A multistage model of hospital bed requirements.

This article presents a model for projecting future hospital bed requirements, based on clinical judgment and basic probability theory. Clinical judgment is used to define various categories of care, including a category for patients who are inappropriately hospitalized, for a large teaching hospital with a heavy indigent and psychiatric workload. Survey results and discharge abstract data are then used to calculate expected discharges and patient days for each clinical category. These expected discharges and patient days are converted into estimated bed requirements using a simple deterministic equation. Results of this multistage model are compared with the results obtained from exercising the simple deterministic equation alone. Because the multistage model removes patients from the hospital if they are deemed inappropriately placed, this model results in the projection of 5.1 percent fewer hospital beds than the simple deterministic equation alone.

Duke Scholars

Author Jane Frances Pendergast Biostatistics & Bioinformatics, Division of Biostatistics