"We are aware of the pitfalls"

Hi Anna! Why are you part of Clincium’s advisory network?

Research is close to my heart, and I can’t think of a more enjoyable job. Being part of Clincium’s advisory network gives me the chance to stay connected with many exciting researchers and to work on a large number of interesting studies. I have worked as a statistician in the pharmaceutical industry since the last century, and I feel that I can contribute a lot when discussing clinical trials with researchers—it is incredibly satisfying!

The fact that I have fantastic colleagues in statistics and learn something new every day is another reason why I enjoy going to work.

What does a statistician do?

Statisticians working with clinical trials always start by trying to understand the intended outcome. We are most useful if we can be involved from the planning stage. That allows us to give advice on study design, how to measure the effect of the innovation being studied, how many patients are needed in a study to be able to show an effect, and how the collected data should be analyzed statistically. Are the patients in your study supposed to be randomized into different treatment arms? No problem — we can give randomization advice and generate randomization lists.

A good tip is to let a statistician review the study protocol and the statistical analysis plan. We have read countless such documents and know where the pitfalls are, and how to write them clearly so that the purpose of the study is obvious, as well as how it should be conducted and later evaluated. We are also good at checking that your data collection form matches the protocol and that all necessary data for evaluating the study is collected correctly.

If you are considering an interim analysis, we can explain the advantages and disadvantages of including one in your study. And finally, when the study results are presented in tables and graphs, it is valuable to have a statistician help interpret the results.

What kinds of questions can I ask you?

I want to show that the new treatment is better than the existing ones — how can I do that?
How many patients do I need to include?
I’m thinking about using a logistic regression model — what do you think?
How about a coffee?