All Times EDT
Keywords: clinical significance; statistical significance; safety; decision-making; benefit risk
In clinical studies, a result with statistical significance is often misinterpreted as a clinically important result. However, statistically significant results are not necessarily clinically significant, and vice versa. Statistical significance is used to measure the probability of a study's results being due to chance while clinical significance is used to determine whether the results of the trial are likely to impact current medical practice. In this talk, Mengchun Li will be sharing some examples, in related to balancing statistical findings and clinical findings in decision-making. The examples include situations where a statistically significant result may in and of itself not be clinically significant because the clinical effect is small but may have achieved statistical significance only because the sample size was so large. There are also situations where two treatments were found to be statistically significantly equivalent in terms of efficacy, a strong result, where it’s hard to imagine the clinical and statistical judgments diverging from efficacy point of view. Then there’s the safety side where there is a trend suggesting a clinically significant difference in a risk but, as is common with safety assessments, statistical significance of the difference in risk could not be established because the study was not powered for the specific safety endpoint. Should you run another study to obtain statistical significance on the safety endpoint, or is a qualitative analysis sufficient to guide the decision confirming the risk? In conclusion, both statistical and clinical considerations are important in clinical trial design. Planning for how the results will be assessed should be done early in the program to achieve efficiency in the benefit-risk assessment, and realize the scientific value of the clinical program. It is also important to have ongoing conversations with regulatory health authorities concerning the decision-making process.