Abstract: PUB052
Prevalence of CKD and Associated Factors in Adults in Panama (2022-2024) Using Artificial Intelligence (AI) for Data Analysis
Session Information
Category: Artificial Intelligence, Digital Health, and Data Science
- 300 Artificial Intelligence, Digital Health, and Data Science
Authors
- Courville, Karen, Caja de Seguro Social, Panama City, Panama
- Bustamante, Norman, Caja de Seguro Social, Panama City, Panama
- Lezcano, Jorge, Maxia Latam, Panama City, Panama
Background
Chronic Kidney Disease (CKD) represents a significant global health burden. Understanding its prevalence and associated risk factors is essential for developing effective public health strategies and clinical interventions. This study aims to estimate the prevalence of CKD and identify associated factors in the adult Panamanian population aged 18 years and older.
Methods
This observational study utilizes a large dataset of laboratory results from individuals aged 18 and above in Panama, spanning from 2022 to 2024. Artificial Intelligence (AI) algorithms were employed for the comprehensive review and analysis of these extensive laboratory data, specifically focusing on serum creatinine and albuminuria measurements from the MODULAB system. Estimated Glomerular Filtration Rate (eGFR) was calculated using the CKD-EPI 2021 equation, adjusted for sex and age. CKD progression was defined as a sustained eGFR less than 60 mL/min/year, confirmed by at least two measurements over three months. Chronicity of CKD was assessed using a Q0-Q4 scale based on KDIGO criteria. Patients were classified into KDIGO G (eGFR) and A (albuminuria) stages (G1-G5, A1-A3) to determine severity and progression risk. The Theil-Sen method was applied for robust regression to estimate eGFR change slopes, with 95% confidence intervals derived via bootstrapping. Finally, a KDIGO risk stratification matrix was applied to assign risk levels based on combined G and A stages.
Results
Preliminary analysis of the dataset, significantly facilitated by AI, reveals diverse eGFR values, chronicity classifications, and corresponding KDIGO risk stratifications across the studied population. The AI-driven approach efficiently processed a vast number of laboratory records, enabling detailed classification and risk assessment of individuals for CKD.
Conclusion
The application of AI in analyzing large-scale laboratory data is instrumental in accurately estimating the prevalence of CKD and identifying associated factors in Panama. This methodology offers an efficient approach for epidemiological studies, facilitating timely identification of at-risk populations and informing targeted public health interventions to mitigate the burden of CKD. The results underscore the potential of AI for enhanced clinical decision support and discovery research in nephrology.
Funding
- Government Support – Non-U.S.