Conference Paper/Proceeding/Abstract 672 views 74 downloads
Forecasting Branded and Generic Pharmaceutical Life Cycles
International Symposium on Forecasting
Swansea University Author: Sam Buxton
PDF | Author's OriginalDownload (418.25KB)
This paper will look at modelling and forecasting of branded and generic pharmaceutical lifecycles with a 1 year forecasting horizon. The focus will be on pharmaceutical life cycles around the time of patent expiry as the sales of the branded pharmaceutical decline and the sales of the corresponding...
|Published in:||International Symposium on Forecasting|
No Tags, Be the first to tag this record!
This paper will look at modelling and forecasting of branded and generic pharmaceutical lifecycles with a 1 year forecasting horizon. The focus will be on pharmaceutical life cycles around the time of patent expiry as the sales of the branded pharmaceutical decline and the sales of the corresponding generic equivalent increase. Understanding the patterns of decline and the associated generic growth is increasingly important and the market is currently worth over £5bn in the UK in 2013 and while it is greater than any other industrial sector in the UK it has declined from £7bn in 2009. The number of ‘blockbuster’ drugs also continues to decline. As a result the pharmaceutical industry makes efforts to extend the commercial life of their brands and the ability to forecast sales is of increasing importance in this regard. The paper also provides for effective governance because the use of a branded drug when a generic equivalent is available ultimately results in wasted resources. The pharmaceutical prescription data comes from a database known as JIGSAW. The prescription drugs that were modelled were those that had the highest number of prescriptions within the database. There were five models originally used to model and forecast this data. These were: Bass Diffusion, Repeat Purchase Diffusion Model, Moving Average, Exponential Smoothing and the Naïve. Based on previous research it was expected that the more complex models would produce more accurate forecasts for the branded and generic life cycles than the simple benchmark models. As none of the complex models yielded results more significant than those of the Naïve model, it was thought to be appropriate to add additional models to the analyses. The additional models added were: Holt Winters Exponential Smoothing, Auto-Regressive Integrated Moving Average (ARIMA), Robust Regression, Regression over t, Regression over t-1 and Naïve with drift. The empirical evidence presented here suggests that the use of the ARIMA provided the most accurate and robust method of modelling and forecasting branded pharmaceuticals. For the generic equivalents the empirical evidence suggests that the Naïve model with the addition of a 70% trend would provide the most accurate and robust modelling and forecasting method.
Forecasting; Diffusion Models; Pharmaceutical Lifecycles; Branded drugs; Generic drugs.
Faculty of Humanities and Social Sciences