Materials Index

Active Death Reporting By Key Informants

Active Surveillance of Maternal & Pregnancy-Related Deaths

Autopsy

Bennett-Horiuchi Method

Birth and Death Record Linkage

Capture-Recapture

CEMD

Census

Chandrasekaran-Deming Technique

Civil Registration

Confidential Enquiry into Maternal Deaths

Death Certificates

Decennial Census

Demographic Surveillance Systems

 Direct Sisterhood Method

DSS

Dual Method

Growth Balance

Health Facility Records

HIS

ICD Classification/Coding

Indirect Sisterhood Method

International Classification of Diseases Coding

Key Informants

Key Informant Reports of Death

Modified Death Notification Form

N
 Non-Probability Samples

Parity to Cumulated Fertility Ratio

P/F Ratios

Population-Based Household Survey

Prospective Studies

RAMOS

RAPID

Regression Models

Reverse Projection Techniques

Routine Health Information Systems

Sample Vital Registration with Verbal Autopsy

Sampling at Service Sites

SAVVY

Sisterhood Method

SSS

T
 TRACE

Verbal Autopsy

MMM-Resource: Materials
This section provides an index of methods, techniques and approaches to measuring maternal
mortality. Some allow for the calculation of national or sub-national maternal mortality
indicators, while others describe essential elements for identification of maternal deaths such
as death certification and classification of causes of death.

See the menu on the left for a Diagrammatic Overview of the opportunities and options for
measuring maternal mortality.

Each item in the Materials Index has a brief description including measurement
requirements, a listing of advantages and limitations, an example of questions (if relevant),
and links to guidelines, scientific articles and other key links. New! These pages are now
available to save or print - just click on the button at the top of the page.

For help in determining which materials or methods might be relevant to your research needs,
please see the Guide section of this website.

Routine Health Information Systems (HIS)

Also: Health Facility Records

Routine HIS: BACKGROUND

Health information system (HIS) data on deaths are generated from health facilities including
hospitals and health centres by medical professionals. Facilities compile statistics on maternal
deaths and births and report these to the central level. HIS data are generally available, in
some form, in most settings. However, quality can be variable and its scope is limited.

Please also see the related topic of ICD Coding

Identification of death

Medical professional/ facility records death at time it occurs.

Ascertainment of maternal/pregnancy related status

• medical professional gives cause (death certificate)

• may use diagnostic aids (e.g. pregnancy or laboratory tests)
• may use autopsy (post-mortem)

Advantages:

• on-going systems that therefore do not require special data collection activities
• can be used as a starting point for in-depth case reviews
• provides annual estimates of maternal mortality
• provides cause-specific mortality and trends

Limitations:

• community and private facility-based deaths are often excluded

• access to private facility data can be problematic
• the chain of moving data to central level is subject to problems
• not appropriate for calculation of population-based maternal mortality ratios unless
close to 100% of deliveries/deaths take place in health facilities
• direction of bias in denominator/numerator is difficult to determine
• deaths that occur in emergency rooms are often missed, as are women who are “dead-
on-arrival” or “discharged to die”
• may need to include record review since deaths within hospital often missed by
hospital statistics
• records may not be kept

Measurement requirements:

• sub-causes need to be grouped as maternal deaths (ICD coding)

• data need to be compiled as part of hospital/health centre statistics
• need to feed into health information statistics and/or civil registration
• data on births are also needed to calculate the maternal mortality ratio (MMR) and
lifetime risk (LTR)

Routine HIS: GUIDELINES

Health Metrics Network. Strengthening country health

information systems: assessment and monitoring tool.
Version 1.96: 04 November 2006. Geneva: World Health
Organization.

Routine HIS: REFERENCES

 Deneux-Tharaux, C, Berg, C, Bouvier-Colle, MH; Gissler,
M; Harper, M; Nannini, A; Alexander, S; Wildman, K;
Breart, G; Buekens, P (2005) Underreporting of pregnancy-
related mortality in the United States and Europe. Obstetrics
& Gynecology. 106(4):684-92.
De Sousa, MH; Cecatti, JG; Hardy, EE; Serruya, SJ (2007)
Declared maternal death and the linkage between health
information systems. Revista de Saúde Pública, 41(2): 1-8.
Graham, WJ; Hussein, J (2006) Universal reporting of
maternal mortality: an achievable goal? International Journal
of Gynaecology and Obstetrics, 94(3): 234-42.
Graham, W; Hussein J (2004) The right to count. The
Lancet, 363(9402): 67-8.
Sombie, I.; Meda, N.; Hounton, S.; Bambara, M.; Ouedraogo,
T.W.; Graham, W. (2007) Missing maternal deaths: lessons
from Souro Sanou University Hospital in Bobo-Dioulasso,
Burkina Faso. Tropical Doctor, 37(2): 96-8.

Routine HIS: ADDITIONAL RESOURCES

Health Metrics Network:

http://www.who.int/healthmetrics/en/

Additional descripive information from WHO

Key Informants
Key Informants: BACKGROUND

Key informants are people within communities who may have access to detailed information
about deaths or maternal deaths. Non-medical key informants may be particularly useful in
settings where many maternal deaths occur outside health facilities. Key informants may be
used in a range of studies, and are particularly common in non-probability samples in which
they are targeted for their expertise or community access, and may lead to the identification
of other key respondents. They may also be used in Reproductive Age Mortality Studies
(RAMOS), Active surveillance of pregnancy-related and maternal deaths and
Confidential Enquiries into Maternal Deaths.

Key informants may include:

• community health workers

 • traditional birth attendants (TBAs)
• health care providers
• village leaders
• teachers
• political cadres
• cemetery workers
• coffin makers

Key informants can either be interviewed on a one-off or periodic basis and asked to recall
deaths or maternal deaths, or they can be used as part of active surveillance systems.

Accuracy may be improved if two networks of informants or two data sources are used with
capture-recapture adjustment.

Key Informants: ACTIVE DEATH REPORTING BY KEY INFORMANTS

Active and regular reporting of deaths by key informants is a useful way of capturing
pregnancy-related deaths in the community. Immpact developed a tool for capturing maternal
deaths in the community using such reporting, called “MAternal DEaths from Informants” or
MADE-IN. Piloted in Indonesia, MADE-IN uses a network of village health workers as
informants who report on deaths of women of reproductive age that have occurred in a
community in the two years prior to survey. Pregnancy-related deaths are then further
explored with the use of the MADE-FOR (Follow-On Review) tool. MADE-FOR involves
interviewing a relative of the deceased woman and collecting information on the symptoms
and circumstances that preceded the death as well as socioeconomic information (see Verbal
Autopsy).

Identification of death:

Key informants report deaths. May be passive or active.

Ascertainment of maternal/pregnancy related status:

Key informants report pregnancy-related deaths. May be passive or active. When using a
network of community health workers, this approach has been termed MADE-IN.

Advantages:

• potentially more efficient - if key informants have a special reason to know about
deaths or pregnancy related deaths, then fewer individuals are needed to identify
deaths

Limitations:

• deaths may be missed, especially for marginalized women who the key informant
may not know
 • in active systems, maintaining quality of reporting may be difficult long-term
• may require follow-up interview with family and verbal autopsy to confirm maternal
death

Measurement requirements:

• if verbal autopsy questions used they must be coded and classified as maternal
deaths
• duplicate deaths need to be excluded
• data on births are needed

Key Informants: REFERENCES

Lisa M. Koonin, M.N., M.P.H. Hani K. Atrash, M.D., M.P.H.

Roger W. Rochat, M.D. Jack C. Smith, M.S. Maternal
Mortality Surveillance, United States, 1980-1985. MMWR
December 01, 1988 / 37(SS-5);19-29
Walker, GJA; McCaw-Binns, A; Ashley, DEC; Bernard,
GW (1990) Identifying maternal deaths in developing
countries: experience in Jamaica International Journal of
Epidemiology.
Singh, P; Pandey, A; Aggarwal, A. (2007) House-to-house
survey vs. snowball technique for capturing maternal deaths
in India: A search for a cost-effective method. Indian Journal
of Medical Research, 125: 550-6.

Key Informants: ADDITIONAL RESOURCES

Immpact website:
www.immpact-international.org
Immpact/Population Reference Bureau. Measuring maternal
mortality: challenges, solutions and next steps. Washington
DC: Population Reference Bureau. February 2007

Immpact/Population Reference Bureau. Delivering safer

motherhood: sharing the evidence. Policy brief.
Washington DC: Population Reference Bureau. February
2007

Indirect Sisterhood Method

Indirect Sisterhood Method: BACKGROUND
The original (indirect) sisterhood method was developed in the late 1980s as an efficient
means of measuring maternal mortality through population-based surveys, generating a
variety of indicators: the proportion of maternal deaths among female deaths (PMDF), the
maternal mortality ratio (MMR), the maternal mortality rate (MMRate), the lifetime risk of
maternal death (LTR), and the adult female death rate.

