Wednesday, August 29, 2012

Texting to Save Lives in South Asia (repost)

Texting to save lives in South Asia

​by Marc Ellison

Watch and listen to the audio slideshow

Imagine a world where mobile technology helps to save lives by quickly alerting health authorities of new outbreaks of H1N1, chickenpox, and malaria. Such an early warning system was recently piloted, by a technological think tank LIRNEasia, non-governmental organization Sarvodaya, both from Sri Lanka, and the Rural Technology and Business Incubator in India.
Supported by IDRC, the Real-Time Bio-Surveillance Program tested a modern alternative to the paper-based process introduced by the British over a century ago. Using the current system, local data on infectious diseases takes three to four weeks to make its way to national epidemiology centres, at which point an escalation of common symptoms can be detected. By then, a disease may have caused much harm. In 2003, for example, Sri Lanka’s Central Province faced a fever-like disease that went unnoticed until it claimed three lives.
A LIRNEasia survey found that many healthcare workers only learned of outbreaks through the media, by word of mouth, or from peers.
“We need to be able to view cases in real-time to detect outbreaks swiftly. Otherwise it takes several days before the hospitals send the notification paper forms. By that time the patient may be dead or discharged,” says one public health inspector.

Streamlined disease surveillance
To help make real-time disease detection a reality, LIRNEasia tested a system using mobile phones in an IDRC-supported pilot study.
“Mobile phones are the most affordable technology, with the widest reach in India and Sri Lanka. Data is submitted instantaneously, compared with the current ‘snail mail’ system. It’s also much cheaper. It costs a small fraction of a cent to send data, whereas sending forms via regular mail costs 5 cents,” explains Nuwan Waidyanatha, a LIRNEasia project director.

The concept is simple, but the technology behind it is cutting edge.

Community healthcare workers record a patient’s diagnosis using software installed on a mobile phone. They then submit the data directly to national epidemiology centres in Colombo, Sri Lanka and Chennai, India. A data-mining software — developed at Carnegie Mellon University, United States —analyzes this data on a daily basis, allowing epidemiologists to visualize potential epidemics by using mapping tools.

The specialists can then use the system to return messages to health inspectors, alerting them to potential dangers. These messages can then be translated into the local dialect, relayed to the communities, and placed on bulletin boards in village centres.
The research team tested the system in 28 facilities in the southern state of Tamil Nadu in India, and 12 hospitals in the North Western province in Sri Lanka.

A better way to spot outbreaks

The bio-surveillance system has already proved its value. During the 15-months testing phase, the system identified more than a dozen instances of potential disease outbreaks. Four of those (chicken pox, acute diarrheal disease, respiratory tract infection, and mumps) were confirmed by health authorities.

Referring to the chickenpox outbreak in Kurunegala district, Sri Lanka, Waidyanatha says, “The platform was able to detect this outbreak much faster than the paper-based system. The divisional Medical Officer of Health found out about it the next day.”

LIRNEasia found that the new system could reduce operational and archiving costs by 30-50%. For example, public health inspectors would no longer need to travel to town once a week to compile their data; thus eliminating the need for travel subsidies. By limiting potential outbreaks, the new platform could lessen the financial strain on the Indian and Sri Lankan health systems.
Health officials involved with the project indicated that it could be a useful tool to support long term planning and allocation of health resources. The new system can even be used to identify everyday issues in local communities. For example, it “identified that men in Tamil Nadu were complaining of pain during harvest season,” Waidyanatha says. “This highlighted how farmers needed better tools.”

Challenges ahead

LIRNEasia has identified challenges still facing the new system in India and Sri Lanka.

The new process may eliminate the risk of manual-system errors from clerks deciphering the handwriting on paper forms or manually copying patient information into logs. But new problems arose. Submission error rates ranged from 23 to 45% — mostly from different spellings of medical terms such as tuberculosis or the misuse of synonyms like dementia and memory loss. These errors affected the system’s complex statistical analysis, resulting in false predictions.

Although the healthcare workers in this first pilot easily learned how to enter the patient data, some of them saw the new system as a bureaucratic hindrance, while others feared the system would take away their jobs.

A sustainable solution?

Waidyanatha calls the new system a “usable solution,” but one that requires further enhancement.
Subsequent testing in North Western province, by the Sri Lankan Ministry of Health, identified more areas for improvement. The infectious disease control nurses involved in the trial found it difficult to enter data with the mobile keypad, leading researchers to conclude that more sophisticated phones, or tablets, with touch screens or the capability of reading handwriting may be better.

It is clear that technological will and capacity are important factors – but that they need to be examined in the broader context of political will and social acceptance by the different users.
The project has raised awareness of the important gains that technology could bring to public health surveillance in developing countries like India and Sri Lanka.  More research is needed to find the right tools for busy healthcare workers.