News Results & StatisticsI'm not going to bash this stock (as most of you right away will presume I'm doing).... but.... anyone interested in statistics will know that in order for results to be "scientifically significant" you need to have a "p-value of <.05" but if you want to have a greater degree of RELIABILITY (a statistic & scientific term), you need to have a "p-value of <.02 or below".
Having not seen any of these in the pre-liminary data, it is hard to state whether the results obtained are truly "of value". Furthermore, it becomes more difficult to statistically prove the efficacy of the treatment when only 20 patients are enrolled with a 3:1 ratio (i.e. 15 therapy:5 placebo).
The results also didn't indicate whether the placebo group experienced the same type of results. In fact, you can't call these patients "a group" as there was ONLY 1 patient in Type 1 diabetics used as "a control" group.
Basically I'm trying to state these points:
* Sample size is too small to generate "generalizable" conclusions
* Clinical significance of the drug based on statistical significance is as of yet not available. Furthermore, clinical significance comes A Lot later than statistical as patients have to be followed for a number of years to evaluate the drugs' overall efficacy.
On the plus side:
* This type of study is usually used as a "pilot" study for a larger one. Should the results of this study prove to be "significant" a larger, more diverse study design becomes warranted.
* The double-blinding in this experiment is EXTREMELY important as neither the participant, nor the experimenter knows which group a specific patient belongs to. Thus, eliminating major various sources of experimenter bias and "learning effects".
* 35% MAY in fact be a significant value as the appropriate statistical analyses are undertaken. Furthermore, 35% reduction may pose a SIGNIFICANT improvement in QoL (Quality of Life) for individuals with diabetes.
* 14.3% drop out rate (as I read was already calculated) may or MAY NOT be a large number. However, the importance of this number will depend on the "generalizability" of the findings.
Therefore, as a final note, the results are encouraging, but based on the stage of the study and the sample size, it may be a bit short-sighted to start screaming, "Hooray!".
Thank you for your time.
Sincerely,
D-man