Adult respondents are asked four questions about the survival of all their adult sisters born to
the same mother (see below).The method reduces the need for large sample sizes because
there may be more than one respondent per household, more than one sister per respondent,
and because the time period of death is not restricted.

Identification of death:

Adult respondent reports deaths of his/her adult sisters retrospectively to interviewer by

responding to questions 1-3 of the 4 questions.

Ascertainment of maternal/pregnancy related status:

Adult responds to the 4th (pregnancy-related) question

It is not possible to obtain sub-causes of maternal death, as respondent may not know the
signs and symptoms that preceded the death.

Advantages:

• when used on a probability sample it provides data that represent the population and
can be used with relatively small sample sizes
• the four questions can be easily added to multipurpose questionnaires/surveys
• it is particularly efficient in high fertility settings where respondents have large
numbers of siblings
• it can be used at national and sub-national levels
• data processing and calculation of estimates is comparatively simple

Limitations:

• it gives retrospective estimates (usually for 10-12 years before the survey
• a longer reference period than with the direct sisterhood method)
• it assumes no major trends in fertility
• it is less appropriate in settings with substantial migration
• it is difficult to get additional information about deaths (causes, risk factors, timing,
etc as sibling may not have such details)
• it generally has large confidence intervals (as a function of reduced sample size)
• like other survey-based methods, there are no established demographic techniques to
permit evaluation and adjustment, if necessary

Measurement requirements:
Questions are used to derive proportions of adult sisters dying during pregnancy, childbirth or
puerperium. Standard adjustment factors convert these proportions into LTR. LTR can be
converted to MMR and MMRate given assumptions about the Total Fertility Rate (TFR). The
TFR is needed.

Sisterhood questions:

Four questions are asked in addition to the age of the respondent:

1. How many sisters have you ever had, born to the same mother, who ever reached the
age 15 (or who were ever married) including those who are now dead?
2. How many of these sisters reaching age 15 are alive now?
3. How many of these sisters are dead?
4. How many of these dead sisters died during pregnancy or during childbirth, or during
the six weeks after the end of the pregnancy?

Indirect Sisterhood Method: GUIDELINES

WHO (1997) The sisterhood method for estimating maternal

mortality: Guidance notes for potential users. Geneva: World
Health Organization.

Indirect Sisterhood Method: REFERENCES

Graham W, Brass W, Snow RW (1989) Indirect estimation

of maternal mortality: the sisterhood method. Studies in
Family Planning, 20(3):125-35.
Danel, I. Graham, W. Stupp, P. and Castillo, P. Applying the
sisterhood method for estimating maternal mortality to a
health facility-based sample: A comparison with results from
a household-based sample. International Journal of
Epidemiology, 1996, 25 (5): 1017-1022.
Hanley JA, Hagen CA, Shiferaw T (1996) Confidence
intervals and sample size calculations for the sisterhood
method of estimating maternal mortality. Studies in
Family Planning, July/August 27(4).
Rutenberg N, Sullivan JM (1991) Direct and indirect
estimates of maternal mortality from the sisterhood
method. Washington DC: IRD/Macro International Inc.
Shahidullah M (1995) The Sisterhood method of
estimating maternal mortality: The Matlab experience.
Studies in Family Planning, March-April 26(2): 101-6.
Trussell J, Rodriguez G (1990) A note on the sisterhood
estimator of maternal mortality. Studies in Family Planning,
 21(6): 344-6.
WHO/UNICEF. The Sisterhood method to estimate maternal
mortality. Report of a technical meeting, 5-6 December 1996.

Indirect Sisterhood Method: ADDITIONAL RESOURCES

SSS Tool: Calculator for number of survey respondents

required

International Classification of Diseases (ICD) Coding

ICD Coding: BACKGROUND

The International Classification of Diseases (ICD) is the international standard for diagnostic
classification. It is developed and regularly updated by the World Health Organization and is
currently in its 10th version. The ICD provides a comprehensive coding system for diseases
and health conditions to be used on death certificates, hospital and other vital records.

ICD-10 defines a maternal death as:

the death of a woman while pregnant or within 42 days of termination of pregnancy,

irrespective of the duration and site of the pregnancy, from any cause related to or
aggravated by the pregnancy or its management but not from accidental or incidental causes
*

ICD-10 differs from earlier versions in that it also provides a definition for late maternal
death:

the death of a woman from direct or indirect obstetric causes more than 42 days but less than
one year after termination of pregnancy*

While the definition of maternal deaths is further broken down into direct and indirect deaths,
misclassification and under-reporting of maternal mortality can still be a problem as the
precise cause of death (even if identified as pregnancy-related) can be missed.

*International statistical classification of diseases and related health problems, 10th Revision.
2nd edition, Volume 2 Instruction Manual. Geneva: World Health Organization

ICD Coding: GUIDELINES

 International statistical classification of diseases and related
health problems, 10th Revision. 2nd edition, Volume 2
Instruction Manual. Geneva: World Health Organization

ICD Coding: REFERENCES

Hoyert DL (2007) Maternal mortality and related concepts.

National Center for Health Statistics. Vital & Health
Statistics 3(33).
Laurenti, R; Buchalla, CM (1997) Maternal and child health
indicators: implications of the tenth revision of the
International Classification of Diseases. Pan American
Journal of Public Health, 21(1): 13-7.
Office of National Statistics (England and Wales).
Comparing trend data in ICD-10 and ICD-9. 2004.
Starzyk P, Frost F, Kobayashi JM (1986) Misclassification
of maternal deaths: Washington State. MMWR Morb Mortal
Wkly Rep; 35:621-623.

ICD Coding: ADDITIONAL RESOURCES

International Classification of Diseases. Geneva: World

Health Organization.

ICD-10 Online. Geneva: World Health Organization.

Bennett-Horiuchi Method
Bennett-Horiuchi Method: BACKGROUND

Several demographic techniques have been developed to assess and adjust information on
deaths that come from civil registration, demographic surveillance, censuses or
population based surveys that ask about deaths of household members during a specified
time before the survey period.

Some methods compare data from independent sources (direct capture-recapture approaches)
while others are analytic indirect methods based on assumptions on the population age
distribution. The basic idea behind these indirect methods is that everyone who reaches a
given age must die at an older age. By making some assumptions it is possible to compare
deaths by age to the population by age.

The Bennett-Horiuchi Method is an indirect approach to estimating the completeness of death

registration data. For any population, the entry rate minus the growth rate must be equal to
the exit or death rate. Systematic differences between the entry rate and the growth rate
(which is a residual estimate of the exit rate calculated from the census age distributions) and
the exit rate (calculated from information on deaths by age) is used to identify differences in
reporting of population and of deaths. The magnitude of the inconsistency can be interpreted
as a measure of completeness of death reporting relative to population reporting and can then
be used to adjust the mortality estimates calculated from the original data.

Bennett-Horiuchi uses the population age distributions from two separate enumerations to
estimate the growth rate as well as data on deaths registered in the intercensal period.

Its advantage is that it does not assume the population is stable and so can be used when the
population distribution is changing.

Its limitation is that it is however vulnerable to changes in the coverage of the two censuses.

All indirect approaches assume that misreporting of deaths is constant across the age groups
in the population or part of a population under consideration, such as people of a given age or
older. Since this is unlikely to be true for child deaths, the approach is usually applied to adult
mortality. The method provides information on when this assumption is violated and thus
when the method becomes inappropriate.

An alternative way of avoiding the assumption of population stability is provided by

Courbages and Fargues (1979). Their approach assumes constant under-reporting of deaths
by age and compares model life table measures to the extent to which reported deaths are
concentrated in old age. It can be applied to populations affected by migration.

Bennett-Horiuchi Method: REFERENCES

Bennett, NG & Horiuchi, S (1984) Mortality estimation from registered deaths in

less developed countries. Demography, 21(2): 217-34.

Bennett, NG & Horiuchi, S (1981) Estimating the completeness of death registration

in a closed population. Population Studies; 47(2):207-21.

Hill, K (1987) Estimating census and death registration completeness. Asian and
Pacific Population Forum; 1(3):8-13.

Courbage, Y & Fargues, P (1979) A method for deriving mortality estimates from
incomplete vital statistics. Population Studies; 33: 165-80.
 Timaeus, I; Graham, W (1989) Measuring adult mortality in developing countries: a
review and assessment of methods. Policy, Planning and Research Working Papers.
Population and Human Resources Department, World Bank. Washington DC:
World Bank.

Bennett-Horiuchi Method: ADDITIONAL RESOURCES

Example exercise from WHO

Mortpak computer software including the Bennett-Horiuchi technique and other methods can
be purchased from the United Nations:

Mortpak Description and order form

MortPak Manual

Growth Balance
Growth Balance: BACKGROUND

Several demographic techniques have been developed to assess and adjust information on
deaths that come from civil registration, demographic surveillance, censuses or
population based surveys that ask about deaths of household members during a specified
time before the survey period.

