5 keys to unlock the full potential of big data in the life sciences
This article first appeared in Pharmaceutical Online
In a previous article, we wrote about the increasingly data-driven nature of the life sciences industry. We proposed that companies that harness new technologies and Big Data to generate insights will continue to have a competitive advantage for the foreseeable future. Here we explore some of these concepts in more detail.
Data has always been important to pharmaceutical companies. Think about all the data generated in drug discovery and in identifying promising candidates, through to clinical trials data that demonstrates the safety and efficacy of a drug. Without the right data, it would be impossible to drive a new molecule or technology through to market.
Due to a confluence of factors, the role of data and technology will increase significantly over the coming years and will shape the future of the life sciences industry. These developments include:
- An exponentially increasing volume of real-world data generated by wearables, along with the Internet of Things, consumer technologies and social media are all driving an improved understanding of patient needs and behaviors, disease progression, and markets
- Gene sequencing and virtual reality are leading to an improved understanding of the human body and how diseases impact it
- Vastly superior data processing capabilities driven by the cloud, artificial intelligence (AI), robotic process automation, and, ultimately, parallel and grid computing are enabling this wealth of Big Data to be consumed and used in an efficient and meaningful way.
Our new report explores four potential future cell and gene therapy scenarios using our FutureWorlds™ approach
It is an exciting time! Some successes with technologies are already being publicized and many more are in trials. For example, this year, Janssen Research & Development, part of the J&J family of companies, used AI to increase efficiency of drug discovery by up to 250 times. Amgen is running an AI pilot to try to better diagnose osteoporotic fractures. AI is also being piloted to speed up the laborious process of redacting personal data from clinical submissions and to help reconcile clinical trial adverse events with safety databases.
The common thread is that all of these technologies are relying on data to derive insights. All major pharmaceutical companies are now exploring real-world evidence, which promises to unlock a treasure trove of insights on care pathways, phenotypes, disease progression, market behaviors, etc. For instance, this year Roche purchased Flatiron Health for $2 billion to acquire data insights from Flatiron’s network of clinics and research facilities as well as its core data capabilities. This will be increasingly useful for personalized medicine, data on drug usage, and a much broader set of patients than were involved in the clinical trial.
There are five steps necessary to thrive in this ever-increasing data and technology driven world:
- Establish the right data capabilities and let them flourish. Don’t overburden your data scientists. While pharmaceutical companies may have hired an army of data scientists, they are invariably overworked running uninteresting or business as usual reports or trying to source or clean data. Unburden your data scientists and let them ask the interesting questions.
- Encourage collaboration between clinical scientists and data scientists. Context is required to drive meaningful insights. By encouraging your clinical scientists and data scientists to work together for a common purpose, you can drive more meaningful and usable insights. This is not always easy to achieve. Perhaps it is time to start measuring data scientists based on the insights that they generate — a variation of the prisoner’s dilemma?
- Look beyond the usual data. While the “usual” data will continue to yield some value, the real value will be in identifying, integrating, or using different data sets or combining them in innovative ways. This may be data already in-house that is not being used, data generated from real-world evidence, virtual registries, digital biomarkers, or data sourced from external parties and partnerships. The company that achieves an end-to-end (cradle to grave) view of a population’s health data, incorporating different data types at the right level of granularity while also remaining compliant will have an enormously valuable asset.
- Foster a culture of innovation and experimentation. It is a cliché, but failing fast and creating an environment where it is acceptable to fail is the only way to encourage data scientists to harness new data and technologies and push the limits of what these can provide. This also requires a team that is passionate about the cause and is incentivized to own and explore their ideas and hypotheses. Adopting an agile approach can help to drive these behaviors.
- Be brave and actively manage the risks. Many organizations are paralyzed by fear of regulations, changing guidelines and standards, data security, and the pace of technological change. Running away from this or putting it in the “too difficult” bucket won’t help. Collaborate with the people that understand this world. Work with your compliance and security teams to understand the guard rails and requirements, and establish a safe and compliant space for your data scientists to do their thing.
The convergence of data and technologies provides an opportunity for companies to make a step change in innovation and performance. Those that successfully apply the strategies to all parts of the business, including drug discovery, clinical trials, and sales statistics, will thrive in an increasingly complex sector.
Neil Saward and Mariko O’Neill are life sciences experts at PA Consulting.