Based on published collaborative research co-authored by Dr Arash Ardalan.
The research question
How could time from surgery to recurrence be analysed alongside patient and tumour characteristics?
Study context
The study examined records collected between 2010 and 2021 at a breast cancer centre in Tehran. Of 2,246 records examined, 171 cases met the study's inclusion criteria.
Statistical approach
- Compare parametric and semi-parametric approaches to modelling time to an event.
- Assess model fit using the Akaike Information Criterion (AIC).
- Use Cox regression to examine associations with patient and tumour characteristics.
What the analysis provided
The publication compared model fit and reported associations involving age, progesterone receptor status, tumour grade and tumour stage in the included cases.
These were observational data from a selected study sample. Reported associations require interpretation in that context and do not establish causation.
Practical lessons for your project
- Define the event carefully
- Agree the starting point, the event of interest and how follow-up and censoring will be handled.
- Choose models deliberately
- Model choice should reflect the research question, assumptions and diagnostics alongside fit statistics.
- Explain the findings clearly
- A research report should connect estimates and uncertainty to the question, with a clear account of the study's limitations.
Original publication
Monireh Shahmoradi; Ahmad Fazilat; Mostafa Ghaderi-Zefrehei; Arash Ardalan; Ali Bigdeli; Nahid Nafissi; Ebrahim Babaei; Mahsa Rahmani
Unveiling Recurrence Patterns: Analyzing Predictive Risk Factors for Breast Cancer Recurrence after Surgery
Cancer Informatics · 23, 11769351241297633 · 2024