Some methods compare data from independent sources (direct capture-recapture

approaches) while others are analytic indirect methods based on assumptions on the
population age distribution. The basic idea behind these indirect methods is that everyone
who reaches a given age must die at an older age. By making some assumptions it is possible
to compare deaths by age to the population by age.

Growth balance is an indirect approach to estimating the completeness of death registration

data. For any population, the entry rate minus the growth rate must be equal to the exit or
death rate. Systematic differences between the entry rate and the growth rate (which is a
residual estimate of the exit rate calculated from the census age distributions) and the exit rate
(calculated from information on deaths by age) are used to identify differences in reporting of
population and of deaths. The magnitude of the inconsistency can be interpreted as a measure
of completeness of death reporting relative to population reporting and can then be used to
adjust the mortality estimates calculated from the original data.
The original Brass Growth Balance assumes a stable, closed population which in
demographic terms means there have been no changes in fertility or mortality and no
migration.

Its advantage is that it only requires the distributions of the population and of deaths by age.

Its limitation is that the assumption that a population is stable is often inappropriate, and
means that the approach cannot be used with sub-national populations.

Where two census distributions and a distribution of intercensal deaths are available, a
reformulation eliminates the need for assuming a stable population.

All indirect approaches assume that misreporting of deaths is constant across the age groups
in the population or part of a population under consideration, such as people of a given age or
older. Since this is unlikely to be true for child deaths, the approach is usually applied to adult
mortality. The method provides information on when this assumption is violated and thus
when the method becomes inappropriate.

An alternative is the Preston Coale (1980) method. This requires detailed information on ages
at death and a measure of the growth rate. It is more vulnerable to age misreporting
(exaggeration of age at death) but less sensitive to mortality decline (a violation of the
assumption of stability).

Growth Balance: GUIDELINES

United Nations: Methods for estimating adult mortality. Population Division,

Department of Economic and Social Affairs UN Secretariat. P21-26.
ESA/P/WP.175 2002.
United Nations: Manual X: indirect techniques for demographic estimation. New
York, United Nations; 1983.

Growth Balance: REFERENCES

Bhat, PNM (2002) General growth balance method: a reformulation for populations
open to migration. Population Studies; 56(1): 23-34.

Brass, W (1971) On the Scale of Mortality, in W. Brass (ed.), Biological Aspects of

Demography. London: Taylor & Francis.

Hill K (1987) Estimating census and death registration completeness. Asian and
Pacific Population Forum; 1(3):8-13.
 Martin, LG (1980) A modification for use in destabilized populations of Brass'
technique for estimating completeness of registration. Population Studies, 34(2):
381-95.
Preston, SH; Coale, AJ; Trussell, J; Weinstein, M (1980) Estimating the
completeness of adult deaths in populations that are approximately stable.
Population Index; 46: 179-202.
Preston SH: Use of direct and indirect techniques for estimating the completeness of
death registration systems. In: Data Bases for Mortality Measurement. Edited by:
United Nations. New York: United Nations; 1984.
Timaeus, I; Graham, W (1989) Measuring adult mortality in developing countries:
a review and assessment of methods. Policy, Planning and Research Working
Papers. Population and Human Resources Department, World Bank. Washington
DC: World Bank.

Growth Balance: ADDITIONAL RESOURCES

Example exercise from WHO

Capture-Recapture: BACKGROUND

Capture-Recapture methods were originally developed by animal scientists for estimating

wildlife populations but are now being used increasingly in epidemiology and public health.
The general idea is to use two separate methods to identify cases of enquiry (e.g., maternal
deaths) and to examine the proportion of cases identified by the second method that were also
identified by the first. From this, it is then possible to estimate the coverage of the first
method and hence to estimate the total number of cases. For example, Buescher et al (2001;
see below) used capture-recapture to identify pregnancy-related deaths by using death
certificates and enhanced surveillance with birth-death record linkage. Capture-recapture has
been used to look at maternal mortality in the USA, Indonesia and Bangladesh.

For capture-recapture estimates, each death included must be confirmed (no false positives)
and it must be clear which method ‘identified’ the death. The identification must be
independent and the method assumes all cases are equally likely to be identified. Missing
some cases is not a serious concern, but it decreases precision.

Capture-recapture can be used to assist in monitoring trends over time. Trends can be
problematic to interpret if there are changes in coverage of deaths identified as well as in the
risks of death. Capture-recapture approaches estimate coverage, and hence population totals,
even when coverage changes.

Capture-Recapture: REFERENCES
General:

Amstrup SC, McDonald TL, Manly BFJ. Handbook of

Capture-Recapture Analysis: Princeton University Press, 2005
Hook, EB; Regal, RR (1992) The value of capture-recapture
methods even for apparent exhaustive surveys. American
Journal of Epidemiology; 135: 1060-7.
LaPorte, RE (1994) Assessing the human condition:
capture-recapture techniques. BMJ; 308: 5-6.
Watts, C; Zwi, AB; Foster, G. (1995) How to do (or not to
do).Using capture-recapture in promoting public health.
Health Policy & Planning; 10(2): 198-203.

An example for maternal mortality:

Buescher, PA; Harper, M; Meyer, RE. Enhanced surveillance

of maternal mortality in North Carolina. Raleigh: Center for
Health Informatics and Statistics, April 2001.

Direct Sisterhood Method: BACKGROUND

The direct sisterhood method is a variant of the indirect Sisterhood Method, and is currently
used by Demographic and Health Surveys (DHS). While it requires larger sample sizes than
its predecessor, the advantage of the direct method is the targeting of a more limited reference
period for sister deaths. Using a more detailed set of questions, in particular to ascertain
deaths among all sibling and then those that are pregnancy-related as well as when the death
occurred, point estimates for maternal mortality usually relate to 0-6 years and >6 years prior
to the survey. A retrospective maternal mortality ratio (MMR) can be calculated for the
reference period in question. The method is not recommended for duplication at short time
intervals due to large confidence intervals.

Identification of death:

Adult respondent reports deaths (and sex and age at death) among all his/her brothers and
sisters born to the same mother in response to sibling survivorship questions and direct
sisterhood method questions.

Ascertainment of maternal/pregnancy related status:

Adult responds to pregnancy-related questions (see below).

It is not possible to obtain sub-causes of maternal death, as respondent may not know the
signs and symptoms that preceded the death.

Advantages:

• yields a more recent estimate than the indirect sisterhood method (generally with one
reference period 0-6 years prior to the survey and one 7-13 years prior to the survey
versus 10-12 years for indirect sisterhood method)*
• provides data that represent the population
• can be added to multipurpose surveys

* Rutstein, SO; Rojas, G. (2006) Guide to DHS Statistics. Calverton: ORC Macro J. p134

Limitations:

• requires larger samples than the original indirect sisterhood method

• should not be replicated at short intervals owing to overlapping reference periods
• more complex data collection (more questions including age at death, the year in
which the death occurred and the years since the death) and data analysis
• less appropriate in settings with substantial migration
• difficult to get additional information on deaths (causes, risk factors, timing ,etc as the
sibling may not have such details)
• no established demographic techniques to permit evaluation and adjustment, if
necessary, of survey-based data

Measurement requirements:

Involves conducting a sibling history, with dates of births and deaths, and for sisters dying of
reproductive age the 4th question of the indirect sisterhood method is also asked. Data on
births are also needed to derive the maternal mortality ratio.

Direct sisterhood questions:

1. How many children did your mother give birth to?

2. How many of these births did your mother have before you were born?
3. What was the name given to your oldest (next oldest) brother or sister?
4. Is (NAME) male or female?
5. Is (NAME) still alive?
6. How old is (NAME)?
7. In what year did (NAME) die? OR How many years ago did (NAME) die?
8. How old was (NAME) when she died?

For dead sisters aged 15-49 only:

9. Was (NAME) pregnant when she died?

10. Did (NAME) die during childbirth?
 11. Did (NAME) die within two months after the end of pregnancy or childbirth?

Direct Sisterhood Method: GUIDELINES

Direct Sisterhood Method: ADDITIONAL RESOURCES

SSS Tool: Calculator for number of survey respondents

required

Measure DHS website: http://www.measuredhs.com

DHS Questionnaires and Modules

DHS Maternal Mortality Questionnaire

Demographic Surveillance Systems (DSS)

Demographic Surveillance Systems: BACKGROUND

Surveillance involves actively following populations to detect births and deaths.

