This paper addresses the Patient Admission Scheduling (PAS) problem. The PAS problem deals with assigning elective patients to beds, satisfying a number of soft and hard constraints. The problem can be seen as part of the functions of hospital management at an operational level. There exists a small number of dierent variants on this problem. We propose an optimization-based heuristic building on branch-and-bound, column generation, and dynamic constraint aggregation for one of the variants. We achieve tighter bounds than previously reported in the literature, and in addition we are able to produce new best solutions for ve out of six instances from a publicly available repository.