Missing data is a common problem in clinical trials despite all our best efforts to minimise it through design. It is likely to occur in most randomised controlled trials. When missing data is present, the ability to conduct intention-to-treat analyses, which require the complete inclusion of all data from all randomised patients, is compromised and can influence results. For this reason, much research is focussed on analytical techniques to estimate unbiased effects in the presence of missing data, including imputation-based methodologies. PHASTAR statistician Zara Ghodsi has recently had a paper published where she proposes a new method for the imputation of time series data based on singular spectrum analysis (SSA).[1]
When designing a clinical trial, one of the biggest factors that needs to be thought about is blinding. In clinical trials, there is a risk of expectation influencing findings so if a patient knows which treatment group they have been allocated, there is a risk that this could bias results. In controlled trials, the term “blinding” refers to keeping study participants, those involved with their management, and those collecting and analysing clinical data unaware of their assigned treatment so that they cannot be influenced by that knowledge. If patients are not blinded, awareness of group assignment may affect their behavior in the trial, and their responses to subjective outcome measures.
Blinding may not always be easy or possible, and it goes much further than just keeping the name of the treatment hidden. Differences in appearance of the drug used in the study could give a clue to its appearance. Differences in taste, smell, mode of delivery, or even colour may also affect perceived efficacy, so these aspects should be identical in each treatment group. PHASTAR Senior Statistician, Stephen Corson, recently published a paper with the results from a study aiming to investigate whether adding levomenthol to an ibuprofen gel could reduce the time taken for an analgesic effect to occur in patients with soft-tissue injuries.[1] Menthol produces the sensation of cooling without reducing skin temperature and following topical application, menthol has an anaesthetic effect. Menthol can also enhance the skin penetration of topical analgesics, potentially increasing their effectiveness in relieving pain. But how do you blind menthol? That was the challenge that Stephen faced when designing the study procedures for this trial. Here he explains how these issues were addressed.
I happened to fall into statistics by accident. By my third year at Lancaster University I was starting to panic about what I wanted to do and after selecting mainly statistics-based modules for my third year, I realised this was something I enjoyed. I stayed at Lancaster to complete my MSc in statistics, and chose the medical pathway for my module options, studying topics such as principals of epidemiology, clinical trials, and longitudinal data analysis. I found this offered me a good foundation of statistical knowledge and have often referred back to my lecture notes while at work.
I am now one year into my graduate role at PHASTAR, and it has been one of the quickest years of my life. I joined PHASTAR after graduating from the University of Strathclyde with a Masters’ degree in Mathematics and Statistics. I sought a job where I would be challenged, and I was not going to lose all the knowledge and skills I gained from University. And I was not disappointed.
When I applied to work at PHASTAR, the entire process, from initial application to accepting the job offer, was very straightforward. A group of us took part in an assessment centre which involved several different tasks and a final interview. The day was, overall, very laid back – which was very different to assessment centres I had attended previously. I could tell from this day alone that the company was a very social company – with managers and employees happily chatting to all of us and making the experience as relaxing as possible. The final interview was obviously very daunting for me, as most graduates would agree when interviewing for a graduate job. However, the interview, like the rest of the assessment centre, was uncomplicated and not intimidating whatsoever. I heard back shortly after, and happily accepted the job offer.
The day will explore the theme of 'how statistical analysis and data science can enhance drug and diagnostic device development' and will feature talks given by PHASTAR staff and external experts.
Featuring talks from:
'What is Health Data Science? One biostatistician’s answer' by Professor Peter Diggle, Distinguished University Professor at CHICAS, Lancaster University Medical School.
'What sample size? 101 other ways statisticians influence trial design' by Andrew Lloyd, Head of Statistics - PHASTAR.
'Extracting value from real-world data: an example with wearables' by Professor Jennifer Bradford, Head of Data Science - PHASTAR.
'You've developed your AI tool - now how do we test it?' by Professor Jennifer Rogers, Head of Statistical Research - PHASTAR.
'Ensuring sponsor oversight, and quality of data and interpretation for early phase trials' by Dr Susan Lovick, Statistical Technical Lead - PHASTAR.
The event will take place on Tuesday 26th November, 9:30am-4pm at the Alderley Park Conference Centre. Registration will start at 9:30am and the sessions will run from 10am-4pm, followed by a networking reception at 4pm-6pm.
If you are interested in attending sign up is free and can be found here!
