Prescription and drug switching dynamics in oncology

Background and objectives

Lifescience Dynamics was commissioned to carry out a patient diary study in multiple myeloma. Our client had a chemotherapy product and its sales were not matching their expectations. They wanted to understand the market at a granular level to diagnose the low sales issue. The objectives included measuring the frequency of behaviours and results within the key steps (brand choice and re-initiation) in the treatment decision process:
In addition they wanted to know:

• Why patients "leak" or drop-out?
• What are the clinical patient characteristics that influence treatment choice?
• Where are the high value opportunities in terms of potential our product?

Approach

We designed a qualitative phase with focus groups and IDIs to understand and catalogue all parameters for decision making in different stages of cancer in both treatment native and pre-treated patient groups.

This helped us to design pre-coded field materials for an on-line survey in the USA and EU5 markets in the form of a 30 minute patient medical note collection.

Results delivered

Our report covered all the objectives of such patient referral patterns for un-diagnosed and diagnosed patients. It included data analysis (according to analysis plan) such as switching and churn rates, conclusions and recommendations.

• Market dynamics simulation model from conjoint
  • The resulting perceptual model is held in an Excel-based modelling structure developed by Lifescience Dynamics which, given its user-friendly Windows interface, was supplied to the client for their own internal use in future model simulations with minimal training.
  • The Excel-based interface is vastly more user-friendly than the CBC simulation programme, yet more powerful because perceptions are incorporated into the model. There is, in fact, a facility for relative adjustment of current perceptions, as well as for the blanket imposition of new product alternatives.
• Patient record form database
• Most of the interpretation of the model structure relied on visual representation of the relationships discovered. To this end a 'results investigation tool' was developed by the LSD team and provided to the client.
• The tool was written in 'Visual Basic' and ran within Excel.
  • There was a front-end tool, which enabled the client's personnel to visualise graphical results from the mined data sets.
  • The tool consisted of user-friendly software, which is menu/buttons driven and easy to use with 'click and drop'.
  • The tool automated the graphing and reporting routines. It was possible to make graphical representation of various possible combinations of fields/attributes within the mined data set.