Unstructured data in healthcare: The missing link to interoperability
While achieving interoperability between disparate IT systems remains a priority among healthcare providers, integrating new technology with legacy systems and accessing unstructured data throughout the healthcare ecosystem continues to be a challenge.
The recent HIMSS study, Connected care and the state of interoperability, which surveyed approximately 118 clinical and IT leaders from healthcare providers across the United States, found data that detailed this pain point quite clearly.
According to the study’s results, most providers felt they were making progress when it came to establishing an interoperable health IT framework. For example, 75% of respondents ranked their interoperability progress as excellent or good.
Respondents were asked to further rate their progress based on four different “levels” of interoperability that ascend based on complexity. Once again, the results here are promising.
Respondents ranked their organizations as being very successful or somewhat successful at achieving the following levels of interoperability.
- Foundational (93%): Allows data exchange from one system to be received by another, but does not require the ability to interpret data
- Structural (95%): Ensures data exchanges between IT systems can be interpreted at the data field level
- Semantic (82%): Provides interoperability at the highest level at the data element level, enabling the ability of two or more systems to exchange and use information
- Organizational (75%): Includes clear policy, social and organizational components to facilitate communication and use of data within and between organizations and individuals
Legacy systems, unstructured content and interoperability challenges
Based on these figures, one might think the healthcare industry figured out how to achieve true interoperability throughout the care continuum. However, the data shows key obstacles remain. One is integrating new technology solutions with legacy systems. Of the respondents, 59% noted the challenge associated with connecting new standards-based technology systems to legacy solutions that don’t have the capacity to support these new standards.
Another is managing unstructured data and content throughout the healthcare ecosystem. This includes identifying this information as well as making it easily accessible to key stakeholders throughout the healthcare enterprise. This challenge was identified by 53% of respondents as a key obstacle to true interoperability. Even more concerning, respondents report 73% of the unstructured patient data and content in their organizations is inaccessible by key clinical stakeholders for review and analysis.
Why unstructured data in healthcare matters
Unstructured data in healthcare is a critical problem because it is a key part of a patient’s overall medical history. Examples of unstructured data include things like clinical documents and charts, clinical narratives, lab results, consents, referral documents and a wealth of medical images — even images from digital cameras, smartphones and scopes.
If this vital unstructured information is not included in the core clinical systems healthcare providers use every day, then the patient picture is woefully incomplete. In fact, industry estimates from analysts such as Gartner and IDC state that, even today, nearly 80% of the information that exists on a patient is in an unstructured format.
The providers themselves seem aware of the problems this unstructured data causes, especially when it comes to medical images. For example, 90% of the HIMSS study respondents stated it was critical or very important for clinicians to have access to patient images at the point of care. However, they also stated an average of 18% of their images are captured offline (e.g., encounter-based, via mobile device, etc.). And 65% of respondents cited a direct integration with their radiology picture archiving and communication system (PACS) was their primary means of integrating medical images with their EMR. This leaves a potentially vast array of non-radiology images inaccessible from this core clinical system.
When unstructured content and images aren’t included in the IT systems that are central to interoperable infrastructures, then these interoperability efforts are incomplete.
Harnessing unstructured content for true health IT interoperability
Considering the importance of unstructured content in creating a comprehensive patient record, how can this information be harnessed to ensure it’s included in interoperability initiatives? While definitely a complex undertaking, the process can be boiled down to the following three steps:
First, you need to identify where your valuable unstructured data resides. Often this information exists in multiple departmental content management systems, third-party cloud archives, paper medical documents and charts, faxes, email, specialty imaging systems, isolated modality stations, mobile devices, physical storage media and more.
Next, steps need to be taken to consolidate this data as much as possible. Consolidate electronic documents from existing repositories and scan, index and archive paper documents into a single repository. For medical imaging, enterprise imaging solutions such as a vendor neutral archive, universal image viewer and image connectivity tools can be implemented to complement or replace some PACS. This approach enables centralized management, access and sharing of all types of images — regardless of format and source. It also eliminates vendor lock and block that often accompanies proprietary systems, giving organizations all-encompassing ownership and control of their imaging data. These content services and enterprise imaging repositories can be consolidated further by establishing a single enterprise platform with a shared storage and presentation (i.e., viewing) layer.
Finally, these consolidated repositories need to be integrated with core clinical systems such as the EMR. Do this by using accepted industry standards such as HL7, Fast Healthcare Interoperability Resources (FHIR), DICOM, Cross Document Sharing (XDS) and more. The goal is to present unstructured and structured patient data in a single view that clinicians can access within their workflow via the EMR throughout the enterprise and beyond.
If done well, you’ll take interoperability to the next level by sharing complete patient information for improved care decisions.
This post was originally published on himssconference.org.