Determining Maternity Case Load by Means of a Poisson Process

While planning a new maternity teaching hospital, we encountered the problem of relating number of delivery rooms to expected case load. The flow of patients in the teaching and research sections of the hospital could be predicted on the basis of previous experience, but for an attached "sector unit" the problem was much more difficult. The unit was intended to serve a new function, and therefore no case load data existed. The existing maternity teaching hospital contains 85 beds. Most routine deliveries are excluded from this hospital and admitted elsewhere. A special unit of 45 beds for the professor of obstetrics is located some 2 miles away in the main university teaching hospital. It is intended to combine both these units in the new maternity teaching hospital. In addition, a special "sector unit" is planned. This will provide for the routine maternity needs of