The document discusses methods for measuring vital events like births and deaths. It describes the SAVVY (Sample Vital Registration with Verbal Autopsy) approach used by MEASURE Evaluation to strengthen civil registration and vital statistics systems in countries. SAVVY involves a census, registration of vital events, and verbal autopsies to determine causes of death. The document discusses case studies of SAVVY implementation in Tanzania and Zambia, noting achievements like improved data quality and integration into national health information systems. It also covers maternal mortality estimation using surveys and census data, challenges in ascertaining causes of maternal death, and a study of indirect causes like HIV/malaria in Mozambique.
8. Strengthening Health Information
and Civil Registration/Vital Statistics
Systems – SAVVY– A Promising Approach
Robert Mswia
MEASURE Evaluation
End-of-Phase-III Event, May 22,
2014
9. Stepping Stones to a Vital Statistics System
Source: WHO/Health Metrics Network
DSS
SAVVY
10. SAVVY Data Collection: Census Enumeration
A census enumerator using a Tablet PC, and a
pen-and-paper when needed (Source: SAVVY
Tanzania)
11. SAVVY Data Collection: Verbal Autopsy Interview
A SAVVY Coordinator
conductingVA interview
for a probable cause of
death (Source: SAVVY
Tanzania)
12. Active Reporting of SAVVY Vital Events
A SAVVY Facilitator training Key Informants on using mobile
phones for reporting birth and death events in their communities
13. Different Scenarios – Examples
From 3 Countries Currently
Implementing SAVVY
Zambia
Tanzania
Malawi
15. Case Study 2 – Zambia
SAVVY seen as key strategy for:
Improving quality and coverage of vital
events
Building national system of vital
registration
16. 5.0
9.9
1.2
1.3
1.5
1.8
2.4
3.1
3.8
5.6
5.7
5.9
6.2
6.5
9.8
10.9
19.3
4.3
10.2
1.5
0.4
1.6
2.3
1.4
3.1
3.2
4.1
4.8
4.9
9.0
5.9
5.3
4.4
12.2
21.5
25 20 15 10 5 0 5 10 15 20 25
Ill-defined & undetermined causes
All other remaining causes
Maternal causes
Disorders of the kidney
Meningitis
Senility/oldage
Diabetes mellitus
Neoplasms
Stillbirth
Diarrhoeal diseases
Perinatal and neonatal conditions
Pneumonia/ARI
Disease of the circulatory system
Tuberculosis
Malnutrition
Injuries & Accidents
AFI / Malaria
HIV-related diseases
Females
(N=1,282)
Males
(N=1,474)
Example: Output from SAVVY
Zambia – Leading Causes of Death
17. SAVVY and the Saving Mothers,
Giving Life Initiative (SMGL)
in Zambia
SAVVY approach adopted for M&E of
SMGL initiative in 4 districts.
Document change in maternal deaths
before and after implementation of
program interventions
18. SMGL Zambia: Percent Distribution
of Maternal Deaths by Main Causes
(N = 91 maternal deaths)
19. SAVVY Implementation in Zambia
and Tanzania – Achievements
Change in practice: Cause of death
certification based on ICD
SAVVY data to complement HMIS data
SAVVY incorporated into national M&E
strategic plans as key data source
21. Key Issues in Maternal
Mortality-Estimation and
Cause of Death
Kavita Singh
MEASURE Evaluation
End-of-Phase-III Event, May 22, 2014
22. A Human Rights Issue
Motherhood and
childhood are
entitled to special
care and
assistance.
UN General Assembly, 1948.
Article 25 of The UN Declaration
of Human Rights
23. Maternal Mortality Estimation
Challenges
Incomplete vital
registration systems
Large sample sizes
needed for surveys
Two key activities:
2010 Bangladesh Maternal
Mortality Survey
2014 Workshop on
Estimating Maternal
Mortality from Census Data
24. 2010 Bangladesh Maternal
Mortality Survey (2010 BMMS)
175,000 households surveyed
Precise estimate of maternal
mortality
Rich data on socio-
economic and biological/
health services factors
25. Ascertaining Cause of Death
from the 2010 BMMS
Sisterhood Method
Married womenareaskedaboutsurvival status
ofsisters
Questions on Household Deaths
Householdswere askedaboutdeaths since 2006
Verbal Autopsy follow-up
Householddeaths ofwomen13to49 were followed upwith a
verbal autopsy
26. Estimation from Census Data
Largesamplesizesavailable
Questionscanbeincludedonrecent
deathsandspecificallypregnancy-
relateddeaths
Yields pregnancy-related mortality
rather than maternal mortality
estimates
Sub-national level estimation is
possible
27. Estimation from Census Data
Need to carefully assess quality of
census data first
Reporting of adult (female) deaths
Reporting of pregnancy-related deaths among all
deaths of women of reproductive age
Reporting of births
Key Resource: WHO (2013). WHO Guidance
for Measuring Maternal Mortality from a
Census. WHO, Geneva
28. Cause of Maternal Death
WHO (2010) A global overview of Maternal Mortality.
http://www.childinfo.org/maternal_mortality.html
Cause of Maternal Deaths Globally 1997-2007
29. Indirect Causes of Maternal Mortality:
A Study of Mozambique
Post Census Verbal Autopsy Survey Data 2007-2008
Indirect Causes
18.2%: HIV; 23.1%: Malaria
Study also found
Provincial-level variation in cause of death.
Population-level data revealed more indirect deaths
than facility-level data.
Singh,K, Moran, A, Story, W, Bailey, P, Chavane, L. Acknowledging HIV and Malaria as Major
Causes of Maternal Mortality in Sub-Saharan Africa: A Study of Mozambique (forthcoming)
Curtis, S, Mswia, R, Weaver E. Application for Measuring Maternal Mortality: Three Case Studies
using Verbal Autopsy. Presentation at the 2013 IUSSP International Population Conference.
Busan, Republic of Korea.
31. Summary
Survey Methods: Provide a rich source of
information
Census Method: A promising method of
pregnancy-related mortality estimation
From a program and policy perspective
Indirect causes of maternal mortality need attention in
Africa
Cause of death and sub-national population-level data
are crucial