Dear Friends, Family, and Supporters,

As the summer of 2013 approaches, the Village Zero Project is one year past our first phase of data collection in Matlab, Bangladesh. In one year, we have moved from being data analysists to being informed actors. The data  collected last year has become compelling visualizations of the progression of cholera in space and time. The visualizations show that the number of cholera cases show clear trends both geospatially and temporally.

Over the course of the last few months, we focused on creating new visualizations to help us understand the areas most affected during each month with greater statistical precision. Our next step was to investigate why we were observing the trends that we observed. By doing extensive research on the causes of endemic cholera, we learned a great deal about the strong link between climate and cholera. We collected climate data, and used this data to understand which factors are strong predictors of disease occurrence.

We are in the process of publishing our findings in a peer-reviewed journal. In addition to writing up our findings, we have also given two presentations of our findings at the Water: Systems, Science and Society Sympossium, and the Education for Public Inquiry and International Citizenship Symposium: Global Health and Security.

Apart from our findings, we also presented briefly about another project currently in progress: an interactive applet that enables data discovery. We are currently in the process of designing the interactive system that would allow the user to find connections between climactic data and disease incidence data. This applet will be posted on our website in the coming months as it is developed.

With so much that has already been achieved in a year in the way of data analysis, we look towards a future of informed action. Now that we have conclusive evidence for the geospatial and temporal trends in endemic cholera, the next step is to increase the scope of Village Zero. We plan on using mHealth (mobile health) to collect realtime data about disease incidence to help create a prediction system for cholera based on all the data analysis we have done this year. This prediction system will eventually allow resources to be diverted to the right place at the right time.

We have gotten through the data analysis phase of the Village Zero project with the support of our donors. The next phase of our project involves implementing an mHealth data collection system, and creating a framework for resource allocation. As we now move into the action phase, we need more support then ever.

Thank you for your support, and help us transform data analysis into informed action!

-The Village Zero Project

So exciting! A friend of ours is currently helping us develop the shapefile that we need in order to produce our visualizations. To create our shapefile, we’re converting a PDF map of Matlab, Bangladesh into a vector map that contains geographic information regarding our location of interest. For the time being, here is a snapshot of our PDF map:

Courtesy of the ICDDR,B

As you can see, the map is broken down into many small sections. These sections each represent individual villages. The cholera data that we have is georeferenced using this map. Each entry is associated with one of these individual villages; moreover, each entry corresponds with a specific date. In this sense, our data consist of two parts: spatial and temporal. Once the shapefile is finished, we plan to plot our cholera data in space and time. Using a very cool software called StatPlanet, we’ll essentially be able to how how cholera moves from one village to the next from one date to the next!
We’re hoping to have the shapefile complete by the end of the month, at which point I’ll post a snapshot of it to give you guys an idea of what the “base” of our visualizations will look like.

As soon as we hear back from the AGU, we’ll be able to publicly post the entire abstract, as well as the math behind our statistics! Check back in early October for the overview. But for now, here’s a little excerpt:

Using 10 years of geographically referenced cholera data, cases were plotted in time and space. Preliminary patterns suggest that villages closer to the Meghna River experience the majority of the area’s cholera outbreaks and that case count is highest in late spring and late fall. April/May and November/December represent 25% and 23% of total annual case counts respectively. Moreover, villages further from the coastline demonstrate 57% higher relative prevalence in fall than in the spring.

Looks promising, don’t you think? I think we have a pretty sound hypothesis brewing… Just in time to start mapping! In the coming weeks, David and I will be posting about our mapping process for those of you at home that are curious about geographic information systems (GIS) and the wonders of spatial (and temporal) epidemiology. We hope to have a prototype of our interactive map, using the data we have presently, in the fall, too.

Signing Off,
M. Majumder

David is still in South Asia, working in Nepal with Nyaya Health on another mHealth-related project, while Kate and I begin data analysis here at home. Thanks to help from our friends at the ICDDRB, we have 10 years of Matlab’s cholera hospital data at our disposal that has been geographically-coded! We will begin mapping the data in space and time over the next several months. We hope that this preliminary visualization will be helpful in obtaining the large research grants we need to begin our real-time data collection in and around Matlab next spring. Moreover, if the visualization produces results that are worth publication (which we think it will!), we plan to submit an abstract to the American Geophysical Union’s Fall Conference!

Initial analysis suggests some very interesting geographic progressions of cholera over the course of different seasonal transmission cycles, and we can’t wait to see what our preliminary visualization will yield. Expect it to be up (and interactive, so our users can see how we plotted the data retrospectively) by early Fall 2012! We hope it’ll be indicative of the maps we’ll be producing and updating once we begin real-time data collection, and we look forward to keeping all of you posted as we progress.

Signing Off,
M. Majumder