Commentary: The database revolution happening in healthcare
Call it a “tech-tonic” shift measuring 7.0 on the disruption scale.
We’re talking about the enterprise Relational DataBase Management System (RDBMS) industry. The battle lines are being drawn between startup technologies wrapped around specialized Not only Structured Query Language (NoSQL) hardware, software, and services against the traditional RDBMS vendors — and healthcare business executives, hospital CIOs and federal leaders around the Washington Beltway would be wise to understand the disruptive implications.
Several reasons for this type of NoSQL approach include greater simplicity of design, horizontal scaling, better data availability controls, lower cost development and maintenance, and greater interoperability. The NoSQL data structure is typically a key-value, graph, or document, versus the traditional relational database row/column structure, and, as such, some searches can be faster in NoSQL while some may still be faster in the RDBMS environment, depending on the type of problem.
Currently, Venture Capitalists are pouring money into NoSQL and related Hadoop ecosphere investments, realizing that the new technology offers a way to develop data-driven applications faster and at lower cost, address structured and unstructured data, and enable access to silos of disparate data. Many early Big Data projects have utilized some version of the Hadoop operating environment in their solution.
Starting to shake
Federal agencies appear to be gaining interest in at least exploring Hadoop solutions based on their open source, simplified approach to data management, and potentially lower cost approach for delivering comparable capabilities to more traditional proprietary database or high performance computing (HPC) environments.
Enterprise software developers have designed and built increasingly sophisticated and complicated layers of software for nearly 30 years. In fact, there is an entire class of software solution providers who are proficient at coding software following a traditional n-tier architecture approach to enable enterprise level access to data stored and curated in relational databases. The standard approach to accessing this data has been through a Structured Query Language, referred to as SQL.
[See also: Top 5 Government Health IT stories of the summer.]
Now, the new NoSQL databases are providing a refreshing technical alternative and exciting new approach to the storage and retrieval of data. Big Data applications can be built as simple “apps” on top of these NoSQL-based operating platforms, which provide access to data stored across large, distributed multiprocessor and virtual computing environments. There are about a half dozen platforms that are competing for market share, and there is also an emerging tools vendor market that provides hundreds of templates as a springboard to further accelerate the agile development of Big Data “apps” leveraging the Hadoop operating environments.
Today it seems that the typical Washington, D.C.-oriented business executive or Federal agency looks at this new NoSQL technology and thinks that these early use cases are not likely to impact or benefit current enterprise data services anytime soon. This view seems particularly true in the healthcare field, where recent surveys have shown very low levels of agency adoption in healthcare Big Data projects, as noted recently in the MeriTalk “Big Data Cure” study, which surveyed 150 Federal healthcare executives. [[Full Disclosure: EMC, the employer of one author of this article, sponsored the MeriTalk research.]]
Disruption ahead
There are signs that this new NoSQL technology is creating a market ready for disruption: Hadoop may represent the biggest paradigm shift in database technologies in a generation. The power of Hadoop is that it significantly simplifies the process for accessing and analyzing large sets of distributed data. The technology also reduces the level of technical support needed to deal with the analytics, network, processing, and data storage components. A single analyst can work independently on the data and not require the services of database administrators when implementing a Hadoop processing and storage approach.
It also appears that many of the big Federal Systems Integrators (FSIs) that are supporting the government either don’t yet realize or are slow to appreciate the shift that is occurring in their relational database technology-related business base.
New database technologies will require different skills and new business processes. The current FSIs have an army of software developers and database managers who are going to need to be retrained to support the Hadoop and NoSQL platforms. In addition, their software development processes need to be adapted to new agile methods, leveraging tools to get started fast and provide quick iterations, similar to how smartphone applications are developed.
If federal agencies and FSIs aren’t careful, they may miss this magnitude 7.0 “tech-tonic” shift that is occurring, and the results could be disastrous for their business and their ability to support their government customers.
There is a technology disruption going on, and the healthcare technology executives who are focused on the federal government market inside the Washington Beltway need to be vigilant and exercise caution.
Related articles:
10 things DoD wants in its next EHR
Does Apple's HealthKit prove FDA guidance is working?
Q&A: Why true security requires 'a philosophical shift'