Understanding health data in the context of the patient journey can support addressing specific challenges that can only be met with purpose-built data sets.
It's been six months since we did our first installment of this article series on healthcare terminology tactics! In that time, we have been very busy at Health Language working on bringing high value terminology to the healthcare ecosystem.
Previously, Sarah Bryan and I had the privilege of presenting a webinar titled Taming the data tsunami: five healthcare terminology tactics for success in the post-COVID ecosystem. Since then, we have been taking a deep dive into each of the five tactics in a series of articles, including linking clinical and claims data and creating a single source of truth. In this installment, I want to dig a little deeper into some very specific challenges that can only be met with purpose-built data sets.
Health data in the context of the patient journey
In the webinar, Sarah posed some specific questions as examples of what healthcare providers and health plans are trying to solve for with data:
- How do I find just the information regarding my patient’s COPD journey?
- How do I keep sensitive information from being shared inappropriately?
- How do I help my members understand their patient record?
- How do I understand a patient’s social determinants of health?
- How do I capture the right information for oncology?
- How do I combine both claims and clinical data for analysis?
- How do I cohort my patient population for meaningful population health analytics?
So much to talk about, and so little time, but let’s focus on four of these questions today. Reach out to me to dive deeper into the others, I’m always happy to talk terminology.
If you think about data in the context of the patient journey, a patient first sees the clinician in an office or inpatient setting and, ideally, that encounter is documented in an electronic health record (EHR). According to a survey conducted in 2019 by the CDC, 89.9% of office-based physicians were using an EHR, 72.3% were using a certified EHR. Looking deeper into this data reveals some interesting trends, I encourage you to look at.
One question I found particularly interesting was about the amount of time spent documenting solely for billing purposes, not clinical purposes, increasing the amount of time a clinician spends in the EHR. 41% of respondents, representing 80% of physician’s surveyed, agreed or strongly agreed with that statement. Wow, that’s significant! What if we could do something to help these clinicians same time by facilitating greater semantic interoperability?
1. How do I capture the right information for oncology?
In our webinar, Ninja tactics for curating, cleansing, & enriching clinical data for high value use cases, Mark Wozny from MD Anderson Cancer Center spoke about their Oncology Data Foundations commitment to standardizing the capture of oncology specific data elements at the point of care. We have been privileged to take this journey with them not only as their single source of truth but as a partner in creating a specific data set, based on the International Classification of Oncology (ICD-O-3).
Traditionally, problem lists contain terms that are mapped to SNOMED CT for interoperability and ICD-10 for billing. That’s great for many use cases, but oncology has some very specific needs around documenting histology and mapping to ICD-O-3. I will never forget when one of their oncologists told me that the difference in the specific histology of a cancer can make a monumental difference in the treatment and outcome for the patient, and that having it documented on the problem list saves him time and gets him to the correct treatment protocol quickly. Now that’s a goal I can sink my teeth into.
Perhaps as valuable is having the problem list terms based in and mapped to ICD-O-3 to assist in registry reporting. As some of you may know, gathering the right data for registry reporting is a time-consuming task. Capturing the ICD-O-3 code at the point of care, reduces that time and can increase the accuracy of reporting.
2. How do I keep sensitive information from being shared inappropriately?
Moving along in the patient (data) journey, often a medical record contains sensitive information that is protected by the HIPAA privacy law and 42 CFR part 2 (Confidentiality of Substance Use Disorder Patient Records). As the interoperability rulings that went into law in 2020 are implemented and the data tsunami continues to grow, the problem of masking sensitive information that a patient has not given consent to share grows, the harder this challenge will become.
It is imperative that your organization have a strategy to protect the information that you are required to protect, while sharing the data you are required to share. Health Language can help you segment that data through the use of our sensitivity codes, designed to mask data that is captured using many different terminologies, and kept up to date by our team of expert terminologists.
3. How do I help my members understand their patient record?
I think I need to make two more stops on the patient journey, one stop is the communication that a provider or payer is going to have with their patient or member. As a patient myself, and especially as an advocate for my family, I often use patient portals to understand the conditions and treatments that my family members have – with their permission of course. The industry has come a long way in making these portals user friendly, but the language in them can sure be frustrating to make sense of.
For instance, one of my family members had a test for sensory impairment performed at the neurologist’s office, but the medical record and explanation of benefits listed that procedure as, “QUANT SENSORY TEST&INTERPJ/XTR W/ VIBRJ STIMULI.” WHAT? I’m not sure anyone would know what that translates to. I’m in the terminology world myself, but without tapping into my Health Language tools, I wouldn’t have the slightest clue what that meant. Health Language has the ability to translate that medical, coding jargon into more consumer friendly language by using our consumer-friendly descriptors for Diagnosis and Procedures.
4. How can I use data for meaningful analytics?
Finally, the questions about analytics cannot be overlooked, especially in the age of value-based care. As the population in the US ages, chronic conditions become more and more complex, and the cost of healthcare rises, we simply must address population health and care management strategies using data driven principles. It is critically important to be able to identify the patients with the highest risk of poor outcomes and intervene early.
Much of the information needed to stratify patient populations for inclusion in programs that utilize telehealth and remote patient monitoring strategies is found in the unstructured text of the medical record. Many long and tedious hours are spent combing through records to identify important pieces of information that never make it onto a claim. Clinical natural language processing (cNLP) can help to find that information, and as importantly codify it to a standard terminology so that it can be leveraged in high value use cases. Once the information is codified, you can apply value sets to cohort patients into high-risk groups, identify what interventions have been performed, and close gaps in care.
This all requires that you have value sets that include all relevant terminologies and concepts related to a condition, as well as the relevant treatment protocols. Some of these value sets can be found in quality measures and other industry standard value sets, but those may not be up to date with the latest and greatest terminology, or they may not meet the needs of your specific program. A solid terminology management platform with a cohort management tool can save you time and improve the consistency of your data over time.
If you've made it this far, I have to assume that you are just as passionate about the quality of data currently found in the data tsunami as Sarah and I are. Please stay tuned for the fifth installment of this riveting series and reach out with questions, comments or requests for more information. We’d love to hear what’s on your mind in this crazy data world we are all living in.