There are four main approaches:

• Demographic Surveillance Systems (DSS )

• Prospective studies
• SAVVY
• Sample Vital Registration Systems

Demographic Surveillance Systems (DSS) began in the 1960s as a means of tracking

longitudinal demographic changes to populations in developing countries. Unlike prospective
(cohort) studies, DSS are able to monitor entire populations and are usually larger and longer
term. Field sites collect data on births, deaths (including causes) and migration which provide
an important resource for evaluating health care interventions. They also offer a starting point
for new studies. The INDEPTH Network (“An International Network for the Continuous
Demographic Evaluation of Populations and Their Health in Developing Countries”), is an
international network of 31 DSS field sites in 17 countries spanning Africa and Asia.

Identification of death:

Enumerators make initial census/count of population. Subsequently, an adult family member

reports household deaths since the last update round (ranging from every two weeks to annual
rounds). Enumerators probe on all births (pregnancies), deaths, and migrations.

Ascertainment of maternal/pregnancy related status

 • Adult responds to Pregnancy-related questions; or
• Appropriate respondent gives cause of death (from health facility, death certificate, or
Verbal Autopsy); or
• Birth-Death Linkage is used to identify pregnancy related deaths

Advantages:

• useful for monitoring and evaluation of public health interventions

• more complete enumeration of vital events than available from Civil Registration in
most countries with weak infrastructure
• provide comparable data over time for analysis of trends
• researchers associated with the DSS are on-site for data analysis

Limitations:

• data from DSS field sites may not be generalisable to regional or national populations
• costly and research intensive
• covers relatively small populations
• women who leave the surveillance area to deliver and subsequently die outside the
area may be missed, as might women who migrate into the surveillance area to deliver

Measurement requirements:

• if verbal autopsy questions are used they must be coded and classified as maternal
deaths
• data on births are needed to calculate the maternal mortality ratio (MMR)

Demographic Surveillance Systems: GUIDELINES

INDEPTH Resource Kit for Demographic Surveillance

Systems (Beta Version 0.9)

Demographic Surveillance Systems: REFERENCES

Baiden, F; Hodgson, A; Binka FN (2006) Demographic

surveillance sites and emerging challenges in international
health. Bulletin of the World Health Organization, 84(3):
163-4.
Ngom, P; Binka FN; Phillips, JF; Pence, B; Macleod, B
(2001) Demographic surveillance and health equity in sub-
Saharan Africa. Health Policy and Planning, 16(4): 337-44.
Nhacolo, AQ; Nhalungo, DA; Sacoor, CN; Aponte1, JJ;
Thompson, R; Alonso, P (2006) Levels and trends of
demographic indices in southern rural Mozambique:
evidence from demographic surveillance in Manhiça district.
 BMC Public Health, 6(291).
Tatem, A.J; Snow, R.W.; Hay, S.I. (2006) Mapping the
environmental coverage of the INDEPTH demographic
surveillance system network in rural Africa. Tropical
Medicine & International Health, 11(8):1318-26.

Demographic Surveillance Systems: ADDITIONAL RESOURCES

INDEPTH Network: An International Network for the

Continuous Demographic Evaluation of Populations and
Their Health in Developing Countries: http://www.indepth-
network.org
INDEPTH Network (2002) Population and health in
developing countries: Volume1: Population, health and
survival at INDEPTH sites. Ottawa: International
Development Research Centre.

Demographic Surveillance Systems (DSS)

Demographic Surveillance Systems: BACKGROUND

Surveillance involves actively following populations to detect births and deaths.

There are four main approaches:

• Demographic Surveillance Systems (DSS )

• Prospective studies
• SAVVY
• Sample Vital Registration Systems

Demographic Surveillance Systems (DSS) began in the 1960s as a means of tracking

longitudinal demographic changes to populations in developing countries. Unlike prospective
(cohort) studies, DSS are able to monitor entire populations and are usually larger and longer
term. Field sites collect data on births, deaths (including causes) and migration which provide
an important resource for evaluating health care interventions. They also offer a starting point
for new studies. The INDEPTH Network (“An International Network for the Continuous
Demographic Evaluation of Populations and Their Health in Developing Countries”), is an
international network of 31 DSS field sites in 17 countries spanning Africa and Asia.

Identification of death:
Enumerators make initial census/count of population. Subsequently, an adult family member
reports household deaths since the last update round (ranging from every two weeks to annual
rounds). Enumerators probe on all births (pregnancies), deaths, and migrations.

Ascertainment of maternal/pregnancy related status

• Adult responds to Pregnancy-related questions; or

• Appropriate respondent gives cause of death (from health facility, death certificate, or
Verbal Autopsy); or
• Birth-Death Linkage is used to identify pregnancy related deaths

Advantages:

• useful for monitoring and evaluation of public health interventions

• more complete enumeration of vital events than available from Civil Registration in
most countries with weak infrastructure
• provide comparable data over time for analysis of trends
• researchers associated with the DSS are on-site for data analysis

Limitations:

• data from DSS field sites may not be generalisable to regional or national populations
• costly and research intensive
• covers relatively small populations
• women who leave the surveillance area to deliver and subsequently die outside the
area may be missed, as might women who migrate into the surveillance area to deliver

Measurement requirements:

• if verbal autopsy questions are used they must be coded and classified as maternal
deaths
• data on births are needed to calculate the maternal mortality ratio (MMR)

Demographic Surveillance Systems: GUIDELINES

INDEPTH Resource Kit for Demographic Surveillance

Systems (Beta Version 0.9)

Demographic Surveillance Systems: REFERENCES

Baiden, F; Hodgson, A; Binka FN (2006) Demographic

surveillance sites and emerging challenges in international
health. Bulletin of the World Health Organization, 84(3):
163-4.
 Ngom, P; Binka FN; Phillips, JF; Pence, B; Macleod, B
(2001) Demographic surveillance and health equity in sub-
Saharan Africa. Health Policy and Planning, 16(4): 337-44.
Nhacolo, AQ; Nhalungo, DA; Sacoor, CN; Aponte1, JJ;
Thompson, R; Alonso, P (2006) Levels and trends of
demographic indices in southern rural Mozambique:
evidence from demographic surveillance in Manhiça district.
BMC Public Health, 6(291).
Tatem, A.J; Snow, R.W.; Hay, S.I. (2006) Mapping the
environmental coverage of the INDEPTH demographic
surveillance system network in rural Africa. Tropical
Medicine & International Health, 11(8):1318-26.

Demographic Surveillance Systems: ADDITIONAL RESOURCES

INDEPTH Network: An International Network for the

Continuous Demographic Evaluation of Populations and
Their Health in Developing Countries: http://www.indepth-
network.org
INDEPTH Network (2002) Population and health in
developing countries: Volume1: Population, health and
survival at INDEPTH sites. Ottawa: International
Development Research Centre.

Decennial Census
Decennial Census: BACKGROUND

A national population census involves collecting, evaluating, analysing and disseminating

demographic and socio-economic data on all persons in a country (or in a well-delimited part
of a country) at a specified time. Most countries do a national census every 10 years.

Where countries already ask about deaths in the household in the year before the census (or
some other reference period), adding questions on whether the deaths were pregnancy-related
can be an efficient way to get national and subnational estimates of maternal mortality,
including the proportion of maternal deaths among female deaths (PMDF), the maternal
mortality ratio (MMR), the maternal mortality rate (MMRate) and the lifetime risk of
maternal death (LTR).

However, if the census is used for the measurement of maternal mortality, it is essential that
these data be evaluated and adjusted, if necessary. Demographic techniques have been
developed to adjust for under/over reporting of births (reverse projection techniques or parity
to cumulated fertility ratios) and deaths (growth balance).
Data requirements include:

• the distribution of the male and female population by age

• the distribution of male and female deaths by age (preferably from the current and the
previous census)
• the number of live children ever born by age of woman
• the number of live births in the household during the year preceding the census

Identification of death:

Adult household members retrospectively report to the enumerator the age and sex of all
household members who have died within a specified period, usually 1 or 2 years prior to the
census. These are referred to as the Direct Mortality Questions.

Ascertainment of maternal/pregnancy related status:

For all female deaths to women aged 15-49 identified in the household, an adult respondent
reports if the death occurred during pregnancy, childbirth or within the six weeks following
the termination of the pregnancy. This series of questions is referred to as Pregnancy-related
Questions.

Advantages:

• it is a routine undertaking every 10 years (with the possibility of every five years if
there is a large inter-censal survey)
• estimates are more recent than those resulting from many other methods of maternal
mortality measurement (the reference period for maternal mortality estimates from the
census is generally one to two years before the census)
• the census provides data on births, deaths and socio-economic characteristics of the
household for the entire population - enabling reporting of maternal mortality by
urban/rural residence, region and other socio-economic characteristics
• the addition of pregnancy-related questions lengthens the duration of the census
interview only negligibly

Limitations:

• it is a large-scale undertaking
• data processing for the census may take considerable time
• it will not give detailed causes of maternal death (without the addition of a later,
separate interview including verbal autopsies)
• space on the questionnaire may be limited
• training of enumerators may be poor
• disintegration of households after death can mean there is no “household” to report
the death
 • there are no standard measures for quantifying uncertainty due to non-sampling
errors, and currently methods have yet to be developed to quantify uncertainty of the
adjusted estimates of maternal mortality

Census Questions

When information is collected on household deaths (by age and sex) in the past 12 months (or
some other reference period), the following question is asked for any woman who died
between the ages of 15 and 49:

• Did the death occur while she was pregnant, during childbirth or during the six weeks
(40 days) after the end of pregnancy?

