Hospital unit demand forecasting tool

1. A method of forecasting a demand for a particular hospital unit comprising:

programming a computer readable medium with steps comprising:

logging a total number of beds in the particular hospital unit and available nursing slots to determine a capacity for the particular hospital unit;

analyzing data for patients scheduled to stay in the particular hospital unit to predict stochastic arrivals in order to estimate a total inflow;

predicting a length of stay of a patient in the particular hospital unit using a survival analysis based on physician orders to estimate a total outflow; and

executing an algorithm designed to use the capacity, total inflow, and total outflow to determine the demand for the particular hospital unit.

2. The method of claim 1 wherein estimating total inflow further comprises analyzing predictive data related to a time of the day for the demand for the particular hospital unit.

3. The method of claim 1 wherein estimating total inflow further comprises analyzing predictive data related to a season of a year for the demand for the particular hospital unit.

4. The method of claim 1 wherein estimating total outflow further comprises analyzing predictive data related to the ages of patients currently staying in the particular hospital unit.

5. The method of claim 1 wherein estimating total outflow further comprises analyzing predictive data related to a source of arrival of patients currently staying in the particular hospital unit.

6. The method of claim 1 wherein estimating total outflow further comprises analyzing predictive data related to a time of the day for determining the demand for the particular hospital unit.

7. The method of claim 1 further comprising determining the demand for the particular hospital in six hour intervals.

8. The method of claim 1 further comprising determining the demand for the particular hospital unit for a 72 hour period of time.

9. The method of claim 8 further comprising determining the demand for the particular hospital unit every six hours within the 72 hour period of time.

10. The method of claim 1 further comprising predicting the stochastic arrivals using a feedback mechanism whereby the probability of stochastic arrivals being admitted to the particular hospital unit is a function of difference between forecasted demand and available capacity.

11. The method of claim 1 further comprising using a Poisson regression to model the relationship between arrival counts and predictor variables from a stochastic source.

12. The method of claim 1 further comprising executing the survival analysis using one chosen from the group consisting of discrete-time logistic regression, semi-parametric hazard regression, and parametric unbiased, stable estimates.

13. The method of claim 1 further comprising grouping the physician orders as medication orders, breathing support orders, and feed orders.

14. The method of claim 1 further comprising using physiological measures from the patient as a predictor of length of stay.

15. The method of claim 1 further comprising basing the available nursing slots on predictive data generated every 12 hours.

16. The method of claim 15 further comprising updating the available nursing slots every three hours based on trigger events.

17. The method of claim 1 further comprising grouping the length of stay as a short stay of less than three days or a long stay of more than three days.

18. The method of claim 1 further comprising updating the demand for a particular hospital unit in real time.

19. The method of claim 1 further comprising defining the particular hospital unit as a Pediatric Intensive Care Unit.

20. The method of claim 1 further comprising outputting information representative of the demand for a particular hospital unit.