The MODE Study “One Sheet of Paper” Analysis Day

SARAH PRICE Senior research fellow in the Cancer Discovery Group, University of Exeter

Background to project

Cancer can be difficult to diagnose, especially when patients have symptoms that have a range of causes, many of which are more common than cancer. Family doctors might reasonably start with a common non-cancer diagnosis that matches the patient’s symptoms. In people who are diagnosed with cancer soon after, these initial non-cancer diagnoses might be called an “interim diagnosis”. Sometimes interim diagnoses are good clinical practice, considering what was known at the time; however, sometimes they may lead to delayed cancer diagnosis and poorer outcomes. In the MODE Study (NIHR SPCR grant 680), three SPCR institutions are taking a mixed-methods approach to help understand when interim diagnoses might represent missed opportunities to diagnose cancer.  The University of Exeter team is characterising patterns of interim diagnoses using large observational datasets. The Queen Mary University of London and The University of Oxford teams are using qualitative interviews to understand the perspectives of patients and of family doctors and other general practice staff, respectively. We got together in April 2025 for a fantastic day of results sharing and used the One Sheet of Paper method and insight from experienced primary care clinicians to start to understand what our rich data are showing.

Who was there

Researchers from University of Exeter were MODE Study co-leads Professor Richard Neal and Dr Sarah Price, along with Anne Spencer, Professor of Health Economics, Professor Willie Hamilton, Dr Luke Mounce and clinical academics Dr Deepthi Lavu and Dr Judit Konya. Also from Exeter was Angela King, the study PPI co-applicant and Chair of the PPI Advisory Group. From Queen Mary University of London, we were joined by Professor Fiona Walter, Dr Georgia Black and Dr Mel Ramasawmy. From University of Oxford, we had Dr Claire Friedemann Smith and Dr Luke Robles.

Aim of the day

Our main aims of the day were to review the findings from the three studies and get clinical input to understand patterns in the data. This would allow us to start integrating the three datasets and characterise how interim diagnoses might arise.

What we did and what we learned

We focused on getting clinical input on initial themes emerging from the data. We talked about how our findings might sit alongside Jess’s Law and its message of “Three strikes and we rethink”. We discussed how people deemed at “low risk” of having an undiagnosed cancer are treated and what might trigger general practitioners to increase their suspicion of cancer. We also talked about the options that general practitioners have for people whose presentations do not meet criteria in NICE suspected-cancer recognition and referral guidelines, and what happens if a person does have an undiagnosed cancer but is sent on an urgent referral pathway for a different cancer site. The discussion extended to covering how non-cancer diagnoses are made – clinically or based on tests or on imaging – and how this affects the certainty placed on that diagnosis. The feeling was that clinical diagnoses may be less certain than those made based on test results or imaging. That being so, “clinical” diagnoses may be more likely candidates for missed opportunities and associated with delays in cancer diagnosis. We also talked about how interim diagnoses might be questioned in the context of diagnostic certainty, covering topics including self-advocacy, health literacy, deprivation, false reassurances around non-cancer diagnoses, and patient—doctor mismatches.  Continuity of care was a big topic too, both what it means – always seeing the same person or continuity of information in the notes – and its benefits versus a fresh pair of eyes. Finally, something that was important to patients was the variation in service provision and the difficulties in getting appointments.

With these ideas buzzing in our heads, we worked in groups to review each team’s findings. Georgia Black, Richard Neal, Anne Spencer, Deepthi Lavu and Judit Konya worked on the Exeter quantitative findings. This work has produced many, many results and, so far, it had been hard to choose the right framework for interpreting them. With the clinicians’ help, all potential interim diagnoses were grouped by how they are diagnosed – clinically, based on test results, or on imaging. It's rare to have a whole day with GPs who have collectively more than 100 years of clinical experience (if not more, I suspect) all focused on the results, and we were keen to make the most of it.

Next, we mapped all the interim diagnoses to their potential impact on patients, for example advanced-stage cancer diagnosis or poor 1-year survival. Watching Georgia apply her expertise of qualitative analyses (such as the One Sheet of Paper method) and knowledge of the patient interview findings to this work was amazing.

Meanwhile, Mel and Angela were applying the One Sheet of Paper approach to the results of University of Oxford interviews with clinicians and Claire and Luke Robles took a similar approach to the QMUL data from transcripts of patient interviews.

Together, we weaved the results together and soon found that our theory that clinically made interim diagnoses were more likely to be associated with delays in cancer diagnosis did not hold water. Instead, we saw different patterns in the data, and these gave insights into how interim diagnoses might arise. By the end of a long and incredibly productive day, we had the foundations of our synthesised results – patterns for how interim diagnoses may come about along a person’s journey to a cancer diagnosis. As someone whose focus is quantitative research, just one day working this closely with qualitative researchers, alongside so many clinicians and Angela enriched my understanding of the data beyond expectations. I wish that I had had this experience much earlier in my career.

We continue to refine this work, illustrating the different patterns of interim diagnoses with examples from the qualitative and quantitative data. We held a stakeholder event in October to share our findings with clinicians, patients, members of the public, cancer charities, cancer alliances, community action groups, pharmacists, allied health professionals and others with an interest in missed opportunities to start testing for and diagnosing cancer.

Please follow us for updates: 🦋 Bluesky: @apexcomms.bsky.social