A project that used sophisticated analytical tools on data sent on mobile phones by village-level nurses has provided a promising model of harnessing mobile technology for disease reporting, surveillance and health strategy planning in Tamil Nadu.
The “Real-Time Bio-surveillance Programme” piloted in Thirupathur block in Sivaganga district demonstrated that affordable mobile phones could be an effective, accessible and cheaper technology option for advancing public health in rural scenarios.
“Seeing the success, we propose to scale up to cover Sivaganga district with a network of 47 PHCs by year-end. That experience would give us the confidence to look at entire Tamil Nadu and perhaps the country as a whole,” Nuwan Waidyanatha, Senior Research Manager, LIRNEasia told The Hindu .
The pilot project was jointly undertaken by LIRNEasia, Sri Lanka-centred not-for-profit think-tank on ICT policies for Asian Governments, the Rural technology and Business Incubator at the IIT-Madras and the Carnegie Mellon University, U.S.
A similar project has also been concurrently undertaken in Sri Lanka with promising results. As part of the pilot, 29 grassroots-level nurses in Thirupathur block were provided mobile phones loaded with custom software for messaging out-patient data in prescribed format to a central server at the Directorate of Health Services. The digitised data is analysed by sophisticated statistical tools to map real-time morbidity scenarios. “The attempt was not merely to computerise data reporting processes but to apply advanced analytics to look at a gamut of statistical relationships between all combinations and permutations of gender, locations, ages, symptoms and signs,” said Artur Dubrawski, Director, Auton Lab at Carnegie Mellon.
The T-Cube Web Interface with algorithms to instantly detect unusual morbidity spikes and stratify alerts was developed at Carnegie Mellon while the IIT-Madras produced the mHealthSurvey mobile application to be downloaded on phones to detect a set of 25 infectious diseases as well as monitor non-communicable diseases.
The application that used a standard reporting format was also tailored to reduce errors. The T-Cube can reduce extremely time-intensive calculations to a matter of minutes and the costs of computing are independent of the volume of data. “What matters is not the amount of data but how many dimensions are used for matching criteria,” said Mr. Dubrawski.
The Sivaganga pilot project used a framework of about 50 dimensions to achieve granularity in the automated analysis of data. The researchers believe that m-health solutions could work better in a day-to-day application environment at the last-mile typical of mobile usage than e-health initiatives that are still hampered by costs and connectivity. Apart from the base handsets that cost about Rs.4,000, the pilot project showed that data from each partner PHC reviewing about 100 patients every day could be transmitted for Rs.200 a month. “The m-health paradigm is not being advocated as a substitute to existing systems but rather is intended to complement the health services,” said Mr. Waidyanatha.
