A recent survey of health system IT, informatics, business, and clinical staff revealed that less than 40% of healthcare executives believe their organizations successfully share clinical data with external health systems, payers, and other partners. When asked what steps their organizations are taking to address interoperability challenges, 60% of respondents reported their organizations were moving to a single, integrated EHR. In addition to costing tens or hundreds of millions of dollars, changing EHRs to homogenize across a system does not solve the myriad challenges that stand in the way of seamless data exchange—neither within an organization nor with outside entities. Consider the following reasons why moving to a single EHR will not remove all barriers to achieving true interoperability in healthcare:
1. Associated small practices are unlikely to change
Moving to a single system does indeed make information easier to share from a technical perspective. The EHR market has some big players, particularly in the hospital market. So, expanding use of Epic or Cerner within a hospital system makes sense from an IT and change management perspective. But among physician practices of 1-3 doctors, 14.40% use Epic, 10.56% use eCW, and after that, market share drops to single digits. CMS reports that 30% of physicians use systems from small vendors or ones that are self-developed.
And these smaller practices are least likely to change and least able to afford to do so. Trying to solve the issue of interoperability by making everyone move to a single EHR across a network of both employed and associated physicians is simply not practical given the cost and disruption it would cause.
2. There are several ways to code within the same EHR
Health systems that have the same EHR may be using different code systems or medical annotations for the same clinical concept. For example, consider “Glycohemoglobin” or “Glycolated Hgba1c” versus “HbA1c” and its associated LOINC code 4548-4. A business partner recently reported that during a consulting engagement with a hospital system that had implemented a single EHR, the CMIO reported to her that he had identified 18 different codes to document a mammogram. Her analysis uncovered 123 different codes within that same health system.
3. Different providers document differently
How individual users document clinical care adds even more variation. One clinician might document an entire episode including diagnosis and treatment by typing a narrative into a note. Another might dutifully input information into structured data fields or select from a pulldown menu displaying discrete choices. Even if structural interoperability ensued (that is, the information could be interpreted at the data field level), if a field is empty, no relevant history is transferred.
4. Semantic interoperability remains elusive
Just because I can open a document doesn’t mean I – or a software system – can understand it, as there is more to exchanging knowledge than simply being able to open a file. Although human beings can often read between the lines, software systems have limited or no ability for interpretation. Interoperability is limited by common issues in clinical documentation.
These include the use of local versus standard terminology libraries; entering free text instead of capturing data in structured fields; and mistakes such as using the wrong unit of measure such as a percentage instead of a count. These variations that limit downstream use by both software and human users all need to be accounted for in our interoperability strategies. This is a fundamental issue that moving to one EHR won’t solve.
5. The entire ecosystem cannot be overlooked
Another consideration is that interoperability solutions should support healthcare processes beyond the single-use case of care coordination among clinicians. This requires accepting data from pharmacy systems and other parts of the supply chain, and communicating information captured at the point of care with disease registries, clinical trial systems and more. Adopting a single EHR will not help with interoperability across this broader ecosystem of systems and applications.
Physicians will continue to document as they will, and different institutions and individuals will pick the software they want, as well they should. As healthcare IT vendors and service providers, it’s our job to instead adopt open standards that seamlessly translate from one system to the next. Homogeneity is not a long-term solution to fluidity and flexibility in the exchange of clinical information across the healthcare ecosystem. Continual work behind the scenes to liberate data from proprietary systems – not lock it down – is what we owe healthcare stakeholders.
About Kimberly Howland, Ph.D.
Kimberly Howland, Ph.D., is Chief Product Officer at Diameter Health, the leader in normalization, cleansing, deduplication and enrichment of clinical data from across the care continuum.
