In late 2018, the Global Big Data Analytics in Healthcare Market report released some eye-opening information about big data (BD) in healthcare: it is “expected to generate revenue of around USD$68.03 billion by 2024, growing at a CAGR of around 19.34% between 2018 and 2024.” Additionally, the report highlights how BD analytics can be beneficial in healthcare by curing diseases, avoiding epidemics, and even cutting down costs. Given its magnitude, let’s take a look at the applications and challenges of BD in healthcare.
At its simplest, BD refers to excessive amounts of information created by the digitization of… everything. It often requires more advanced storage and management software and/or infrastructure. Healthcare-related data, such as health conditions or medical claims, has increased exponentially in less than a decade. In 2013, there were 153 exabytes in medical data, with one exabyte being equal to one billion gigabytes. In 2020, there will be an estimated 2,314 exabytes—or 2.3 trillion gigabytes—in medical data, representing a 48% annual growth rate since 2016. In fact, healthcare data growth is considered to be one of the fastest across most industries.
So, why has BD become a central focus in healthcare? Physicians’ decisions are relying more and more on evidence-based research and clinical data. IBM’s Big Data & Analytics Hub reports that 80% of healthcare data is unstructured and stored in hundreds of different forms, such as lab reports and medical images. If mishandled, overwhelming amounts of unorganized data can lead to dire patient circumstances due to clinical errors. This is why data collection and management is so crucial to the development of BD analytics and its use cases. Currently, there are a number of data aggregation and management solutions on the market. However, applications of BD analytics are still being developed and improved upon. As mentioned in the BD Analytics report, developing these applications for BD will contribute to cost reductions and improved patient outcomes results through predictive analytics and remote patient monitoring, respectively.
Application: Predictive Analytics
Organizing medical data into operational silos to identify at-risk patients more efficiently is an important consideration in BD. Let’s consider online shopping for a second. When you shop for pants online, you might add a “bottoms” filter to make your search more efficient and avoid seeing anything other than what you need. Now, substitute clothing filters with medical symptoms. A physician could enter patient characteristics and symptoms into a search engine, which would then give a list of possible conditions relating to the inputs. Processes like these can then be automated to fast-track diagnostics and allow for quicker response times to urgent medical situations. This is essentially advanced data analytics within healthcare: “using data-driven findings to predict and solve a problem before it is too late, but also assess methods and treatments faster, keep better track of inventory, involve patients more in their own health and empower them with the tools to do so.”
Application: Remote Patient Monitoring & Telemedicine
Predictive analytics and remote patient monitoring (RPM) go hand in hand, especially with regards to chronic disease management. As highlighted above, using BD analytics is beneficial for risk prediction, diagnostic accuracy and patient outcomes. RPM and telemedicine give patients the freedom to be at home or go to work—rather than staying in a hospital—while being monitored continuously by their healthcare provider. In the event of any irregularities, physicians would medically intervene by asking the patient to come in or making a home visit. This is the current reality in digital health and chronic disease management through the advent of medical-grade wearables. Integrating predictive analytics into RPM or telemedicine tools can further improve access to care, quality of care and patient outcomes. Considering the popularity of virtual care, it’s only a matter of time till BD analytics becomes standard practice in delivering medical care.
Healthcare—and its data—is highly regulated, which further complicates anything having to do with medical data. The biggest undertaking of BD in healthcare is turning data assets into data insights. This requires infrastructural consideration for various data analytics applications, most significantly data collection, storage, and security. Despite the growing dependence on electronic health records and digital tools, many processes used by health systems are paper-based and/or require manual data entry, leading to potential man-made errors. Investing in the appropriate infrastructure and staff is the first challenging step in utilizing BD analytics to its fullest potential. Currently, quantum computers are the most popular option for housing and handling abundant amounts of data. But, it’s not that simple. A single quantum computer costs tens of millions and they’re still not ready for a commercial roll-out. On top of this, organizations with quantum computers will need to hire the right technicians to use these machines. This is such a concern that the US government passed the National Quantum Initiative bill to invest in training and staffing.
Challenge: Policy & Regulation
With the increase of data breaches—Facebook and Cambridge Analytica being the latest publicized scandal—privacy has become an important conversation in technology. More specifically, we have to ask ourselves: Who has access to the data and how will they access it? How do we ensure accountability from patients who give out their data and providers who might collect and manage this data? What steps are required to keep the data safe and secure? The obvious implications for data security, as of today, have to do with privacy compliance, such as HIPAA in the US and PIPEDA in Canada. Moving forward, policies should be created around each step of the BD security life cycle: data collection, data transformation, data modeling and knowledge creation. Additionally, security safeguards should be emphasized for any digital health tool that can collect and store patient information. For example, virtual appointments should be done on compliant platforms that offer full encryption with their services, instead of using Skype or Facetime.
Clearly, BD is an unavoidable conversation relating to the future of healthcare, like other digital health trends such as these other digital health trends. Its applications are promising and could revolutionize how healthcare is delivered. But, in order to be proactive, we must start considering and incorporating the infrastructure, overhead and regulations necessary to benefit from BD analytics.