Analyzes and interpretation of the results of a trial
Statistics compare groups of figures and indicate the probability for the difference observed between two groups to be not due to chance.
Statistics do not give any indication on the origin or the cause of these differences, nor on the superiority of a drug over another. The conclusions depend only on the logic on which the methodology is based. In addition, a low variation, for example attenuation of a symptom of a few percent, statistically significant, can be clinically insignificant.
The interpretation of the results of only one parameter is generally difficult because if, for example, a drug causes a lower decrease of blood pressure than another, this drug can be regarded as less effective, or less brutal and thus more suitable. Moreover, this comparison of two drugs is valid only at the dosage used and it can happen that the optimal dosage of one or the other was not well defined.
The interpretation of the results of many parameters, in the sight of an overall assessment, is much more difficult. At the time of the comparison of two products, “the best” would be either most effective, or that which will have long lasting effect, or best tolerated, according to the weight that is given to each criterion. Indeed, a drug is seldom without any adverse effect and it is always necessary to take account of the benefit/risk ratio. This ratio is particularly difficult to establish for a drug which is used for several years for example. A cholesterol-lowering drug which decreases blood cholesterol but increases mortality could be regarded as beneficial if one took into account only the cholesterol lowering effect.
These observations explain the difficulties encountered to highlight in an objective way the differences (advantages, disadvantages) between two or several drugs in order to lead to a preferential classification for a given indication.
The method which consists in gathering and studying simultaneously the results coming from several independent studies about the same subject, is called meta-analysis. It is frequently used in medicine, in particular to try to evaluate the effectiveness of drugs.
The use of complex protocols does not mean that isolated observations of doctors, general practitioners or specialists, of nurses and of patients themselves, are without interest. These observations, if it is paid attention to, can allow to find a new indication or an ignored adverse effect of a drug.