Choice modelling for medical research agency
The brief to customise choice modelling
Our client, a leading medical research agency, tasked us with undertaking a highly complex trade-off between branded and biosimilar products in the Ophthalmology marketplace. The study included multiple countries with personalised designs for each market, particular trade-offs at different time points and links with current prescribing. The profiles involved discounts and various other indications for the tested products. These parameters had to be integrated into an all-in-one simulator.
The complexity of the task
This project was one of the most highly customised choice modelling exercises we have ever undertaken. We had the challenge to incorporate a staged trade-off at different hypothetical points in time and carry out customisation by market and products to develop a simulator integrating all stages of modelling. There was a requirement for the simulator to have a dashboard add-in.
Steps to achieve results
- During the initial stage of the project, we spent several weeks with our client formalising the requirements of the design at each hypothetical time point and crystallising key parameters of the model with the ability to change them for each product.
- After that, we had to internally work out the best way to manage the modelling during each step, considering the complexity of the brief and keeping the design as simple as possible.
- Finally, the development of the custom simulator started, followed by many iterations to find a configuration meeting the client’s needs.
The brief
To undertake a highly complex trade-off between branded and biosimilar products in the Ophthalmology marketplace.
The requirement
There was a requirement for the simulator to have a dashboard add-in.
The challenges we faced
That was a highly complex brief requiring ongoing dialogue with our client. One of the main challenges was finding pragmatic ways of accommodating client requests while managing the modelling and simulator development. We had to find solutions to continuous new requests not easily fitting into a choice modelling framework, including making simulations directly from observed choices in some scenarios as well as outputs from our model. The data and simulator prepping works were enormous, but we were able to accommodate all requests.
The outcome
The models and simulators were developed incrementally using interim data. The result is a highly customised excel simulator and dashboard addressing all the client’s objectives.
The challenge
To find pragmatic ways of accommodating client requests while managing the modelling and simulator development.
Results
The result is a highly customised excel simulator and dashboard addressing all the client’s objectives.