Unlocking Precision Medicine: The Need for Genomic Health Records

As the era of precision medicine begins, digital systems must be harnessed to manage the transformation. Enter the Genomic Health Record (GHR), an Electronic Health Record (EHR) designed for genomic integration.

The healthcare landscape is undergoing rapid transformation as the integration of clinical and genomic data begins, paving the way for an era of precision medicine. At the forefront of this transformation stands the Genomic Health Record (GHR), an Electronic Health Record (EHR) designed for genomic integration, envisioned by Marc S. Williams in 'The Genomic Health Record: Current Status and Vision for the Future' (2019). Today, we’ll be diving into the challenges and limitations of current EHRs and explore the value that GHR systems bring, from both a technology and human perspective.

As genomic testing becomes more mainstream the volume of relevant data will continue to increase, but several factors complicate straightforward, structured integration of genomic data in EHRs that would ultimately provide direct insights to clinicians.

Why Do Genomic Data Belong in (or linked to) EHRs?

The Clinician's Perspective: Unleashing the Power of Genomic Insights

For clinicians, improved availability of genomic data represents a transformative leap in patient care. The incorporation of genetic test results can equip clinicians with a detailed view of a patient's genetic makeup, enabling more informed diagnosis and treatment decisions. Combining clinical data with genetic test results empowers clinicians to tailor interventions with extreme precision, including the foresight of drug-gene interactions, early detection of treatable conditions, and improved insight into prognosis. This not only streamlines decision-making but also enhances the efficiency of clinical workflows, ultimately holding the power to improve patient outcomes. Additionally, incorporating genetic information into the EHR reduces barriers to sharing data with additional providers within the patient’s circle of care, ensuring coordinated care.

The Patient's Perspective: Empowering Informed Healthcare Choices

With both patient portals and genomic data linked to the EHR, patients can access their genetic information independent of their healthcare provider. The continuous update of interpretations ensures that patients can access the most current and relevant information, fostering a sense of control and involvement in their healthcare decisions. With improving data portability allowing patients’ data to travel with them through their lives, the insights derived from genomic testing add measurable value to the overall picture of a patient’s current and future health.

Challenges in Current Electronic Health Record Systems (EHRs)

EHRs have opened the door to better patient care by putting a timeline of the patient’s journey at the fingertips of healthcare providers. No longer do physical charts have to be manually reviewed. The nature of the structured data within electronic health systems lends itself to layers of decision support tools aimed at aggregating meaningful data and trends to assist patient care. From allergy alerts to workflow guides, to the identification of similar cases across a patient population, EHRs have revolutionized the way that providers can access patient data. As genomic testing becomes more mainstream the volume of relevant data will continue to increase, but several factors complicate straightforward, structured integration of genomic data in EHRs that would ultimately provide direct insights to clinicians. These factors include challenges with:

• how genomic data flows from laboratory information (management) systems (LIS/LIMS) to EHRs
• the structure of genomic data
• which data should be stored
• how EHR systems can add insights to aid in the interpretation of results and streamline care delivery

Where Does Genetic Data Come From Anyway?

To understand the challenges healthcare is facing when attempting to integrate genomic data into EHR systems, it’s important to realize how test results are produced. A standard sequencing pipeline starts with generating a raw sequencing read aligned against a human reference genome. This file is then analyzed to identify variants, producing a format called a VCF (variant call format) file. Then, specialists at the lab compile their interpretations based on the VCF and produce a report, typically a text file or PDF (Walton et al., 2019). During each stage of this process, some data are lost (Walton et al., 2022), and so the results that reach the clinician are a filtered subset of the original dataset.

While this data loss is required to allow for interpretation by providers without a specialty in genomics and to allow for scalable storage (raw results are quite large, posing storage issues across large populations), it also has some drawbacks. The nature of the rapidly evolving genomics field means that reanalysis happens regularly, so the raw data must be routinely reviewed to generate new interpretations (Grebe et al., 2020). Because this data is not accessible from the EHR itself, there may be challenges if the raw data needs to be retrieved and sent to a different lab for analysis.

