Carrying out the trial
At this point our patients should have lost the effect of most drug therapies which they had been on and should be in a fairly suitable condition for the trial. Use of pain-killers will be helping somewhat with aggravated conditions that have arisen through lack of treatment. The reason for specifying that the pain-killers should not be of the aspirin type is that this type is also effective as an anti-arthritic treatment and would influence our results.
Our thirty patients are divided at random into two groups of fifteen and are administered either placebo or active substance over a period of eight weeks. Weekly checks on the condition of each patient are made and recorded. At the end of eight weeks we reverse the treatments, so that those who had been receiving placebo would now be treated with the active substance, and vice versa. Of course, no one other than the holder of the code knows which is which. Weekly conditions of each patient are recorded for the next eight weeks and the trial then ceases. We now have to analyze the data collected and compare active and placebo results for each patient. This is where the weakness of this trial will show up.
Assessing the results
As we are conducting an imaginary trial we will not have any actual results. It will therefore be necessary to look at the trial critically and see what sort of results we could get from it.
This patient receives no treatment at all (the pain-killers are not treatment for the disease) for the period of eleven weeks. By this time the patient’s condition has deteriorated to a considerable degree. Now the active substance is given. Suppose that it is effective for this person, and that it begins to rectify the situation. It will probably take the whole eight weeks to bring the patient back towards the condition he or she was in at the start of the trial. Remember, we have at least fifteen patients who will have been treated with placebo first and whose circumstances may well be very similar. If we now do what is required by this trial process, that is, measure the difference between patient condition on active treatment and on placebo, we will find in all probability that there is no significant difference.
Now let us take the case of patient ‘B’, who is also representative of fourteen other patients. This person, after the three weeks of no treatment, begins the trial on the active substance. Again, let us assume that the substance works for our patient and an improvement in condition begins to take place about six weeks from starting the trial. This is looking good: how can we fail with this one? Unfortunately we can, because at eight weeks from the start of the trial we change the patient to placebo. Because our active substance is effective for a long time it continues to maintain the patient in good condition throughout the eight weeks of placebo treatment. Now we come to study our results and find that in this case the patient seemed to do as well, if not better with placebo as with the active substance! Thus, the conclusion must be drawn that the trial has demonstrated no significant difference in the therapeutic effect of the active substance compared with placebo.
Quite so: this is precisely what the trial has demonstrated, but it is not ‘Seapower’, our active substance, that has failed: it is the trial procedure itself.
It is possible to select suitable procedures for trials of substances like our imaginary one. It is also important that the people responsible for setting up such trials do select appropriate procedures. The very reason that this subject has been emphasized so strongly here is that personal experience has shown that this is not always done. The ones who lose out through such negligence are the people awaiting effective new treatments.
Some doctors are now advocating changes in the trials system to take account of failures like the one just described. Not only are better trial designs being suggested, but also shorter periods of involvement are being recommended. Whilst the need for thorough testing of new substances is appreciated, the lengths to which some of the governmental agencies go, possibly in the interests of self-protection, are becoming ridiculous. The banning of saccharin by the United States Food and Drug Administration on the grounds that it is cancer-producing demonstrates this trend. Many things are, or can be shown to be, dangerous if used to excess but are quite safe in normal use. The sun is carcinogenic (cancer-producing): let us hope they do not ban that! It is probable that many of the more practically-minded scientists involved with trial techniques for the safety, and efficacy of materials realize the negative effect that results from some of our present systems and will effect a change for the better.
*30/48/5*