season, the local prevalence of HIV and malaria,
section 3, 4, and 5 are essential information for the analytical software that assigns causes
of death. No questions must be removed from these sections. The numbering of the
questions must remain unchanged. Questions added locally will not be used by the currently
available analytical software.
Other components of the personal information and the respondent can be adjusted to the local
legal requirements.
The instrument consists of the following sections:
1. Preset HIV-Malaria mortality and season (the project office may make this question
2.
3.

4.
5.

6.

hidden to the interviewers in the electronic format);
Information on the respondent and background about interview;
Information about the deceased and vital registration;
a. Information on the deceased
b. Civil registration numbers.
History of injury/accidents;
Health history;
a. Duration of illness
b. Medical history associated with final illness
c. General signs and symptoms associated with final illness
d. Signs and symptoms associated with pregnancy and women
e. Neonatal and child history, signs and symptoms
f. Health service utilization
g. Background and context
h. Death certificate with cause of death.
Open narrative (text field).
a. Check list of additional items to record in the narrative open space.

Section 1
collects information about the prevalence of malaria and HIV in the area
where the deceased lived and whether death occurred in rainy or dry season. This
information is essential for selecting the appropriate algorithm used by some software
for assigning the cause of death. In most settings this information will be pre-completed
by study staff or supervisors.
Section 2
collects information about the respondent, consent if required in certain
contexts and time the VA interview was started.
Section 3
contains key identifying and socio-demographic information and data fields
necessary for the management of completed forms.
Section 4
provides essential information for assigning the cause of death due to
accidental and intentional injuries.

11

Application and implementation of the 2016 WHO VA instrument

Section 5
contains several sub-sections that collect information required for
assigning causes of death. Section 5a) has questions to determine the duration of the final
illness; 5b) history known past or present diseases that would give clues to the causes of
death; 5c) contains symptoms and sings that are relevant for all deaths; 5d) contains
symptoms and signs specific to maternal deaths; 5e) contains symptoms and sings
relevant for neonatal and child deaths; and 5f) contains questions about the utilisation of
health services and contextual factors. Section 5g has fields for recording information
from a medical certificate of cause of death if this is available.
Section 6
is an open narrative text field that allows for comments and adding
additional information. This section is particularly useful for quality control and for
providing additional information for physician assessment of the cause of death if needed.
While its use is optional, it is recommended that this question be asked, even if it is not
recorded, in order to complete the checklist of some indicators (section 6a) that are
required for assigning causes of death using Tariff 2.0.

5.2 Technical description of the Table of Indicators (ODK XLS)
The Table of Indicators of the 2016 WHO VA instrument (separate file mentioned in Appendix
2) consists of one table containing all indicators for all age groups with relevant details
describing each indicator. The questions are grouped by sections, as is described above.
Relevant skip patterns by age and sex are defined for each indicator.
Questions, hints and skip instructions are listed in the sheet “survey”. Selectable values are
listed in sheet “choices”.
Quick overview of the columns in the sheet “survey”
type

Describes the kind of question, e.g. yes/no, multiple choice, integer, or
text

name

language independent identifier of the question

label::English

question in a specified language: here English. You may add a column
with the title “label::mylanguage” for the language of your choice.
Having several language columns active allows to create multilingual
forms.

hint::English

hint for the question in the specified language. You may add a column
with the title “hint::mylanguage” for the language of your choice.
Having several language columns active allows to create multilingual
forms.

relevant

here you have the information for the skip patterns.

required

determines whether the question must be answered, if asked

appearance

describes appearance of questions in the form

calculation

specifies calculations using the values of preceding questions. It is used
to determine the age group

default

12

2016 WHO verbal autopsy instrument

constraint

add constraints to the data fields;

constraint message

used to display a message why the entry is not accepted

v 1.1

Details about the format are available online at http://xlsform.org/ and more generally at
https://opendatakit.org .

