Dr. Chepenik is with the Department of Psychiatry and the Department of Emergency Medicine and Dr. Pinker is with the School of Management, Yale University, New Haven, Connecticut (e-mail: lara.chepenik@yale.edu ).
OBJECTIVE: The study illustrates the use of approaches based on queuing theory to understand bottlenecks in patient flow. A queuing simulation was used to predict the potential benefits of additional clinical personnel on patient flow through a psychiatric emergency service (PES).
METHODS: A discrete-event simulation model was calibrated on the basis of two months of data collected in a PES. This model examined the effects of adding between.5 (half-time) and three additional providers to the 8 a.m. to 4 p.m. shift.
RESULTS: The model showed that an addition of a half-time clinician produced the biggest change in patient flow metrics. Average length of stay was predicted to drop from 38.1 hours to 33.2 hours for patients who were awaiting hospitalization and from 13.7 to 9.0 hours for patients who were eventually discharged. The number of patients waiting to see a provider decreased by two-thirds between 8 a.m. and 4 p.m., and it decreased by one-half during the rest of the day, even though the number of clinical staff remained the same. Adding more providers failed to reduce these numbers much further. Adding more than a half-time provider also failed to significantly reduce boarding (remaining in the PES while awaiting hospitalization).
CONCLUSIONS: Queuing simulation predicted a dramatic benefit to patient flow with a fairly small addition in clinician time, a benefit that persisted beyond the time during which the additional staff was on duty, especially when this staff was added during a period of high demand.