Dynamic simulation modeling of ICU bed availability

Abstract

The intensive care unit accounts for nearly 30% of all inpatient expenditures while representing only 8% of the patient population. All healthcare systems must balance the need for access and availability of intensive care unit beds (ICU) versus excess capacity that wastes increasingly limited healthcare resources including bed space and personnel. The Cincinnati VA Medical Center is an acute care, university affiliated 220-bed facility serving eligible veterans with medical, surgical, neurological and psychiatric care needs. ICU beds are unavailable nearly one third of the time, eliminating new ICU admissions, and requiring diversion of ambulance traffic. Diverting ambulance traffic adversely impacts patient satisfaction and community perception of quality of care delivered at this center. Phased construction to relieve the problem was planned, including additional telemetry beds, move of ventilator dependent patients out of the ICU to a Respiratory Care Unit (Tele/RCU), and development of ICU swing beds in the emergency room area (Heart ER). We assessed the likelihood that the planned changes would result in the desired outcomes. A computer model representing medical bed utilization at this facility was developed using dynamic simulation software (Arena). This model analyzed the flow of patients through the ICU, telemetry and medical floor beds under current bed allocation. The model was then used to evaluate the effects of the planned phased construction. The model demonstrated improved availability of ICU beds with the addition of the telemetry and respiratory care unit beds. Resolving ICU access problems required addition of Heart ER beds. Unexpectedly, increased ICU bed availability resulted in increased telemetry and medical floor bed utilization downstream and increased length of stay on the medical service as the proportion of post-ICU patients increased on the floors.