13

Application and implementation of the 2016 WHO VA instrument

5.3 Sample questionnaires
The 2016 WHO VA instrument is designed for use with an electronic data collection
platform. Excel- and XML-files with the required instructions are posted on the WHO
website at: www.who.int/healthinfo/statistics/verbalautopsystandards. A set of sample/paper
questionnaires that demonstrate the layout of the questions, as described in the Table of
Indicators (Appendix 2), are also provided as a guide to the user.
The sample questionnaires may also serve as a guide for data collection using paper-based
questionnaires. However, to facilitate 1) the use of automated analytical software for the
assignment of cause of death, and 2) the comparison of the responses across VA studies, all
data (regardless of data collection method) should be entered into a database following the
instructions provided in Section 4.8.2. It is especially important to retain the variable IDs as
defined in the Table of Indicators in Appendix 2 review and.
Sample questionnaires are provided for three age groups (under four weeks; 4weeks-11 years,
12 years and above), as shown in the Appendices 3-5.

5.3.1 Sample VA questionnaire 1: death of a child aged under four weeks
Sample VA questionnaire 1 is designed to determine causes of early neonatal deaths, late
neonatal deaths, perinatal deaths and stillbirths. In addition to a “signs and symptoms noted
during the final illness” list, the questionnaire contains questions concerning the history of the
pregnancy, delivery, the condition of the baby soon after birth, and the mother’s health and
contextual factors.

5.3.2 Sample VA questionnaire 2: death of a child aged four weeks to 11 years
Sample VA questionnaire 2 is designed to ascertain the major causes of post-neonatal child
mortality (i.e. starting from the fourth week of life), as well as causes of death that may be seen
through 11 years of age. Questionnaire 2 includes all the common sections and questions
described above, as well as questions related to causes of death in children aged four weeks to
11 months. The skip pattern is indicated by references to the next question.

5.3.3 Sample VA questionnaire 3: death of a person aged 12 years and above
Sample VA questionnaire 3 is designed to identify all major causes of death among adolescents
and adults (i.e. starting at age 12), including deaths related to pregnancy and childbirth.
Questionnaire 3 includes a section for all female deaths, in addition to the above mentioned
common sections and questions.

5.4 Guidelines on augmentation and local adaptation
The indicators contained in the 2016 WHO VA instrument address the most relevant causes of
death in most populations where the use of VA is a necessary means to obtain cause of death
information. The 2016 WHO VA instrument by design allows for evolution of the instrument.
Users may add questions but under no circumstances should questions be removed from the
list because of the resulting impact on the comparability of the causes of death information and
the further data based evolution of the instrument.
NOTE: Age, sex, information about the season, the local prevalence of HIV and malaria,
section 3, 4, and 5 are essential information for the analytical software that assigns causes

14

2016 WHO verbal autopsy instrument

v 1.1

of death. No questions must be removed from these sections. The numbering of the
questions must remain unchanged. Questions added locally will not be used by the
analytical software.
It is acknowledged that there may be a desire to expand the instrument to address locally
relevant conditions. However, adding signs and symptoms to the 2016 WHO VA instrument
needs to be carried out with much caution because alteration can compromise the
comparability of VA data between populations. In particular, the addition of new questions
about particular diseases of interest may bias results if a disproportionate amount of
information about only one condition is available in the cause of death assignment process.
Modifications may be necessary if there are emerging or locally important causes of death for
which there are no questions on the 2016 VA questionnaires. In these circumstances, advice
may be sought from WHO for making such modifications. If modifications are necessary, they
should be carefully documented and distinguished from the 2016 questionnaire sections and
variables. In general, only changes to the wording of existing variables for the purposes of
enhancing local comprehension or ensuring cultural acceptability of questions are to be
undertaken. The definitions in the 2016 WHO VA instrument may provide some guidance
about the meaning that needs to be preserved in such changes. Any need for modification
should be shared with WHO together with the rationale for modification. The reporting of
modifications made to WHO will inform future revisions of this instrument.
Examples of modifications that are unlikely to affect the comparability of results include:
•
•

Adding questions or sections about household characteristics or environmental or
behavioural risk factors;
Adding or changing questions about usage of a particular health context.

Examples of modifications that may affect the comparability of results include:
•
•

Changing or adding to response categories in the checklist of “signs and symptoms
noted during the final illness”;
Adding new questions about diseases of particular interest (e.g. malaria, HIV/AIDS,
diarrhoeal disease).

