For as long as organizations have been using data to help make business decisions, data analytics has been a mostly linear assumption. A leads to B, B leads to C and so on. It’s straightforward, process-oriented (and depending on the format), relatively easy to work with. When businesses started considering data as a strategic asset to make decisions, data analytics began to explode. All types of data began to infiltrate organizations, including: predictive analytics, big data analytics, cognitive analytics, enterprise data management and much more.
Within healthcare, adoption of new technologies and methodologies has been slower on the uptake. Part of this is because data comes through in huge volumes from many disparate sources, including: medical images, claims data, electronic health records, health surveys, and more recently non-traditional data, such as purchasing, transportation and geography. All these sources directly correlate to Social Determinants of Health. Data in the healthcare industry is abundant, but that abundance has its drawbacks. Healthcare data is not being leveraged to its greatest potential.
Data, data everywhere
Healthcare organizations have struggled to create a solid foundation that will support their digital enablement. There are more sources of healthcare data today than ever before, including EMR/EHRs, medical devices, claims processing systems, eligibility systems, and wearables to name just a few. Fortune notes that healthcare data is being generated “… on the order of 750 quadrillion bytes every day — or some 30% of the world’s data production.” While most healthcare organizations have an enterprise data management program, healthcare-related data continues to grow at such an exponential rate that just trying to maintain current capabilities can lead to a feeling of drowning.
Navigating the data lake
Healthcare organizations are, however, making significant strides forward. Many healthcare organizations have started to embrace the concepts of big data through the implementation of a data lake. While this results in the ability to accept unstructured and semi-structured data, it doesn’t always lead to improved usage of the data even when using advanced analytics, such as machine learning, automation and other forms of artificial intelligence. Interoperability is also still a significant concern, but with the recent release of the FHIR® v4.0.0 standard by HL7.org, and the February 11 release of proposed updates to the 21st Century Cures Act, progress is being made on this front as well.
Seema Verma, CMS Administrator, stated during her keynote speech at HIMSS 2019 that information blocking, one of the key barriers to interoperability, would no longer be acceptable. Further, she stated that, “Our agenda is to create a future where the underlying databases that power our health system are connected using APIs and other modern technology. To not just share data, but incorporate it into a single record.” The proposed rule from the Office of the National Coordinator (ONC) and the CMS proposed rule both provide new considerations for health care organizations in their handling of data as a strategic asset. Included in the ONC proposed rule is a removal of the definition for the “Common Clinical Data Set” (CCDS) and the introduction of the “United States Core Data for Interoperability” (USCDIv1). The challenge for healthcare organizations will be to segment the vast amount of data available to them into the various data classes included in this new standard.
Grabbing the life preserver
Under value-based care, healthcare organizations are mandated to lower costs and improve quality and access to care. With a solution like the Business Insights Engine (BIE) from NTT DATA Services, there’s a clear plan for enabling digital data and intelligence to support that need; a plan that addresses many of the points raised in Administrator Verma’s speech. It puts the power of data in the hands of the business. The Business Insights Engine is focused on helping organizations successfully manage and mine the huge mountains of data created across the healthcare continuum.
The BIE facilitates the adoption of big data, AI, and machine learning strategies by ingestion, storage and usage of multiple data types and sources, including structured, semi-structured and unstructured data in a single repository. Using the BIE to conduct analytic activities is enabled via the API-ready nature of the solution. Machine learning tools use Rest APIs to access the data and create predictive and/or prescriptive models that output actionable insights that are then sent to downstream applications such as care/case management, member portals, provider portals, etc. With the flexibility, scalability and sustainability provided solutions like the BIE, healthcare organizations will have the necessary foundation to facilitate future growth and transformation.
Swimming to shore
Once healthcare organizations have grabbed onto the life preserver, broader usage and access to the data is enabled. Consider this real-world scenario: A person with asthma is seen in the ER at least three times a year. Testing indicates that the asthma is triggered by dust and rescue and maintenance inhalers are prescribed. However, this doesn’t prevent the person from visiting the ER multiple times the following year. It is here that the insight and the “one record” concept become valuable. Through inclusion of claims data, pharmacy data, social determinants of health and other non-traditional data sources such as weather data and wearables, it is found that:
- This person’s asthma is triggered by dust
- The person lives in an area where dust storms are prevalent
- There is little public transportation and the nearest pharmacy is 5 miles away
- Claims data shows that the last time the person filled a prescription was 6 months ago and that the lot number is one that has expired
- Another dust storm predicted in 3 days
Through use of all these data sources, a new rescue inhaler can be drop-shipped to the person, along with outreach from the provider or care manager telling them to expect the shipment and why. It’s care that is predictive, preventative, personalized and participatory. Solutions such as the Business Insights Engine are harnessing the power of the data — structured and unstructured — to form value-based, cost-saving and member-personalized decisions; allowing healthcare organizations swim safely to shore.