Q: Do we have to adjust our hopes
in terms of what can be reached
I don’t think so. I am convinced that
the fact that telemedicine hasn’t made
it into regular care in most places in
Europe so far has more to do with the
inertia of political will in healthcare
systems than with a lack of effectiveness data. And I am not sure that we
are doing ourselves and telemedicine a
favour by focusing too much on RCTs.
Q: RCTs are the gold standard for
proving efficacy, aren’t they?
They are the gold standard for
proving efficacy of therapies that have
not proven efficacy before and that
carry a certain risk for the patients.
This is what they were invented for.
But telemedicine is not a drug. There
are different types of telemedicine,
with different needs for evaluation.
Furthermore, technologies and care
concepts change over time. An RCT
can easily take years to conduct. The
risk is that in the end you get answers
to already outdated questions. In this
case you would need another RCT, and
so on. This is unsatisfactory, and it is
why at ZTG we are advocating a stratified model for telemedicine evaluation.
And we are increasingly convinced
that this model will find its way in one
way or the other into eHealth legislation and reimbursement discussions.
Q: How exactly would such a
stratified evaluation model look?
We would suggest three different
scenarios. Number one is a scenario in
which telemedicine is implemented on
top of an already evidence-based medical model and does nothing more than
replace the traditional way of information transfer. A good example is
telemedicine for home blood pressure
monitoring (HBPM). HBPM is strongly
recommended in the European guideline as a result of randomised studies
that have shown that it is better than
office-based blood pressure monitoring. These studies were done with pen
and paper, but data transmission could
as well be done with telemedicine. This
doesn’t need an RCT, because the medical concept was proved years ago. It
should be perfectly fine to do a feasibility study in this case.
Q: What are the other scenarios?
Scenario two is a situation in which
the medical model is evidence-based,
but telemedicine changes the way care
providers co-operate. A typical scenario
is a heart failure telemonitoring project
that involves a telemedical service
centre. In this case you don’t need to
prove the medical model, but you need
to perform a health-economic analy-
sis. After all, it could well be the case
that telemedical care is vastly more
expensive than alternative ways to get
to the same results. Health-economic
analyses can be done with an RCT, but
normally a cohort study with scientifi-
cally sound patient matching should
be enough. Scenario three, finally, is
about telemedicine applications that
work with unproven medical concepts.
A typical example is a new telemedical
implant for disease monitoring. These
implants usually rely on parameters
that need to be evaluated properly,
and they also carry risks that noninterventional telemedicine do not. In
these cases, an RCT is
“Don’t Focus Too
Much On RCTs”
By Philipp Grätzel von Grätz
>> Two major randomised controlled trials (RCTs) of telemedicine
systems, the English Whole-System-Demonstrators and the German
TIM-HF study, failed to meet expectations. Rainer Beckers, a leading
expert on telemedicine in Germany, argues that an RCT is not the best
method for telemedicine evaluation, in many cases.