Statisticians Essential To Success of Clinical Trials

May 20, 2013

in At Work With Statisticians

By Dr. Jennifer Rogers

I have been working at the London School of Hygiene and Tropical Medicine as a research fellow in the Medical Statistics department for the last two years. My research is primarily concerned with the planning, statistical analysis and reporting of clinical trials in cardiovascular disease alongside other statisticians, clinicians and industry collaborators.

My involvement in the analysis of clinical trial data is mainly focussed on secondary analyses. These analyses allow us to identify areas where standard statistical practice becomes unsuitable and new methodology needs to be developed in response to interesting characteristics of the datasets with which we are presented. The constant development of new statistical methodology is essential to ensuring that future clinical trial data collected is utilised to its fullest so that meaningful clinical conclusions can be drawn.

My current research has been focussed on exploring existing methods for the analysis of recurrent hospitalisations for worsening heart failure, as an alternative to composite endpoints that combine fatal and non-fatal outcomes and typically analyse time to first event. An increase in heart failure hospitalisations typically results in an increased risk of death, so I am also developing novel statistical models for the analysis of hospitalisations, where the associated time to death is a potentially informative dropout time.

One important role that I have is my involvement on data and safety monitoring boards (DSMBs). A DSMB is an independent group of experts who monitor patient safety and treatment efficacy whilst a clinical trial is ongoing.

I typically act as the independent statistician who prepares the DSMB reports to be shared with the rest of the committee, and my primary mandate is to examine any differences in adverse events between the two treatment groups.

I also collaborate with pharmaceutical companies, offering advice for the planning of clinical trials. My role as a statistical consultant mainly has been to provide guidance in choosing a suitable primary endpoint and associated statistical methods for examining the efficacy of a new treatment or device. I also have been involved in simulation studies that have been used to assess sample size considerations.

There are many tools and skills that are crucial for successfully carrying out my roles in clinical trial research. I must possess the skills to identify areas where future research is needed and have the necessary mathematical and computational competency to develop novel methodologies.

An appreciation of the application and clinical impact of new methods that are developed is essential. Sound communication skills ensure successful dissemination of new methodology to both statistical and non-statistical audiences. An ability to communicate complex novel statistical models to clinicians is crucial for their utilisation in future clinical trials.

Rogers is a research fellow at the London School of Hygiene and Tropical Medicine and a member of the Young Statisticians Section of the Royal Statistical Society.