The Optimizing Pharmaceutical Research Resource Allocation (OPRRA) Software was created by Oxford Professor John C. Gittins and his research team, including his final student, Dr. Anne Marie. The aim of OPRRA is to assist pharmaceutical companies in taking decisions on rates of resource allocation during the discovery, or pre-clinical, phase of pharmaceutical research so as to improve profitability.
Discovery is very different from clinical trials, on which there is a huge literature. Two of the important differences are that it mainly involves chemists and biologists rather than clinicians, and that there are fewer regulatory constraints on how it is conducted. OPRRA focuses on this relatively unstructured area. It is used as an aid to inform the periodic meetings at which pre-clinical research projects are selected and prioritized.
OPRRA is based on a stochastic optimization model. For each project the inputs are estimates, possibly in the form of probability distributions, for
the success probabilities, costs and durations of the various stages of pre-clinical and clinical research,
the net present value of a new drug when it emerges from clinical trials,
the likelihood of pre-emptive discoveries by competitors, and their likely impact on the value of any new drug, and
the relationship between the duration of a research stage and the level of effort allocated to it.
Estimates are also required for the company’s weighted average cost of capital and for the annual cost of employing a senior scientist.
Given these inputs OPRRA will calculate
the most profitable effort allocations for the successive pre-clinical stages of each project,
the sensitivity of profitability to variations in model parameters and in allocations,
the probability distributions of profitability, both for individual projects and for a portfolio of projects, for any given allocation plan, and
projections of the future effort allocations required to implement an allocation plan.
So OPRRA is a tool which enables a planner to explore the consequences of different possible allocation plans in terms of their overall effort requirements, their profitability, and their varying degrees of risk. Of course a model can only ever be an imperfect representation of reality, so apparently good allocations derived from OPRRA can only be suggestive. However, some movement of the resource allocation plan in the direction suggested by OPRRA deserves serious consideration, especially when the apparent scope for increased profitability is high, as is often the case.
Effort allocations are assumed to be reduced to a one-dimensional scale. They should be aggregated over the different types of expenditure concerned. There is no modelling aimed at determining the appropriate mix of chemists and biologists or of in-sourcing and out-sourcing. The assumption is that for any rate of effort allocation these splits will be made in an appropriate way. The unit of resource on this scale may be expressed in any convenient way, for example as a rate of expenditure of $100,000 per annum, or as one scientist or senior scientist together with a normal share of support staff and overhead costs. One possibility would be to use an in-house scientist as the unit, scaling this up for those of star quality, and translating outsourcing expenditure into these terms by dividing the annual cost by the inclusive annual cost of an in-house scientist.
OPRRA is written in C++ and uses the Visual Studio environment.
Demos of OPRRA, as well as case studies using the OPRRA product, can be furnished upon request. Contact us to get a free demo.