The ideal response is “yes/no”, however the only available answer may be “unknown” or
“not sure”. *

Alternatively, this question can be broken into parts, which may be easier to understand:

• Did the death occur while she was pregnant? (“yes/no”; “unknown/not sure”)
• Did the death occur during childbirth (“yes/no”; “unknown/not sure”)
• Did the death occur during the six weeks (40 days) after the end of pregnancy
(“yes/no”; “unknown/not sure”)

*UN (2007) Principles and recommendations for population and housing censuses, Rev 2,
Draft. Page 131-2. New York: United Nations

Decennial Census: GUIDELINES

Hill K, Stanton C, Gupta N (2001) Measuring maternal

mortality from a census: Guidelines for potential users.
Chapel Hill: MEASURE Evaluation

Decennial Census: REFERENCES

Stanton C, Hobcraft J, Hill K, Kodjogbe N, Mapeta WT,

Munene F, Naghavi M, Rabeza V, Sisouphanthong B,
Campbell O (2001) Every death counts: measurement of
maternal mortality via a census. Bulletin of the World Health
Organization, 79(7): 657-64.
Hill K, on behalf of participants at the Expert Panel on
Capturing Maternal Mortality in the 2010 Census Round
(2006) Measuring maternal mortality. Lancet, December
368(9553): 2121

Decennial Census: ADDITIONAL RESOURCES

 Immpact/Health Metrics Network meeting Census meeting
(September 2006)
United Nations Statistics Division

Free downloadable software, US Census

Special topic: Health Surveys in Difficult settings

MMM-R has been developed with the support of Immpact

Initial concept and content developed by Oona Campbell and Lauren Foster
Website designed, built and maintained by Elizabeth Hauke

All rights reserv

Death Certificates
Death Certificates: BACKGROUND

Death certificates are vital records which document the medical causes and circumstances of
all deaths. A standard death certificate form asks for primary and underlying causes of death,
which may or may not capture those related to pregnancy. As a result, even in settings where
vital records are routine and complete, pregnancy-related deaths may be missed. A Modified
Death Notification Form, which includes additional questions and tick-boxes related to
pregnancy, is preferable. Please see the sample death certificates below from the United
Kingdom (does not include pregnancy-related questions) and the USA (a modified death
notification form) for examples.

In some settings, such as South Africa, a separate Maternal Death Notification Form (MDNF)
is used. This supplements a standard death certificate with more detailed information about
the circumstances of a maternal death. See below for a sample MDNF.

Death Certificates: GUIDELINES

Instructions for Classifying the Underlying Cause-of-Death,

ICD-10, 2007
Instructions for Classifying the Multiple Causes of Death,
ICD-10, 2007

Death Certificates: REFERENCES

 Harper, M; Parsons, L (1997) Maternal Deaths Due to
Homicide and Other Injuries in North Carolina: 1992-1994 -
A forty-year experience. Obstetrics and Gynecology; 90(6):
920-3.
Identifies maternal deaths solely through death certificates.
Horon, IL (2005) Underreporting of maternal deaths on
death certificates and the magnitude of the problem of
maternal mortality. American Journal of Public Health,
95(3): 478-82.
Hoyert DL (2007) Maternal mortality and related concepts.
National Center for Health Statistics. Vital & Health
Statistics 3(33).
MacKay AP, Rochat R, Smith SC, Berg CJ. (2000) The check
box. Determining pregnancy status to improve maternal
mortality surveillance. American Journal of Preventive
Medicine; 19: 35-39.

Death Certificates: ADDITIONAL RESOURCES

Sample Death Certificate, United Kingdom. Office of

National Statistics.
Sample Modified Death Notification Form, USA. Centers for
Disease Control and Prevention

Sample Maternal Death Notification Form, South Africa.

South African Department of Health.

Pattinson, RC. Guidelines for completing the maternal death

notification form. 2nd edition. Pretoria: South African
Department of Health. 1999.

Confidential Enquiry into Maternal Deaths (CEMD)

Confidential Enquiry: BACKGROUND

Confidential enquires into maternal deaths are intended primarily to assess the levels, causes
of and contributors to maternal mortality and to learn lessons to address these. A key feature
of the method is confidentiality, which is maintained at all levels keeping the women, health
care providers and institutions anonymous at all times. Very often, confidential enquiries
involve active surveillance of pregnancy-related or maternal deaths.
Confidential enquiries make extensive efforts to identify all maternal deaths. If confidential
enquiries are ongoing, they provide useful information on maternal mortality trends.

The United Kingdom has the longest history of confidential enquiry into maternal deaths,
dating back to the 1950s. National enquiries, either continuous or undertaken at regular
intervals, and sub-national enquiries have also been carried out in Australia, some states in
the USA, and in some European countries. Confidential enquiries have been started in parts
of Suriname, Malaysia, Israel, Indonesia and South Africa by adapting the United Kingdom
methodology. Time-limited enquiries have also been carried out in Jamaica, the Netherlands
and Egypt.

Another tool which modifies the traditional CEMD approach is called TRACE (Tracing
Adverse and Favourable Events in Pregnancy Care). Developed by Immpact, TRACE differs
from CEMD by including cases of severe obstetric morbidity (“near misses”) as well as
maternal deaths, by examining favourable as well as adverse factors associated with obstetric
care (useful for preserving morale in low-resource health facilities), and by assessing
community factors which may have contributed to a maternal death. For more information on
TRACE please see the Additional Resources section of this page.

Identification of death

Generally not applicable (only potential pregnancy-related or maternal deaths are reviewed)
but may also use a RAMOS approach.

Ascertainment of maternal/pregnancy related status

Varied, and sometimes multiple, sources used to identify and record all potential pregnancy-
related deaths.

Existing records for confidential enquiry can include:

• civil registration
• facility records
• morgue records
• burial/cemetery records
• newspapers
• census/survey of households
• survey of key informants, or Active reporting by Key Informants

Maternal deaths are usually ascertained by medical professionals (with or without diagnostic
aids and/or autopsy) or by verbal autopsy.

Advantages:

• provides additional data on quality of care associated with maternal deaths

• often has medical diagnoses of cause of death
Limitations:

• requires complete death reporting and good cause of death data

• if multiple sources are used, duplicate deaths need to be avoided
• can be complex and time consuming to undertake, especially on a large scale
• data on community level risk factors often missed when medical records relied on
• method is essentially an observational case series and risk factors can only be
quantified when contrasted with an appropriate comparison group

Measurement requirements:

• duplicate deaths need to be excluded

• sub-causes of maternal deaths must be coded and classified as maternal deaths
• data on births are needed to calculate MMR
• data on population of women aged 15-49 are needed to calculate MMRate

Confidential Enquiry: GUIDELINES

Lewis G (2004) Confidential enquiries into maternal deaths.

In: WHO (2004). Beyond the Numbers: Reviewing Maternal
Deaths and Complications to Make Pregnancy Safer. Geneva:
World Health Organization.

Confidential Enquiry: REFERENCES

D'Ambruoso L, Achadi E, Adisasmita A; Izati Y;

Makowiecka K, Hussein J. (2007) Assessing quality of care
provided by Indonesian village midwives with a confidential
enquiry. Midwifery. In press.
Drife, J (2006) Fifty years of the Confidential Enquiry in
Maternal Deaths. British Journal of Hospital Medicine,
Vol. 67(3): 121-5.
Drife, J (1999) Maternal mortality: lessons from the
confidential enquiry. Hospital Medicine. 60(3)3: 156-7.
Godber, G (1994) The origin and inception of the
Confidential Enquiry into Maternal Deaths BJOG: An
International Journal of Obstetrics and Gynaecology. 101
(11), 946-7.
Hussein, J (2007) Improving the use of confidential enquiries
into maternal deaths in developing countries. Bulletin of the
World Health Organization, 85(1): 68-9.
 Rankin J, Bush J, Bell R, Cresswell P, Renwick M (2006)
Impacts of participating in confidential enquiry panels: a
qualitative study. BJOG;113:387-92
Schuitemaker N, van Roosmalen J, Dekker G, van Dongen
P, van Geijn H, Gravenhorst J.B (1998) Confidential enquiry
into maternal deaths in the Netherlands: 1983-1992.
European Journal of Obstetrics & Gynecology and
Reproductive Biology, 79(1): 57-62.
Salanave B, Bouvier-Colle M-H, Varnoux N, Alexander S,
Macfarlane A, The MOMS Group (1999) Classification
Differences and Maternal Mortality: a European study. Int J
Epidemiol; 28: 64-9

Confidential Enquiry: ADDITIONAL RESOURCES

The Confidential Enquiries into Maternal Deaths in the

United Kingdom (2002) Why mothers die: 1997-1999:
Midwifery summary and key recommendations. London:
RCOG Press.
The Confidential Enquiries into Maternal Deaths in the
United Kingdom. Why mothers die: 2000-2002 report.
London: RCOG Press.