Adding and removing questions will impact the comparability of the data but also may
compromise the usability of analytical software for assigning cause of death. It may either
not be possible to use the existing analytical software for assigning the newly added causes of
death at all, or the outputs from the software become unreliable.

5.5 Translation
The specific terminology used for indicators and interviewer and translator notes (in the Table
of Indicators) aims to convey the highest level of clarity about the intent of a question.
Indicators, instructions and data collection tool need to be translated or adapted for local use
(even if administered in English, as lay language differs across English speaking regions). The
notes in the Table of Indicators are intended to guide translators in the translation process; both
the questions in the instrument and the hints in the Table of Indicators should be translated, as
the notes will also provide guidance to interviewers. Translators may need to adapt the wording

15

Application and implementation of the 2016 WHO VA instrument

of the questions to the local terminology used in the locations where interviews will be
conducted. For quality assurance, a second translator should carry out a back-translation to
English.
For any translations, please use the XLS form for the 2016 WHO VA instrument that you can
download at www.who.int/healthinfo/statistics/verbalautopsystandards.
Fill in your
translations in the sheets “survey” and “choices” adding columns and do it the same way it was
done for the existing translations. The given identifiers and the resulting multilingual file will
allow users to 1. re-use a translation as necessary for other projects, and 2. load translations into
the WHO software for immediate use, if so desired.

5.6 Vital registration
For use of the 2016 WHO VA instrument in routine vital registration, two options are possible:
1) A death has been reported and notified, and an interviewer is sent to query the cause
of death. In this case, the personal data are known and the interview will be conducted
only to identify the cause of death. The “information on the deceased” section will be
prefilled before the interview based on information from the death notification form,
and a death registration number (or a similar identifier) will allow the user to link the
VA outcome with the related entry in the death registration registry.
2) A death is reported and the certification and interview are conducted at the same time.
In this case, the personal data are not known, or are known only in part, and need to
be recorded at the time of the interview, using the “information on the deceased”
section. A registration number will be required to ensure the vital registration linkage
between the death registry and the VA data.

5.7 Age categories of death
Some projects may be interested only in particular age categories of death, such as perinatal,
maternal, child or adult deaths. In this case, the relevant subset of questions can be extracted
from the list of indicators of the 2016 WHO VA instrument. The three age-group specific
questionnaires in the appendices for three age groups (under four weeks; 4weeks-11 years, 12
years and above) may serve as examples here. Where data are captured electronically, the
embedded skip patterns will ensure that only the relevant subset of questions is applied.
Where interviews are conducted for all age categories of deaths, ideally the interview data
should be captured electronically using the embedded skip patterns. Otherwise (eventually
using paper if there is no other way), the interviewers should always be sure to have
questionnaires available for all three age group during house visits for VA interviews.

5.8 Infrastructure
In routine surveillance contexts, information needs to be timely and linked to a response that
involves effective dissemination mechanisms, appropriate use of data, and periodic evaluation
of the surveillance system. The latter can trigger formulation of recommendations for a revision
of the components of the 2016 WHO VA instrument to WHO and local workflows of the VA
system in use. Thus, use of VA in routine surveillance involves monitoring, accountability,
planning and programming.

16

2016 WHO verbal autopsy instrument

v 1.1

To develop an adequate instrument for application in large-scale surveillance, the VA
instrument needs to have a system that synergizes with other national mechanisms that are
already in place. A permanent team that conducts Verbal autopsies on a representative sample
of deaths is likely the most feasible way.
The infrastructure will depend largely on the given setting. Data should ideally be stored in a
centralized location but at a level that facilitates necessary queries. Usually this is likely to be at
the district level. All verified district data would then be forwarded to a central database at the
national level.
Arrangements regarding collection, storage and handling of the vital registration and cause of
death data should be made between the offices that are involved. Depending on the national
infrastructure these offices are most frequently the Statistical Office, Ministry of Health,
Ministry of Justice and Ministry of Interior.
Different arrangements may be necessary where VA is conducted in another context, as for
research or disease specific programmes.

