Planning for both the short and long-term future in healthcare is vital. Allocating too many resources to a service can lead to waste and overspend, while under provision can lead to long waiting times and even the potential for loss of live.
Introduction of a new service or treatment presents further issues in the sense that there may be little or no historic data on which to base projections of cost or usage on.
In this module, participants will learn about tried and tested ways to both generate forecasts in the healthcare setting and analyse their accuracy. In particular, participants will develop an understanding of how to select and justify the appropriate forecasting technique.

What did previous attendees say?

“Very useful and packed with applicable learning. Examples very suited to health attendees”
Chris Morris, Principal Information Analyst.

“Well presented and structured with good practical examples”
James Tilbury, Information Analyst

“Course was very good, very well presented and made you think about how to break down and make sense of forecasting problems”
Ellie Parker, Senior Information Analyst.

“Very useful and informative, I can see lots of applications of this course at work”
David Rugman, Principal Information Analyst.

Aims/Learning objectives

At the end of this module the student should be able to:

  • Appreciate the role and potential benefits of forecasting for estimating future spend and resource use in the healthcare setting.
  • Generate forecasts using time series methods, including Simple Moving Average, Exponential Smoothing and Trend Adjusted Exponential Smoothing.
  • Create forecasts using econometric-based methods, including the Linear Regression Model.
  • Analyse and compare the accuracy of different forecasting methods by analysing their errors.
  • Identify outliers and know when they should be removed to improve forecast accuracy.
  • Incorporate changes factors such as seasonality, changes in population, lifestyle factors and alternative service plans.
  • Perform sensitivity analysis to identify factors which are statistically significant in helping us to understand the future.


  • Introduction to forecasting in healthcare
  • Forecasting using time series methods
  • Forecasting using econometric methods
  • Identifying and incorporating changes in external factors
  • Comparing forecast accuracy and model section
  • Identifying and removing outliers
  • Discussion: uses of forecasting, reservations and challenges

Case material

Three case studies presented in background material.

  • To practice some elementary time series forecasting techniques using a sample healthcare data set: GP practice new demand projection.
  • To practice using exponential smoothing and double exponential smoothing using a sample healthcare data set: cost projections of COPD and patient numbers over the entire stages of the care pathway for COPD.
  • To practice using econometric based methods for forecasting using a sample healthcare data set: hospital demand projection.

Course delivery

In general each module of the course is associated with one day’s face-to-face training in our Central London premises supported by additional material and resources. Distance learning might be considered in the future depending on need.
We can also deliver customized in-house courses.

Module Format

For face-to-face delivery, a mixed structure is generally adopted. There is a blend of sessions introducing the topics and techniques of interest and use examples to illustrate key themes, and of sessions engaging participants in interactive and hands-on experience of tools and techniques drawing on case material of direct relevance to commissioning or provision of services.


Overall in order to take part in this module, potential participants will need mathematical skills to include simple algebraic manipulation (GCSE grade B or better) and some understanding of basic statistical techniques.
Basic IT skills are required, e.g. use of excel spreadsheet.

Software Support

The software needed for this module are SPSS (predictive analytics software) and Microsoft Excel 2007/2010.


Prof. Thierry Chaussalet or Philip
Tel: +44 (0)20 3506 4575

Download module data