Student profile: Kelsey Grantham

Dr Kelsey Grantham

PhD: Statistical methods for the design and analysis of longitudinal cluster randomised trials

 Kelsey Grantham

My PhD was really about how researchers can work smarter, delivering trials that are complex and scientifically robust, but are lean on budgets and reduce resource wastage.

Dr Kelsey Grantham undertook her PhD within our Biostatistics team, graduating in 2021. Her projects were methodological, and focused on trial design, specifically around cluster randomised trials (CRTs).

Armed with a solid foundation of mathematics and a Master of Applied Statistics, Kelsey was always  interested in the application of statistics to health and healthcare. She was able to explore this under the supervision of Professor Andrew Forbes, Associate Professor Jessica Kasza and Professor Stephane Heritier.

CRTs are clinical trials testing an intervention that is administered to groups of people, called “clusters”, such as hospitals, schools or villages. They often lend themselves to testing new practices or approaches to care at a system level, rather than testing a new drug for physiological effects on individuals.

She says, “My PhD was really about how researchers can work smarter, delivering trials that are complex and scientifically robust, but are lean on budgets and reduce resource wastage.

“For instance, say we wanted to design a trial to test an intervention that can be repeatedly applied and removed, such as the use of new equipment for patients in intensive care units. A cluster randomised crossover design may be appropriate, where each ICU in the study would cross over between the intervention condition (use of the new equipment) and the control condition (usual practice) a certain number of times. I sought to address a number of questions that arose in the planning of these types of trials:

“Is there a benefit in terms of statistical power to crossing over between the intervention and control conditions more frequently? If so, how many crossovers would give the majority of the benefit?

“I derived mathematical results and wrote code to compare the efficiency of different designs. I also developed a web app so researchers can input their own trial parameters and cost constraints and explore the efficiency of designs with different numbers of clusters and crossovers. Other projects within my PhD focused on guiding researchers in making realistic modelling assumptions and using appropriate methods to analyse trials with a small number of clusters.”

Kelsey’s findings have been cited by several published studies as part of their justification for selecting specific trial designs and making certain modelling choices. They are also serving as stepping stones for further methodological developments.

The diversity of research undertaken by her fellow PhD students was something that enriched Kelsey’s experience with the School. “I was really inspired hearing about the ways in which other students were tackling complex questions in health, and learning more about public health and medicine through some fascinating projects such as assessing the impact of clean water initiatives in rural areas, modelling tuberculosis transmission, and forecasting cancer incidence. It was a really collegial environment.”

Find out more about our PhD program.

Search for current PhD opportunities within our School.