5.8.1 Interviewers
Interviewers should be trained on using the instrument and on conducting interviews with
persons who may still be in mourning and may become upset during the interview.
Interviewers should be given enough time to prepare and carry out VA interviews. It is
proposed that at least one VA interview per month should be conducted by each VA interviewer
to retain their proficiency in conducting VA interviews. Interviewers involved in the
application of the VA should have the following minimum qualifications:
•
•
•
•

Have completed at least secondary school and have good working knowledge in the
relevant local language(s);
Be acceptable to the local community; where possible, selected by the local
community;
Have good training in conducting VA interviews;
Know very well the content and uses of the VA instrument.

The following steps are recommended to train VA interviewers. First the trainee interviewer
should review all elements and the flow of questions with the help of an expert VA trainer. After
discussing any questions that need clarification with the VA trainer, the interviewer can role
play a VA interview using the VA trainer as proxy VA respondent. For this role play the VA
trainer should have different case scenarios. Such scenarios would include the different age
groups of the deceased (maternal, perinatal, child, adult) and thus varying relationship of the
respondent to the deceased and also probable local behaviour on the interview per se, and on
sensitive questions. In the next step, the interviewer would conduct real VA interviews in the
presence of the expert trainer. The number of VA interviews to be conducted in the presence
of the VA trainer required to certify proficiency of individual trainee VA interviewer will vary
depending on the skills and abilities. Nevertheless typically it would take at least five VA
interviews to become confident in doing VA interviews.
The interviewer guide that explains the meaning and importance of each indicator and how to
ask each question included in the 2016 WHO VA instrument is posted on the WHO website
at www.who.int/healthinfo/statistics/verbalautopsystandards.

17

Application and implementation of the 2016 WHO VA instrument

5.8.2 Data collection software, database, technology and staff
The 2016 WHO VA instrument is designed to be compatible with electronic data collection
platforms and automated analytical software to assign cause of death.
The 2016 Instrument form is provided in a format that is fully compatible with the Open Data
Kit (ODK - http://opendatakit.org). This allows one to quickly start collecting data using
Internet cloud-based services as well as local servers. Data is output in a table format and at
any time ready for analysis. A link to the demo version for testing and download is available
at www.who.int/healthinfo/statistics/verbalautopsystandards. The electronic data collection
form includes embedded skip patterns that automatically navigate the various combinations
of age-, sex-, maternal- and perinatal-specific indicators within a single, comprehensive
instrument.
Any software can be programmed for data collection using this instrument, as long as all
answers to all questions, assigned cause of death and method of assigning cause of death are
recorded in a database. Further, any tool for assigning the cause of death can be used, as long
as the data collected are what is required for the specific analytical tool.
The ODK format has been adopted by several data collection software systems. Software with
wide implementation are OpenSRP and DHIS2. Both can handle ODK standard instruments
and have the ability to conduct batch processing and output of results. For DHIS, this requires
setting up a separate server that feeds the data into the DHIS server.
Regardless of the data collection method (electronic or paper), answers to all questions and the
cause of death assigned to each case should be recorded in a database. The database should
retain the cause of death together with a variable that identifies the method of assigning the
cause of death. The name of the interviewer and date, time and duration of the interview should
also be retained in the database. If data are reported electronically this information can be
generated automatically. If physician review is used to assign the cause or causes of death, then
all assigned causes and the identity of the physician who assigned each should be recorded.
In order to facilitate the use of data collection and analytical software and to simplify the
interviews, most questions follow a simple yes/no pattern. However some questions address a
time interval or a frequency. All continuous variables should be recorded as continuous
variables in the database and will be categorized in a second step using a recommended
threshold value. Categorization depends on the analytical software that is used to identify the
cause of death. The converter tool will include the necessary algorithms.
The compilation of the information above into a database will provide a tool for reviewing cases
as well as enable sharing of results with WHO to facilitate further improvements to this
instrument.
In addition to the questions, a narrative can be helpful if physician assessment is a possibility,
for quality assurance, and for later review. The full verbatim narrative should be stored in the
database as well.
In order to use the existing analytical software, the data collected with the WHO VA Instrument
need to be converted into the formats that can be processed by the analytical software that
determines the cause of death. WHO makes available conversion algorithms in collaboration
with the University of Washington and the Swiss TPH. These include a simple conversion

18

2016 WHO verbal autopsy instrument

v 1.1

(programmed in R) and an all-in-one R package (openVA - https://cran.rproject.org/web/packages/openVA/index.html) that does the conversion and runs the
analytical software and outputs the cause of death. Links to the packages are also available from
the WHO website www.who.int/healthinfo/statistics/verbalautopsystandards.
VA, using ODK or other electronic data collection methods, requires at a minimum a central
server, and mobile devices for data collection. The data will be stored on the central server as
soon as a mobile or wired internet connection to the server is available. Sufficient maintenance
staff with IT administration skills is necessary in order to set up and maintain the technical
infrastructure. Data and software staff will ensure that data collection is complete and
workflows work. Statistical and epidemiological staff will use the outputs from the system for
tabulation and analysis.

