Elevating Clinical Research in Otolaryngology: Reg-ent’s New Research Platform
New data model is designed to enhance data quality, completeness, and abstraction.
Samuel Kim, BS, AAO-HNSF Reg-ent Registry Analyst, Jocelyn Meyer, MPH, AAO-HNSF Director, Reg-ent Registry, and Kristine Schulz, DrPH, MPH, AAO-HNSF Senior Director of Research and Quality
Reg-ent’s growing integration of advanced AI models has further enabled us to extract and standardize clinical information from previously untapped data sources such as physician clinical notes, radiology and lab results, patient-reported outcomes, and more. By applying natural language processing technologies, Reg-ent can now systematically convert free-text documentation into structured formats, significantly expanding the breadth and depth of the available data.
Introducing the New Reg-ent Research Platform
The prior data studio lacked the necessary detail to analyze patient populations and measure outcomes effectively. For instance, a researcher could previously verify that a thyroid-stimulating hormone test was performed by confirming the presence of a CPT code within the platform, but they could not access the actual lab result. Going forward, we will have the ability to retrieve lab values, along with a wide range of other valuable clinical data.
The enhancements achieved through our collaboration with MatrixPPA have paved the way for the launch of the Reg-ent Research Platform: a dedicated environment built for researchers seeking to generate insights from the largest and most comprehensive otolaryngology-specific real-world dataset in the United States.
In short, the new platform empowers researchers to explore critical questions, uncover novel insights, and make meaningful contributions to the advancement of otolaryngology through real-world data.
A New Data Model, Purpose-Built for Research Excellence
A major advancement accompanying the launch of the Research Platform is the creation of Reg-ent’s new custom data model and data dictionary, a critical step toward maximizing research utility and flexibility.
Under our previous model, data abstraction and housing were handled by two separate vendors. Clinical data was abstracted by one vendor, transferred to a second vendor, and curated into research datasets using a standard data model that Reg-ent did not control. As a result, data elements often had to be mapped into preexisting structures not optimized for otolaryngology-specific research. Valuable clinical details were sometimes lost or fragmented during these transitions, limiting the depth and completeness of available data for research applications.
Recognizing the need for a more researcher-centered approach, AAO-HNSF made the strategic decision to consolidate all data abstraction, curation, and housing under one vendor. By eliminating the need for data transfers between vendors, we have streamlined data handling processes, minimized information loss, and preserved the integrity of the original clinical data.
Most importantly, we worked collaboratively with MatrixPPA to design and implement a new data model purpose-built for the Reg-ent Registry. Compared to the previous vendor’s data dictionary, the updated version includes over twice the number of defined data elements. This model not only incorporates a wider array of clinical fields from EHR systems but also continues to expand to previously inaccessible domains, including:
- Audiograms and sleep study results
- Laboratory results
- Radiology results
- Oncology data
- Social determinants of health data
Moreover, the data model is flexible, allowing for future expansion as new types of clinical data become relevant to research priorities.
Plan your Next Research Project
We invite researchers, clinicians, and institutions to join Reg-ent in this new era of discovery and explore the opportunities enabled by our new data model. Together, we can continue to advance the science of otolaryngology and deliver meaningful improvements in patient outcomes across the nation.
