Sharpen your research with Clinicum's methodological support

Stephen Nash, who do you think should apply for Clinicum’s support?

Everyone! Anybody doing research at KI or for Region Stockholm can apply – and I think the service will benefit anyone involved in research, whether they are a junior researcher starting out in their scientific career, or a professor with years of experience. It’s always useful to get an outside view, especially with something like statistics, where many people feel under-prepared and there is often a nervousness about doing the right thing. Although it’s only a few hours of support, statisticians excel at honing in on the important questions to ask and focusing on the key aims of the research. And we’re not just statisticians, we’re epidemiologists and bioinformaticians as well, and all of us have years of experience. We can help make your research better: more impactful, more precise, and more relevant.

At what stage of their research do people usually apply to Clinicum?

Any stage, but the earlier the better! Solving issues before you seek funding is easier than when the budget is in place, and fixing data issues is easier at the design stage than the analysis stage. We’re always happy to discuss ideas and plans, which can be much more fruitful than trying to fix things after the event. I’ve been involved with very early ideas, helping people put together grant applications, as well as advising on analysis plans and how to interpret results. The earlier we speak, the bigger the impact we can have on the project.

What are the most common methodological challenges you see researchers have, and how can they be prevented?

For me, it’s often about the design of a clinical trial, and in particular being really precise about the question you’re asking. In a clinical trial, vagueness at the design stage can cause real problems later on, so it’s great to think hard about definitions: of the intervention, of the outcome, and of the population you’re working with. The best method I’ve ever found to prevent problems later is to think really hard! That’s an important part of our job, and asking as many people as possible can help with that. On a more practical note, there are some useful guidelines from the European Medicines Agency about how to define an “estimand” in a clinical trial. It’s not a word that everyone will be familiar with, but basically it’s a precise definition of what you’re trying to measure. There are some great guidance papers which provide step-by-step practical help for people setting up a clinical trial, to help them write an estimand for their primary outcome – I’ve found using this framework has really shifted how I think about trials and helped me to better understand and explain the trials I’m working on in my own research.

How can you optimize a Clinicum consultation?

I like to read as much as possible before we meet, so that we don’t spend too much time on background when we meet. I find that my brain often works slowly in the background, so having a chance to digest things for a couple of days can improve the advice I offer. I find that’s more important than people having a clear question for me, or concrete aims about what they need – sometimes, after discussing, those aims can shift to a different (and better!) direction. So provide background material in advance, be open minded, and be prepared to discuss all elements of your work – statistical thinking can provide clarity to so many areas of scientific research.