Researchers at the Nuffield Department of Primary Care Health Sciences have published an important update to their BMJ study on unexpected weight loss as a sign of cancer, after identifying and correcting a methodological issue in their original 2020 analysis. The revised findings, based on a much larger dataset, provide stronger evidence about when GPs should investigate unexpected weight loss.
The updated analysis, published in the BMJ, examined records from over 326,000 patients - a significant increase from the 64,000 in the original study. The researchers discovered that 4.8% of patients with unexpected weight loss received a cancer diagnosis within six months, compared to the 1.4% previously reported.
"During further work to validate our original findings in another dataset, we identified an unintended selection bias in our original analysis," explained Dr Brian Nicholson, GP and Associate Professor at the Nuffield Department of Primary Care Health Sciences. "We worked with the BMJ to correct this issue promptly, leading to these updated findings which provide a more accurate picture of cancer risk in patients presenting with unexpected weight loss."
The revised results show that unexpected weight loss should trigger cancer investigation in all men aged 50 or over and women aged 60 or over, as well as younger patients with additional symptoms. This represents an important change from previous findings that suggested investigation was only warranted in more limited circumstances.
"These updated results have important implications for clinical practice," said Dr Nicholson. "We hope they will be taken into account when national and international guidelines are updated regarding which patients should be urgently referred for cancer investigation."
This update demonstrates the importance of rigorous scientific validation and the commitment to ensuring research findings are as accurate as possible to support clinical decision-making.
___