Artificial Intelligence in Medical Diagnosis
Artificial intelligence (AI) has revolutionized medical diagnosis, improving its accuracy and speed. However, its implementation poses legal challenges that require careful analysis to ensure a safe and fair application in healthcare.
Main Challenges
It is essential to consider that AI feeds on data to train its algorithms. If there are biases in the data, natural or intentional, this will be reflected in the diagnosis. Although patients are not necessarily aware of a physician’s biases, in the case of an AI it is important to control the data that feed the algorithms to ensure the highest possible quality and safety in the diagnosis.
Liability in Cases of Erroneous Diagnosis
The General Health Law and the Federal Civil Code regulate professional liability in health services in Mexico, but do not provide clarity on the use of AI in medical diagnoses. The lack of specific regulation leaves doubts as to whether software developers, physicians or institutions would be liable in case of misdiagnosis. At the international level, the European Commission has proposed the European Union AI Law, which classifies the use of AI in healthcare as “high risk” and requires AI developers to meet safety and quality control requirements.
Transparency and Explainability of Algorithms
NOM-004-SSA3-2012, which establishes requirements for electronic medical records, and the General Law on Transparency and Access to Public Information seek to ensure that patients have access to information about their diagnosis and treatment. However, when “black box” AI algorithms are used for diagnosis, patients’ right to understand medical decisions is limited. Regulations need to include specific requirements for algorithms to be explainable and accessible to professionals and patients. The WHOAI Ethics Guideline also recommends transparency in AI systems used in healthcare, allowing patients and professionals to have confidence in the results generated.
Informed Consent
The General Health Law and NOM-004-SSA3-2012 establish that patients should receive clear and accurate information about their treatments. In the context of AI, it is crucial that informed consent addresses the specific risks and benefits of using AI-based technology in diagnosis. This includes providing patients with detailed information about the use of AI in their diagnostics, as stated in the Declaration of Helsinki for Health Research, which emphasizes informed consent in innovative medical procedures.
Equity in Access to Technology
Mexico’ s General Health Law and the Law on the Rights of Health Services Users aim to guarantee equitable access to quality services for all patients. However, AI in medical diagnostics is mainly found in high-level private institutions. To reduce the access gap, the National Health Plan 2020-2024 for the integration of advanced technologies into the public health system, a policy that should include access to AI diagnostic tools for the general population.
Artificial intelligence offers great potential to transform medical diagnosis, but its implementation must be considered due to its ethical and legal challenges. At ECIJA Mexico, we guide our clients to comply with current health regulations and protect patients’ rights.
Health Area of ECIJA Mexico
socios.mexico@ecija.com
+52 55 5662 6840
www.eciija.com
