Núria Buil-Bruna, Alienor Berges
Clinical Pharmacology Modelling and Simulation, GlaxoSmithKline, UK
Objectives: As pharmacometricians, we have all experienced the frustration when first trying to communicate technical items to multidisciplinary teams which can translate into less impactful decision making. In fact, it is recognised that most pharmacometricians lack strategic and effective communication skills [1]. Several previous efforts have been made to emphasise the importance of effective communication to improve our impact in drug development or influence regulatory decisions [1-3], but there is still a long way to go. Here we aim to provide a guide to help pharmacometricians communicate in a common language easily understood by non-technical audiences.
Methods: Based on discussions within our Clinical Pharmacology Modelling and Simulation department we identified key messages that we most commonly struggle to communicate effectively. In addition, we’ve also consulted key stakeholders to understand their view of our messaging to get further insight of areas for improvement. Inspired by online guides to de-code British euphemisms [4], we have created a communication cheat sheet to illustrate:
- What pharmacometricians say: including common technical language within the pharmacometrics community,
- What others understand: showing what non-pharmacometricians can interpret/misinterpret from it,
- What pharmacometricians mean: to recommend alternative phrases that communicate the same message more effectively.
Results: The key common concepts identified as most difficult to communicate efficiently can be grouped into two main areas:
- 1) Utility of the analysis (e.g. what a model can(not) do, strengths and limitations, suitability for regulatory decision, changes in the label),
- 2) Interpretation of results (e.g. between subject variability vs. uncertainty, model evaluation, significant covariate vs. predictor of effect)
We have created a communication cheat sheet to address common messaging issues in these areas of focus. The example below illustrates the concept in 2) when applied to communicating model/parameter uncertainty, and emphasises the potential pitfalls of getting it wrong.
- What pharmacometricians say: The model predicts a typical effect at 100mg of 35% associated with a wide uncertainty (-25 to +55%)
- What others understand: Great! The effect with 100mg will be 35%!
- What pharmacometricians mean: The model shows that if we do the study at 100mg, our best guess is a modest effect, but we may see no effect at all.
In this case, confidence intervals should be seen not only as numbers in parentheses but as scenarios of potential truth. The consequences of ineffective communication in this example may lead to team members thinking of larger confirmatory studies rather than exploratory studies to refine the dose prediction.
In addition to the cheat sheet, we have also generated some widely used plots (e.g. VPCs, covariate forest plots, etc) with added legends that can help non-pharmacometrician understand the key message of each picture without superfluous technical details.
Conclusions: In our experience, the chance of influencing clinical teams is higher when they understand pharmacometrics strategy and outputs. Therefore, it’s worth taking the time to find the best way to communicate technical findings to multidisciplinary teams. Here we’ve proposed tools aiming to facilitate communication for other pharmacometricians.
References:
[1] Mehrotra, S., & Gobburu, J. (2016). Communicating to Influence Drug Development and Regulatory Decisions: A Tutorial. CPT: pharmacometrics & systems pharmacology, 5(4), 163-172.
[2] Krause, A., & Gieschke, R. (2010). Interactive visualization and communication for increased impact of pharmacometrics. The Journal of Clinical Pharmacology, 50(S9), 140S-145S.
[3] Leinfuss, E. (2015). Be a model communicator and sell your models to anyone. CPT: Pharmacometrics & Systems Pharmacology, 4(5), 275-276.
[4] https://www.economist.com/johnson/2011/05/27/this-may-interest-you
Reference: PAGE 29 (2021) Abstr 9796 [www.page-meeting.org/?abstract=9796]
Poster: Methodology - Other topics