Eliciting and incorporating patients’ opinions about missing data in randomised controlled trials

Every year thousands of clinical trials take place to determine how safe and useful medical interventions are for patients. Missing data is something that all clinical trials have. Simply put, it is when measurements meant to be collected from participants are not. This can pose a problem when trying to understand the results from the clinical trial.

To understand this, imagine that a clinical trial is a puzzle. During the trial participants have measurements taken to assess their health, with each measurement being a puzzle piece. Despite careful planning, their healthcare can be disrupted, preventing them from giving all their puzzle pieces. This means that at the end of the trial the clinical trial puzzle has gaps, making it hard to understand what the clinical trial results mean, making it harder to know how useful or safe that medical intervention is for patients.  While there are ways to handle these missing pieces, decisions are made without input from patients and the public. 

This research team have been working with members of the public to design a way to get patients involved in decisions regarding missing data. The research is still ongoing with two current projects ongoing. 

• Project 1:  What did I miss? (a series of workshop to develop a method)
• Project 2:  Rethinking missing data with Patients  (test application of this method)

Alongside the patient focus this research is primarily methodological. Our research question can be written as so:

- “How can we adapt an existing method so that patients can help us “fill-the-gaps” in our clinical trial puzzles?”. 

The method that we have been focusing on is called “expert elicitation”, which has been used previously to “fill-the-gaps” of historical clinical trials. An exploration of the literature around this method was conducted at the start of the thesis with results to be published. 

The intended output of this thesis is to have an adaption of this method, which can be used to produce a patient-informed approach to missing data modelling. Alongside the method there will be a series of resources for implementation and also a series of recommendations for applying this method in this patient context.