by Soumik Ghosh

Technology that helps safeguard our little ones

Jul 19, 2016
AnalyticsApplication SecurityAuthentication

India sees over two million kids die every year, and a million more disabled for life. Reason: Kids not being vaccinated in time. Project Khushi steps in to the rescue by integrating mobile health, NFC technology, and cloud computing.

An Immunize India report reveals that the neonatal mortality rate (death within the first month of the child’s birth) in India is 27.7 per 1000 live births. The global average stands at 19. In addition to our disconcerting stats, we have to battle with the fact that the current data collection methods are outdated, cumbersome, and lack specific patient information.

Also, many mothers in our country do not understand the importance of getting their child vaccinated on time, and chose to skip immunization clinics. NGOs are unable to judge the efficacy of vaccinations as they aren’t able to track real-time, actionable data.

A silver lining

A non-profit organization, Khushi Project, is harnessing technology to give these kids a shot at a healthy life. The project has integrated wearable NFC technology, mobility, and cloud computing under one roof to fix the vaccination problem.

The project bagged the UNICEF ‘Wearables for Good’ challenge, following which, UNICEF is considering using the technology to benefit a larger base beyond India.  

A tête-à-tête with Ruchit Nagar, Director at Khushi Project, reveals how the fledgling organization has tied up with a Rajasthan-based NGO, Seva Mandir to roll out the tracking system across 96 camps in India.

So far, the system has tracked about 16,000 vaccination events for 1500 children in Udaipur, Rajasthan. Project Khushi is currently working in 60 villages with an NGO called Seva Mandir. This technology will be rolled out in 400 more villages, and the data collected will be compared directly with the government status quo mechanism.

The tech that powers the project

The tracking system involves a rewritable NFC pendant, which uses the N-Tag 216 chip that can be worn around the neck. The chip has re-writable memory, and this makes it a valuable record in the size of a paper stamp.

What takes the cake, though, is the fact that this technology is unpowered. The device draws power from a phone or a tablet, using near frequency to temporarily activate the chip by induction. These chips are password protected on read and write. So, you can program it such that only your app can edit/read this chip after the first time you write on it.

What the project is trying to achieve is to straighten out the manner in which data is stored. And that’s where the Mamta card steps in (maternal and child healthcare card for pregnant women through the first year of the child’s life and immunization records).

The challenge Khushi faced was in storage – reducing the number of bytes they were taking up (government records contain 120 columns of data), and this compelled the team to think out of the box. The team needed to record the dates that immunizations were given on. Now, instead of recording every single date for every single vaccine, they have an offset date from a reference date, which is usually the date of birth, or the date of registration.

“So, these are the ways we can be parsimonious with the space that we have, and store all that data in the RCH register. We’ve different curtifications for each different field, and we use them to represent the data more efficiently,” explains Nagar.

Cloud to the rescue

The first intent was to build an offline-ready system. This ensures that health workers can access patient records without internet connectivity. When the health worker returns to the office, or where there is internet connectivity, they have the ability to then upload the data to the cloud server, and this is done using the Heroku sequel database.

The data is stored in three places – starting with the highest level of severity. First is the necklace, which is the patient’s copy of the record. So, that is considered to be the most recent updated information. The second level is what is stored in the mobile device/tablet, on its local memory. And finally, the data is also stored on the cloud. Most other mobile health-type applications have the latter two – on the local device, and the cloud.

“The issue that we’re trying to work around is to not simply add another level of redundancy, but we need a tablet to access health records. Or in cases where the tablet has been wiped, or the data has been lost, or the coding that used to work is no longer on the tablet,” says Nagar.

But, as long as they have the app, and as long as they are in possession of that pendant, they can still retrieve that data and act upon it. And the paradigm here is that it’s truly decentralized, instead of relying on pulling data from the cloud, or look up identifiers in a centralized location, or backed-up data on a tablet.

The Analytics Angle

There’s no denying the fact that Big Data is all-pervasive right now. The database actually has over one million records of children and new pregnant women, and more are added to the system each year.

The project uses PCTS (Pregnant Women & Child Health Tracking System). The updation of this centralized system starts with the nurse recording the data on a register. She then fills out monthly forms. This then gets passed on to a data-entry worker, who enters the data into an access database, which then gets uploaded to the centralized place.

Now, once these records were loaded into the database, the team found that many of these records were incomplete, the fields not validated, and there were many gaps. This presented a huge opportunity for big data analysis to look for predictors of stillbirth, or critical health outcomes from the vast amounts of data coming in.

What the project is aiming for is to make a robust, comprehensive data collection system that will feed in this data, so hand-offs can be eliminated. The data can be directly synced from the app to the central server.

So, the app can fill certain fields, or explain why it was unable to. For example, there are certain mobile health apps that record what treatment is given to somebody. But very few of them actually go into the detail of saying ‘I couldn’t give the vaccine today and here’s why.’ Or details like the child could’ve been sick, or there might have been a shortage of vaccine vials, or maybe wastage.

Sounds swell, but how secure is the data?

Security starts right at the field. “You need to have an authorized user working on the apps, otherwise the data could be manipulated. The tag itself has to be protected. The connection over which you’re sending the data from the tablet to the cloud needs to be secure, and finally the cloud itself needs to be secure,” says Nagar.

From the database in the cloud, the team sends voice call reminders to mothers. So, the system is actually being used for both data collection, and for community engagement. These reminder calls to the mother ensures that she knows when to come in for the next checkup or when the next vaccination is due.

Now, each one of these tests need to be protected, and the team is building in encryption at each level. For its next deployment, the team is also integrating biometric authentication.

“What we’re working on is a closed-circuit biometric authentication system, where you can store the digital template of the biometric, either thumbprint, or any other info on the tab – it doesn’t take up much memory,” says Nagar.

The project uses tablets that are equipped with inbuilt biometric sensors. Part of the system design needs to be fairly platform agnostic, but in cases where biometrics is available, the team has the ability to use that as well.

It’s imperative to ensure that toddlers in far-flung corners of our country are vaccinated in time, and solutions like this give them a leg up by providing last-mile connectivity.