Community-based Verbal Autopsy linked to Kenya’s Civil Registration and Vital Statistics System

Project Name: Community-based Verbal Autopsy linked to Kenya’s Civil Registration and Vital Statistics System

Project Status: Proposed

Point of Contact: Erin Nichols

Center: National Center for Health Statistics

Keywords: Verbal autopsy, mortality surveillance, vital statistics, civil registration

Project Description: Verbal autopsy (VA) is the process of interviewing the caregiver of a deceased person to capture signs and symptoms experienced by the deceased prior to death, resulting in a probable cause of death. The process has emerged as the leading method targeted to enhance mortality surveillance efforts among the more than 2/3rds of the world’s population for which low quality or no mortality data are available. Despite recent advances in tools to streamline and automate the VA data collection and analysis processes, successful implementation within a community-based (i.e., non-research) setting and in the context of a country’s legal registration and vital statistics (CRVS) system has not been achieved. Building from an existing CDC CRVS improvement project in Homa Bay County, Kenya, a county plagued by a national high HIV prevalence of 27% and under-five mortality rate of 130 per 1,000 live births, this activity proposes to demonstrate for the first time community-based VA implementation linked to a national CRVS system.

  • Potential impact of project if successful:
    Successful VA implementation in Homa Bay will for the first time provide the county with information on causes of community deaths and enable mortality surveillance for public health decision making. On a greater scale, the demonstration will inform Kenya’s national verbal autopsy agenda, which will enhance national mortality surveillance, providing routine information on community deaths.
  • Applicability of project for reuse within other public health programs:
    Beyond strengthening mortality surveillance in Kenya, this project will serve as a first model for the many partners around the world that have recently invested in strengthening mortality surveillance and CRVS systems, including implementing and supporting routine VA, particularly in Sub-Saharan Africa.  These partners include the WHO VA Working Group, the African Program on Accelerated Improvement of CRVS (APAI-CRVS), the UN/World Bank Global Financing Facility, and the $100 million Bloomberg Data for Health Initiative which is investing in 20 countries, including four CDC project countries.  Lessons learned will also inform other CDC initiatives using verbal autopsy, including CHAMPS (Child Health and Mortality Prevention Surveillance Network), the Division of Reproductive Health, and the Global Health Security Agenda.
  • Methodology and planned implementation of project:
    When a death occurs at the household, a volunteer Community Health Worker (CHW) assigned to the household reports the death and notifies the local civil registration authority and the CHW’s supervisor—a Community Health Extension Worker (CHEW). The CHEW conducts the verbal autopsy interview at the household using an ODK-based data collection platform programmed in the local language on a locally-acquired smart phone. The interviewer uploads the data to Google Cloud. The analysis team then downloads the data and uses an automated verbal autopsy analysis software (Inter-VA) to provide a probable cause of death. The output of the VA process yields cause-specific, population level mortality statistics that are processed and reported through various levels of the Ministry of Health and Civil Registration Department.A phased approach of testing and implementation is scheduled.  Phase 1, proof of concept testing, was successfully completed in December 2015.  The project proposed for funding comprises Phase 2, a demonstration in one sub-county in Homa Bay County. Assuming a successful demonstration, determined by a six-month evaluation, Phase 3 would be gradual roll-out to selected sub-counties throughout Kenya, with a built-in, iterative process for review and revision. Roll-out would be informed by the experience of Godefay, et al. in Tigray, Ethiopia to reach both regionally and nationally representative information.
  • Innovation of expected results with how you will determine overall project success:
    The innovation in this model is two-fold: 1) the application of the mobile IT platform among the government-supported Community Health Unit structure; and 2) the inter-agency linkage of the verbal autopsy process, conducted within the Ministry of Health, with the Civil Registration Department to officially notify and register community deaths.
    • Success of the demonstration will be measured through periodic evaluation of the system to assess usability and acceptability of the mobile-IT platform, registration coverage of reported deaths, and quality of data reported by interviewers. Success will also be marked by the completion of the first report on cause of death statistics for community deaths.
    • Success of the innovation model will be measured by acceptance and adoption of the model by the national Mortality Subcommittee and CRVS Technical Working Group.
  • Stakeholders of project including STLT, Academic, extramural programs:
    • Within Kenya: Government of Kenya (national and subnational Ministry of Health and Civil Registration Department); CRVS improvement partners, including Measure Evaluation/PIMA, UNICEF, MSF, UNFPA, and WHO.
    • Outside of Kenya: CRVS improvement partners, including Measure Evaluation, UNICEF, WHO, UNECA’s African Program on Accelerated Improvement of CRVS (APAI-CRVS), and Bloomberg Data for Health Initiative.

For more information about this project, please contact the CHIIC at or Brian Lee at

Page last reviewed: February 15, 2019