Kidney Week

Abstract: SA-PO022

Bridging the Future: Perspectives of a Multinational Group of Nephrologists on Artificial Intelligence (AI) in CKD Management

Session Information

• Augmented Intelligence, Large Language Models, and Digital Health
  October 26, 2024 | Location: Exhibit Hall, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Augmented Intelligence, Digital Health, and Data Science

• 300 Augmented Intelligence, Digital Health, and Data Science

Authors

• Silva, Eliana, Diaverum AB, Malmo, Skåne, Sweden

Eliana Silva, MSc

• Santos, Carla, Diaverum AB, Malmo, Skåne, Sweden

Carla Santos, MD, PhD, ScM

• Lucas, Carlos, Diaverum AB, Malmo, Skåne, Sweden

Carlos Lucas, MD

• Rodrigues, Luis, Diaverum AB, Malmo, Skåne, Sweden

Luis Rodrigues, MD

• Garrido, Jesus, Diaverum AB, Malmo, Skåne, Sweden

Jesus Garrido, MD

• Santos, Lidia, Diaverum AB, Malmo, Skåne, Sweden

Lidia Santos, MD

• Macario, Fernando Jose Gordinho Rocha M, Diaverum AB, Malmo, Skåne, Sweden

Fernando Jose Gordinho Rocha M Macario, DrMed

Background

Artificial intelligence (AI) techniques, particularly machine learning (ML) and deep learning, have made significant strides in predicting and diagnosing chronic kidney disease. This study aims to explore the perceptions of a multinational group of nephrologists regarding the application of AI in clinical practice. Understanding their views is crucial for the integration of AI in nephrology.

Methods

A prospective, observational study was conducted in March 2024, involving nephrologists from 17 countries across four continents, all affiliated with a large hemodialysis provider. The survey, validated by 6 nephrologists, comprised 8 technical questions using a 5-point Likert scale, with 1 corresponding to "totally disagree" and 5 to "totally agree." Demographic data were also collected. Results presented as mean±standard deviation or proportions, as appropriate. T-test was used for statistical analysis, with a p-value below 0.05 considered statistically significant.

Results

Among the 196 valid responses, 80% of nephrologists recognized the term "AI," and 65% were familiar with "ML." Nephrologists acknowledge the potential of AI in the future but currently rely more on traditional tools to support the clinical decision-making process (3.74±1.14 vs. 3.12±1.07, p<0.05). This reliance on traditional tools is more evident among nephrologists not familiar with the term “ML” (3.90±0.93 vs. 3.09±1.01, p<0.05). When asked about the advantages of AI and ML in supporting clinical decision-making, the highest score was attributed to “simplification of decision algorithms” (3.82±1.10). Interestingly, nephrologists without previous contact with AI-supported decision-making tools considered the role of AI less significant in “assisting in making decisions, allowing concentration on high-value activities”, compared to those with previous experience in this area (3.38±1.22 vs. 3.88±0.97, p<0.05).

Conclusion

Enhancing the understanding of AI can facilitate the integration of these technologies into nephrology practice, benefiting patient care. Given the perceived potential of AI in managing different types of CKD patients, efforts to promote AI literacy in the nephrology community may be fundamental for the adoption of these tools in regular clinical practice.