Shahid Akhter, editor, ETHealthworld, spoke to Dr. Vikram Venkateswaran, Co-chair of the healthcare working group at the IET Future Tech Panel, to figure out the role of technology in disease surveillance that can culminate in identifying the spread of diseases much in advance.
Disease Surveillance: Trends
Disease surveillance is probably the most important aspect that is evolving in the global health-care industry. We look at data today that has proliferated across mediums, especially digital mediums, and there is a huge opportunity to use that data to not only prevent diseases but also predict the next outbreak. Such data has been used sporadically, like in combating ebola or HIV in certain cases, but the pandemic has put a new twist on this entire scenario. Today we have the opportunity of not only collecting the data at source but also have a high level of compute that is available across technologies like cloud, that can be leveraged to predict the next epidemic. Though we caught up late during the COVID pandemic, the lessons globally learned have been that we can accelerate this development and use it very effectively to identify and combat the next pandemic.
Disease Surveillance Programme: Idea & Inspiration
The disease surveillance programme that we are running as a pilot as part of our IET Healthcare working group was inspired by similar attempts made by researchers globally to understand the spread of diseases. A simple case of using infectious diseases, especially during the monsoon season in India, was a good opportunity to not only see how far we have progressed in our adoption of technology but also see if this could be used in a real-world scenario like predicting the next epidemic or the next disease spread. The idea was to capture the information at source, triage it, pinpoint exact locations and see if we can triangulate as to where these mentions and data sources were coming and could the local authorities be alerted to the very spread of the infection.
Social Analytics for Rapid Transformation of Health for India (SARTHI) project
We are calling this pilot “SARTHI,” which is “Social Analytics for Rapid Transformation in Health for India.” What we have done is we have looked at publicly available data sources, this includes consumer forums, blogs, news websites, as well as social media data sources like twitter and facebook wherever an API was available, we have plugged it into this system. Then we have written queries to suppress all the false positives, called junk information, to get to the real key information that provides insights. This is then rendered on a front-end application called Flutter, which is then visible to the user as dashboards, which not only show you locations or where the diseases are spreading but also give you the numbers of where exactly and how the trend has been over the last six months.
Disease Surveillance Programme: Challenges
Some of the opportunities or challenges that, as you may say, we came across on the road were: one was, did we have informed consent from patients? Now, consent is a big part of privacy, and as you can see, the new draft data privacy bill also talks a lot about consent. As IET works under GDPR, it was very important to have informed consent from patients to use their data. We overcame that by using publicly available data sources rather than data sources where patient consent was required. The second challenge we came across was the volume of data. So the volume was high, and a lot of it was not relevant data; it was misleading data, false positives, or noise that we call in the system, and we overcame that by writing specific extraction queries, where we identified the false positives and wrote queries to reduce them. The last aspect was on languages. India doesn’t use just one language; we use a variety of languages, so we went with the ten most spoken languages in India.
Relating data to predictive analytics
What we have done now is take a few use cases and correlate them with local vector-borne information data released by the various state governments. We’ve done it in three cases: Delhi, Uttar Pradesh, and Karnataka. But what we’re going to do next is use our association with organisations like CHRI, who are also participating in this project. To do analysis, correlate the data that we have on the digital platform to what they’re seeing in the local primary health centers and that way we want to look at the correlation and see what is the percentage of that correlation. That ground-level validation is the most important step.
The next step above what we’re doing is that we’re going to start exploring avenues to do predictive analytics. To see can I predict trends, can I predict patterns, can I give you an analysis that if you come to Bangalore in November of 2023 should you be carrying your bottle of water or should you be carrying a mosquito repellent, because that is the infection that you are going to face. So that is going to be our next step, and the last step eventually will obviously be to add more data sources. Now one of the most interesting data sources available in India, which is not explored much, is weather data. Most of these infectious diseases are very weather-based. For example, Dengue only happens with fresh water, and malaria happens with stagnant water. So monsoons play a big role in the spread of these infectious diseases.
Disease Surveillance Programme: NCDs and Mental Health
The other challenge we’re facing in India, if you move away from infectious diseases like TB, dengue, and malaria, is non-communicable diseases. Today we do see a big spike in cases of hypertension and diabetes, which are not diseases by themselves but are conditions that lead to the propagation of some other diseases and conditions.
I think in the future we’ll have to look at non-communicable diseases. The last aspect is mental health. I think the COVID pandemic has exposed us to the gaps we have in our mental health framework, and today we have programs, both nationally and regionally run by states, that look at things like suicide prevention and depression. But I think, like the Honorable Former Minister, Suresh Prabhu, said that these are conversations that give you the symptoms, and while it’ll take a while for symptoms to register in a hospital. This gives us an opportunity to intervene in this case right before the person actually becomes a patient. So, we’re looking at the health consumers and providing them support so that they do not have to even enter the hospital in a certain condition. So there is huge scope for this, and the impact both on society as well as the healthcare community can be immense. And what we need is collaboration—building the ecosystem together like we have done in this case.