|Year : 2021 | Volume
| Issue : 5 | Page : 1513-
Statistical Significance and Clinical Importance
Venugopalan Y Vishnu1, Pulikottil W Vinny2,
1 Department of Neurology, AIIMS, New Delhi, India
2 Department of Neurology, INHS Asvini, Mumbai, Maharashtra, India
Venugopalan Y Vishnu
Department of Neurology, AIIMS, New Delhi
|How to cite this article:|
Vishnu VY, Vinny PW. Statistical Significance and Clinical Importance.Neurol India 2021;69:1513-1513
|How to cite this URL:|
Vishnu VY, Vinny PW. Statistical Significance and Clinical Importance. Neurol India [serial online] 2021 [cited 2022 Jan 17 ];69:1513-1513
Available from: https://www.neurologyindia.com/text.asp?2021/69/5/1513/329535
Dr. Prasad has elegantly described the importance of the difference between “clinical importance/significance” and “statistical significance/detection.” We want to emphasize that apart from others, statistical significance depends profoundly on the sample size of the study. So, when the sample size is enormous, any clinically trivial difference (like a 1-day difference in headache-free days in migraine or five days difference in survival between two anticancer drugs) can turn out to be “statistically significant/detectable.”
There are four possibilities while considering clinical and statistical significance.
1. Statistically significant and clinically significant
2. Statistically and clinically not significant
These two scenarios are easy to interpret as they are self-explanatory
3. Statistically significant but not clinically significant
If one increases the sample size considerably, any trivial difference can become statistically “significant/detectable.” For example, before eating, if one uses a very high-power microscope, one can always find “dirt” in hand/spoon, however clean it is. It's up to the individual (in our case, clinician and patient) to decide whether it is clean enough to eat.
4. Statistically not significant but likely clinically significant
This scenario could be due to underpowered study (small sample size) or due to flaws associated with design or execution as described by the author.
We eagerly look forward to future series of articles under the “Know your vital statistics” section.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
|1||Prasad K. Statistical significance versus clinical importance. Neurol India 2019;67:1513-4.|