Comparison of Cause-of-Death Classification Methods for Verbal Autopsies in Mozambique: 2017 Inquérito Sobre Causas de Mortalidade (INCAM)-2 Pilot


PDF document icon wp-20-240.pdf — PDF document, 548 kB (562,154 bytes)

Author(s): MEASURE Evaluation

Year: 2020


MEASURE Evaluation. (2020). Comparison of Cause-of-Death Classification Methods for Verbal Autopsies in Mozambique: 2017 Inquérito Sobre Causas de Mortalidade (INCAM)-2 Pilot. Chapel Hill, NC, USA: MEASURE Evaluation
Comparison of Cause-of-Death Classification Methods for Verbal Autopsies in Mozambique: 2017 Inquérito Sobre Causas de Mortalidade (INCAM)-2 Pilot Abstract:

In advance of the August 2017 Population and Housing Census, the Mozambique National Institute of Statistics (INE) and Ministry of Health (MISAU) began discussing the implementation of a post-census mortality survey. This would be the country’s second such survey. Following the 2007 Population and Housing Census, INE and MISAU conducted a post-census mortality survey (Inquérito Sobre Causas de Mortalidade [INCAM]) in 2007/2008 using verbal autopsies (Mozambique National Institute of Statistics, et al., 2012). Like the first INCAM, the 2017 Population and Housing Census included questions on household deaths in the previous 12 months, collecting the name, sex, age, and date of death for the deceased. A sample of census enumeration areas representative of the national and provincial levels would then be selected, and all deaths reported in the census in those areas would be visited to administer a verbal autopsy (VA) for each death.

In 2008, INCAM used physician-based certification of cause of death based on the VAs. Given the cost of using physicians to determine the cause of death from verbal autopsies and the development in the past 10 years of alternative methods for interpreting VAs, it was decided that the pilot for INCAM-2 would focus on a comparison of multiple methods for determining the cause of death. The goal of the comparison is to help the government select the most appropriate interpretation method for full implementation of INCAM-2. Based on other studies, it was determined that a minimum of 300 completed VAs with at least 30 neonatal deaths would be needed (King, Lu, & Sibuya, 2010).