ML for Predicting Incapacitated Patients’ Treatment Preferences

Think before you code.

ML for Predicting Incapacitated Patients’ Treatment Preferences

Lecture on the Use of Machine Learning for Predicting Incapacitated Patients’ Treatment Preferences (November 30, 2019, Munich)

Patients’ values and treatment preferences must be observed even if a patient is incapacitated. The corresponding measures provided by German civil law are, however, unsatisfactory: Most notably, surrogates predict patients’ treatment preferences only slightly better than chance. The recently proposed “Autonomy Algorithm” promises to improve the accuracy of such predictions and, thus, to support surrogate decision-makers: Machine learning-based predictive analytics could be trained on electronic health records and social media data in order to estimate the confidence of the prediction that a patient would make a specific treatment decision. Even though this algorithm is expected to enhance patient autonomy, ethical and legal questions arise from machine learning technology itself on one hand and its application within the context of healthcare decision-making on the other.
The lecture is part of TUM’s symposium on “New Ideas for Medicine” that will take place on November 29 and 30, 2019, in the Auditorium of the Faculty of Medicine, Biedersteiner Straße 29, 80802 München. For further information, please refer to the attachment.

Leave a Reply

Your email address will not be published. Required fields are marked *