Authors:
Paul Harper
1
;
John Minty
1
;
Sujit Sahu
2
;
Bernard Baffour
2
and
Christophe Sarran
3
Affiliations:
1
Cardiff University, United Kingdom
;
2
University of Southampton, United Kingdom
;
3
Met Office, United Kingdom
Keyword(s):
Forecasting Demand, Hospital Capacity Management, Weather, Simulation.
Abstract:
Improved short-term predictions of hospital admissions and bed occupancy offer the potential to plan resource needs more accurately and effectively. The MetSim project explores the relationship between weather and health, building novel Bayesian models that are more sensitive to fluctuations in weather. Short-term forecasts of the numbers of admissions, categorised by age, gender and medical condition, are produced. In turn, coupled with predictions on length of stay and information on current occupancy, MetSim uses hazard ratios embedded within a simulation framework to provide forecasts of short-term bed needs. MetSim is a collaboration between Cardiff University, the University of Southampton, and the Met Office. Cardiff and Vale University Health Board and Southampton University Hospitals NHS Trust have guided the development of MetSim, provided data and piloted the tool.