The responsibility for providing appropriate healthcare till now was vested with the government. With the entry of corporate hospitals and availability of health insurance, the demand for better health services from both private and government hospital has been increasing. Whether the health care services are provided by the state or privately run hospitals, it is becoming more and more important to optimally utilize the available resources. Two important indicators of effective health service are length of stay in the hospital and hospital related expenses. Identification of demographic and environmental variables impacting these indicators will immensely help the government in formulating long-term policies. Since most of these variables are categorical in nature, it is not easy to apply the usual mathematical and econometric models for prediction purpose. This paper uses neural prediction models to predict the duration of stay in the hospital and hospital related expenses using demographic, environmental and hospital related factors and identify the profiles of the patients in various segments, which will help in policy formulation.