RESUMEN
OBJECTIVES: To describe utilisation of verbal autopsy as one of the data collection approaches in cancer registration in an Indian setting. We aimed to estimate the proportion and epidemiological characteristics of malignancies identified by the Varanasi population-based cancer registry (PBCR) using verbal autopsy between 2017 and 2019 and to develop a thematic network for implementing verbal autopsy. METHODS: This was a cross-sectional mixed-methods study. Quantitative methods were applied to analyse information from PBCR proforma of the verbal autopsy-confirmed cancers; qualitative methods were applied to evaluate verbal autopsy conducted by field staff from key informants. In-depth interviews of field staff for the challenges and potential solutions during verbal autopsy were assessed. RESULTS: Of 6466 registered cancers, 1103 (17.1%) were verbal autopsy-confirmed cancers, which had no other source of information. The majority of verbal autopsy cases were from vulnerable populations who were aged >50 years (721, 65.4%), female (607, 55.1%), from rural backgrounds (853, 77.3%), illiterate or just able to read and write (636, 57.7%), and from lower and middle-income groups (823, 74.6%). Verbal autopsy helped provide information about symptoms and site of disease, diagnostic and treatment details, and disease status. Major challenges during verbal autopsy described by field staff were incomplete cancer treatment, destruction of medical records and non-cooperation by the community and lack of support from the local workforce as cancer is not notifiable. CONCLUSION: Verbal autopsy helped identify cancers that would have been missed during active case finding from available resources. The majority of verbal autopsy-confirmed patients belonged to vulnerable populations. Non-cooperation from community and local health systems was major challenge during verbal autopsy. Developing robust cancer awareness, patient navigation, and social support programmes will strengthen verbal autopsy. Integration of standardised and reproducible methods of verbal autopsy in cancer registry and digitalization of health information, especially in limited-resource settings with weak vital registration, will facilitate completeness in cancer registration.
