Big Data Analytics In Healthcare

Big data drives the need for the continual use of analytics to improve services and enable informed patient care decision making.

By John Backhouse, executive director of the Omni Program, Information Builders

There are a number of questions surrounding big data. Some healthcare providers argue the use of it is the key to achieving a sustainable competitive advantage and optimizing clinical outcomes. Others wonder if big data carries more risk than reward.

But healthcare organizations all must find new approaches to analytics to fully harness the available data points across the patient journey and transform it into actionable insight that can truly enhance care quality.

For healthcare providers and payers alike, more powerful and robust tools are needed to successfully leverage big data and analytics to improve patient care delivery, accelerate accountable care business models, reduce costs, and drive organizational change.

Data is never at rest. As patients see physicians and receive care, information about those interactions flows between systems, providers, insurance companies, and healthcare networks. Other data may also be required, including details about enrollments, provider credentialing, or fee schedules. Information collected from medical collaborators and other external sources, such as electronic health records, medical images, and care management documentation, also factors in. This data in motion drives the need for the continual use of analytics to improve services, reduce costs, and enable informed patient care and decision-making.

Integrating Big Data
Big data needs a place to reside. It needs powerful hardware, backed by solid software, so that the variety and volume of information can be harnessed. The data also originates from a vast array of sources – personal medical records, radiology images, clinical trial data, FDA submissions, human genetics, and population information, among others.

State-of-the-art hardware and software allow for efficient processing of big data, and for support of analytical workload complexity and agility. A high percentage of big data is often described as multi-structured to distinguish it from the structured operational data used to populate a data warehouse. In most organizations, multi-structured data is growing at a considerably faster rate than structured data.

The challenge lies in unifying these disconnected data sets and fully leveraging the potential of existing information. Technology enables this integration, as well as the generation, capture, and analysis of the new healthcare data streams such as genomics, sensor readings, and population and disease statistics.

As the volume and variety of data diversifies, so too does the velocity at which it is generated and the speed with which it must be retrieved, analyzed, and compared to facilitate better decision-making.

Storage and analytics platforms provide the ability to create and leverage a data warehouse to cost-efficiently bring all that information together and make it readily available for reporting and analysis.

Data Quality is Primary
A common mistake is to presume that all data within a warehouse is clean. Data quality issues are of particular concern in healthcare because they can impact decision-making related to quality of care and patient safety.

The quality of healthcare data, especially unstructured data, is highly variable as it comes from a multitude of sources – written prescriptions, doctor’s notes, patient charts, etc. That means veracity in healthcare data is imperative. Are the patient, hospital, payer, and reimbursement codes correct? Are diagnoses, treatments, prescriptions, procedures, and outcomes properly captured? Variety, volume, and velocity all impact veracity by fueling the cycle for continuous improvement and analytics. The highest value in healthcare data will be realized with enhanced data quality and representation of relevant measures.

The collection, standardization, integration, and consumption of this rapidly growing structured and unstructured data defines big data analytics. A software solution that integrates, normalizes, and validates clinical data from across the continuum of care can not only improve outcomes and reduce costs, but also help identify disease trends, coordinate rules-driven patient registries, and drive performance management strategies – a key ingredient for meeting Accountable Care Organization (ACO) requirements.

The IT Challenge
Gaining an enterprise-wide view of healthcare data helps organizations provide fact-based answers to questions ranging from which therapeutic approaches work best, to which patients are at highest risk for re-admission, to how physicians are performing in relation to quality and cost. The challenge is getting that data into a format that allows clinicians to make decisions more quickly and accurately.

While traditional analytics focuses on the use of data warehouses, big data calls for immediate synchronization. Big data analytics is more iterative and predictive, anticipating what may happen in the future so providers can take a more proactive approach to patient care.

However, eliminating data silos and bringing together fragmented information in real time doesn’t have to be so difficult. Imagine the value of identifying approaches to use leading indicators, instead of trailing ones, so that interventions could be made more rapidly – and with more predictive power behind them.

The move to big data analytics for healthcare is an opportunity to explore and seek competitive advantage through greater insight. Like any investment, it comes with risks and considerations. Whether or not more data will likely yield valuable insights must be determined, but in the end, it’s about using data more effectively.

The best way to approach big data analytics, and support the heavy processing involved, is to deploy optimized hardware and software solutions for processing different types of big data workloads. These solutions can then be combined with the existing enterprise data warehouse to create an integrated information supply chain.

Once an organization has a sense of its data, it must determine what the most critical business and clinical decisions are and whether or not the required data and analytics are available to make them. How many of areas of the healthcare organization would be further supported with better analytics? Is the organization structured in a way to capitalize on better decision-making – especially with more data being created every day?

With powerful big data analytics tools, healthcare organizations have the storage and computation needed to ingest and digest information from across their data-rich enterprises. Through techniques such as machine learning and statistical modeling, organizations can learn from the details in a more comprehensive manner – capturing and analyzing all the data. Whether the focus is on patient care delivery or clinical research for disease management, these systems will help successfully drive any analytics strategy forward.

The Path to Data-Driven Healthcare
The promise of big data in healthcare is revolutionary. Use of big data will ease the transition to authentic data-driven healthcare, allowing healthcare professionals to improve the standard of care based on millions of cases, define needs for subpopulations, and identify and intervene for population groups at risk for poor outcomes.

Big data and analytics can create immediate value for an organization. Analytics can improve an organization’s performance by making information more transparent and measurable, while exposing variability as well as potential issues and opportunities. This transparency, through operational dashboards or performance management, fosters better decision-making across an organization and enhances patient care delivery at a lower cost and improved efficiency, while also satisfying benchmark and regulatory issues.

A successful business and clinical analytics strategy requires a singular mechanism for gathering, consolidating, preparing, and storing data from any source, in any location. Only with the right comprehensive platform – one that can enable integration of all information assets, dramatically improve the integrity of the data contained in those assets, and transform that data into powerful intelligence for wide-scale use – can companies ensure that their enterprise information is truly analytics-ready.

About the author
John Backhouse is executive director of the Omni Program for Information Builders, directing the Global Healthcare practice within the iWay division, and is responsible for driving application development for the healthcare market.