Colloquium on Real-Time Biosurveillance For Early Warnings in Sri Lanka

Posted on October 12, 2006  /  1 Comments

Assume a scenario where among the chief complaint strings of two unrelated patients in the same District on the same date there was a mention of bloody stools in pediatric cases. The multiple mentions of “bloody stools” or “pediatric” might not be surprising, but the tying together of these two factors, given matching geographic locations and timings of reporting, is sufficiently rare that seeing only two such cases is of interest. This was precisely the evidence that was the first noticeable signal of the tragic Walkerton, Canada, waterborne bacterial gastroenteritis outbreak caused by contamination of tap water in May 2000. That weak signal was spotted by an astute physician, not by a surveillance system. Reliable automated detection of such signals in multivariate data requires new analytic approaches.

The socioeconomic see the key problem is not software but accurate and timely entry of data by Medical Practitioners. Hence the project will extend the user interfacing to the last mile using information communication technology (ICT) networks that already span the island of Sri Lanka. They would be the GSM and CDMA Wireless Local Loop (WLL) markets that are far beyond the fixed phone market, it is intuitive to introduce WLL applications opposed to traditional personal computer applications in order to increase the early detection and warning of diseases outbreaks in Sri Lanka.

The proposed Biosurveillance project is an extension of the Last-Mile HazInfo project. The intent of this colloquium is to validate the research and objectives of the Biosurveillance project proposal.

The presentation Nuwan is making can be downloaded HERE

Skype participants: Dr Artur Dubrawski (AutonLabs), Dr Gordon Gow (University of Alberta) and Shanmugarajah

Nuwan: 80% of the Haz-Info project completed.
The research question is: Can Biosurveillance Algorithms coupled with Wireless Local Loop Network Applications increase the early detection and warning of Communicable diseases in Sri Lanka?

Biosurveillance as defined by the Ministry of Health. WLL includes GSM and CDMA Wireless Local Loop (WLL).

To be more specific, we need to look at reliability of WLL ICTs in communicating health information, reliability of Biosurveillance Algorithms, the contribution of organizational level, gender specific responses and the degree of integration of ICTs.

DG: Is this proposal geared towards IDRC and if it is going to be for the Gates foundation?

Nuwan: That is undecided so far.

The surveillance and alerting system was described, step by step as seen in Slide 7 of the presentation.

DrDubrawski: Need for feedback from MoH at 9.

RS: It wont make sense to get feedback after 9. Step 7 would be the feedback point.

Nuwan: Message verficiation and feedback takes place at 7 anyways.

Slide 9: The data algorithms to be used will be What’s Strange About Recent Events (WSARE 3.0), SpatialScan and TIPMON.

For WLL data exchange, we will be making use of GSM mobile phones and CDMA fixed phones.

Examples of patterns that can be detected using this system: diarrhea cases among children, respiratory syndrome cases among females, Botulinic syndrome cases and number of wrist injuries in a Base Hospital, etc.

Checking with archived records (last year’s record) and recent records (yesterday’s patient records) with current records (today’s information) is a costly affair.

The WSARE software obtains recent and baseline datasets, searches for rules, determines the significance level of best scoring rule, and reports all rules that are highly significant.

The SpatialScan statistics looks for over densities – the regions where counts are significantly higher than expected, given the underlying population.

Dr Dubrawski: The SpatialScan system collects data from pharamcists on drug purchases on a daily basis. This can be indicative of any impending outbreaks of disease.

The NoisyCopy model was built for the USDA for screening possible food complaints. Basically, the consumers submit complaints through phone interface and the system scans the complaints for similarities. If there are similarities, this may indicate that the product problem has generated by the same underlying process.

Nuwan explained the research matrix as described in slide 23.

Slide 24 – Project Task Matrix
EPID, Sarvodaya, Micro-image, AutonLab, Vanguard, and LIRNEasia will be partners in this project. Tasks have been allocated for each partner.

Slide 25 – Next steps
Proposal to be completed by November 2006.

We hope to seek $200,000 from the Gates Foundation or IDRC.

Target kick-off date March 2007.

1 Comment

  1. ‘Mobiles ‘to help track diseases’

    Mobile phone technology is being developed to help manage the spread of diseases such as HIV and bird flu.

    for dull strory —