Maternal Death Data Collection Form, 2006-2008.

CEMACH.

NCCEMD, National Committee for the Confidential Enquiry

into Maternal Deaths (1998) Saving mothers: report on
confidential enquiries into maternal deaths in South Africa.
NCCEMD, National Committee for the Confidential Enquiry
into Maternal Deaths (1998) Interim Report on the
confidential enquiry into maternal deaths in South Africa.
Ministry of Health. Evaluation of implementation of the
confidential enquiries into maternal deaths in the
improvement of maternal health services. Kuala Lumpur,
Ministry of Health, 1998.
Immpact, University of Aberdeen (2007) Tracing adverse
and favourable events in pregnancy care (TRACE), Module
4, Tool 3. In: Immpact Toolkit: A guide and tools for
maternal mortality programme assessment. Aberdeen:
University of Aberdeen.
Civil Registration
Civil Registration: BACKGROUND

Civil registration is a routine, permanent, nationally mandated data source that captures vital
life events (namely live births, deaths, foetal deaths, marriages and divorces). It provides
legal documentation of such events and is the ideal source for vital statistics. However, even
in contexts where civil registration is complete, underreporting and misclassification of
maternal deaths is a frequent challenge. Civil registration can also be done on a sample basis,
identifying deaths either actively or passively - Sample Vital Registration or SAVVY
(Sample vital registration with verbal autopsy).

Identification of death

Family member and/or facility notify deaths to civil registration system at time they occur
(actively).

Ascertainment of maternal/pregnancy related status:

• medical professional or family gives cause

• ideally medically certified according to ICD reporting requirements
• occasionally, a Modified Death Notification Form is used to explicitly ascertain if
death is pregnancy-related
• occasionally, Birth-Death Linkage may be used to find pregnancy-related deaths

Advantages:

• serves multiple purposes and is representative of entire population

• on-going system that therefore does not require special data collection activities
• provides annual estimates of maternal mortality, regional differentials
• provide cause-specific estimates of maternal mortality
• provides data on births
• ideal for monitoring trends

Limitations:

• susceptible to poor quality - underreporting of deaths, incorrect cause of death

attribution (leading to misclassification of maternal deaths), untimely/late reporting of
deaths and births
• coverage may not be complete (routinely reported in only 78 countries/areas or 35%
of world population)
• rarely provides socio-economic data on determinants of maternal death

Measurement requirements:
 • sub-causes need to be grouped as maternal deaths (see: ICD Coding)
• under-reporting of deaths needs to be assessed/adjusted using demographic techniques
- see: Bennett-Horiuchi Method, Chandrasekaran-Deming, and Growth Balance
• quality of cause-of-death certification and coding needs assessment
• maternal death ascertainment can be corrected using Capture Recapture (Dual
method)
• data on births are also needed (and may need to be corrected) to calculate the
Maternal Mortality Ratio (MMR) and Total Fertility Rate (TFR)
• data on the population of women aged 15-49 are needed to calculate the Maternal
Mortality Rate (MMRate)

Civil Registration Forms

Sample Death Form:

Sample Death Certificate, USA. Centers for Disease Control

and Prevention

Sample Birth Certificate Form:

Sample Live Birth Certificate, USA. University of California,

San Francisco

Also see:

Sample Modified Death Notification Form, USA. Centers for

Disease Control and Prevention

Civil Registration: GUIDELINES

UN (2001) Principles and recommendations for a vital

statistics system. Revision 2. Department of Economic and
Social Affairs. Statistics Division. New York: United Nations.
ST/ESA/STAT/SER.M/19/Rev.2

UN (2002) Handbook on training in civil registration and

vital statistics systems. New York: United Nations.

Civil Regsitration: REFERENCES

AbouZahr C, Wardlaw T (2001) Maternal mortality at the end

of a decade: signs of progress? Bulletin of the World Health
Organization, 79(6): 561-73.
 Deneux-Tharaux, C, Berg, C, Bouvier-Colle, MH; Gissler, M;
Harper, M; Nannini, A; Alexander, S; Wildman, K; Breart, G;
Buekens, P (2005) Underreporting of pregnancy-related
mortality in the United States and Europe. Obstetrics &
Gynecology. 106(4):684-92.

Songane FF, Bergström S (2002) Quality of registration of

maternal deaths in Mozambique: a community-based study in
rural and urban areas. Social Science and Medicine, 54(1):
23-31.

Sombie, I.; Meda, N.; Hounton, S.; Bambara, M.; Ouedraogo,

T.W.; Graham, W. (2007) Missing maternal deaths: lessons
from Souro Sanou University Hospital in Bobo-Dioulasso,
Burkina Faso. Tropical Doctor, 37(2): 96-8.

Civil Registration: ADDITIONAL RESOURCES

Other Handbooks and Training Manuals

MMM-R has been developed with the support of Immpact

Initial concept and content developed by Oona Campbell and Lauren Foster

Website designed, built and maintained by Elizabeth Hauke

All rights reserved. Maternal Mor

Chandrasekaran-Deming Technique
Chandrasekaran-Deming Technique: BACKGROUND

Several demographic techniques have been developed to assess and adjust information on
deaths that come from civil registration, demographic surveillance, censuses or
population based surveys that ask about deaths of household members during a specified
time before the survey period.

Some methods compare data from independent sources (direct capture-recapture approaches)
while others are analytic indirect methods based on assumptions on the population age
distribution.
Preston (1984) reviews the former, which are based on the Chandrasekaran-Deming
technique. Chandrasekaran-Deming is an early demographic method designed to estimate
population birth and death rates from different demographic data collection systems.
Originally developed in 1949, it estimates the total number of events or cases (e.g. deaths)
from the numbers of cases reported by the two reporting systems independently. These can
include the death registration system, censuses, household surveys or active reporting by key
informants. For example, deaths reported in an independent survey of mortality are compared
to deaths reported in the death registration system for the same population. Unmatched and
unrecorded deaths can thus be identified and estimated.

Limitations:

• lack of independence
• incompatible areas of reference
• incompatible reference periods
• difficulties in identifying matches in the deaths reported
• can be an expensive means to assess data completeness since a sufficiently large
population sample is required

Chandrasekaran-Deming Technique: REFERENCES

Chandrasekaran, C; Deming WE (1949) On a method of estimating birth and death

rates and the extent of registration. Journal of the American Statistical Association
44(245): 101-15.
Greenfield, CC (1983) On estimators for dual record systems. Journal of the Royal
Statistical Society. Series A (General), 146(3): 273-80.
Preston SH: Use of direct and indirect techniques for estimating the completeness of
death registration systems. In: Data Bases for Mortality Measurement. Edited by:
United Nations. New York: United Nations; 1984.
Seltzer, W; Adlakha, A (l974), On the effect of errors in the application of the
Chandrasekaran-Deming Technique. Laboratory for Population Statistics Reprint
Series No.14, University of North Carolina, Chapel Hill.

Birth and Death Record Linkage

Birth and Death Record Linkage: BACKGROUND

Birth and death record linkage is an analytical approach that aims to identify pregnancy-
related deaths by linking records of births with those of reproductive-age female deaths.
Deaths that can be linked to within a prescribed time-period of a birth (or recorded
pregnancy) are deemed to be pregnancy-related deaths. These can be further investigated to
determine whether they are maternal.
This approach has been applied with historical data by using parish records of baptisms and
burials. It has also been used in settings with unique person identification systems such as
Finland, and in settings such as the US with good vital registration. Brazil has also
experimented with this approach using civil registration and it has been used in Demographic
Surveillance System (DSS) sites such as Matlab, Bangladesh.

Advantages:

• it uses existing records and overcomes some problems with misclassification of

pregnancy-related deaths
• it can be easy if records are computerized and identifiers are good

Limitations:

• it misses deaths associated with abortions

• it may miss deaths associated with stillbirths (if only records of live births exist)
• it requires high coverage of births and deaths
• if unique identifiers do not exist and matching has to be done based on mother’s
name, this may be difficult

Other data requirements:

• the need to be able to link birth and death records by date and by either a unique
identifier linking the mother with the foetus/infant or by the mother’s name (or
soundex of surname, first initial, date of birth, sex, or postcode)

Birth and Death Record Linkage: GUIDELINES

No guidelines for Birth and Death Record Linkage to identify maternal death exist to our
knowledge. Existing records of births (including stillbirths), pregnancies (if available) and
deaths are obtained from routine Civil Registration sources (or historically using parish
records of baptisms and burials) or from Demographic Surveillance Systems.

