I am quite familiar with statistical significance, but do not very understand the clinical significance I consulted the wiki but still felt confused with this concept. Could anyone explain for me what is clinical signivicance and what is the difference between clinical significance and statistical significance?
Statistical significance allows us to explore quantitatively whether differences between groups sampled from a population are due to chance or some true difference. While statistical significance occurs along a continuum, a p-value of less than or equal to 0.05 is generally considered significant evidence against the null hypothesis of no difference to suggest that true quantitative differences exist. In a rudimentary form, statistical significance answers a question related to the play of chance.
Clinical significance pertains to whether an observed effect is important in the treatment of a disease, determination of diagnosis or other comparative/descriptive statistic. Clinical significance asks: "Does this observed difference affect the approach to treatment or diagnosis for patients?"
An hypothetical example of statistical significance and clinical insignificance would be a new drug that reduces mortality by an absolute 0.1% compared to placebo (p=0.001) for a rare disease. The p-value is sufficiently low to achieve statistical significance, but routinely choosing the new therapy over placebo is not likely to improve health meaningfully given the small effect size of 0.1%.
An hypothetical example of statistical insignificance but clinical significance would be a new drug that reduces mortality by an absolute 55% compared to placebo (p=0.12). Here, the very large clinical effect may demand further study to determine whether a true effect actually exists.
I would also observe that the magnitude of effect determining clinical significance is influenced by where the problem addressed lies on the continuum between rare diseases and public health. Generally, the closer a disease is to a public health problem, smaller effect sizes will be accepted as clinically significant (think aspirin for heart attack).
Statistical significance is, well, statistical. But, clinical significance is practical.
There can be a high degree of certainty that one drug is minimally more effetive than another drug in attaTo make matters worse, its minimal primarry effect may be far outweighed by its side effects.
This is one reason that we in medicine and public health do not do some screening tests. They may secure only minimal benefit but lead to many false positives, so many false positives that more patients would be injured in investigations and therapies in pursuit of the false positives than would be injured by missing a few results by not doing screening.
There is no difference between statistical and clinical significance. The misconception is due to lack of statistical understandings. statistics is not hypothetical system, rather it is a real world system. if statistical significance doesn't match to clinical significance then no reason to accept statistical methodologies. if you have sufficient sample size, appropriate study design, reliable and appropriate tools; all together, if you satisfy scientific considerations of methodologies then both significance are equivalent, otherwise, not certain to say...