Psychometric sensitivity analyses for clinical trials

Background

Patient-reported outcome measures (PROMs) are questionnaires that measure aspects of a person's health, like knee function, or impact of back pain. They are often used in medical research to compare how well different treatments work. These research studies are important for changing what treatments are offered in the NHS.

But they are usually funded by the taxpayer and are expensive. They may have risks for people taking part (for example, you might get a treatment that turns out to not work). It is important that we can trust the results of the PROMs used.

Not all PROMs provide good measurements. For example, some PROMs may ask questions that are not relevant or meaningful. Some PROMs may make different scores that mean different things in different populations (for example, a PROM question "is it hard for you to feed yourself?" might depend on whether it is normal for you to use a knife and fork, chopsticks, or to eat with your hands). PROMs can significantly vary in quality, though this is often overlooked in trials.

We think that we need to use the best-quality PROMs in research to ensure accurate results, but we have not tested this yet. We may need to re-look at the results of some trials after "tuning up" the PROMs used in them already. We could also suggest techniques for researchers to test and account for problems in PROM measurements. Also, if the quality of a PROM does not matter that much in the reality, this is also important to find out.

Aim

The aim of this project is to understand whether the quality of PROM measurement affects the results of research studies. If it does, then by how much? We also want to produce guidance on how to check whether a PROM is producing trustworthy results in research. Methods

I am going to look at research studies that have been published in the last 5-years, and funded by the National Institute for Health and Care Research. I will use existing checklists to judge how well PROMs are being used and reported in these trials.

I will look closely at the results of some studies that have already been completed. I will tune up the PROMs used in them. Then I will repeat the analyses in these studies. I will then check whether my results match up to the original results that have already been published.

I will then work with experts that make PROMs, and scientists who use PROMs, to develop guidance on whether and how we should check the quality of PROM measurements in medical research. Patient and Public Involvement

I am going to work closely with patients and the public as they may have insights that I lack. For example, I will need to consider whether PROMs are asking relevant questions to patients in the trials. I will also need to consider whether questions are interpreted in the same way by different groups, and whether questions have been grouped together correctly to measure the same aspect of health.

My patient partners have lived experience of treatment and disease, and will be able to help me make judgements about this. Impact of Studentship

Through this sudentship I will develop advanced quantitative skills, patient-centered research approaches, and foster collaboration with leading experts in the field. Expanding my current skillset through this studentship will make me a well-rounded academic and support my journey towards research leadership.