A general manual on record linkage (aka data integration):

Statistics New Zealand (2006) Data Integration Manual.

Wellington: Crown Copyright.

Birth and Death Record Linkage: REFERENCES

Canadian Perinatal Surveillance System. Progress Report

1997-1998. Maternal mortality and morbidity study. Ottawa:
Public Health Agency of Canada.
 Dye, TD; Gordon, H; Held, B; Tolliver, NJ; Holmes, AP
(1992) Retrospective maternal mortality case ascertainment
in West Virginia, 1985 to 1989. American Journal of
Obstetrics and Gynecology; 167(1): 72-6.

Gissler, M; Hemminki, E; Lonnqvist, J (1996) Suicides after

pregnancy in Finland, 1987-94: register linkage study. BMJ,
313: 1431-4.

Jocums, S; Mitchel, EF Jr; Entman, SS; Piper, JM (1995)

Monitoring maternal mortality using vital records linkage.
American Journal of Preventive Medicine; 11(2): 75-8.

Misclassification of maternal deaths - Washington State.

Morbidity and Mortality Weekly Report; October 03, 1986,
35(39);621-3. Centers for Disease Control and Prevention.

Georgia, 1990-1992. Morbidity and Mortality Weekly Report;

February 10, 1995, 44(05);93-96. Centers for Disease Control
and Prevention.

An historical example:

Hammel, EA; Gullickson, A (2004) Kinship structures and

survival: Maternal mortality on the Croatian-Bosnian border
1750-1898. Population Studies-A Journal of Demography.
58(2): 145-59.

Birth and Death Record Linkage: ADDITIONAL LITERATURE

Atrash, HK; Ellerbrock, TV; Hogue, CJR; Smith, JC (1989)

The need for national pregnancy mortality surveillance.
Family Planning Perspectives; 21(1): 25-6

A review of record linkage focusing on infant mortality. Contains useful general information:

Machado, CJ (2004) A literature review of record linkage

procedures focusing on infant health outcomes. Cadernos de
Saúde Pública; 20(2): 362-71.
 Ahmed K. Elmagarmid, Panagiotis G. Ipeirotis, Vassilios S.
Verykios (2007) Duplicate record detection: a survey.
Transactions on Knowledge and Data Engineering; 19(1): 1-
16

Birth and Death Record Linkage: ADDITIONAL RESOURCES

Free probabilistic record linkage programme developed for

Cancer Registries, yet helpful for general information. US
Centers For Disease Control and Prevention (CDC) for Cancer
Registries

Parallel Large Scale Techniques for High-Performance Record

Linkage. Australian National University

Autopsy
Autopsy: BACKGROUND

Also known as a post-mortem examination, an autopsy helps to determine primary and

underlying causes of death by examining the body and organs of a deceased person. Tissue
and other samples may be taken during autopsy and tested by pathologists to provide further
information. Findings are detailed in autopsy and associated pathology reports.

Autopsy: REFERENCES

Hutchins, GM; Berman, JJ; Moore, GW; Hanzlick, R; and the

Autopsy Committee of the College of American Pathologists
(1999) Practice Guidelines for Autopsy Pathology Autopsy
Reporting. Archives of Pathology & Laboratory Medicine;
123: 1085-92.

Le Coeur, S; Ronsmans, C; Halembokaka, G; Augereau-

Vacher, C; Khlat, M (2006) Comparison of family reporting
of pregnancy status with a post-mortem ß-HCG test in
deceased women: a study in Pointe-Noir, Congo. Tropical
Medicine and International Health; 11(4): 528-31.
 Millward-Sadler, GH. Pathology of maternal deaths. In:
Kirkham, N; Shepherd, N (eds). Progress in Pathology,
Volume 6. pp. 163-185. London: Greenwich Medical Media
Ltd, 2003.

Autopsy: ADDITIONAL RESOURCES

Hoyert D. The autopsy, medicine, and mortality statistics.

National Center for Health Statistics. Vital Health Stat 3(32).
2001. Centers for Disease Control

The Royal College of Pathologists. Guidelines on autopsy

practice: best practice scenarios. Scenario 5: maternal death.
2005. London: The Royal College of Pathologists.

Active Surveillance of Pregnancy-Related and Maternal

Deaths
Active Surveillance of Pregnancy-Related and Maternal Deaths: BACKGROUND

Active surveillance into pregnancy related and maternal death is most often done as part of
Confidential Enquiries into Maternal Deaths, but can be used to obtain an estimate of the
maternal mortality ratio. A key feature is the ongoing effort to actively identify potential
maternal deaths as they occur. In the UK confidential enquiry, for example, health staff
including midwives, obstetricians and general practitioners are encouraged to report any
possible maternal deaths to a confidential enquiry assessor. In the USA, active surveillance is
done using a modified death notification form, birth and death record linkage and active
notification by individual practitioners, maternal mortality committees and even newspaper
reports. In Egypt, maternal mortality surveillance is ongoing based on a modified death
notification form and interviews with the families of all pregnancy related deaths.

Identification of death

Not Applicable (only potential pregnancy-related or maternal deaths are reviewed)

Ascertainment of maternal/pregnancy related status

Varied and sometimes multiple sources are used to identify and record all potential
pregnancy-related deaths:

• Active reporting by Key Informants

 • Maternal deaths ascertained by medical professionals (with or without diagnostic aids
and/or autopsy)
• Verbal autopsy

Advantages:

• provides additional data on quality of care associated with maternal deaths

Limitations:

• requires complete death reporting and good cause of death data

• if multiple sources are used, duplicate deaths need to be avoided
• can be complex and time consuming to undertake, especially on a large scale

Measurement requirements:

• duplicate deaths need to be excluded

• sub-causes of maternal deaths must be coded and classified as maternal deaths
• data on births are needed to calculate MMR
• data on population of women aged 15-49 are needed to calculate MMRate

Active Surveillance of Pregnancy-Related and Maternal Deaths: REFERENCES

Daniel J. Pallin, MD, MPH, Vandana Sundaram, MPH,

Fabienne Laraque, MD, MPH, Louise Berenson, MS, and
David R. Schomberg, Rpa-C (2002) Active Surveillance of
Maternal Mortality in New York City. Am J Public Health,
August; 92(8): 1319-1322.
Centers for Disease Control and Prevention. Surveillance
Summaries, February 21, 2003. MMWR 2003:52(No. SS-2).
Drife, J (2006) Fifty years of the Confidential Enquiry in
Maternal Deaths. British Journal of Hospital Medicine, Vol.
67(3): 121-5.
Drife, J (1999) Maternal mortality: lessons from the
confidential enquiry. British Journal of Hospital Medicine,
60(3)3: 156-7.
Godber, G (1994) The origin and inception of the
Confidential Enquiry into Maternal Deaths BJOG: An
International Journal of Obstetrics and Gynaecology 101
(11), 946-7.

Active Surveillance of Pregnancy-Related and Maternal Deaths: ADDITIONAL

RESOURCES
 Maternal Death Data Collection Form, 2006-2008.
CEMACH.

Saving mothers: report on confidential enquiries into

maternal deaths in South Africa 1998

Ministry of Health. Evaluation of implementation of the

confidential enquiries into maternal deaths in the
improvement of maternal health services. Kuala Lumpur,
Ministry of Health, 1998.

News
February 2008 News

Monday 25th February 2008

Incessant alcohol abuse killing more in the Upper West

The high maternal and juvenile mortality in the Upper West
Region is the result of incessant alcohol abuse among expectant
mothers, the Regional Director of health, Dr Erasmus Agongo has
said.

Sunday 24th February 2008

Negligent medics to be interdicted

The Ministry of Health is to interdict negligent health workers to
check maternal deaths. The assistant commissioner for
reproductive health, Dr. Anthony Mbonye, said maternal death
audits were being conducted and a report would be ready soon.

Friday 22nd February 2008

Commissioner advises Nigerian men on maternal mortality

Dr. Fred Achem, the commissioner for health in Kogi, in Lokoja
advised Nigerian men to endeavour to know more about maternal
mortality rate in the country. Achem gave the advice at the
opening of a two-day advocacy meeting of stakeholders on
strategies to reduce maternal mortality rate in the state. He said
more knowledge was necessary not only for the men to appreciate
the enormity of the problem, but to contribute towards the
 reduction of the incidence.

Thursday 21st February 2008

Express vehicles for moms - Emergency transport for

expecting women
In an attempt to encourage institutional delivery and make
transportation available 24 hours for pregnant women, National
Rural Health Mission and health and family welfare department
would launch a scheme called Janani Express.

