The nephrologist in the present and near future: between algorithms and autonomy

Area of Science:

• Nephrology
• Artificial Intelligence in Medicine
• Health Informatics

Background:

• Artificial intelligence (AI) is poised to significantly alter the field of nephrology over the next 20 years.
• The integration of AI tools aims to enhance patient care while maintaining the central role of the nephrologist.
• Current healthcare systems face challenges in managing complex kidney diseases and administrative burdens.

Purpose of the Study:

• To explore the prospective impact of AI on nephrology practice and the evolving role of nephrologists.
• To identify key AI technologies, including prediction models and large language models (LLMs), and their applications in kidney care.
• To outline the necessary competencies and ethical considerations for nephrologists in the age of AI.

Main Methods:

• Review of current and emerging AI applications in nephrology, including prediction models for acute kidney injury and chronic kidney disease.
• Analysis of multimodal tools integrating electronic health records, biomarkers, imaging, and wearable data for patient-level risk estimation.
• Examination of the capabilities and limitations of large language models (LLMs) as assistive tools for clinical tasks.

Main Results:

• AI-driven prediction models and multimodal tools will offer continuous, patient-specific risk assessments.
• Large language models (LLMs) show potential to reduce administrative workload through assistance with documentation, triage, and patient communication.
• AI implementation carries risks, including model hallucinations, bias, and variable accuracy, necessitating careful clinical oversight.

Conclusions:

• Nephrologists require new skills in AI model calibration, fairness assessment, communication, and ethical decision-making to critically evaluate algorithmic recommendations.
• Special attention is needed for vulnerable patient populations, including the elderly, frail, and those with multiple comorbidities.
• Effective AI integration for proactive, person-centered kidney care depends on clinician involvement in AI design, governance, evaluation, and supervision within learning healthcare systems.