The materials management group at any hospital is responsible for ensuring that their inventory policies provide a good service in delivering products. The group also needs to be aware of their own costs for distribution, in terms of the frequency of delivery. With future changes to the hospital's infrastructure, function and size, the group require that these policies are reviewed and prepared for the changes taking place. For this, inventory policy models need to be built and used to anticipate the effect of these changes. Due to the importance of many products, high service levels are essential, yet there are often space and delivery constraints, limiting the amount of stock which can be held and delivered at each location. We present in this paper a new constraint-based model for determining optimal stock levels for all products at a storage location, with restrictions on space, delivery and criticality of items taken into account. We validate this model on sterile and bulk items in a real-life setting of an intensive care unit within Cork University Hospital, Ireland.