Tuesday 19th February 2008

Many maternal deaths worldwide are preventable

Women who die during pregnancy and childbirth in sub-Saharan
Africa, more may die from treatable infectious diseases than from
conditions directly linked to pregnancy, revealed by researchers.

Wednesday 13th February 2008

Malawi Midwives body hits at female legislators

Association of Malawi Midwives (AMM) and White Ribbon
Alliance for Safe Motherhood (WRAM) has voiced concern that
female parliamentarians have failed to articulate issues that affect
women in parliament and sees no need for having a deliberate
policy on more women representation in the August
House...Kamwendo was wary that Malawi is losing many women
due to complications during or after giving birth as mothers are
not protected by law.
60 million dollar USAID funded family planning programme
launched in Pakistan
...according to international studies Pakistan’s health indicators are
among the worst in the world. For every 100,000 children born,
350 to 500 mothers die annually.
Nigeria: Quack Midwifes to Blame for Maternal Mortality -
Experts
The Society of Gynaecology and Obstetrics of Nigeria yesterday
attributed the increase in maternal and neonatal mortality in the
country to quack midwifes undertaking 90 percent of deliveries.

Friday 1st February 2008

 2008 to be celebrated as year of LHWs
The Ministry of Health has declared 2008 as the ‘Year of Lady
Health Workers.’ The decision has been taken to celebrate the
contributions of LHWs as a cadre that safeguards the health needs
of the community, and to capitalize the outreach of these ‘lady
soldiers on foot’ to overcome the deficits of infant-, child- and
maternal mortality rates in Pakistan.
Efforts to curb maternal deaths doubled
Health officials in Capiz are doubling their efforts to decrease and
even curb maternal deaths. Maternal and Child Health
Coordinator Cristita Grandflor of the Provincial Health Office of
Capiz said that they have been conducting a series of trainings for
health workers in the province.

January 2008 Archive

December 2007 Archive

November 2007 Archive

October 2007 Archive

News
Announcements

Tuesday 16th October 2007

Call for collaboration - Sampling at Service Sites

Immpact would like to form a collaborative network to undertake
further development of Sampling at Service Sites (SSS).

General News

Friday 28th March 2008

Gordon Brown’s wife Sarah writes in the Mirror

There is no excuse for letting mums die as they give birth.
Across the world half a million mothers’ lives will be lost
between now and this time next year.

Thursday 27th March 2008

 Sarkozy’s wife rallies British women to help end
childbirth tragedies
Carla Bruni-Sarkozy yesterday called on women to unite
in the battle to help mothers who die needlessly in
childbirth.

Wednesday 26th March 2008

Nigeria: Gynecologist Decries Lack of Radiotherapy

Equipment
A gynecologist, Dr Emily Nzeribe, says Maternal
Mortality Rate (MMR) in the south-east will reduce if
radiotherapy equipment are installed in the hospitals in the
area.

Tuesday 25th March 2008

Nigeria: And Our Women Continue to Die

It is estimated that about 50,000 Nigerian women die
annually due to maternal mortality.

International maternal mortality figures still too high

The Irish Forum on Global Health has expressed its
concern that many countries are not on track to reach the
UN target of reducing maternal mortality figures.

Monday 17th March 2008

Global shortfalls in health workers

UNFPA urges action
“To reduce maternal mortality, we must make sure that
these women receive the specialized care they need during
delivery, as well as pre- and postnatal services.”

Voices for Malaria-free Future launched

“Most frightening however, is the fact that 18% of all
under-five deaths are caused by malaria and 9% of
maternal deaths are also caused by the disease”, she
revealed.

Monday 10th March 2008

 Nigeria: High Maternal Mortality Worries Sultan
The Sultan of Sokoto, Alhaji Muhammad Sa'ad Abubakar
III has expressed concern over the high rate of maternal
mortality in the country and has stressed the need to check
the trend.

Liberia: Maternal Health Worsened Since War Ended

New statistics showing an increase in maternal mortality
since the end of Liberia's civil war in 2003 have created
alarm among health workers who say the country's almost
non-existent healthcare system is to blame.

Ghana: Unicef Stresses Importance of Investing in

Maternal Health
UNICEF marked International Women's Day 2008 last
Saturday, 8 March by drawing attention to the need for
improvements in maternal health care.

77,000 Indian women die every year at childbirth

At least 77,000 mothers in India die every year during
child birth, Minister of State for Women and Child
Development Renuka Chowdhury said Monday.
Chowdhury told the Rajya Sabha that the latest survey
report of the Registrar General of India published in 2006,
the maternal mortality ratio (MMR) for India is 301 per
100,000 live births.

February 2008 Archive

January 2008 Archive

December 2007 Archive

November 2007 Archive

October 2007 Archive

About MMM-R
Maternal Mortality Measurement Resource (MMM-R) provides a central source of maternal
mortality measurement information and resources.

It is targeted towards:

• research scientists
• programme managers
• health professionals
• policy makers
• others involved in safe motherhood work

Information is geared both towards those who are new to the field of maternal mortality
measurement and those who are experienced in the area.

We welcome new material and links on maternal mortality measurement. We give full credit
to the originating groups or individuals.

Please email us your suggestions.

Funding for MMM-R is through Immpact. Immpact is funded by the Bill and Melinda
Gates Foundation, the Department for International Development, the European
Commission and USAID.

The funders have no responsibility for the information provided or the views expressed on
this website.

We use logos next to hyperlinks and materials that we cite in order to indicate the originating
source. We are in the process of obtaining permission for these. If you would like us to
remove your logo, please contact us and we will do so right away.

About Maternal Mortality

Maternal mortality is one of the most discrepant public health indicators between developed
and developing countries.

Starrs A. The Safe Motherhood Action

Agenda: Priorities for the Next Decade.
Report on the Safe Motherhood

Worldwide, over half a million women die each year of largely preventable maternal causes.

UNFPA (2000) The state of the world

population 2000: Lives together, worlds
apart. New York: UNFPA
The disproportionate burden of excess maternal mortality in developing countries is
recognised as a major international problem and is included as Millennium Development
Goal 5, which will be judged based on success in reducing maternal mortality by 2015.

MDG 5: to reduce maternal mortality by 75%

by 2015

While considerable advances have been made in increasing international commitment to

reducing maternal mortality, the challenges associated with measuring it in low-resource
settings remain a barrier to progress. There is currently an unprecedented need for estimating
maternal mortality in developing countries.

Since the launch of the Safe Motherhood Initiative in 1987, new and enhanced approaches for
measuring maternal mortality have been developed, tested and used. This site shares these.

Guide
Welcome to the Guide.

This interactive tool will help you to determine which sets of materials will be the most
helpful for your work. Simply answer the following questions and let us search out a short-
list of materials for you to review.

If you already know what you need, go straight to the Materials section and use the index.

These basic questions will help to determine the tools best suited to your measurement needs:

Question

Are you interested in methods and techniques for:

generating population-based estimates of levels and trends or

causes of maternal mortality

adjusting data quality

Background Information
Maternal mortality is complex and multi-dimensional, as are the approaches used to measure
it (for more advanced information please see the Materials section of this website).
The common foundation among all measurement techniques, however, is an understanding of
core principles about maternal mortality research such as:

• why is it important to measure?

• how is it different from other mortality indicators?
• how can it be measured in challenging contexts?

The Basic Information page is designed to provide information to users new to the area of
maternal mortality measurement.

The History page contains points of interest regarding the history of Safe Motherhood, and
of the measurement of maternal mortality over the years.

MMM Resource
Welcome to the Maternal Mortality Measurement Resource website!

Maternal Mortality Measurement Resource (MMM-R)

is a web-based collection of methods, techniques and
approaches for measuring maternal mortality.
This website has been divided into four sections -
each accessed via one of the buttons in the top menu.
Within each section, please use the menu bars on the
left to navigate between pages.

Our ‘home’ section contains useful information about this website. You
can also find a brief introduction to maternal mortality here. For more
indepth information, please see the ‘Background’ section. Please use
the contact us page to send us an email if you have any questions, or
would like to add any material to our wesbite.

Our ‘background’ section contains lots of history and background

information about measuring maternal mortality. If you are new to
this field, the ‘background’ section will provide you with an excellent
foundation before you move on to look at specific methods and
resourecs.

The ‘guide’ is an interactive tool to enable you to find the most

relevant materials and resources for your needs. This is especially
useful if you are not sure which materials to look at first. Answering a
few simple questions about the work you wish to do, will allow us to
produce a shortlist of methods and resources for you to start with.

In the ‘materials’ section you will find an extensive index of tools,

materials, resources and sources of further information. Each page
contains a wealth of information and links to references, examples and
further resources. In addition, this section contains some materials that
provide an overview of the vast range of methods that exist.

New! All the tool pages are now available as PDFs to save or print.