2016 ICT4D Conference

Even with the marked expansion of ICTs in developing contexts, many rural and low-income communities worldwide continue to rely on paper-based systems for management of health goods and services. While paper is familiar, reliable, and inexpensive, these paper-based systems require extensive time and human resources for collection, aggregation, and reporting of essential health data. The ODK Scan application was developed as a bridge between these paper and digital systems, utilizing custom image recognition technology and a smartphone's built-in camera to automatically extract and digitize handwritten data from paper forms. Part of the Open Data Kit set of tools, ODK Scan is developed by VillageReach and the University of Washington specifically for use in low-resource communities and requires no internet connection. In field testing with community health workers in rural Malawi focused on maternal and neonatal health, the ODK Scan alpha technology showed accuracy rates of 83% for numbers and 96% for fill-in bubbles. Further improvements in data processing are currently being tested in a Pakistan-based field trial focused on monitoring of tuberculosis patients. (Preliminary data from that trial will be available at the time of the conference.) At its core, ODK Scan aims to create an efficient paper-to-digital process that can improve data visibility from the last mile and improve decision making at all levels.