Validity of verbal autopsy in determining causes of adult deaths

A verbal autopsy (VA) questionnaire, which had an open-ended description of the deceased's illness followed by a series of close-ended questions on specific symptoms and signs, was used by three trained field workers to interview relatives of the deceased who had died in a Chandigarh hospital in previous year. The sensitivity and specificity of the cause of death assigned by physician from the verbal autopsy was calculated against the cause of death derived from the hospital records. Of the 262 verbal autopsies, 60% were males and 23% belonged to rural area. Specificity of VA cause of death was high (>95%) for all broad cause groups except cardiovascular (79%) diseases. Sensitivity was highest for injuries (85%) and it was in the range of 60% to 65% for circulatory diseases, neoplasms, and infectious diseases. Sensitivity was low (20% to 40%) for respiratory, digestive and endocrine diseases. Cause specific proportionate mortality for major cause groups by the VA were statistically similar to the causes of deaths derived from the hospital records (p>0.05). Therefore, VA can be used for determining causes of adult deaths.