Welcome back! Today, we're taking a brief detour from our regularly scheduled content to dive into the nitty-gritty of lab data—how it’s structured, why it’s tricky to manage, and some tips to make it easier.
When we think about lab data, there are many use cases. This blog will focus on business analytics, how data is displayed to providers, how it shows up in a patient's portal, as well as how it is used in population health settings.
Understanding Lab Workflows and Data Structure
First, let’s break down the basics of how lab data is created and structured. Here is a general lab workflow and how each data element is related:
Orders: This is where everything starts. A provider orders a specific lab test.
Specimens: Samples (like blood or urine) are collected based on these orders.
Tests: These samples are tested in the lab. A specimen can have multiple tests!
Results: The data from these tests are recorded and often coded with LOINC (for lab tests) and SNOMED (for microbiology organisms). Tests can have multiple results!
Here are some other terms that are important in understanding lab data:
LOINC codes: Logical Observation Identifiers Names and Codes (LOINC): are used to identify lab tests. These codes ensure that tests are universally understood and standardized across different systems and organizations.
SNOMED codes: Systematized Nomenclature of Medicine (SNOMED): codes are used for more detailed clinical terms, including diagnoses and findings. In the lab context, SNOMED codes help standardize the reporting of microbiology results.
Reference ranges: Reference ranges are the values considered "normal" or "abnormal" for different populations. They provide context for interpreting lab results and can vary based on age, sex, and other factors.
This workflow creates data points along the way that we can use for reporting. Generally speaking, when people discuss "lab data" or "test results" they are referring specifically to the "results" outlined above.
The Challenges of Managing Lab Data
Variability in Methods and Units:
Different tests use various methods and units, complicating data standardization. For example, glucose levels might be reported in mg/dL or mmol/L, depending on the lab or instrumentation. This variability makes it difficult to compare results across different labs, especially when these differences in units lead to variation in reference ranges.
When managing this in a platform like Beaker, multiple records are needed for the same test. This can complicate how tests are used in analytics, how they appear in the chart, and ultimately how they are presented to patients and providers.
Microbiology Data:
Microbiology results can be particularly tricky because they include workflow specific qualitative data, like identifying bacteria, which needs detailed interpretation. Unlike straightforward numerical results, these qualitative results require specialized knowledge to interpret and report accurately. Tools like SNOMED help with standardization, but different organizations might follow unique workflows, creating unstructured, dissimilar data that is hard to use across different hospital organizations.
Qualitative Data:
Many lab results are reported as text values, which can lead to significant variation. In Beaker, you can implement measures to limit this variation, but inconsistencies may still arise. For example, one organization might report results as "Positive," while another might use "Detected." for the same test. Similarly, "Not Detected" can be reported differently, such as "Negative" or "Not Found." This complexity increases when additional details are included, like "Positive for XXX" or abbreviated forms like "Pos."
External Reference Labs:
Integrating data from outside labs can be challenging due to different reporting standards and formats. External labs might use different codes or units, making it hard to maintain consistency with in-house testing. These external result components are often for less common tests, making them time consuming and difficult to manage.
The Reporting Struggle
Here are the three most pressing issues related to reporting lab data:
Data Volume and Complexity:
Labs handle vast amounts of data daily, making it tough to keep everything consistent and accurate in Beaker. Each test result must be meticulously recorded and stored, creating a massive data management task for any organization. A single result component may need logic for various age groups and genders, as well as meticulous management of things like LOINC or SNOMED codes. Most organizations on Epic Beaker LIS will have tens of thousands of results components and thousands of tests that they need to manage.
Getting Results to Providers and Patients:
Result components need to be manually plugged into tools to display correctly to both providers and patients. Organizations must decide how to display results with different units, which results make sense to view together, and how these results get pulled into EHR tools.
Standardization Issues:
Even with standard codes like LOINC and SNOMED, differences in data entry and interpretation can result in inconsistencies in reporting. Variations in how data is recorded can lead to discrepancies, such as using different decimal places for specific result components or recording numeric values as text values in certain cases. For example, when doing a manual kit urinalysis, it isn't uncommon for some organizations to use a preset list of numeric values that can be resulted (stored as text) instead of numeric values.
Best Practices for Beaker Reporting:
To navigate these challenges, here are some practical tips:
Focus on Standardizing Common Orders:
Start with the top 100 most common orders. Standardizing these can help ensure accuracy and quick access. By focusing on the most frequently used tests, you actually take care of the majority of your order volume. Think about the 80/20 rule here: by getting 20% of your lab results configured correctly, you will hopefully cover 80% (or more!) of your volume.
Once you have fully configured and managed your top 100 orders you can then work on tests with the highest impact on patient care. Create governance groups to help prioritize and provide feedback to your Beaker team to organize and structure lab components correctly.
Get Feedback Often:
Creating governance groups allows you to get feedback from providers on how data is displayed and how it can be improved. Additionally, getting feedback from your reporting team is also beneficial as they will be the ones in the weeds creating reports. With the complexity of how Lab/Epic Beaker LIS data is structured, it is recommended to have a strong working relationship between your application and reporting teams.
Advance Your Team's Skills with Reporting Tools:
Understand the tools available in Epic and master them! Several robust dashboards and reports can be used out of the box but Epic also offers several customizable robust report templates and logic. SlicerDicer in Epic is also great tool for extracting insights on your organization's lab data. Getting your analyst team exposure to these tools will help them support your end users more confidently while also building a better understanding of how Beaker works.
Navigating the complexities of lab data is challenging, but manageable with the right strategies. By focusing on common orders, gathering feedback, and using Epic's advanced tools, you can streamline your lab analytics. These strategies not only improve data quality but also help improve patient care and operational effectiveness. Stay tuned for more insights on optimizing lab data with Epic Beaker LIS!
More about Kyle:
With 7 years of Beaker experience, I have been deeply involved in Beaker across 11 different healthcare organizations. My journey has taken me from entry-level roles to leadership positions, managing projects that span from initial implementation planning to post-live optimizations and ongoing maintenance. My expertise lies not only in the technical aspects of Beaker but also in leading teams to ensure successful a successful project.
While my experience has been rich with diverse challenges and learning opportunities, the insights I share in this article are drawn from my personal expertise and do not reflect the views or opinions of my employer or the healthcare organizations I have supported. They do not include or represent any proprietary information from Epic Systems Corporation. The content presented here is my own intellectual property, intended to guide and inform others in the healthcare IT community as they embark on their own Epic Beaker lab transformation journeys.
Comments