Kidney Week

Abstract: PUB087

Real-World Data Acquisition Methodology for Studying Cardiovascular Functional Capacity in CKD

Session Information

• Publication Only

Category: Augmented Intelligence, Digital Health, and Data Science

• 300 Augmented Intelligence, Digital Health, and Data Science

Authors

• Groninger, Nolan, Indiana University School of Medicine, Indianapolis, Indiana, United States

Nolan Groninger

• Samanani, Nikita Firozali, Indiana University School of Medicine, Indianapolis, Indiana, United States

Nikita Firozali Samanani, MS

• Altherr, Cody A., Indiana Center for Musculoskeletal Health, Indianapolis, Indiana, United States

Cody A. Altherr, MS

• Narayanan, Gayatri, Indiana University School of Medicine, Indianapolis, Indiana, United States

Gayatri Narayanan, MS

• Moe, Sharon M., Indiana University School of Medicine, Indianapolis, Indiana, United States

Sharon M. Moe, MD, FASN

• Lim, Kenneth, Indiana University School of Medicine, Indianapolis, Indiana, United States

Kenneth Lim, MD, PhD, FASN

Background

Cardiopulmonary exercise testing (CPET) is a diagnostic tool that generates hundreds of data points from ventilatory, cardiac and hemodynamic assessments during graded exercise. Unfortunately, there currently exists a paucity of epidemiologic datasets incorporating CPET data in patients with chronic kidney disease (CKD). Translational bottlenecks also exist to obtain and process the data for research purposes. Herein, we developed a methodology to create a novel centralized “CPET data hub” for capturing, automating data transfer and creating data linkages with important clinical information.

Methods

At Indiana University Health, CPET data is acquired by a specialized software (Breeze) that pushes a summary image report onto our electronic medical record system. To address this, we created a unique CPET data hub—a centralized data acquisition and processing hub enabling real-time capture of CPET output (linked to our hospital SQL server). The CPET data hub houses a digital platform (involving the development of algorithms via data mining methods and query modules) to process and calculate computed variables uniformly. CPET data on the hub is then linked to clinical information obtained through the Regenstrief Institute Informatics and the Indiana Network for Patient Care. CPET data is merged with clinical information including demographics, comorbidities, imaging, laboratory data and longitudinal outcome data.

Results

To date, our CPET data hub has captured comprehensive CPET data from n=1,953 adult ambulatory referral patients who underwent CPET testing. The cohort includes n=773 males and n=712 females. Approximately 37% have CKD stage 3; 4% have CKD stage 4; and 3% have end-stage renal disease or a history of kidney transplantation. The CPET data hub and linkage to important clinical outcomes (ongoing work) has led to the development of the “Cardiorespiratory Fitness in Patients with Chronic Kidney Disease in Indiana (FIT-INDY) cohort”.

Conclusion

As automation processes becomes readily available through the integration of computer science into medical research, it is necessary that we incorporate data acquisition methodologies into nephrology to assess real-world health outcomes for patients with CKD. The feasibility of epidemiologic studies incorporating CPET data is enabled by the CPET data hub.

Funding

• Other NIH Support