During the reflection on the allocation of intensive care resources in the context of COVID- 19, the question received more attention whether a person’s current treatment should be interrupted in favour of a third person with a better prognosis, if the existing resources did not allow treatment of both (so-called progression triage). The present article attempts to make suggestions for ethically responsible treatment termination under conditions of acute resource scarcity by means of a differentiated analysis of the possible types of action and a problematisation of the medical indicators of physician’s judgement.
Using machine learning (ML) technologies holds numerous potentials for improving medical practice and the treatment of patients. In particular, this is to be achieved by providing more or less automatised decision recommendations. However, some hopes for ML-based systems seem to be at least partially compromised in case of intransparent generation of such recommendations. This has significant consequences for the doctor- patient relationship and can severely limit forms of shared decision-making between patients and doctors. Thus, a successful implementation of ML-supported technologies will be measured substantially by its successful integration into the communication between doctors and patients.