Skip Navigation

Institute for International Programs

Malawi Verbal Autopsy and Social Autopsy Studies (VASA)

The National Statistics Office of Malawi (NSO), working with JHSPH, conducted the 24,000 household Rapid Mortality Monitoring (RMM) Reference Survey from October 2011 to February 2012. This survey, which included a full birth history to identify neonatal and child deaths, was conducted as a “gold-standard” reference survey to help evaluate methods for the rapid monitoring of child deaths. Balaka and Salima districts were selected for this effort based on their high mortality and fertility, accessibility to the study team, full coverage of community health workers, and average population size.

JHSPH/CHERG worked with the NSO to conduct a VASA study by revisiting households with a recent death identified by the RMM Reference Survey to conduct a VASA interview. The RMM survey identified about 1300 under-five deaths with a 4-year recall period, among whom about 450 were neonatal deaths and 900 were post-neonatal deaths. The VASA study aimed to provide estimates of the proportions of deaths due to important causes for neonates and 1-59 month olds with the following three specific objectives;

  1. Estimate the cause distributions of death, separately for neonates (0-27 days old) and young children (1-59 months old), in Balaka and Salima districts, Malawi;
  2. Estimate the proportions of deaths, by cause, of neonates and young children that occur in health facilities in Balaka and Salima districts, Malawi; and
  3. Estimate the prevalences of social factors contributing to deaths of neonates and young children in Balaka and Salima districts, Malawi.

A total sample of 1289 under-five deaths in the 4-years prior to the RMM survey, including 470 neonates and 819 1-59 month olds, were identified for a VASA follow-up interview. Of those, VASA interviews were successfully completed for 1,090 child deaths, including 79 stillbirths (identified, mainly, as neonatal deaths by the RRM), 320 neonatal (0-27 days) deaths and 691 young child (1-59 months) deaths. The majority of the parents of the deceased children had some primary level of education, and agriculture was the dominant occupation in most of the households. A very small percent of the households used piped water and it took on average 2 hours to reach the nearest health facility from the household. The results showed that one-half of caregivers of deceased newborns did not seek care for the fatal illness, while careseeking was delayed for almost all sick children even after recognition of a possibly severe or severe illness symptom. Generally, there was low coverage of key interventions along the continuum of care for pregnant women (4+ quality ANC visits), neonates (thermal care, kangaroo care for preterms) and young children (full immunization). Detailed findings, conclusions and recommendations can be found in the study report.


Understanding misclassification between neonatal deaths and stillbirths: empirical evidence from Malawi. Liu L, Kalter HD, Chu Y, Kazmi N, Koffi AK, Amouzou A, Joos O, Munos M, Black RE. PLOS ONE. 2016. doi: doi:10.1371/journal.pone.0168743

A social autopsy of neonatal mortality suggests needed improvements in maternal and neonatal interventions in Balaka and Salima districts of Malawi. Koffi AK, Mleme T, Nsona H, Banda B, Amouzou A, Kalter HD. JOGH. 2015. doi: 10.7189/jogh.05.010416