Algorithms are typically seen as tools to be used by clinicians to make decisions for patient care. However, machine learning may be able to find latent patterns that may be relevant for clinical decision making above and beyond the single risk prediction. Trust affects both initial adoption and continued use of any technology, and is particularly relevant in high-stakes settings such as healthcare. The reasoning behind generated predictions could could contribute to shared decision-making; however, this is predicated on trust in the algorithm to provide something useful. In order to understand this, the role of expectation violation, degree of algorithmic transparency, and impact of hierarchical decision-making should be explored.