Pragmatic Randomized Controlled Trials in Health Care | KIx on edX | About Video


[MUSIC PLAYING] DR. MERRICK ZWARENSTEIN:
Why do we spend so much on health care both
individually and as societies? Because we think the health care works. But how do we know that? For some drugs, like penicillin,
when it was first introduced it was obvious that it was effective. And for some treatments
like appendisectomy, it’s long been obvious that if
a patient had appendicitis then surgery would almost always save
them from death due to Peritonitis. But for most treatments
for most diseases the effects were less dramatic. And so it was not so
obvious that a new treatment was more effective than an old one. And so we imported from
agriculture the randomized trial as a technique for
evaluating interventions to tell us whether or
not they were effective. To do a randomized trial
we form two or more groups that there are alike in
every factor that might affect the outcome of treatment,
except the treatment itself. One group gets the new treatment
that we’re interested in. And the others get an alternative–
perhaps an older treatment, perhaps usual care, perhaps
an inactive placebo. Because the groups are otherwise
identical any difference in outcomes is then due to the
difference in treatments. So if the treatment
group did better then it’s the treatment that’s
responsible for that improvement. Comparison of the outcomes
thus reveals the effects of the different treatments. This was electrifying for health
care and triggered a golden age for the invention of new treatments. We now have around half a
million randomized trials. And knowing an enormous
amount about what does and doesn’t work in
health care and medicine. As time went by though,
we were trying to detect even smaller effects
and smaller effects. And in order to make our trials
sensitive to these small differences we started to eliminate the patients
whose conditions, perhaps they had several conditions,
or who’s advanced age or less than perfect
adherence to their medication might even slightly reduce
the impact of the treatment we were interested in evaluating. And as a result the
people in these trials, the institutions in which
the trials were done, even the clinicians providing
the care within these trials became less and less similar. So the usual patient’s
communities and practitioners in whom we hope to use
these treatments once we’ve shown that they’re effective. As a result, trial results have over
time become less and less applicable and that’s less relevant
to or use in deciding whether or not to implement a
particular intervention or treatment. I’m Merrick Zwarenstein from Western
University in Ontario, Canada and my colleagues at the
Karolinska Institutes from Aberdeen University
in Scotland and elsewhere, will be providing a course of six
weeks on randomized control trials. And we hope that we’ll be able to
help you design your own trial. And through this process
understand how best to design a randomized trial that
answers the decision that you are trying to make in health care. [MUSIC PLAYING]

Leave a Reply

Your email address will not be published. Required fields are marked *