Gill Worthy recently joined the Mtech Access evidence team as a Consultant Statistician, specialising in network meta-analyses and systematic literature reviews. In this interview, Gill tells us more about what inspired her to pursue a career in the life sciences and shares her insights into network meta-analysis.
What led you to a career in the field of evidence generation?
I was working as a Clinical Trial Statistician and I saw an advert for a Trainee Systematic Reviewer at the Centre for Reviews and Dissemination at University of York. I applied, and that was the start of my career in reviews! It was an accidental career move, but I enjoy finding evidence and working out the best way to clearly present it so that others can interpret it; it’s like being a detective.
What motivated you to join Mtech Access?
I was looking for a more varied role where I could use and develop my statistical knowledge, particularly in methods such as indirect comparisons and network meta-analysis. Previously I was part of an Evidence Review Group and reviewed many NICE submissions, so I am keen to use this knowledge to assist in the preparation of submissions rather than their critique.
I am also looking forward to working in different therapy areas such as cancer, as my background is mostly wound care and cardiovascular disease, and also to collaborating with other colleagues in health economics, market access and medical writing.
What have you been doing in your first few weeks with Mtech Access?
Whilst getting to know the team and the different aspects of the business, I have been working with Stephen Mitchell and Sarah Batson from the Systematic Reviews team on a client project to determine how feasible it would be to perform a complex analysis known as a matching adjusted indirect comparison (MAIC). A feasibility assessment is essential to ascertain whether a MAIC is possible and enables us to advise our clients on the most appropriate evidence to include.
A MAIC analysis is performed in the absence of comparative trials. The patient data from one study are matched to the baseline characteristics of another trial to make them more comparable, and then the reweighted data are used in the analysis of the outcomes. This enables a comparison of treatments that may not be possible via traditional analysis methods, and can provide the evidence required to achieve reimbursement with bodies like NICE.
Your specialism is in Network Meta-Analysis, what different insights does this offer?
Network meta-analysis differs from standard meta-analysis in that multiple treatments can be compared in the same analysis. Network meta-analyses use Bayesian methods that combine prior assumptions about the data with the actual study data. These are now easier to perform thanks to software such as WinBUGS. In recent years, network meta-analyses have been more widely used, particularly in regulatory submissions where there are a number of eligible comparator treatments under consideration.
When choosing a method of analysis, there are important assumptions to consider including whether or not the studies have similar methods, patient populations and outcomes to enable them to be pooled to provide one overall result. All forms of meta-analysis need careful assessment and interpretation.
There have been some recent innovations in network meta-analysis, what is your view on these approaches?
In my previous role I reviewed submissions to NICE and in the last 2 years, population adjusted meta-analysis methods, such as MAIC have become more popular, before that I didn’t see any. This increase in popularity is particularly the case in rare diseases where there is often only a single-arm trial, i.e. one that is evaluating one treatment and not comparing it with any other treatments.
As highlighted above, MAIC are more complex than standard meta-analyses and require assistance from a statistician. They also require the individual participant data for the treatment of interest, unlike a network meta-analysis, which uses published data. This is an interesting area and research is ongoing.
When you are not at work, what do you like to do?
In my spare time I look after my 10-year-old twins and enjoy music; I sing in the local choir and have been learning to play the piano.