Accurate and credible cause-of-death (COD) data are critical to understand, interpret, and address the burden of diseases and tailor public health policymaking at subnational, national, and regional levels. A complete diagnostic autopsy (CDA) is the gold-standard procedure for determining COD. When full autopsy is not affordable or feasible, medical certification of COD (MCCOD) is often conducted using all medical information relevant to the terminal illness. In low- and middle-income countries (LMICs), CDAs are very rarely conducted due to cultural, religious, and infrastructural constraints, whereas MCCOD has suboptimal coverage that is usually limited to deaths that occur in health facilities and is of variable quality.1 For settings without the capacity to conduct CDA or where they are infrequently done, a nonclinical approach called “verbal autopsy” (VA) is commonly used. VA is a systematic postmortem interview of the relatives of the deceased on the health history, signs, and symptoms of the fatal illness that can potentially identify the COD. Although the reliability of the VA at the individual level is questionable,2 it is often the only feasible option and has become a key source of COD data in LMICs that do not have fully developed civil registration and vital statistics systems with MCCOD information.3 In addition, VA-based results are often useful for studying population-wide trends of cause of death.
