JMIR Publications

ABSTRACT

Background:

Verbal Autopsy (VA) has been a crucial tool in ascertaining population-level cause of death (COD) estimates specifically in countries where medical certification of cause of death is relatively low. World Health Organization (WHO) has released an updated instrument, i.e. Verbal Autopsy Instrument 2022 that supports electronic data collection methods along with analytical software for assigning cause of death. This questionnaire encompasses the primary signs and symptoms associated with prevalent diseases across all age demographics. Traditional methods have primarily involved paper-based questionnaires and physician-coded approaches for cause of death assignment which is time-consuming and resource-intensive. Over the years, computer-coded algorithms have evolved to assign a cause of death based on verbal autopsy, however, data collection in densely populated countries like India has always been a challenging task.

Objective:

In this article, we present the development of an Android application specifically tailored for streamlining VA data collection by utilizing the existing public health system workforce. The application has been designed to integrate real-time data collection by frontline health workers, information transmission, and reporting on the cause of death by physicians in a seamless digital manner aimed to enhance the efficiency and accuracy of cause of death assignment through VA.

Methods:

The application was developed on Android Studio, also the primary Integrated Development Environment (IDE) for developing Android apps using JAVA. The Frontend interface was developed using XML whereas SQLite and MySQL streamlined the complete data storage on the local and the server database respectively. The communication between the app and the server was done through the PHP API to sync the data from the local database to the server database. The complete prototype was built focused on reducing the manual intervention and automating the VA data collection.

Results:

The developed mobile application was critically reviewed by the subject experts and modified until their final approval. Experts’ recommendations included content and information pathways such as details of the deceased data, health workers, and their demarcated health centers and the information pathways for frontline workers and Medical officers. The app also includes both English and Hindi language to ease the language barrier. These recommendations were implemented to make the app user-friendly and also enhanced the automation of the verbal autopsy process.

Conclusions:

The development of a unified mobile app could streamline the VA process generating accurate national and sub-national cause of death estimates. This development can act as a baseline and can be further extended to different regions to integrate the automated VA model within the existing public health system for generating mortality statistics in India.