CIO Insider

CIOInsider India Magazine

Separator

Dark Data Is The Holy Grail Of Healthcare

Separator
Srinivas Iyengar, Vice President, EVRY,,

EVRY delivers business-critical solutions to companies and public sector organizations at the national and local government level. The company has 8,800+ employees across Norway, Sweden, Denmark, Finland, India, USA, Great Britain, Ukraine, Latvia and Bulgaria.

‘Dark data’ will soon be the new buzz word in healthcare industry. It is the next biggest challenge that we will all solve! While it is important to understand what Dark data is, we also need to understand how the world is moving towards patient centricity and the concept of hyper personalization in healthcare. This will help us understand the magnitude of the problem that we are trying to solve.

What is Dark data?
Dark data is the data that is digitally present, but not available for the care provider when needed contextually. It could be a simple case wherein a patient had visited a private hospital a few months ago for a check-up and that encounter is not digitally available with his treating physician for clinical assessment. The digital data that is available with the private hospital becomes dark data to the treating physician.

This example is aged! Today, many healthcare thought leaders are talking about hyper personalization of healthcare wherein your lifestyle is linked to better health. Just to give you an example, a doctor once said, if I knew that the patient had visited this specific remote part of the world where there was a Ebola outbreak, clinical assessment would have taken a different route and I probably could have saved patient easily by giving right treatment rather than treating for cholera or malaria.

In the world of hyper personalization, your lifestyle data which is digitized also becomes dark data for your care provider as those are not easily available for care provider to do an informed diagnosis.

What do statistics say?
Industry leaders today are trying to quantify importance of dark data and quite a lot of statistical analysis is being done. I’m listing a few of them to give you a glimpse

1. Post digital transformation initiative in healthcare, both clinical and non-clinical data are

growing by more than 100 percent year on year
2. Less than 5 percent of available digital data is being analyzed today
3. Improved patient outcome and reduced sick days by 20 percent
4. More than USD 200B in healthcare savings by 2030 utilizing dark data

Why are we talking about Dark Data now?
We are now living in a world where we have more digital data than ever before and more than we had imagined earlier, and in an era, that has computing powers to generate insights from exabytes of data generated from various forms of data sources.

In today’s world I can confidently say, much of the digital data is available somewhere in some format which can tell us why a person fell sick. This is the reason why Dark data is becoming important. You would understand it better it you had watched English TV serial “Person of Interest” where the AI based super computer has access to public data and performs the required assessment to proactively save people from dying.

The digital data that is available with the private hospital becomes dark data to the treating physician


The process of digitization of healthcare started couple of decades ago. At that point, patient health record was mostly used for clinical assessment of patient condition. Today we have gone beyond patient health record to person information to clinical analysis. There seem to be no end points to what data can be used for in the healthcare industry. This is what we call as hyper personalization.

If we can build such a system where your care provider can proactively prevent you from falling sick, then we have reached the holy grail of healthcare! For this to occur, care provider must have access to all digital footprints created by a patient in one place. Having such consolidated data will allow clinicians to review all digital data that is available for that person to nail down possible reasons of sickness. Hypothetically, somebody could have fallen sick wearing a nylon shirt! Even if we have access to such digital data, it is humanly impossible to mine the data and list down possible reasons for falling sick.

What are the next steps towards efficient utilization of Dark Data?
1. Centralization of Person/Patient Data is an important step towards efficient utilization of person data for better care. This cannot be done at provider/hospital/municipality level. This needs to be taken up as part of country’s population health strategy.
2. Data Lake Infrastructure needs to be built to store the data along with automated methods for data classification. Data management policy on what data to be stored, archived and discarded should be clearly laid out.
3. Data Security and information protection should be top priority. Since we are moving towards hyper personalized data, any data pilferage can be a national security issue
4. Data Mining and Analytics should be taken up by an ecosystem of organizations including government and private which will aid in better patient care. Hence policy on data access, scrubbing and data sharing should be in place.

EU is leading today in consolidating patient data and moving towards patient centricity and better population health management. Most of the EU countries have clearly defined data strategy today including Norway, Sweden, Denmark and Finland since they are largely public funded and follow single payer concept for healthcare. It will be interesting to see how countries such as USA, India and China will approach their data consolidation strategy as they have a multi payor/provider driven healthcare system.

Current Issue
Technopilot: A Train of Security Solutions