TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

Fighting COVID-19 with Frictionless Data Sharing

Fighting the COVID-19 pandemic is going to take data -- lots of it -- and that data will need to be shared easily and seamlessly. Here are the top six steps that we will need to take.

• By Birju Shah, Nick Jordan
• May 19, 2020

It’s hard to believe it has only been three months since the initial COVID-19 pandemic began. Seemingly, our lives have changed overnight. Socially, we’ve had to change the way we interact, and economically, we’ve come to a near halt with most businesses impacted. Data leadership is now in the spotlight. Global epidemiologists, data experts, and operational gurus have become the information armed forces, creating a feedback loop of information to supply leaders around the world with data that can help empower decision making.

After the Industrial Revolution we entered the Information Age, which promised to enable us to share information freely and have it be consumed around the globe. In the world of healthcare, we are at the early stages of such information sharing, but by being able to ingest, normalize, layer, and extract insights from data, we can solve problems at scale. The impact of COVID-19 is accelerating and flattening the world of information across the healthcare supply chain so we can rapidly innovate.

However, solving the COVID-19 crisis is going to take data -- lots of it -- and it’s going to require frictionless data sharing. Here are the top six steps that we will need to take.

Step 1: Create a single data repository

The first step will be to flatten the information structure of a complex healthcare supply chain. Fragmented hospital systems, insurance claim IT systems, clinical trial information, and pharmaceutical clinical protocols will have to be in one central repository for all researchers to access.

Next, we must establish frictionless key sharing across a list of archaic and newer IT systems. The data from the United States Veteran’s Administration systems is very different than data from Kaiser Permanente and even new primary care providers such as One Medical. Creating an “IP protocol” for easy data sharing across these systems to enter into one repository is a must.

Step 2: Clean up the data

All types of data (including patient, cellular protein, clinical protocol, phenotypic, and environment data) need to be normalized to a standard record format that can be used by everyone with unimpeachable reliability, accuracy, precision, and speed.

Step 3: Normalize and layer the data

Much like baking a multilayered cake, we must first assemble the ingredients (the data), form the ingredients into a sheet cake (a record), repeat these steps for another sheet (normalization), then make sure the multiple sheets of the cake are perfectly lined up to stack on top of each other (the layering).

Data resides in the human body (genomics, proteomics, metabolic, phenotypic) with attributes of when and how you were measured (diagnostics), the data record is your patient medical record, normalization is anonymized normalization (for example, “Indian male, age 35-45, in athletic shape”), and layering is taking all the layers of data in your normalized record and adding other similar people on top of your record. This process will allow scientists to understand broad communities and the likelihood of disease and/or antibody discovery.

Step 4: Create work teams

Once data is in a central repository and normalized, we can expect to see epic healthcare teams form across enterprises. These are the Roche and Novartis, the J&J and Sanofi, and the Gilead and Biogen partnerships. Companies that compete can team-up to accelerate innovation.

It would be ideal to also start getting pharmaceutical companies as well as government entities to share data. Governments across the globe value pharmaceutical discovery because it drives a significant percentage of GDP for them as an exporter and also affords strong leverage when negotiating agreements with other countries.

Step 5: Revisit regulations

The epic teams and a strong desire from governments to solve the crisis and restart their respective economies will also lead to new fast-path commercialization through regulation.

The saving grace for this industry now is compassionate use by the FDA that allows untested drugs to be used as a lifeline for end-of-life patients. Compassionate use may become the norm more often.

Step 6: Build an information repository

With frictionless data, we can and will overcome COVID-19, and we will have accelerated healthcare’s ability to innovate rapidly because of it. We are at a unique point in time with the opportunity to build a massive, normalized repository of human and animal layers of healthcare data that global researchers, scientists, and doctors can access.

Such a repository will allow global access on a scale never seen before and it will set up the ability to backtest, forgo having to repeat experiments, and develop an understanding of characteristics, such as the environmental impact on an unknown disease.

The potential is enormous for attaining goals such as computerizing vaccination testing with a new virus and recommending solutions within a month instead of 18.