Forecasting for the - sadrabiotech.com

1 Forecasting for the Pharmaceutical IndustryThis page intentionally left blank Forecasting for the Pharmaceutical IndustryModels for New Product and In-Market Forecasting and How to Use ThemARTHUR G. COOK, ASSOCIATES Arthur G. Cook 2006 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the byGower Publishing LimitedGower HouseCroft RoadAldershotHampshire GU11 3 HREnglandGower Publishing CompanySuite 420101 Cherry StreetBurlington VT 05401-4405 USAA rthur G. Cook has asserted his moral right under the Copyright, Designs and Patents Act, 1988, to be identified as the author of this Library Cataloguing in Publication DataCook, Arthur for the pharmaceutical industry Forecasting Research '761519'0112 ISBN-10: 0 566 08675 1 ISBN-13: 978-0-566-08675-5 Library of Congress Cataloging-in-Publication DataCook, Arthur for the pharmaceutical industry : models for new productand in-market Forecasting and how to use them / by Arthur G.

2 Bibliographical references and indexISBN 0-566-08675-1 (alk. paper)1. Pharmaceutical industry-- Forecasting . 2. Pharmaceutical [DNLM: 1. Drug Industry--trends. 2. 736 C771f 2006] '4--dc222006012223 Printed and bound in Great Britain by TJ International Ltd, Padstow, of Figures viiList of Tables xiList of Abbreviations xiiiAcknowledgements xvForeword xvii1 THE PAST AND THE PRESENT 1 The Inaccuracy of Forecasting 1 Forecasting in the Pharmaceutical Industry 3 The Current State: Influences Across Functions 5 The Time Horizon for the Forecast 8 In Summary 92 THE Forecasting PROCESS 11 Define the Forecast 11 Select a Forecast Method 14 Enable Analytic Insights 20 Present the Results 31 Final Considerations 333 NEW PRODUCT Forecasting 35 Tools and Methods 35 New Product Forecast Algorithm 35 Modelling the Market 36 Forecasting the Product 49 Converting Patients to Revenue 65 Final Considerations 694 IN-MARKET PRODUCT Forecasting 71 In-Market Product Forecast Algorithm 71 Trending Historical Data 72 Applying the Effects of Ex-Trend Events 78 Converting Trended Data into Forecast Outputs 91 Final

3 Considerations 100 CONTENTSvvI5 THOUGHTS FOR THE FUTURE 101 Era of Revitalisation 101 Create Stories, Not Spreadsheets 102 Hire a Science Fiction Writer 104 Holistic Forecasting 1046 PHARMACO A Forecasting CASE STUDY 107 Suggested Answers 121 Appendix: Forecast Techniques 131 References and Further Reading 135 Index 139 CONTENTSList of Forecast links to other functional areas Time horizons in Forecasting The process of Forecasting Forecasting is a balancing act Forecast methods Alignment of forecast methods with the product lifecycle Combining qualitative and quantitative methods The quantification process Methods in forecast evolution The linkage of resources to the forecast Analytic tools in Forecasting Risk and uncertainty Calculating risk as a probability of launch Uncertainty as an expected value Uncertainty as a

4 Range of potential outcomes Linking planning to forecast uncertainty Simulation methods Simulation methods generate output ranges Scenario and simulation techniques Combining scenarios and simulations for portfolio planning Decision analysis framework Tornado diagram Waterfall diagram Calculation of the productivity multiple 30 LIST OF Lessons learned in presenting forecasts Generalised new product forecast algorithm Components of the new product forecast algorithm Patient- versus prescription-based models Filters used to convert potential patients into treated patients Prevalence- and incidence-based epidemiology Patient-based and patient-flow models An example of a systems dynamics forecast model Patient segments may be collapsed at any point in the algorithm The effects of consumer education and direct-to-consumer marketing The effects of comorbidity on patient pool calculations Methods used in product share projection The attribute approach to share calculation Examples of scales used in attribute scoring methods The distribution of users of new products Conversion of users into product revenue Mathematical representation of an S-shaped adoption curve Compression of adoption time leads to a fast uptake curve Lengthening of adoption time leads to a slow uptake curve Criteria used in selecting product analogues

