Theme aim

To develop methods to ensure that RCT funding is targeted at the questions where it is most needed and that RCTs, once funded, are designed to collect complete and unbiased evidence on the cost-effectiveness of the intervention.

Major achievements

• We reviewed Resource Use Measures (RUMs) and used a Delphi survey to identify key items (e.g. GP appointments or outpatient visits) that should be included in a standardised RUM. This RUM is being developed and validated in ongoing PhD project work. The aim is to make economic evaluations alongside RCTs simpler and more comparable.
• We surveyed health economists to develop a template for a Health Economics Analysis Plan (HEAP) for use alongside RCTs.  HEAPs, drawn up in advance of the analysis phase, are an accepted means of reducing bias and increasing transparency of RCTs.
• We developed methods from financial mathematics to improve efficiency of value of information (VOI) calculations. The aim is to increase the use of VOI methods in practise by improving computational feasibility.
• We reviewed empirical evidence on biases in RCTs, and developed RoB 2 - a revised tool for assessing the risk of bias in RCTs, and ROBIS - a new tool for assessing the risk of bias in systematic reviews. We audited practice and reviewed current tools for assessing reporting bias.
• We explored how meta-analysis of previous RCTs can be accounted for in the design and analysis of a new RCT, surveying current practise and proposing a framework for future practice.
• We developed Model Based Network Meta-Analysis (MBNMA) methods to incorporate dose and time-course information in meta-analysis of phase-II studies, to inform design decisions for phase-III RCTs and Health Technology Assessments.
• We developed a method to assess sensitivity of recommendations based on network meta-analysis to potential bias and errors in the RCT evidence, and successfully piloted the method on a NICE clinical guideline. 
• We reviewed and critiqued methods to combine individual participant data from an RCT with summary data from RCTs to make population-adjusted indirect comparisons between treatments. We developed a new approach, Multi-Level Network Meta-Regression (MLNMR) , which overcomes some of the limitations of earlier approaches.
• We further developed and validated the ICECAP questionnaires to measure how health and healthcare affects people’s capabilities, with particular emphasis on exploring the use of the ICECAP Supportive Care Measure in RCTs focusing on care at the end of life.