Estimating the risk of patients dying is arguably one of the most difficult and stressful challenges physicians face. This has been especially true in the midst of the global COVID-19 pandemic, with doctors around the world repeatedly confronted with difficult decisions. In the best of cases, they have been able to adjust treatments and save lives. In the worst case scenario, however, physicians have to allocate scarce beds and life-saving machines in intensive care units. An international team led by researchers at the Max Planck Institute for Intelligent System has now developed an algorithm and trained it with machine learning methods to help medical professionals with mortality predictions. The algorithm can also be trained to predict mortality risk for other diseases, and thus support physicians in decision-making processes.