5 Converting patients to volume Compliance and persistence The ubiquitous effects of price and reimbursement Generalised in-market product forecast algorithm Historical patient and total prescription data for a product An example of historical product performance Projection using varied historical trending periods An example of the ubiquitous effects of marketing interventions Techniques used to quantify the effects of marketing interventions A marketing intervention triangle Addition of multiple intervention triangles over time A set of marketing intervention response curves Return on investment curves for marketing interventions 85 LIST OF Promotion response analysis Response curve for personal promotion activity Carryover rates in personal promotion activity Preliminary demand forecast with trend and ex-trend analyses An example of products with audited and unaudited channels Generalised distributor activity Demand units compared to ex-factory units Pipeline fill and bleed Speculation An illustration of parallel imports Comparing project alternatives Comparing project alternatives with uncertainty ranges The future of Forecasting 105 LIST OF FIGURESiXThis page intentionally left blank List of The success of new product introductions Blockbusters that went bust Beyond expectations Future pharmaceutical blockbusters?

6 An example of patient-flow forecast calculations Calculation of shares based on scores and weights Evaluation of different share projection methodologies Data useful in trending historical market dynamics Share theft matrices Aggregate effect of intervention triangles Forecasting eras 102 LIST OF TABLESxIThis page intentionally left blank List of AbbreviationsCNS central nervous systemCOGS cost of goods soldDOTs days of therapyEU European UnionHIV Human Immunodeficiency VirusNPV net present valueNRx new prescriptionsR&D research and developmentRx prescriptionTRx total prescriptionsUS United States of AmericaLIST OF ABBREVIATIONSxIIIThis page intentionally left blank AcknowledgementsOne of the central tenets of Forecasting is that collaboration leads to better insights and results.

7 The same is true of writing, and I gratefully acknowledge the many co-workers over the years who have helped to shape this text. I also wish to acknowledge Michael German whose artistic prose added much of the human element to the PharmaCo Case StudyMy colleagues over the years, from University days to Syntex Pharmaceuticals to ZS Associates have provided me constantly and consistently with both the intellectual challenge and reality checks that have led to many of the approaches discussed in this text. Balance is a key to success in many ventures, and I have been fortunate to be exposed to people who exemplify dynamic balance. Industry experience at Syntex Pharmaceuticals brought me from the real business into the academic best practices domain.

8 Consulting experience at ZS Associates has translated intellectual rigour into practical application. Without my colleagues educating me in both these approaches this book would not have been particular I am grateful to BN, AW and RB for their friendship, intellectual challenge and practicality checks over the years. Without their constant support, encouragement, friendship and honesty this text never would have been page intentionally left blank ForewordWhat is a forecast? Potential answers to this question is as diverse as the many different functional roles within a company and include: Forecasting is an accurate picture of the represents the best judgement of the serves as a framework for interpreting present identifies factors with which the corporation must provides a sorting rule among corporate forces examination of current strategic sets up guideposts to mark the path into the offers aid in offers directions for is a measure of combines art and is not an exercise in mathematics; it is an expression of the art of which of these views do you agree?

9 The answer will depend on your perspective. If you are looking at resource trade-offs in an R&D pipeline, you may view Forecasting as the keystone of strategic portfolio planning. If you are developing tactical marketing plans, the forecast may be a tool to gauge the financial results of specific tactics. If you are in charge of manufacturing, you may look to the forecast to provide accurate direction for the number of units that must be manufactured. This simple question What is a forecast? illustrates the key challenge of Forecasting : the need to provide inputs for multiple decisions within an book addresses the varied elements of Forecasting . It is organised into six chapters. In Chapter 1, we examine the history of Forecasting and the lessons that can be applied to the practice of Forecasting in the pharmaceutical industry.

10 In Chapter 2, we examine the tools, methods and analytics that are available to the forecaster. In Chapters 3 and 4, we discuss the detailed approaches and algorithms that can be used in new product and in-market Forecasting . In Chapter 5 we conclude the main text with our forecast of Forecasting that is, thoughts on the future of Forecasting . Chapter 6 provides a case study to help the reader apply the Forecasting concepts. FOREWORDxVIIThis page intentionally left blank THE PAST AND THE PRESENT11 CHAPTERThe Past and the PresentThings are more like they are now than they ever were D. EisenhowerThe future is like the present only GossageTHE INACCURACY OF FORECASTINGP redicting the future is difficult. A historical look at Forecasting over time suggests that we have continually tried to predict the future.