No. You may achieve statistical significance simply by having a very larger sample even if the clinical differences are quite small. Large sample means smaller sampling error.

Only the researcher can decide the acceptable clinical difference.

Please see below for my previous discussion on this: 

Dear Editor,

We read with interest the article by Gibbs and Gibbs on the misuse of the word trend to describe ‘almost’ significant p values [1]. However, we believe this discussion of semantics is clouding a somewhat greater issue in the reporting of statistics in the scientific literature. As Gibbs and Gibbs noted, p values are inferential statistics that give the probability of obtaining a value the same or greater than that found if the null hypothesis was true. However, we argue that p values are outdated and for clinical studies in particular, their use should be avoided. This is not a new concept, however, we hope this letter will serve to remind the readership of the flaws in null hypothesis statistical testing. We will first highlight the problems related with the [mis]use of p values, and then discuss the advantages of estimation-based methods, illustrating this with a theoretical example.

Firstly, the use of p values can often detract from a more important issue when conducting a clinical study, the assessment of clinical significance. The misinterpretation of p values means that readers may mistake ‘statistical significance’ as ‘clinical significance’. As the calculation of p values is heavily dependent on sample size, large studies may demonstrate very small p values that are not clinically significant. In fact, such small p values may in fact be evidence against the use of a particular treatment, as it can make us more confident that a treatment will not have a clinically significant effect (see later example). Secondly, the choice of a significance level of p0.05 (if 95% CI are used) [7]. Moreover, confidence intervals can be better used to assess the likelihood an intervention has a clinically significant effect. Consider an example, we wish to know the efficacy of two different analgesic agents (x and y) for treating postoperative pain. We undertake two randomised controlled trials (RCT) with both agents and pre-determine a clinically significant reduction in pain as 15mm (on a 100mm VAS) [8]. The first RCT with agent x enrols a large number of participants and demonstrates a mean difference of -5mm (95% CI -3mm to -7mm; p

Statistical significance means only that the results would be unlikely to occur in the absence of a "real" effect (i.e., by chance alone). It is an interactive product of effect size and sample size (see Robert Rosenthal's classic papers and research methods textbooks). Clinical importance, of course, is entirely a matter of effect size. So if one is working with a large sample, relatively small (even trivial) effects may attain statistical significance. Significance tests are mainly a way of keeping researchers honest (not allowing claims to be made based on potentially random outcomes), but they are most valuable when we are working with moderate size samples. 

Statistically significance does not necessarily mean clinical significance. it mainly reflects the true outcome of a research process as originally designed. It is also a function of other factors such as the sample size, degree of validity, reliability, sensitivity of clinical research instruments

Statistical significance or association between variables does not mean causal effect. 

No. That's why effect size was invented. It gives you the clinical meaning of the difference between groups, not the statistical meaning.

Thank you for all above helpful comments.

I'm no expert, but I'd share what I know.

Statistical significance refers to how certain you are regarding the values of your findings.

How clinically relevant your study is, depends on the effect size, simply speaking, regarding how big the difference really is between the findings. 

You could be 99.999% sure that potassium in group A is 3.75 mmEq/l and 3.77 mmEq/l in group B (p value < 0.0001). You have an statistically significant result, nonetheless a useless clinical/laboratorial diference. 

Pretty good discussion of the the clinical significance and the statistical significance. I take issue, however, with the statement that p-values are outdated. If they are, then so are confidence intervals, because p-values use the same mathematical values as CIs do. One calculation is sort of "inside out" of the other. Both use means and standard deviation to decide statistical significance. If you use a 95% CI then you are using an arbitrary p≤0.05. If you use a 99% CI then you are using an arbitrary p≤0.01.

These days we can calculate very precise p-values. Certainly knowing that the probabilities of results of a study are 0.00018 is more informative than p≤0.05 or even p≤0.001. CI are always based on an arbitrary p-value.

I do agree that CIs are in many instances in medicine more useful than simple p values for precisely the reason that an interval is more useful than a point estimate. But that has nothing to do with the difference between clinical and statistical significance. In every endeavor in science one must understand the question and the data to decide if statistical significance is important. Very large data sets may have very small p-values. But if a very small clinical significance makes a big outcome in the patients outcome then you would need to have a very large sample to detect the difference. For example finding a statistical significant difference in treatment effect for a very rare disease will require a very large sample.

Also I believe that RA Fisher eschewed arbitrary p-value limits and may have jokingly suggested such a device. Also remember that finding exact p-values was rarely possible in Fisher's time.

Except of course for the Fisher Exact Test! Thanks for bringing in the important subject of confidence intervals, Patrice.

Hi., it is depend on the context and study objectives and how the data has been setup ... Results are not necessarily to be important when they are statistically significant ..... !! especially in cohort studies (randomization is not there /matching)...

It depends on the independent variable and dependent variable. And also the study design too

In short, clinical significance is the primary question. So if the statistical significance is based on a very large sample, the clinical difference could very small, too small to make a medical difference. It is the effect size that is clinical meaningful and depends on knowing the science and precisely what you are aiming to find out. The if the desired effect is statistically significant as well you have your answer.