5.8.3 Methods for determining causes of death
In the past decade, methodological developments in automated methods for assigning cause of
death for VA have emerged with significant potential for future application in routine national
and research data collection platforms. These methods create new opportunities for reliable,
timely, and useful cause-specific mortality measurement. These developments have created a
shift away from limited individual-level and clinical paradigms towards population-based
epidemiological thinking and public health.16
The 2016 WHO VA instrument contains information on diseases, signs and symptoms, the age
and sex of the deceased as well as his or her medical history (if available). Additional
information may be recorded in the open text field at the end of the interview. To facilitate
application in routine surveillance systems, the 2016 WHO VA instrument was specifically
developed to ascertain cause of death through automated methods. As a more cost-effective
and feasible alternative to physician-coded VA, the WHO recommends the use of automated
methods for cause of death identification. The use of automated VA cause of death assignment
methods also ensures that causes of death are determined in a standard fashion, removing the
variability inherent with physician coding of VA.
Analytical software tools compatible with the 2016 WHO VA instrument for cause of death
assignment without the use of physicians are listed on the WHO VA website
(www.who.int/healthinfo/statistics/verbalautopsystandards). At present (2014) InterVA4
(University Umea), SmartVA (PHMRC/IHME) and InsilicoVA (University of Washington) are
fully compliant with the 2016 WHO VA instrument and can be linked for batch processing to
the data collection software.
The intended possible use of these software will allow to assess both against the same database
of indicators and contribute to further development of this Verbal Autopsy instrument as well
as of the software.
In case physicians assess the cause of death, ideally two physicians will review the outcome of
an interview and formulate a cause of death independently. If there were a discrepancy, a third
physician would arbitrate the result. The opinion of each physician involved should be
separately recorded in the database, as well as the consensus finding.

19

Application and implementation of the 2016 WHO VA instrument

5.8.4 Legal requirements, privacy, confidentiality, informed consent
Ideally, informed consent should be sought from the respondent. Where the legislation or local
regulations require that VA be conducted, informed consent may not always be necessary.
Personal data, in particular name, geographical information and contact information about the
respondent, should be kept separate from the epidemiological data and ideally be encrypted to
protect privacy and ensure confidentiality. Additional measures to anonymize the individual
record may be subject to the legislation in force. A common case-ID in the person identifiable
VA dataset and the diagnostic VA dataset will allow data linkage between personal and
diagnostic data upon formal request in line with national and international regulations.

5.9 Appropriate respondents and recall period
The respondent who provides information about the deceased and allows the interviewer to
complete the VA questionnaire should be the primary caregiver (usually a family member) who
was with the deceased in the period leading to death or a witness to a sudden death or accident.
This individual is likely to provide the most reliable and accurate account of the signs and
symptoms of importance. It is not uncommon for a VA respondent to require assistance from
other household or family members in answering the VA questions. However, the verbal
autopsy interviews should be conducted in privacy. The VA interviews should be conducted as
soon as practically possible after the report of the event is received, but after any culturally
prescribed mourning period has passed. Recalls of more than one year should be interpreted
with caution. In general, shorter recall periods are preferable.

