What is it about?
A study analyzed 32 statistically significant test results from nuclear medicine clinical trials. It found the typical cutoff between normal and abnormal values was just 0.66 standard deviations from the mean. At that point, 34% of healthy individuals would be misclassified as abnormal, showing statistical significance does not reliably indicate clinical usefulness on an individual level.
Featured Image
Photo by Martin Martz on Unsplash
Why is it important?
This study highlights an important shortcoming of relying solely on statistical significance to guide clinical practice - statistically significant findings may still fail to accurately classify individuals as normal or abnormal. The findings serve as a sobering reminder that statistical significance does not necessarily translate into clinically meaningful differences that benefit real patients.
Perspectives
As a clinician, I found this study eye-opening. It's easy to assume that a statistically significant finding represents an important breakthrough. Yet we must remember - that statistics operate in probabilities, while patients require more definitive assessments. A statistically significant cutoff may still harm 1 out of 3 healthy individuals by mislabeling them as abnormal. This study teaches me to consider clinical context and real-world implications, not just statistical outputs, in applying research findings. Statistical rigor is important, but clinical relevance is the ultimate benchmark.
Thomas F Heston MD
University of Washington
Read the Original
This page is a summary of: How often are statistically significant results clinically relevant? Not often, Authorea, Inc.,
DOI: 10.22541/au.151140118.82849134.
You can read the full text:
Resources
Contributors
The following have contributed to this page
