Niki Arya, Principal Statistician at GlaxoSmithKline
When I tell people I am a statistician, they make comments like, “Wow, you must be really good at math.” Some ask questions like, “So, do you sit behind a computer crunching numbers all day?” It seems statisticians are generally perceived as bookish individuals who do a lot of behind-the-scenes mathematical grunt work for a study, such as performing the power and sample size calculations during the design phase and running a bunch of statistical tests once the results are known. It is understood and well respected that every study needs a statistician, but we often don’t get the glory or glamour that certain others receive.
This is why I was so pleasantly surprised by a reaction I recently got when someone found out I was the statistician on a certain project. I was assigned to be the statistician on a bioequivalence study comparing alternative formulations of an oncology compound to the commercial tablet. Although the commercial formulation is biologically effective, it is not easy for the cancer patients to take because of its large size. The purpose of the study is to find a more easily administered alternative formulation that is bioequivalent to the commercial formulation.
I started to do some sample size calculations. Because of the high pharmacokinetic variability with the drug, the planned sample size for the trial would be very high, potentially leading to logistical difficulties and long delays in trial completion.
So, I had to think outside the box and come up with creative approaches to the design. After doing a bit of research and consultation with some statistical experts, I came up with an adaptive design with interim analyses to potentially stop the trial early for efficacy (to declare bioequivalence) or futility (to fail to declare bioequivalence). These interim analyses can potentially lead to a substantial reduction in sample size, if the rules for early stopping are met. As a statistician, I felt I was just doing my job, coming up with this more feasible adaptive design.
Well, several months ago, I joined a rock-climbing group outside of work and saw someone who looked familiar. She looked back at me, and I got the impression she thought I looked familiar, as well. We started talking and it turns out she was the formulation chemist for the bioequivalence study. We had seen each other at large team meetings and had talked in a group setting, but it was the first time we met one-on-one. When I told her I was the statistician on the study, she got excited and commended me for the study design. Apparently, she and her team of scientists thought I was the ‘hero’ for developing a study design that potentially reduces the sample size and saves the company a lot of money. I was so flattered by her reaction. I never thought statisticians would be seen as heroes!
The next week, when I saw her rock climbing again, she mentioned that one of the other scientists on the study—someone I barely knew—told her to tell me “hello.” I had no idea my study design would generate so much attention. So much for the notion that statisticians are behind-the-scenes number crunchers who often go unnoticed!
This experience shows just how much of an impact statisticians can have in a company. We are more than just the math experts. We are scientists who bring a unique perspective to research because of our quantitative skills. In an age in which pharmaceutical companies are desperately trying to reduce costs, our innovative study designs and methods for analyzing data are critical to improving the efficiency of drug development. I encourage all statisticians to step up to the plate and fully embrace this role. It’s a nice feeling to be considered a hero.
Leave a Reply