Validation of causes of infant death in the community by verbal autopsy

Abstract

Validation of findings of verbal autopsy to assess the cause(s) of death during infancy was tested by comparing the information gathered by lay reporters with that generated by professionally trained personnel. This information was obtained on 280 infant deaths and 107 stillbirths reported from 40 villages in two community development blocks where a reliable information system has been established. During neonatal period, in 156 deaths, 222 underlying or associated causes were identified with 75.7% agreement. In the post neonatal period, 266 causes were recorded in 124 deaths with 83.1% agreement. In 97 of 107 stillbirths reported, agreement was found. Neonatal tetanus and measles (fever with rash) as cause of death matched fully; poor agreement was found in conditions like refusal of feeds; jaundice; septicemia and vomiting. There was full matching of agc at death between 8 days and 365 days and 86.7% agreement below 8 days. Lay reporters can be used for determining cause(s) of deaths during infancy by a recall of sequence of events preceding death and recording easily recognizable signs and symptoms. There is scope for training of lay reporters to obtain even greater correlation between their finding and those of professionals.