Reference Data Management Solution
HealthAugust 20, 2021

Taming the data tsunami: 5 clinical terminology tactics for success

We had a delightful time presenting Taming the Data Tsunami: 5 Terminology Tactics back in June. Sarah Bryan, director of product management for Health Language, and I delivered the presentation from my toasty Colorado home as our Denver offices are still closed. About 15 minutes into the presentation there was an unexpected torrential thunderstorm leaving us running around shutting windows and hoping our audience could still hear us over the thunder. The closed windows then amplified the heat about a degree each minute - we can’t wait to get back to predictability of office acoustics, temperatures, and internet stability for future webinars!

Despite the mayhem, I think we managed to tell our story of what’s driving the massive increase in healthcare data, why unlocking the wealth of information in the often cryptic and complex data is essential for value-based care as well as achieving the “quadruple aim.” As an informaticist, I wish clinical data were nicely structured like claims data. As a member of both HL7 and AMIA, I’m intrigued by the promise of interoperability standards and technologies like Natural Language Processing (NLP) to help bring structure to clinical data sources. As a patient, a caretaker, and a member of the HIT community, I’m passionate about technology that can help clinicians provide optimal care for patients, while at the same time eliminate the burden of administrative duties. These are difficult and challenging topics that sometimes seem to have no answer.

Addressing the challenge: Interoperability and data management in healthcare

Even still, I'm optimistic about the solutions on the horizon to address these difficult challenges. Interoperability mandates and standards have laid the framework for sharing data. It’s exciting to see vendors stepping up to make the promise a reality. We are working together with several data aggregation and master data management (MDM) partners to bring a holistic solution to market that tackles these very complex problems. Stay tuned for more details.

At Wolters Kluwer, Health Language, we play a small, but essential role in addressing these challenges. Healthcare organizations need an overarching data governance and data management strategy that encompasses a broader interoperability strategy, tightly coupled with risk management and compliance. Having a foundational strategy for managing the reference data that all healthcare systems rely on will make everything else a little easier and will ensure that you have the highest quality and enriched patient data for accurate analytics, meaningful data sharing, and other high priority strategic initiatives.

Here are a few of the most impactful terminology tactics discussed in the webinar, and will unpack each tactic in greater detail in a series of Expert Insight articles. 

Tactic #1: Establish a single source of truth for healthcare terminologies

At Health Language, our internal joke is: “What do you mean we don’t have standards in healthcare in the U.S.? We have nearly 200 of them.” Inevitably, it triggers great philosophical discussions about what a healthcare terminology really is. And, how is a terminology different than an ontology, or a classification? And, which one is LOINC®, again?

The reality is that there are hundreds of terminologies in healthcare, and they are always changing (even more rapidly since COVID). Shifts in value-based care and greater realizations about the importance of things like, social determinants of health (SDoH), telehealth, remote monitoring, and more are all contributing to the volatility of reference terminologies.

Some important questions to ask yourself: Who in your organization needs the most up to date codes? If you are a vendor, how important is it to have the latest codes? What happens if you try to run a report on how many of your members/patients/cohorts have had COVID or have been vaccinated, if the codes are out of date? It makes me uncomfortable just thinking about it.

That’s part of this tactic – maintaining a single source of truth for all code sets in healthcare so you know your data and analytics are accurate and up to date. Another part is having that single source for your entire enterprise, especially with payers, large analytics vendors, providers, and all the healthcare organizations growing rapidly either organically or through mergers and acquisitions or system consolidation. Allowing your organization to have multiple sources of truth for this reference data can result in different versions of the same data. We’ll go into more detail in a following article, but this is bad news and can mean you’re losing money just from not having current reference data.

Tactic #2: Validate and enrich clinical data with timely terminologies

There’s no requirement from the interoperability mandates to make sure the codes being sent through your FHIR messages are valid. That description in the CCD message you received from another healthcare entity may or may not be the actual description for the code. Perhaps the most difficult data type to interpret is the unstructured data – the labs, radiology reports, clinical notes, etc. It can be so overwhelming, time consuming, and expensive to review, that many organizations just give up and focus on the structured data. However, the true patient story, nor the population health status cannot be told or determined by structured claims data alone. You must leverage ALL of the data coming your way and in order to do so, it must be parsed, codified, and organized. A terminology server (solution) can help with the variety of data types that you will encounter. Below are some examples of how a terminology server can help:

  • Verify code validity from the structured data
  • Verify code/term matches from structured or semi-structured data
  • Map, or normalize non-codified, semi-structured, or structured data to reference terminologies
  • Realize value from unstructured text through clinically trained NLP tools that also codify clinical entities to reference terminologies

Tactic #3: Leverage purpose-built data sets to address your specific use case

Some terminologies are used in almost every use case in health care. SNOMED, for example, is a rich ontology loaded with multiple domains of clinically relevant concepts and is leveraged in most data aggregation and analytics use cases. But many use cases also have unique terminology requirements.

For example: did you know Health Language offers a CPT® to LOINC mapping for linking lab orders billed with a CPT code to the lab result codified in LOINC? As a payer that’s starting or continuing to receive lab results, you can use the mapping to make sure you got the lab results from a lab order you paid for. These lab results are key to understanding quality of care and enabling you to appropriately correct course prior to HEDIS reporting.

Do you need to document both the topology and morphology/behavior codes found in ICD-O3? Health Language has specialized oncology terms for documenting both in the problem list and connecting that to the ICD-10, SNOMED CT, and ICD-O3 codes. What about sensitivity codes for masking sensitive data or consumer-friendly descriptions for translating complex medical jargon into language your patients and members can understand?

We’ll dive into much more detail on the form and function of these purpose-built Health Language proprietary data sets in an upcoming article.

Tactic #4: Use a healthcare terminology management platform

Still using spreadsheets to manage your terminology? It's such a manually intensive process and how do you ensure high data quality and accuracy? If you are authoring med policies, or value sets, or custom terminologies or code sets, or extending standards like SNOMED or RxNorm, you really should be using a terminology platform designed to do these tasks, and has all of the standards available and updated in a timely fashion. Doing this ensures all codes are correct and up to date, which streamlines authoring workflows, and ensures error free data for use downstream. This is a complex subject which we will dive in deeper in an upcoming article.

Tactic #5: Begin with a terminology management strategy

Start with the highest priority use cases for your organization. Are you ramping up a risk adjustment program? Transitioning to alternative payment models? Is telemedicine a good option for you? How well positioned are you to really leverage the data that will be coming your way, if not now, in the near future? Each of these strategic priorities has its own terminology needs, so working with a vendor such as Health Language can help you to determine what is right for your organization now and down the road.

Speaking of interoperability, we had the distinct privilege of having two expert panelists on this webinar to talk about their experience and knowledge of the interoperability standard FHIR. Watch the webinar on demand to hear them share their expert insights.

Overall, it felt like we covered a lot during the webinar but still have so much more to cover. We hope that you tune into this article series and encourage you to engage with us on LinkedIn or directly about these important topics. We love talking about terminology and are here to serve you in your journey to getting the most out of clinical data, analytics initiatives, interoperability, and data management.

Cheryl Mason
Director, Content and Informatics, Health Language
As the Director of Content and Informatics, Cheryl supports the company’s Health Language solutions leading a team of subject matter experts at that specialize in data quality. Together, they consult with clients across the health care spectrum regarding standardized terminologies, data governance, data normalization, and risk mitigation strategies.
Health Language Reference Data Management
Centralize and manage your reference data for clinical, claim, and business data to enable interoperability, drive compliance, and improve operational efficiencies.
Back To Top