Effect of misclassification of causes of death in verbal autopsy: can it be adjusted?

Background Verbal autopsy (VA) is an indirect method of ascertaining cause of death from information about symptoms and signs obtained from bereaved r elatives. This method has been used in several settings to assess cause-spe cific mortality. However, cause-specific mortality estimates obtained by VA are susceptible to bias due to misclassification of causes of death. One w ay of overcoming this limitation of VA is to adjust the crude VA estimate o f cause-specific mortality fractions (CSMF) using the sensitivity and speci ficity of the VA tool. This paper explores the application of sensitivity a nd specificity of VA data obtained from a hospital-based validation study f or adjusting the effect of misclassification error in VA data obtained from a demographic surveillance system. Method Data from a multi-centre validation study of 796 adult VA, conducted in Tanzania, Ethiopia and Ghana, were used to explore the effect of distri bution of causes of death in the validation study population and the patter n of misclassification on the sensitivity and specificity of VA. VA estimat es of CSMF for six causes (acute febrile illness, diarrhoeal diseases, TB/A IDS, cardiovascular disorders, direct maternal causes and injures) were obt ained from a demographic surveillance system in Morogoro Rural District in Tanzania. These were adjusted for misclassification error by using sensitiv ity and specificity values of VA obtained from the validation study in a mo del proposed for correcting the effect of misclassification error in morbid ity prevalence surveys. Results Sensitivity and specificity of VA differed between the three valida tion study sites depending on the distribution of causes of death. These di fferences were explained by variations in the level and pattern of misclass ification between sites. When these estimates of sensitivity and specificit y were applied to data from the demographic surveillance system with a comp arable structure of causes of death the difference between crude and adjust ed VA estimates of CSMF ranged from 3 to 83%. Conclusion Estimates of sensitivity and specificity obtained from hospital- based validation studies must be used cautiously as a de facto, 'gold stand ard' for adjusting the misclassification error in CSMF derived from VA. It is not possible to use sensitivity and specificity estimates derived from a location-specific validation study to adjust for misclassification in VA d ata from populations with substantially different patterns of cause-specifi c mortality.