Applying Gravity model to predict demand of public hospital beds

We used empirical patient days of existing hospitals, and residential addresses to calibrate the “Gravity mass” coefficients of each hospital. We then used these coefficients to predict the new redistribution of market catchment after NTFGH was opened. We also compared the gravity coefficients of different service types and showed the differences. The results indicated some hospitals were more “attractive”. The short-term predictions for redistribution were also fairly good. Hence, with this Gravity model, we are hopeful to estimate future hospitals' demand and make appropriate infrastructure and manpower planning recommendations.