Allison Drosdowsky is a researcher and PhD Candidate in the Department of General Practice and Primary Care and Centre for Cancer Research at the University of Melbourne. Or as she puts it – a statistics nerd who loves numbers.
Allison’s current research, which falls under the VCCC Alliance Data Connect program, examines the role of time to diagnosis and treatment in colorectal cancer to identify where there may be opportunities to make diagnoses sooner, and at an earlier stage, to provide the best chance to improve outcomes for people with cancer.
The Data Connect program enables researchers to link data from primary care and hospital settings, as well as data from clinical registries, to study cancer from pre-diagnosis to post-treatment care.
“Because of the data we are now able to access, thanks to the Data Connect process, this is the first Australian study to look at these questions. We examined data of people who were diagnosed with colorectal cancer after seeing their GP,” Allison said. “And what we were able to see is what is referred to as the ‘waiting time paradox’.”
Simply put, what you would expect to see when looking at cancer data is that the longer you wait for diagnosis and treatment, the longer cancer has to develop, and therefore the harder it is to treat and the lower the rate of survival. No surprises there.
However, what Allison also found was that a group of patients who were diagnosed and recommended for treatment early on had a low survival rate similar to those who weren’t diagnosed until later.
“We hadn’t seen this in Australian data, so that made us sit up and pay attention,” Allison said. “It shows the quick action from doctors/clinicians to diagnose and treat patients, but they don’t see improved outcomes, and we can’t definitively say why.”
While some research projects provide clear answers to some questions, others pose more, but this study did both. “We looked at variables such as gender, socioeconomic status, metropolitan and regional status, whether the person had colon or rectum cancer, and found none of these seemed to make any difference to how long diagnosis and treatment took,” Allison said.
“However, the data about the use of health care in the time before diagnosis and treatment for colorectal cancer patients did show some areas to guide optimal time frames for key events in the cancer pathway – such as a patient detecting bodily changes and deciding to see their GP – and look at potential interventions to reduce times.”
So, where to from here? Well, Allison would like to look at a larger sample size and examine the patterns of care – from presenting to a GP, to being referred to a specialist, to being seen by the specialist and finally, receiving a colonoscopy. She would also like to examine the different pathways to diagnosis, for example: public versus private healthcare, participating in the National Bowel Cancer Treatment program and those who underwent a ‘routine’ colonoscopy as part of their regular healthcare plan.
Allison would need to dig deeper to see the reasons why some who had a quick diagnosis still experienced poor survival outcomes.
“I would love it if I was able to use contemporary techniques such as ‘instrumental variables modelling’, which are methods that would allow me to assess causal relationships in these complex real-world datasets” she said. Spoken like a true data nerd.
