question

I personally think that the Custom Arrivals option for patient arrivals provides the best representation and flexibility to achieve what you are after. The Custom Arrivals table allows you to specify as many time periods of any duration (with a start and end time) and then specify the type of patients and number of patients arriving in each of the periods. The type and quantity of patients arriving in a given period can be entered as a constant value or an expression involving a distribution if necessary; although I usually keep the PCI type a constant and only use a distribution to define the quantity.

The Custom Arrivals table does not need to be chronologically organized, so you are free to organize the rows according to PCI groupings to make it more user friendly. Even the time periods (ie rows) defined for a given PCI type don't need to be distinct and separate - in other words the time periods may overlap.

I rarely use any other arrival type to define unscheduled arrival patterns as I feel I can achieve the most accurate representation with the Custom Arrivals table. It also makes for a nice straightforward import from an Excel workbook containing the information. In the Excel workbook I may have several sheets making it easier to enter and update the information in a user friendly manner, but then have one sheet that is linked to the other sheets so it is built automatically, and it is this single sheet that I import into my model.