5.10 Use of verbal autopsy-generated data
The purpose of VA is to describe the causes of death at the community level or population level
in instances where no better alternative sources of mortality data exist. Therefore, VA serves as
a limited but essential substitute for medical certification. The quality of information of the
assigned cause of death varies depending on the skills of the interviewer and the ability of the
respondents to recognise, recall and report key indicators.
The 2016 WHO VA cause of death list (Appendix 1) is a core mortality classification system,
specifying the most important causes of death in low-income and middle-income countries
where it is deemed feasible to certify cause of death using VA. Coding causes of death using the
ICD coding system facilitates the comparison of data and the retention of as much detail as
needed in local settings.
The context and method of information gathering to assign cause of death from VA is different
from the medical certification of cause of deaths by a physician. The certainty of the cause of
death is much lower in VA, and VA cannot reliably ascertain some causes of death. Thus, causes
of death data obtained from these two systems should not be merged, as it would conceal
differences that may result from these methods and lead to misinterpretation of the results.
ICD-10 provides tabulation lists for mortality and morbidity in volume 1. Other professional
groups have made different lists for grouping diseases and presenting mortality statistics.
Regardless of the list used, deaths should be classified by sex and into the following age groups:
aged < 1 year, aged 1–4 years, and then in 5-year groups from age 5 years to 84 years, followed

20

2016 WHO verbal autopsy instrument

v 1.1

by a group for those aged 85 years or older. Volume 2, section 5.6.1 of ICD-10, contains a full
set of instructions for tabulation.

21

Application and implementation of the 2012 WHO VA instrument

6 Bibliography
1
2
3

4
5
6
7
8
9

10

11
12
13
14

22

Verbal autopsy standards: ascertaining and attributing cause of death. (World
Health Organization, 2007).
InterVA. InterVA,  (2011).
Murray, C. J. et al. Population Health Metrics Research Consortium gold standard
verbal autopsy validation study: design, implementation, and development of
analysis datasets. Popul Health Metr 9, 27, doi:1478-7954-9-27 [pii]
10.1186/1478-7954-9-27 (2011).
Soleman, N., Chandramohan, D. & Shibuya, K. Verbal autopsy: current practices
and challenges. Bull World Health Organ 84, 239-245, doi:S004296862006000300020 [pii] /S0042-96862006000300020 (2006).
Biraud Y. Méthodes pour l’enregistrement par des non médecins des causes
élémentaires de décès dans les zones sous-développées. Geneva: World Health
Organization; 1956. WHO document HS/60
Murray, C. J., Lopez, A. D., Feehan, D. M., Peter, S. T. & Yang, G. Validation of the
symptom pattern method for analyzing verbal autopsy data. PLoS Med 4, e327,
doi:07-PLME-RA-0521 [pii] 10.1371/journal.pmed.0040327 (2007).
Abouzahr, C. Verbal autopsy: who needs it? Popul Health Metr 9, 19, doi:14787954-9-19 [pii] 10.1186/1478-7954-9-19 (2011).
Setel, P. W. Verbal autopsy and global mortality statistics: if not now, then when?
Popul Health Metr 9, 20, doi:1478-7954-9-20 [pii] 10.1186/1478-7954-9-20
(2011).
Leitao, J. et al. Comparison of physician-certified verbal autopsy with computercoded verbal autopsy for cause of death assignment in hospitalized patients in
low- and middle-income countries: systematic review. BMC medicine 12, 22,
doi:10.1186/1741-7015-12-22 (2014).
Desai, N. et al. Performance of four computer-coded verbal autopsy methods for
cause of death assignment compared with physician coding on 24,000 deaths in
low- and middle-income countries. BMC medicine 12, 20, doi:10.1186/17417015-12-20 (2014).
Lozano, R. et al. Performance of InterVA for assigning causes of death to verbal
autopsies: multisite validation study using clinical diagnostic gold standards.
Popul Health Metr 9, 50, doi:10.1186/1478-7954-9-50 (2011).
Oti, S. O. & Kyobutungi, C. Verbal autopsy interpretation: a comparative analysis
of the InterVA model versus physician review in determining causes of death in
the Nairobi DSS. Popul Health Metr 8, 21, doi:10.1186/1478-7954-8-21 (2010).
Scanlon, P., Nichols, E. (2014). National Center for Health Statistics. Hyattsville,
MD Results of the Cognitive Interviewing Study of the 2012 WHO Verbal Autopsy
Instrument in Nyanza Province, Kenya.
Bauni, E. et al. Validating physician-certified verbal autopsy and probabilistic
modeling (InterVA) approaches to verbal autopsy interpretation using hospital
causes of adult deaths. Popul Health Metr 9, 49, doi:10.1186/1478-7954-9-49
(2011).

