The first issue determines whether short-term, medium- term or long term forecasting techniques are appropriate. The second will determine which forecasting techniques are feasible depending on the amount and type of past data available.
Forecasting that is, estimating the future will also have to rely on trends to a great extent, because how far in the future it is possible to estimate is limited by the data available.
1. Forecasting Demand:
Forecasts of demand should be made as far ahead as the maximum lead time the period of time it will take to implement a decision.
Since the past alone is an unreliable basis for predicting the future, especially in longer range forecasts, there is limited value in acquiring large amounts of past data.
Future based forecasting techniques may require minimal data if simple correlations are to be used or a large number of observations if a significant number of influential factors are to be employed.
Quantity of data generally adds to knowledge about the present and past, and may not necessarily increase the accuracy of forecast of the future.
2. Forecasting Utilization:
In forecasting utilisation, it has to be determined what the future utilisation of a specific service will be. A change in utilisation in future may be expected or desired.
Factors which are likely to cause expected change will need to be analysed. Strategies which will affect the desired change should be selected and implemented.
In forecasting utilisation, complex techniques such as multiple regression, simulation, multivariate equations and other quantitative techniques may enable the planners to deal with large number of factors affecting utilisation.
However, such techniques in the Indian context may not always be accurate and may result in expensive scientific looking wild guesses.
Apart from the change in the demographic characteristics, the changes in attitude and behaviour patterns of user population, changes in medical care patterns (e.g. ambulatory vs inpatient care, new technology), and changes in the concurrent factors affecting utilisation (e.g. medical insurance, employee health programme) must be incorporated in a utilisation forecasting.