<img alt="" src="https://secure.lote1otto.com/219869.png" style="display:none;">
Skip to content

Creating a Common Language: Forging Statistical and Clinical Collaborations


In this blog, Paul Terrill, Director of Strategic Consulting at Cytel outlines his blueprint for ensuring smooth communication between statistical and clinical stakeholders. Paul draws upon his 20 years of experience working as a statistician and his training background to share his guidelines for success. Whether you are a statistician looking to hone your project communication skills, or a clinician keen to maximize the benefit of statistical input to your trial, this article will provide helpful pointers.

Don’t Assume We Speak the Same Language
One of the problems is that statisticians and clinicians do tend to talk a different language, both using very different technical terminology. Statisticians sometimes struggle to understand the technical language that clinicians use, and clinicians of course sometimes struggle to understand the terminology we use. None of us should assume, when we're talking about a topic, that the other party already knows what the terminology means. Particularly when working with someone new, we should make an effort to explain concepts and terms in order to establish a common understanding. Only once you’ve built up a relationship and worked with a person for a period, can you begin to understand what they know. For example, when I'm in discussions with one particular close customer, I know they have built up a very strong understanding of statistics over the course of our working relationship. This means that I don't need to set the scene and explain certain concepts whenever we embark on a new project together. However, this isn't going to be the case for every stakeholder you liaise with and so we always need to be mindful of their ‘prior’ (see what I did there?) knowledge and foundation in the concepts under discussion.

Why As Well as What
Quite apart from speaking a different language, there are different worries keeping statisticians and clinicians up at night. As statisticians, we are preoccupied with statistical issues such as appropriate and correct design and analysis methods, multiplicity, missing values, estimands, as well as a host of statistics -focused regulatory guidelines that we need to adhere to.
Clinicians, of course, worry about the medical side of things. Sometimes, when we share a statistical concern, it may not always be immediately obvious to clinical stakeholders why we're worrying about it. We should always clearly explain why we are raising a topic for consideration. Just because it’s obvious to a statistician why the team might need to consider a certain aspect, it might not be obvious to the other stakeholders in the team. That goes both ways, and clinicians may assume that we understand why a certain issue is important when in fact we do not.
You run the risk of your recommendation not being implemented because you haven’t put in the groundwork to help people understand its importance. In my experience, if you don't explain the "why”, your advice may not be acted upon resulting in problems down the road and an issue being picked up by regulators or an ethics committee or another relevant group. This is an outcome that no one wants to see.

training (3)

Use a Training Mindset
When you are communicating complex concepts, quite often, you are actually teaching. You have to explain the concept to someone who doesn't have a pre-existing understanding of it. So if you've developed training skills during your career, or even if you imagine you're in a training situation, that helps communication.
As statisticians, we need to be extremely secure in the ideas we are expressing and present them confidently. For some statisticians, direct communication with a customer isn't their favorite part of the job. It’s very important for the success of our collaborations that we are willing to engage and work on our communication skills to ensure that our messages aren't lost in translation. Even topics that we may consider to be straightforward, like type 1 error and power, are regularly misunderstood (and not just by some non-statisticians; by some statisticians too). Practice makes perfect, and every time you talk through a particular concept you explain it better each time. Remember - keep it simple and relevant.

Develop trust over time
Good communication doesn't come quickly, it's a long-term investment. We often work with sponsors over the course of several projects or an entire program of trials. That allows the trust to be built up so they will say, “Well, you got it right before, I’m going to trust you again”. This really helps to get the kind of positive feedback on projects that we always strive for at Cytel. When you understand your customers’ background and previous knowledge it's much easier to ensure that communication is running smoothly in both directions.

It’s Good to Talk
Face to face communication is often the holy grail to avoid misunderstandings and ensure clarity. It allows you to tune in to whether your teammates are following your points or not. If they are ‘with’ you, you know you can give more information, and if they aren’t you have the opportunity to take a step back and perhaps reexplain what hasn’t been fully understood. That’s obviously much easier to accomplish face to face where body language and expressions provide a lot more information than just what is said. If you can’t be in the same location, making use of video conferencing can be a real benefit. Related to this, we shouldn’t rely too much on email – picking up the phone and ironing out a point verbally is often more efficient and helps relationship building than a long series of emails. By all means, summarise a discussion by email afterward, but it is good to talk.

Get out of your own head
If we focus too much on our own areas of expertise then we sometimes miss seeing the other person's point of view. It’s important to come out of your own perspective and understand the other stakeholders’ areas, not getting so ‘bogged down’ in complicated statistical concepts and terminology that it inhibits a two-way conversation. One area that we should be particularly mindful of is the logistics and implementation that is the responsibility of the clinical stakeholders. If you propose a complex method, you need to ensure that the wider team understands it, and be sure that the implementation is feasible and makes sense from a medical point of view.

The Benefits of Good Communication
When statistics and clinical functions are communicating well you are able to ensure that your approach is following the correct statistical and clinical methodologies and is appropriate for the various regulatory guidelines. It’s critical that what we are working on meets the desired objectives. Without the right input from all parties, you risk reaching the end of the project and realizing a mistake was made at the outset. So, plan to get things right at the beginning, rather than firefighting.
One crucial aspect is when the sponsor has to interact with an organization outside of their own company such as the FDA or EMA. It's obviously much better all round for credibility if the package that's being presented is robust. You can only achieve that credibility if the project team clearly understood the statistical aspects of the package. So when the feedback comes from the regulators that the statistical aspects look fine, or at least is not completely flawed, then we know that we’ve got it right. If that doesn’t happen then we could face having to engage in regulatory negotiations or resubmit the design or analysis. Beyond this, if we are able to effectively communicate what we are trying to achieve from a statistical standpoint and drive the implementation, we are able to influence a great impact on the trial by saving money, time, or reducing the risk. Ultimately, when clinical and statistical stakeholders work well together and communicate effectively, we create a better product and lay the foundations for a better outcome.

Liked this blog?  Click the button below to get more insights and articles direct to your inbox and subscribe for Cytel blog updates.


Do you have a clinical development  project you would like to discuss with our strategic consulting team?  Click the button below  arrange a discussion with one of our team.


About the Author

paulterrillPaul Terrill is a Director of Strategic Consulting at Cytel and has extensive statistical design and analysis skills which he uses to provide valuable statistical consultancy and advice. Paul has excellent training and presentation skills and is praised for his ability to clearly communicate complex statistical concepts to non-statisticians. He started his career working as a statistician in the agrochemical industry at Jealott’s Hill, Berkshire before becoming a statistical trainer for SAS. He moved into the pharmaceutical industry in 2005 and primarily provides support to biotech and small pharmaceutical companies who lack in-house statistical expertise. He has been on the PSI scientific committee since 2014 and joined the PSI Board of Directors in October 2017, taking on the role of scientific committee chair. Paul holds a BSc in Applied Mathematics and Statistics from the University of Wales, Aberystwyth and a PhD in Statistics from the University of Kent, Canterbury.


contact iconSubscribe back to top