Once the lab has completed its analysis, the report is sent to the provider, either via an EHR system or another method such as a text file, HL7 message, or PDF, typically housing the interpreted results as a block of unstructured text, alongside a minimal subset of structured elements such as the testing facility name or ID, the test performed, date, identifiers, etc. This report is then stored as a snapshot that can only be superseded by receiving an amended version of the document from the lab.

With this understanding, it’s natural to wonder why labs are not simply sending reports in a more structured format that could aid long-term healthcare decisions, but this is only one of several factors that stand in the way of deep genomic data integration within EHRs.

The Need for Discrete Genomic Data

The absence of phenotypic and genomic data standards impacts not only the format of the file but also the interpretation of the data itself. With unstructured data, not only is it very difficult to leverage clinical decision support (CDS) models, but it also reduces the ways in which systems can simplify genomic data for providers, especially those in non-genomic specialties, to understand and gain context on what they’re reviewing.

Without the presence of structured data, it becomes nearly impossible to link contextual information to genomic concepts, e.g. displaying information around genes or types of tests. Since variants are recorded as unstructured text, decision support models have difficulty interpreting them to assist in, for example, warning of potential known drug-gene interactions when a provider is looking to prescribe medication (Lau-Min et al., 2020).

Lack of Standards in Genomic Test Reports

One of the challenges with genomic data is the absence of standardized reporting and interpretation practices across laboratories. Analytical pipelines, testing methodologies, data interpretations, and reporting lack uniform standards, making interpreting genomic results complex and variable. The multitude of tests and panels, coupled with discrepancies in how variants are reported, makes it difficult for clinicians to compare results over time (Carter et al., 2022) or implement a data model that leverages standardized data ready for integration into clinical systems.

Because different testing modalities have unique advantages and limitations, providers need to know not only the results as interpreted by the lab, but also what type of test was performed, the most current status of the test, and additional tests completed for the patient. Filtering, comparing, and displaying variant data longitudinally pose additional challenges, hindering the ability to track changes over time (Carter et al., 2022).

The Challenges of Integrating with Laboratory Information Systems

Interoperability challenges persist in the current EHR landscape, affecting the seamless exchange of standardized, coded data between EHRs and laboratories. Limited granularity in coding systems like LOINC (e.g. broad groupings of testing modalities and the lack of hierarchy) and the reliance on HL7v2 messages, a flat file structure, within Laboratory Information Systems (LIS) exacerbate difficulties in exchanging complex genomic data (Carter et al., 2022).

Using the Data: Clinical Decision Support

Clinical decision support (CDS) tools are rapidly being implemented to assist clinicians in providing care, giving them immediate access to the exact data they need and guidance on how to proceed. The pressing need for discrete, or structured, variant standards becomes especially evident in the context of pharmacogenomic results, where healthcare providers would look to a system to recommend the best drug for treatment, or for guidance on potential adverse reactions based on the patient’s sequencing results. Tools that provide guidance on drug-gene interactions require precise, machine-readable genomic data so that accurate decision support can be accessed on demand. Integrating genomic data with decision support systems could guide healthcare providers in making informed decisions, linking genetic insights to patient care.

The Advocacy for Genomic Health Records

In light of the profound benefits for both clinicians and patients, access to meaningful genomic data for healthcare providers needs to be integrated into their workflows. The identified obstacles underscore the need for a dedicated system, such as a GHR, that seamlessly integrates clinical and genomic data, providing a comprehensive solution for precision medicine.

PhenoTips is continuously advancing as a promising solution to address the highlighted challenges. By offering a standardized and interoperable platform for integrating clinical and genomic data, PhenoTips has the potential to revolutionize genetic decision-making in healthcare. The benefits of precision medicine are extensive and will require healthcare providers and organizations to embrace and implement methods of storing these valuable datasets in a structured manner as a fundamental component of modern healthcare practices.

At PhenoTips we are always looking to improve the ways that we can deliver value to our users through structured data, passive decision support, and by championing various interoperability and data standardization efforts around the world.

As we stand on the precipice of a transformative era in healthcare, the integration of clinical and genomic data through a Genomic Health Record is not just a technological advancement but a fundamental requirement to fully empower data-driven decision making in the delivery of care. The potential benefits for precision medicine and patient care are too substantial to be ignored, propelling us toward a future where healthcare is truly personalized and optimized for individual needs.