Difference in clinical trials according to funding source
An interesting article appears in JAMA
, and shows how the source funding of clinical trials on cardiovascular differs between those entities that are funded by non-profit from those so.
It took all editions published by JAMA, The Lancet
and the NEJM
between 2000 and 2005, and analyzed if the trend corresponded to previous studies that showed that year was so.
Of 324 articles, 21 not cite the source. Of the 104 trials funded exclusively by nonprofit entities (ESFN or NFPO), 51 (49%) reported significant evidence for vs. new treatments. standard treatment, while 53 (51%) did not (p = 0.8).
In contrast 92 (67.2%) of 137 clinical trials funded by for-profit entities, clearly favored the new vs treatments. treatment standards (p =;0.001).
Among the remaining 62 trials were funded 35 Overall, a favorable intermediate percentage (56.5%) compared to
For 205 randomized trials evaluating drugs, the ratio in favor of new treatments was 39.5% for studies funded by nonprofit organizations, 54.4% when funding was set, and 65.5% when the Funding
was exclusively for-profit entities. The statistical significance
trend across groups was 0.002).
For 39 randomized clinical trials
evaluating other types of technologies (devices), the proportion in favor of new treatments was 50% in those funded by nonprofit organizations profit, 69.2% in those funded overall, and 82.4 in
funded exclusively by for-profit organizations. The probability trend across groups was 7%, is ie the probability that these results are at random is only 7%.
Regardless of funding source, most of the work used as the final results quantitative results on angiography, intravascular ultrasound, markers biologicicos plasma, and functional measures were more likely be reported positively (67%) than trials using as clinical aspects results. This means that 67% of the jobs are directed to what is known as DOE
studies (disease oriented evidence), vv. other studies known as POEMS (Problem Oriented Evidence that Matters), which were reported in just the 54.1% of the studies (p = 0.02).
As we see is not enough to invent diseases, but also the source of funding is clearly relevant, although
there may be publication bias
(the magazines usually publish more studies are often positive), it is also true that
financing bias also exists, meaning that it tends to fund over that which can provide more revenue, and it is unknown jobs with few adverse outcomes, were never subjected to medical journals.
This brings us back to the theme that science is not neutral, the same nature of clinical studies
is more concerned to validate hypotheses and certainly never to refute them. Induction work even small samples is widespread, and magical way, then becomes widespread, regardless of the particularities of time, space and
individuals. Nothing new to not know, but we are in trouble with this science, apart from all this, taken as the gold standard placebo, which has an average efficiency of 35%.
This would not appear strange that “studies based on evidence “, where diets do not work well, or change
of habits of life. Or the coffee before we caused gastritis Chronic is now that it is exclusively the Helicobacter
Too many factors for a science that attempts mathematize complexity of man, and populations. Hypothesis using monocausal not even exist in physics. Which becomes probabilistic and deterministic from these types of tests more Experimental studies about what happens every day. And that insists on fragmenting and assign us the role of living alone biological, hardly know the complexity that it biology brings together, and wherein the linear models do not exist.
An interesting article from a magazine that occasionally have these fits of editorial clarity and decides to publish this data.