Enhancing Longitudinal Data Analysis with Unstructured EHRs: A Case Study of Renal Function Evaluation in Rare...

Area of Science:

• Nephrology
• Medical Informatics
• Rare Diseases

Background:

• Electronic Health Records (EHRs) offer valuable longitudinal data for disease progression tracking, particularly for rare conditions like ciliopathies with chronic renal decline.
• Structured EHR data often lacks comprehensive patient history and external lab results, which are critical for accurate disease assessment.
• Clinical narratives within EHRs contain vital unstructured information that can augment structured datasets.

Purpose of the Study:

• To enrich structured EHR datasets with unstructured clinical text from rare disease patients.
• To evaluate the impact of this data enrichment on estimating chronic kidney disease progression in ciliopathy patients.
• To enhance the precision and reliability of longitudinal analyses in rare disease research.

Main Methods:

• Utilized natural language processing techniques to extract relevant information from unstructured clinical narratives.
• Integrated extracted data into existing structured EHR datasets.
• Employed linear mixed regression models to analyze estimated glomerular filtration (eGFR) rate trajectories over age before and after data enrichment.
• Quantified changes in patient eligibility, available measurements, follow-up duration, and model precision.

Main Results:

• Data enrichment increased the number of eligible patients for longitudinal analysis by 73.5%.
• The expansion of available measurements reached 189%, and the median follow-up duration extended from 3.2 to 6.6 years.
• Linear mixed regression analysis demonstrated a 30% reduction in standard errors for eGFR trajectories post-enrichment, indicating enhanced precision.

Conclusions:

• Enriching structured EHR data with unstructured clinical text significantly improves the scope and accuracy of longitudinal analyses in rare diseases.
• This approach enhances patient cohort identification and extends follow-up periods, crucial for understanding chronic disease progression in conditions like ciliopathies.
• EHR data enrichment is a valuable strategy for advancing rare disease research and improving the reliability of clinical insights.