2016 WHO verbal autopsy instrument

15
16

v 1.1

Leitao, J. et al. Revising the WHO verbal autopsy instrument to facilitate routine
cause-of-death monitoring. Global health action 6, 21518,
doi:10.3402/gha.v6i0.21518 (2013).
Fottrell, E. Advances in verbal autopsy: pragmatic optimism or optimistic theory?
Popul Health Metr 9, 24, doi:1478-7954-9-24 [pii] 10.1186/1478-7954-9-24
(2011).

23

2016 WHO verbal autopsy instrument

v 1.1

Appendix 1: 2016 cause of death list for verbal autopsy with
corresponding ICD-10 codes (identical with 2014)
2016 cause of death list for verbal autopsy with corresponding ICD-10 codes.
Column 1 contains the code for the verbal autopsy entity. Column 2 lists the related titles.
Column 3 lists the ICD-10 codes that would be used if the condition labelled by column 2
were coded to ICD-10. Column 4 lists the ICD-10 categories that need to be grouped to
match the content of the relevant VA entity.
Verbal
autopsy code

Verbal autopsy title

VAs-01 Infectious and parasitic diseases
VAs-01.01
Sepsis
Acute respiratory infection,
VAs-01.02
including pneumonia
VAs-01.03
HIV/AIDS related death
VAs-01.04
Diarrheal diseases
VAs-01.05
Malaria
VAs-01.06
Measles
VAs-01.07

Meningitis and encephalitis

VAs-01.08

Tetanus
Excludes: Neonatal tetanus VAs-10.05

VAs-01.09
VAs-01.10
VAs-01.11
VAs-01.12

Pulmonary tuberculosis
Pertussis
Haemorrhagic fever
Dengue fever

VAs-01.99

Unspecified infectious disease

ICD-10
code (to
ICD)

ICD-10
codes (from
ICD)

A41
J22/J18

A40-A41
J00-J22

B24
A09
B54
B05
G03; G04

B20-B24
A00-A09
B50-B54
B05
A39; G00G05
A33-A35

A35
(obstetrical
A34)
A16
A37
A99
A90; A91
B99

A15-A16
A37
A92-A99
A90-A91
A17-A19
A20-A38;
A42-A89;
B00-B19;
B25-B49;
B55-B99

A1-1

Non-communicable diseases
Note:
This group covers all non-communicable conditions. Any infection of the systems that are
listed in this section should be assigned to the suitable infectious disease category. Any
maternal and perinatal condition should be assigned to the maternal and perinatal causes
below.

VAs-98

Other and unspecified noncommunicable disease

R99

D55-D89;
E00-E07;
E15-E35;
E50-E90;
F00-F99;
G06-G09
G10-G37;
G50-G99;
H00-H95;
J30-J39;
J47-J99;
K00-K31;
K35-K38
K40-K93;
L00-L99;
M00-M99;
N00-N16;
N20-N99;
R00-R09
R11-R94

C06
C26
C39
C50
C57

C00-C06
C15-C26
C30-C39
C50
C51-C58

C63
C80

C60-C63
C07-C14
C40-C49
C60-D48

Note:
This group covers all non-communicable
conditions that could not be assigned to
another category in this section. There is
a separate category for cases where the
cause of death is unknown.

VAs-02 Neoplasms
VAs-02.01
Oral neoplasms
VAs-02.02
Digestive neoplasms
VAs-02.03
Respiratory neoplasms
VAs-02.04
Breast neoplasms
Female reproductive neoplasms
VAs-02.05
VAs-02.06

Male reproductive neoplasms

VAs-02.99

Other and unspecified neoplasms

A1-2

2016 WHO verbal autopsy instrument

VAs-03 Nutritional and endocrine disorders
VAs-03.01
Severe anaemia
VAs-03.02
Severe malnutrition
VAs-03.03
Diabetes mellitus

v 1.1

D64
E46
E14

D50-D64
E40-E46
E10-E14

I24 (acute
ischemic)
I64
D57
I99

I20-I25

J44

J40-J44

J45 (J46)

J45-J46

VAs-06 Gastrointestinal disorders
VAs-06.01
Acute abdomen
VAs-06.02
Liver cirrhosis

