Agent-based simulation of nosocomial transmission in the medical intensive care unit

Patients receiving care in medical intensive care units (MICU) are at high risk for colonization or infection by pathogenic organisms acquired during their treatment. Hand washing compliance by health care providers is a major determinate of the rate of organism transmission in this setting. We created a computer simulation of the MICU at our affiliated public hospital using the C++ computer language and the Swarm simulation library. The simulation models the colonization of patients with a pathogenic organism and evaluates the effects of infection control techniques. The simulation accurately models the behavior and interaction of nurses and patients in the MICU. Using this application, we altered the hand washing behavior of the virtual nurses to evaluate R/sub 0/ (an epidemiologic measure of pathogen transmission). Our simulation indicates that a hand washing rate of > 60% results in an R/sub 0/ of less than 1. Given the assumptions of the model, this implies when a colonized patient is introduced into the MICU, if hand washing is maintained at a frequency greater than 60%, spread of the pathogen to other patients would extinguish over time and the organism would not become endemic in the MICU.