Validation and application of verbal autopsies in a rural area of South Africa

OBJECTIVE To validate the causes of death determined with a single verbal a utopsy instrument covering all age groups in the Agincourt subdistrict of r ural South Africa. METHODS Verbal autopsies (VAs) were conducted on all deaths recorded during annual demographic and health surveillance over a 3-year period (1992-95) in a population of about 63 000 people. Trained field-workers elicited sign s and symptoms of the terminal illness from a dose caregiver, using a compr ehensive questionnaire written in the local language. Questionnaires were a ssessed blind by three clinicians who assigned a probable cause of death us ing a stepwise consensus process. Validation involved comparison of VA diag noses with hospital reference diagnoses obtained for those who died in a di strict hospital; and calculation of sensitivity, specificity and positive p redictive value (PPV) for children under 5 years, and adults 15 years and o lder. RESULTS A total of 127 hospital diagnoses satisfied the criteria for inclus ion as reference diagnoses. For communicable diseases, sensitivity of VA di agnoses among children was 69%, specificity 96%, and PPV 90%; among adults the values were 89, 93 and 76%. Lower values were found for non-communicabl e diseases: 75, 91 and 86% among children; and 64, 50 and 80% among adults. Most misclassification occurred within the category itself. For deaths due to accidents or violence, sensitivity was 100%, specificity 97%, and PPV 8 0% among children; and 75, 98 and 60% among adults. Since causes of death w ere largely age-specific, few differences in sensitivity, specificity and P PV were found for adults and children. The frequency distribution of causes of death based on VAs closely approximated that of the hospital records us ed for validation. CONCLUSION VA findings need to be validated before they can be applied to d istrict health planning. In Agincourt, a single verbal autopsy instrument p rovided a reasonable estimate of the frequency of causes of death among adu lts and children. Findings can be reliably used to inform local health plan ning and evaluation.