Predictive analytics has the potential to revolutionise Indian healthcare.

In 2015, Hollywood actress Angelina Jolie created a stir in the medical industry when she went public with her double-mastectomy (breast removal). She was carrying a mutated gene BRCA1 which increases the chances of breast and ovarian cancer, a condition that killed her mother some years ago and also impacted her grandmother. It is believed that this condition impacts the Ashkenazi Jewish Women, where 1 in 40 has this mutation and have a higher precedence of both breast and ovarian cancer. While going public with her procedure is a boost in the arm for many women who felt empowered to take this decision, it was a giant leap for predictive analytics in healthcare.

Predictive analytics is a form of advanced analytics that helps predict the future state based on key parameters- historical data, statistical modelling and machine learning. It has been used of late in areas like determining the right candidates for drug trials. One of the biggest reasons why the COVID-19 vaccine came up within a year of the pandemic was because of the advanced predictive analytics capability that was used by most vaccine manufacturers in identifying and then refining the perfect candidates. This branch of data science has huge implications in healthcare and in the rest of the article we will discuss some of the implications and its uses.

The Indian healthcare system is already benefitting immensely from predictive analytics. 

Most major hospitals have been studying the clinical data and making decisions based on the diagnosis and prognosis of patients. As health outcomes are determined by multiple factors including genetics, family history, nutrition, physical activity and mental health, hospitals have started using predictive analytics to understand which treatment modality will work best. 

Some of the key parameters being tracked include 

• Length of stay – Using predictive analytics to reduce the length of stay and faster recovery
• Re-admissions- Looking at patient data on readmissions and identifying patients attributes that require readmission 
• Emergency admissions
• Workflow at departments like radiology and pharmacy 

All this effort has led to improvement in 

• Patient Engagement 
• Clinical Operations 

Another major hospital headquartered in Bangalore has been using a predictive analytics-based AI system to help their oncologists with diagnosis. In a recent study conducted by the hospital involving 600-plus breast cancer cases, it was observed that concordance between the recommendations provided by the AI system and that by the Hospital’s oncologists was as high as 90%.

Apart from hospitals, we also have other parts of the healthcare ecosystem that are investing in predictive analytics. Medical Devices are looking at ways to improve the accuracy of diagnosis while reducing the scan time. They use predictive models to reduce the MRI scan time for example by recreating the anatomy being scanned with predictive software.

The pharmaceutical industry has invested close to 1 Billion USD globally in 2020-2022 in AI and predictive models, in order to reduce time for clinical trials and increase the chances of finding the right candidates. Today a lot of the pre-clinical research work as well post marketing adverse event reporting is managed on predictive analytics systems. 

The biggest boost to predictive analytics came from the pandemic. In India leveraging the Arogya Setu and the CoWIN application the government was able to predict the admissions and the requirement of key resources like hospital beds and vaccines across the country. This data was further supplemented by the local health committees set up for this process. The coordination at national level was helped by predictive models created by institutions like IIT Kanpur as well as National Health Mission. 

While a lot is being done in this area, we need some strong structural changes to make sure predictive analytics is able to scale across the healthcare ecosystem in India. Let us look at some key parameters

• Data- Historical data is the key to the success of the healthcare predictive analytics model. Today we have a lot of it in paper records and even the digital data that is being created is not for use by predictive analytics systems but rather only kept for compliance and ease of storage purposes. Here, a lot is expected from the India Stack and the data standards from the National Digital Health Mission. Once this stack is adopted across all healthcare organisations in India, standardisation of the data will lead to great advancements in care systems. This will also help in extraction of the data from devices, which incidentally store almost 70% of the healthcare data in Indian healthcare ecosystem.

• Standards- While data and standards go hand in hand, it is very important to ensure that there are standards that are established for interoperability across the ecosystem. Again scientists in the Indian bureau of standards are working on standards across the healthcare ecosystem. I am working on a subgroup that is defining the standards for IOT devices measuring stress. This will cover a lot of the user generated health data that comes out of devices like smart watches and out of systems like Ultra Human.

• Talent- No matter how many advances we make in technology there is always a shortage of good talent that understand the ecosystem and are able to apply predictive analytics to solve some of the healthcare sector issues. Today, we severely lack techno-functional leaders, who sit at the convergence of technology and healthcare. We seriously need many medical information leaders who have served patients as well as have a good understanding of how technology can bring a positive impact to them.
• Process Optimisation- Most of our clinical processes and protocols are built around humans and not machines. We will have to revisit many of them to see how these processes are still necessary when predictive analytics starts helping the healthcare system to make decisions.

Predictive analytics is very important for India. As a country with limited resources we need to understand where we need to use them. While we have been reactive in many cases this new thinking can help us change our healthcare posture to be proactive. In conclusion, there are many implications of predictive analytics in healthcare, but in the end the biggest beneficiary is going to be the patient, who can now enjoy a longer life span with a higher quality of life.

 

---

At the Future Tech Congress, we will be discussing AI, Blockchain, Digital Twin and how these future technologies can help businesses gather customer insight and boost growth, in the areas of Manufacturing, Healthcare, Fintech and Supply Chain. Explore our full agenda here.

HEALTHCARE