Rakesh and Mark, thanks for your comments.
Let me re-phrase my claim more precisely: I do not believe that dramatic observations can ever provide such good evidence of benefit that no controlled trial of an alternative intervention would be ethical - and there are always plausibly better alternatives, eg different intervention, form, dose, frequency, duration, route. (Note that there is an asymmetry with respect to the evidence required for dramatic adverse effects - e.g. no controlled trial is needed to test the effects of pulling the trigger of a loaded pistol.)
Mark, I except that a dramatic effect can provide proof of principle, but this leaves open the question of what the optimum intervention is.
Rakesh, your example of a ventriculo-atrial (VA) shunt to replace a ventriculo-peritoneal (VP) shunt when there is persistent ascites may justify a particular clinical decision. But case reports/case series do not provide convincing evidence that this is optimum treatment. For example, a VA shunt introduces new risks such as infection and thromboembolism. If the ascites were caused by a low-grade peritonitis, maybe it would have resolved given more time, or antibiotics, or anti-inflammatories, or peritoneal lavage? (These are examples from my limited knowledge in this area of what might work. I just want to make the point that it is always possible to find a plausible alternative, not suggest that these be tried without further consideration of what has been tried in the past.)
There are, of course, many instances where case reports and case series provide the best current evidence of effectiveness. But, this is not to say that stronger evidence in these instances is not desireable, and not possible.
Michael
|