The University of Queensland’s Clinical Trials Capability (ULTRA)

Building capability for improved trial outcomes

Our mission within the The University of Queensland’s Clinical Trials Capability (ULTRA) program is to accelerate UQ's capability, capacity, coordination and competitiveness in innovative and impactful randomised controlled trials (RCTs).

RCTs are essential to transforming healthcare and public health, however traditional trials are often expensive, difficult to recruit to, and take too long to translate results into clinical practice. The ULTRA program has been created to transform RCTs – building capability and capacity in innovative trial design, conduct, analysis and translation.

Our approach embeds these capabilities within clinical trials to strengthen their design, delivery and impact.

Funded through The University of Queensland’s Vice-Chancellors Health Research Accelerator (HERA) initiative to tackle urgent health and medical challenges, ULTRA is working together with the UQ’s Clinical Trials Centre (CTC) to augment existing expertise and accelerate and innovate RCT capability for UQ staff members and Academic Title Holders.

View the below videos to learn more about ULTRA and the value of embedding clinical trials into the health service.

ULTRA projects

Biostatistics

Joint modelling of longitudinal health-related quality of life and time-to-event data under non-proportional hazards

Health-related quality of life (HRQoL) is an important endpoint in advanced cancer trials. A key challenge is that HRQol assessments often stop at progression, leading to potential informative censoring. Joint modelling approaches explicitly link the HRQoL trajectory with time-to-event (TTE) process, and can therefore account for this type of informative censoring. Using a shared random effects structure, we proposed a joint model that accounts for the endogenous nature of HRQoL trajectories by combining a linear mixed submodel for repeated measures with a flexible accelerated failure time submodel for TTE outcomes. The framework relaxes the proportional hazards assumption and is evaluated through simulation studies under a range of trial scenarios. Application to cancer trial datasets demonstrates its ability to provide robust inference.

Quick estimates in missing data structures with Variational Bayes

This project examines the use of Variational Bayes (VB) as a practical middle ground between traditional methods for handling missing data in clinical trials. The Expectation-Maximisation (EM) algorithm is fast however often limited to point estimates, Markov Chain Monte Carlo (MCMC) provides the exact full uncertainty quantification often desired, however at a high computational cost. VB provides a balance in approximation of the posterior distribution efficiently. Through simulations VB has been shown to capture key features of missing-not-at-random data structures in clinical trials at substantially faster speeds than MCMC, with some parameters (particularly relating to variance) having some difficulty. Various VB methods will be explored testing accuracy, speed and applicability in order to highlight the promise of this approach for exploratory clinical trial analysis.  If you are interested in this field of research please contact Senior Research Assistant in Biostatistics Patrick Maher at ultra@uq.edu.au.

SAFTEP: Semiparametric Accelerated Failure Time Evaluation Project

The Cox proportional hazards (PH) model is widely used in clinical trials but may be unsuitable when the proportional hazards assumption is violated, and alternatives such as stratification or restricted mean survival time (RMST) have notable limitations. The semiparametric accelerated failure time (AFT) model provides a promising alternative, as it does not rely on the PH assumption and offers a more intuitive interpretation of covariate effects. The performance of AFT models relative to PH models is illustrated using ANZUP trials.

Implementation science

Use of implementation science to design site readiness assessments for investigator led multi-site RCTs

Can we create better ways for trial teams to successfully support multiple RCT sites?

In multi-site RCTs, approaches to trial site initiation can be fragmented and site performance unpredictable which contributes to costs and research waste. Optimization of trial site readiness assessments may improve the success and translation of clinical trials. This study will design, validate and evaluate an implementation science informed site readiness assessment protocol to support trialists in RCT site recruitment. If you are interested in piloting this process in your multi-site trial, get in touch with ULTRA Principal Research Fellow in Implementation Science Associate Professor Ingrid Hickman at i.hickman@uq.edu.au to discuss collaboration.

Equity in RCTs

RCTs are the most widely accepted research design for evaluating the effectiveness of health interventions, yet much of the evidence produced through RCTs is drawn from participants with a narrow demographic profile that fails to reflect the diverse demography of the Australian population. ULTRA is tackling the issue of equity in RCTs with a robust implementation science approach and co-design solutions to drive impact and translation. If you are interested in collaborating on this body of work, please contact Postdoctoral Research Fellow in Implementation Science Dr Mark Liu at mark.liu@uq.edu.au. 

Implementation science in hybrid effectiveness-implementation RCTs

Hybrid RCTs incorporate implementation outcomes as well as the typical effectiveness outcomes which should lead to better trial translation. This project will explore the utility of implementation science methods in hybrid trials and the impact on translatability of outcomes. For further information contact Principal Research Fellow in Implementation Science Associate Professor Ingrid Hickman at i.hickman@uq.edu.au.

Qualitative research

Qualitative research within randomised controlled trials

Qualitative research can add value to randomised controlled trials (RCTs) yet little is known about its scale or impact.  We are working towards understanding the current landscape of the use of innovative qualitative methods during RCTs and the impact of its contributions. If you are interested in this topic, get in touch with ULTRA Research Fellow in Qualitative Research Dr Nathalia Costa to discuss collaboration at n.cordeirodacosta@uq.edu.au.

Consumer and community involvement

Bringing the James Lind Alliance Priority Setting Partnership approach to Australia

James Lind Alliance (JLA) Priority Setting Partnerships (PSPs) bring together consumers and clinicians to identify and prioritise unanswered questions (‘uncertainties’), informing researchers and funders of the issues that matter most to those with lived experience of health issues and those who care for them. Each JLA PSP must be supported by a JLA Advisor, who champions inclusivity, transparency & scientific rigor. With no JLA Advisors based in Australia, Australian researchers have limited access to this methodology. As a first step towards establishing an advisory team in Australia, this project will pilot UQ-based facilitation of the final prioritisation workshop in three current PSPs. If you are interested research priority setting, please contact Senior Research Fellow in Consumer and Community Involvement, Dr Bec Jenkinson at r.jenkinson@uq.edu.au.  

Consumer and Community Involvement in randomised controlled trials

Consumer and Community Involvement (CCI; also known as Public and Patient Involvement and Engagement in some contexts) occurs when people with lived experience of a health concern or service collaborate with researchers to prioritise and shape decisions about the design, conduct and dissemination of research. Given the range of benefits of CCI, funders and publishers are increasingly expecting CCI to form part of the research process, providing a useful lever to grow the practice of CCI. This ULTRA program of work aims to generate evidence to grow innovative, sustainable CCI in the context of RCTs. If you are interested in collaborating on this body of work, please contact Senior Research Fellow in Consumer and Community Involvement, Dr Bec Jenkinson r.jenkinson@uq.edu.au