Development of computational models for strategic and tactical management of pharmaceutical R&D pipelines
Over the past century the Pharmaceutical industry has evolved from an "outback garage" industry to a highly competitive, technology-driven, organized industry, churning out highly valuable drug products that have played a key role in vastly improving our quality of life and healthcare. But today this industry faces a challenge of serious proportions---The cost of developing a drug from discovery to market is now estimated at $ 800 MM with times to market ranging between 7 to 13 years. Even worse, most of these drugs have to pass through three clinical trial stages with a cumulative probability of success as low as 10%. Since drug products carry limited patent lifetimes, each day that a drug spends in the pipeline adds to revenue losses as well as loss of 'first mover advantages'. An obvious remedy is to develop as many drugs as possible in order to beat the odds of failure. Unfortunately, this solution is impractical in view of internal resource constraints. The firm only has so many people, testing and manufacturing equipment. Further, resource procurement budgets are tight. As a result the firm can develop only a limited number of candidate compounds. It must select a portfolio of drugs from the available pool of candidate compounds. Further, once a portfolio of drugs has been selected, decisions on which drugs to prioritize for resource allocation and the levels of resource allocations need to be made. Managers making these portfolio selection, prioritization and resource allocation decisions face a major challenge---Drugs that promise high expected returns are usually associated with high risks of failure. Consequently, conventional decision-making methods like charting techniques fail to resolve such reward-risk trade-offs. Neither the strategic finance nor the operations research literature sources provide methodologies that can optimize these decisions in an integrated fashion. This thesis presents new decision-engineering technologies that can allow pharmaceutical managers to build and manage drug development portfolios and corporate resource capacities in a way that maximizes portfolio returns, minimizes portfolio risk as well as minimizes time to market. Decomposition methods combining genetic algorithms, discrete event simulation, mathematical programming, branch-and-bound and heuristics are applied to optimize these decisions in an integrated fashion.
Pekny, Purdue University.
Chemical engineering|Industrial engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our