Healthcare Analytics: Realizing the Value of Health IT
Contributed by Brian Ahier, guest blogger
This is my post on the value of health information technology for National Health IT week. So far we have not yet fully been able to realise the value of health IT. While great efforts at adopting electronic health records and interoperability are beginning to see some results from the billions of taxpayer dollars invested, we still have a long way to go before we see the true value of health IT.
Healthcare has been slowly moving through three waves of digitization and health data management: data collection, data sharing, and data analytics. Data collection and sharing waves have been having some success, spurred on by the HITECH Act and implementation of electronic health records and health information exchanges. They have not yet significantly impacted costs or quality in healthcare. The third wave is ready to crash on our shores and I believe we will actually begin to see an IT infrastructure that support the new payment and care delivery models which are emerging.
This third wave of analytics will enable large numbers of healthcare organizations to realize some significant returns on their IT investments and thrive in the healthcare marketplace of the future. Developing a consensus model for adoption of analytics capabilities could help healthcare leaders and vendors succeed by providing a common roadmap for the deployment of these capabilities. But much of the success of these analytics platforms will depend on the underlying architecture and I think the “late-binding” data warehouse model holds the most promise.
The term late-binding dates back to at least the 1960s, where it can be found in Communications of the ACM. The term was widely used to describe languages such as LISP, though usually with negative connotations about performance. In the 1980s Smalltalk popularized object-oriented programming (OOP) and with it late binding. Alan Kay in History Of Programming Languages 2 laid out the fundamentals of OOP and late-binding architecture in The Early History of Smalltalk section. In the early to mid-1990s, Microsoft heavily promoted its COM standard as a binary interface between different OOP programming languages. COM programming equally promoted early and late binding, with many languages supporting both at the syntax level.
The late-binding data warehouse model is a just-in-time method and is more adaptable to new analytics use cases and data content than those that make use of early binding and tightly coupled enterprise data models. Late-binding is a method of assembling data from disparate sources just in time for particular analytic use cases. The late-binding model of data warehousing is starting to gain traction in healthcare as many provider organizations gear up for population health management. The advantage of this approach is that it allows users to combine disparate data very quickly for targeted analyses without locking data warehouses into a predetermined data model.
This late-binding model for use in healthcare analytics was developed by Dale Sanders, who after witnessing and reflecting upon the failure of several multimillion-dollar data warehousing projects during his work in the US military, saw the same patterns in data engineering as those in software engineering prior to OOP. In the late 1990s, while employed by TRW Inc., he was sponsored by the Pentagon to study advanced decision support in nuclear warfare operations, a project called the Strategic Execution Decision Aid. He turned to the healthcare industry for what he expected to be role-model examples of computer-aided analytics to drive better decisions in time-critical, life-critical situations but instead found almost no examples, with the notable exception of a scattered few such as at Intermountain Healthcare in Salt Lake City.
Last month, I was able to get a peek behind the curtain at the Health Catalyst analytics solution, which Dale has been helping develop. Health Catalyst is a data warehousing company that recently has gained a number of customers and investors among large healthcare organizations. For example, Partners HealthCare System has joined with the investment arms of Kaiser Permanente and CHV Capital in an $8 million equity investment in Health Catalyst. Health Catalyst uses a late-binding data warehouse model. There is a very good Slideshare on Late Binding in Data Warehouses that helps to illustrate some of the concepts.
Last year, The Advisory Board Company acquired clinical analytics vendor 360Fresh. 360Fresh offers two products. Pulse360 uses text and data mining to extract information from EHRs and other systems to provide answers to clinical and quality questions. They also have Track360, which is a clinician care coordination and workflow tool designed to streamline provider handoffs, provide alerts, and improve patient communications. Both of these products are targeted for use by academic medical centers, independent community hospitals, and large-scale ambulatory providers. 360Fresh will augment The Advisory Board Company’s Crimson offering which provides retrospective data review, by adding real time analytics capabilities.
Dale Sanders, along with colleagues Jim Adams, Ernie Hood and Meg Aranow of The Advisory Board Company; Denis Protti of the University of Victoria, British Columbia; Dr. Dick Gibson, Providence Health and Services; Mike Davis of Mountain Summit Partners; along with Dr. David and Tom Burton, both of Health Catalyst, worked to develop an eight-stage Analytic Adoption Model similar to the seven-stage EMR Adoption Model (EMRAM) from HIMSS Analytics. This model was initially described in ElectronicHealthcare in September 2012, and has recently been even further enhanced based on significant health system experience and published as the Healthcare Analytics Adoption Model. The hope is that this model will enable healthcare organizations to fully understand and leverage the capabilities of analytics so that we can begin to see the real value of health IT in laying a technology foundation to achieve the triple aim.
This model borrows lessons learned from the HIMSS EMRAM, and describes a similar approach for assessing the adoption of analytics in healthcare. The Healthcare Analytics Adoption Model provides:
- A framework for evaluating the industry’s adoption of analytics
- A roadmap for organizations to measure their own progress toward analytic adoption
- A framework for evaluating vendor products
HIMSS Analytics has also partnered with the International Institute for Analytics (IIA) to create a benchmarking survey designed to measure and score clinical & business intelligence and analytics maturity in healthcare organizations. They will use the DELTA model (data, enterprise-focus, leadership, targets and analysts) to assess analytic capabilities. I expect we will see a lot more about the Healthcare Analytics Adoption Model as the work continues to progress. This will lead us into the future of health IT where value will ultimately be realized.
View a slideshow of the Analytics Adoption Model here.