The Role of Data Analytics in Healthcare

The increasing digitization of healthcare—from the the evolution of the electronic medical record (EMR) and virtual care to connected medical equipment and smart hospitals—has resulted in a data explosion. According to some industry sources, healthcare produces 30% of the world’s data, with one patient generating “up to 80 megabytes yearly in imaging and EMR data” The increasing volume, veracity, and variety of healthcare data puts it in the category of “Big Data.”

While this is good news for hospitals and other organizations looking to improve patient care and safety and allocate resources more efficiently, adding further complication is the significant amount of digital information still buried in data silos. Healthcare faces the steep challenge of unlocking, ingesting, analyzing, and using this data more effectively.

This is where the power of healthcare data analytics comes in.

An article in the journal Health Services Research and Managerial Epidemiology defines healthcare data analytics as “…a study of methods and techniques to analyze data, discover new information and knowledge, link data in terms of its semantics, and describe data to other informaticians, managers, and other stakeholders.” More simply put, the goal of healthcare data analytics is to pull data from disparate sources and organize it into a cohesive, cloud-based system, making it available to end users in usable formats. Healthcare data analytics is integral to virtually every aspect of healthcare. For Example, using data analytics:

• Physicians can receive real-time patient care insights and alerts via the EMR at the point of care to target and treat high-risk conditions
• Patients’ can access personalized information more readily, including health records, lab tests, and billing
• Hospital revenue integrity departments can predict claims errors and identify missing revenue
• Hospitals can receive real-time data from connected machines, such as CTs and MRIs, to optimize service and utilization and improve uptime
• Facilities can proactively manage building equipment such as boilers and HVAC units to prevent costly breakdowns

Similar to the airlines and manufacturing industries, healthcare is a complex industry that relies on advanced data analytics to manage the intricate connections within and outside of organizations, drive more efficient clinical and business operations, and create alerts that spark specific behaviors. For Example, healthcare analytics is key to creating dynamic staffing and patient scheduling models, managing patient flow inside facilities and across the healthcare continuum, and sharing real-time clinical data and information with the right people and facilities; and much more.

Organizations can also tap into the power of data mining for insights that aren’t intuitive on an individual level to do things such as comb medical records to find and eliminate clinical care gaps; scrub clinical documentation to reduce denials; and look at machine and device data to manage service requirements. A robust data analytics solution can quickly pay for itself with a 3X-4X ROI in hard and soft benefits, including the following:

1. Raises organizational intelligence
2. Drives greater collaboration and efficiency
3. Boosts patient safety, care quality, and care outcomes
4. Increases proactive decision-making
5. Optimizes utilization of fixed assets/increases ROI
6. Reduces costs/improves margins

In healthcare, data analytics is used to capture, ingest, and convert three types of data:

1. Structured data: A vast amount of structured (understandable) data sits in data silos that are difficult to access and share. This data typically resides in transactional legacy systems, such as electronic medical records (EMRs), picture archiving and communications systems (PACS), and radiology information systems (RIS), and others.
2. Unstructured data: Connected machines and buildings create a large quantity of data that is used by healthcare providers, original equipment manufacturers (OEMs), software companies, and others. Much of this data is unstructured, meaning, it’s free flowing information that cannot be readily understood. It must be captured and converted into usable information.
3. Patient data: Patients are generating increasing amounts of data through smart phones and mobile apps. Most of this data resides in different silos either in clouds or via on-premises solutions.

Through analytics, healthcare organizations have an opportunity to look more deeply at the data, drilling down to specific insights. The following are four types of analytics typically used in healthcare.

• Descriptive analytics looks at historical data and patterns, enabling organizations to quickly connect the dots to make daily and weekly decisions. Descriptive analytics us used in simple reports, charts, and graphs and is often referred to as the low hanging fruit. Descriptive analytics is a good starting point for organizations new to data analytics.
• Predictive (proactive) analytics applies algorithms and machine learning techniques to data to make future predictions in certain areas such as hospital bed capacity, patient adverse events, patient scheduling needs, and maintenance requirements on clinical assets. Applying data science at this level enables healthcare organizations to be more proactive in their decisions.
• Prescriptive analytics captures larger amounts of data, providing actionable insights and informing future strategies.
• Learning from user experiences is the newest frontier in healthcare data analytics. Healthcare organizations are now able to crowdsource and learn from user interaction data, including patient ratings. Over time, an organization becomes smarter as it starts to learn what patients and consumers like and don’t like.

Having a strong data analytics platform is essential for all healthcare organizations, and especially those looking to move into smart hospital territory. A comprehensive data analytics solution will enable organizations to ingest and scrub data for insights from three primary areas that are typically challenged by data silos:

1. Healthcare professionals and patients

Dynamic scheduling models: Using data analytics, hospitals and medical practices can predict future scheduling needs and share scheduling information to optimize facility and clinician capacity. Real-time information enables them to book patient appointments with the right provider at the right facility at the right point in time instead of moving the appointment further out. Maximizing scheduling capacity improves patient access and care quality, ROI for fixed assets such as medical equipment, and ultimately revenue.

2. Healthcare machines

Maximizing efficiency and utilization while reducing service costs: Without data analytics many healthcare organizations fly blind when it comes to managing fixed assets such as connected machines, including CT scanners and MRIs. For Example, healthcare data analytics can help organizations avoid an expensive service contract with the original equipment manufacturer (OEM). Analytics also provides key utilization information by facility and OEM type to ensure clinical assets are running at full capacity, while also identifying causes of underutilization such as scheduling or workflow issues.

And predictive intelligence can identify maintenance issues and when a particular piece of equipment is going to fail long before it goes out of service, allowing preventive maintenance and less downtime. Moreover, combining equipment data with real-time locating systems (RTSL) enables organizations to make meaningful decisions. All of this adds up to a richer ROI.

3. Healthcare facilities

Building automation: As smart hospital technology takes off, organizations are creating data frameworks to integrate medical device data with building automation systems.

Managing large amounts of healthcare data against a maze of complicated government legislation and regulations—that oversee patient safety, care quality, security, and privacy—requires tight strategic oversight. Successful healthcare organizations recognize that establishing a framework for a data-driven enterprise is a critical step to ensuring regulatory compliance and improving the following:

• Patient quality care
• Patient safety
• Patient outcomes
• ROI on fixed assets
• Organizational margins

According to a McKinsey report, data-driven enterprises of the near future will share the following seven key components: