Tools for Evaluating Health Technologies: Five Background Papers Page: 70
156 p. : ill. ; 28 cm.View a full description of this report.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
70 I Tools for Evaluating Health Technologies
nificance, whereas the MIAMI result did not. This
difference in the strength of the conclusions that
could be drawn from the two trials resulted almost
wholly from their respective sample sizes.
Another promising area of research in the treat-
ment of acute MI in the early 1980s was the use of
thrombolytic drugs, agents given during the acute
phase of a heart attack to dissolve the clots in the
coronary artery that had precipitated the attack.
By restoring blood flow to areas of the heart
muscle that have been starved of oxygen-rich
blood by the blockage, these drugs can spare the
heart from permanent damage.
By the mid- 1980s, 24 separate trials had tested
the hypothesis that the use of an intravenous
thrombolytic agent (primarily streptokinase)
would decrease the risk of mortality in patients
with acute MI. Of these trials, five reported a sta-
tistically significant benefit on mortality from use
of a thrombolytic drug, 11 suggested a benefit but
were not statistically significant, and eight re-
ported a harmful trend but were also not statisti-
cally significant (50). The discrepancies in the
trials' findings most likely derived from the fact
that the effect of such agents was anticipated to be
modest (on the order of a 10- to 30-percent de-
crease in mortality), and the majority of the indi-
vidual trials were simply too small to detect such a
benefit accurately (none enrolled more than 750
patients).
The uncertainties left by these trials led directly
to ISIS-2, in which more than 17,000 patients
were randomized to the thrombolytic drug strep-
tokinase or placebo as well as to a month-long reg-
imen of daily low-dose aspirin or placebo (33).
With respect to vascular mortality, patients who
received streptokinase experienced a statistically
significant 25-percent reduction in risk, those re-
ceiving aspirin experienced a statistically signifi-
cant 23-percent decrease, and those who received
both treatments experienced a significant 42-per-
cent decrease in vascular death. Thus, this large
and simple trial was able to detect definitively the
modest but clinically meaningful benefits of
thrombolytic therapy in the treatment of acute MI.The reason that larger trials are better able tore-
liably detect modest treatment effects derives not
just from the numbers of randomized participants
but rather from the number of events they experi-
ence. For example, whereas a trial of aspirin in the
primary prevention of heart disease might require
a sample of 22,000 men over the age of 40 in order
to detect a 20-percent reduction in risk, a sample
of 40,000 women over the age of 45 would be re-
quired to detect the same effect, because women
have a lower baseline rate of heart disease than
men do. Thus, trials must be large enough to ac-
crue sufficient numbers of outcome events to
demonstrate either definitive positive results or
truly informative null findings.
The identification of effective treatments for a
condition also affects the sample size require-
ments of future investigations. As the efficacy of
thrombolysis and aspirin has been demonstrated
in large trials, these therapies have become more
common components of the routine management
of MI patients (35,40). Any new therapies, then,
must be shown to confer additional benefits be-
yond those of an expanding regimen of effective
standard treatments. As a result, the absolute mag-
nitudes of any further benefits are likely to be pro-
gressively smaller. Such benefits may be very
worthwhile, since MI is a common and serious
condition, but detecting them will become in-
creasingly difficult and will require trials with
even larger samples.
A second circumstance that affects a trial's
sample size requirements is the need to compare
directly two or more treatments to determine
whether one has clear advantages. It was just such
a question that led to ISIS-3 (34). Randomized
trials had suggested that, in addition to streptoki-
nase, two other thrombolytic agents-tPA (tissue
plasminogen activator) and APSAC (anisoylated
plasminogen-streptokinase activator complex)
were effective in dissolving clots in acute MI and
reducing subsequent mortality. Although throm-
bolytic therapy was clearly a valuable treatment, it
was unclear whether there were any important dif-
ferences in the benefits and risks of the three prin-
Upcoming Pages
Here’s what’s next.
Search Inside
This report can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Report.
United States. Congress. Office of Technology Assessment. Tools for Evaluating Health Technologies: Five Background Papers, report, September 1994; [Washington D.C.]. (https://digital.library.unt.edu/ark:/67531/metadc39759/m1/79/: accessed March 28, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.