Significance of reactions to intradermal injection of autologous granulocytes, mononuclear cells, and serum.

A. Varelzidis,J. Turk
DOI: https://doi.org/10.1136/bmj.2.5510.379
1966-08-13
British Medical Journal
Abstract:It is on this account that we have developed in our clinic the concept of " determinacy " to contrast with " significance." If the result of a trial is determinant we mean that it will determine a course of action, and determinacy therefore depends on the degree of difference between two methods of treatment. Determinacy is not necessarily correlated with significance. For instance, it would be true to say that if the difference between two methods of treatment were significant, and the significantly better treatment was equally acceptable on other grounds, then determinacy and significance would be positively correlated. If, on the other hand, the trouble and anxiety associated with one method of treatment were far in excess of the other, and the difference between these treatments was not significant, then the trial would nevertheless be highly determinant in favour of the second. In planning our trials, therefore, we first decide what levels of significance we are prepared to accept, and, as we have seen, this need not be the same for one of the possible results as for the other, but we also decide what difference we are going to regard as determinant. In the standard " t " test this can readily be done by deciding what difference in means we would regard as determinant. Here again different levels of determinacy may apply, depending upon which of the two methods is to be preferred. Lastly, we must turn to the fallacy deriving from the heterogeneity of the population to be contrasted in a controlled clinical trial. If we took two populations of patients with advanced cancer of the breast, matched appropriately for age, stage of the disease, nature of the deposits, and so on, and conducted a trial on this population to see which was the more effective method of treatment-by androgens or by oestrogens-we might find that androgens were to be preferred, taking the population as a whole, and this result might completely swamp the undoubted preference for oestrogens in the older age group. By suitable stratification, and by examining groups within the population, such an effect would be detected, but this may not be possible. In the Medical Research Council trial to test the relative merits of cortisone and aspirin in the treatment of rheumatoid arthritis, no significant difference emerged between the merits of these two methods. This has often been wrongly interpreted as implying that the M.R.C. trial suggested that in any individual patient it did not matter whether you prescribed cortisone or aspirin. The trial was not in fact designed to test this, and it would be improper, therefore, to draw conclusions of this nature from it. If there are problems-ethical, scientific, and even mathematical-associated with controlled trials it nevertheless remains the case that this technique holds out greater promise for advance in therapy than any yet devised. More important, however, is it that recognition of the scientific basis upon which such trials are constituted will ensure, so far as is possible, that the undesirable state of affairs prevailing in medicine during the first half of this century will never be repeated to the extent of producing so many false trails and so many unnecessary and unworthy modes of therapy.
What problem does this paper attempt to address?