R10
K74

R10
K70-K76

VAs-07 Renal disorders
VAs-07.01
Renal failure

N19

N17-N19

VAs-08 Mental and nervous system disorders
VAs-08.01
Epilepsy

G40

G40-G41

VAs-04 Diseases of the circulatory system
VAs-04.01

Acute cardiac disease

VAs-04.02
VAs-04.03

Stroke
Sickle cell with crisis

VAs-04.99

Other and unspecified cardiac
disease

VAs-05 Respiratory disorders
Chronic obstructive pulmonary
VAs-05.01
disease (COPD)
VAs-05.02
Asthma

I60-I69
D57
I00-I09
I10-I15
I26-I52
I70-I99

A1-3

VAs-09 Pregnancy-, childbirth and puerperium-related
disorders
O00
VAs-09.01
Ectopic pregnancy
O06
VAs-09.02
Abortion-related death
O13 (or
VAs-09.03
Pregnancy-induced hypertension
O15 for
eclampsia)
O46 (ante
partum)
VAs-09.04
Obstetric haemorrhage
O72 (post
partum)
O66
VAs-09.05
Obstructed labour
O75.3 (ante
partum)
VAs-09.06
Pregnancy-related sepsis
O85 (post
partum)
O99
VAs-09.07
Anaemia of pregnancy
O71
VAs-09.08
Ruptured uterus
O05
VAs-09.99

Other and unspecified maternal
cause

VAs-10 Neonatal causes of death
VAs-10.01
Prematurity
VAs-10.02
Birth asphyxia
VAs-10.03
Neonatal pneumonia
VAs-10.04
Neonatal sepsis
VAs-10.05
Neonatal tetanus
VAs-10.06
Congenital malformation

VAs-10.99

A1-4

Other and unspecified perinatal
cause of death

P07
P21
P23
P63
A33
Q89
P96

O00
O03-O08
O10-O16

O46; O67;
O72

O63-O66
O85; O75.3

O99.0
O71
O01-O02;
O20-O45;
O47-O62;
O68-O70;
O73-O84;
O86-O99

P05-P07
P20-P22
P23-P25
P36
A33
Q00-Q99
P00-P04;
P08-P15;
P26-P35;
P37-P94;
P96

2016 WHO verbal autopsy instrument

VAs-11 Stillbirths
VAs-11.01
Fresh stillbirth
VAs-11.02
Macerated stillbirth

v 1.1

P95
P95

P95
P95

V89

V01-V89

V99
W19
W74

V90-V99
W00-W19
W65-W74

X09

X00-X19

X29

X20-X29

X49

X40-X49

X84
Y09
X39
X59

X60-X84
X85-Y09
X30-X39
S00-T99;
W20-W64;
W75-W99;
X50-X59;
Y10-Y98

R99

R95-R99

VAs-12 External causes of death
Note:
The list of questions contains sub
questions that allow for more specificity
for accidents.

VAs-12.01

Road traffic accident

VAs-12.02
VAs-12.03

Other transport accident
Accidental fall
Accidental drowning and
submersion
Accidental exposure to smoke, fire
and flames
Contact with venomous animals and
plants

VAs-12.04
VAs-12.05
VAs-12.06
VAs-12.07

Accidental poisoning and exposure
to noxious substance

VAs-12.08
VAs-12.09
VAs-12.10

Intentional self-harm
Assault
Exposure to force of nature

VAs-12.99

Other and unspecified external cause
of death

VAs-99

Cause of death unknown

Read carefully the explanations given in section 4.1 of this manual, and the instructions in
the remaining sub-sections of section 4.
The sample questionnaires and the excel file for translation and production of ODK forms
are provided as separate files.

A1-5



---

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.3
Linearized                      : No
Page Count                      : 37
Title                           : 01_Manual and  guidelines for application and use of simplified WHO VA tool_2016v1-5
Author                          : WHO
Subject                         : 
Producer                        : Mac OS X 10.13.2 Quartz PDFContext
Creator                         : Word
Create Date                     : 2018:01:25 15:56:23Z
Modify Date                     : 2018:01:25 15:56:23Z

EXIF Metadata provided by EXIF.tools