Once again to the question of lies in scientific journals

Scientists urge not to trust even respected medical journalsABC Magazine
A review published in the journal Pediatrics of almost 150 randomized controlled trials, the participants of which were children, and the results of which were published in respected serious medical journals from 2007 to 2008 (Assessment of Risk of Bias Among Pediatric Randomized Controlled Trials), showed that in 40-60% of cases, the authors either take insufficient measures to minimize the biased assessment the results obtained, or do not describe them in sufficient detail.

For the analysis, the scientists used Cochrane Collaboration methods, which allow us to assess the degree of reliability of the presented data on six critical factors, including randomization and masking in the distribution of placebo and active drugs.

Overall, 41% of the 146 studies analyzed had inappropriate or insufficiently well-described randomization methods. The probability of incorrect randomization for manufacturer-funded studies is 6 times higher than for studies paid for by the state or non-profit organizations. By the way, an earlier study showed that trials sponsored by pharmaceutical manufacturers are 4-5 times more likely to recommend an experimental drug.

Scientists from the Johns Hopkins Children's Center believe that experimental trials funded by pharmaceutical companies or manufacturers of medical equipment, as well as trials not registered in publicly available databases, have a higher risk of payment and bias of the results presented. The same applies to studies evaluating the effect of behavioral therapy, not medication.

Registration of trials is a guarantee of research transparency, which reduces the level of bias of published results, the researchers explain. According to their data, registered trials are almost 70% more likely to have competent randomization than unregistered studies, possibly because during registration, researchers are forced to answer a number of questions related to the goals and progress of the experiments.

Another problem is the violation of the concealment of the order of attribution of test participants to a particular group. This technique was used in 57% of the analyzed studies. But in trials devoted to behavioral therapies, this question arises 4 times more often. In general, almost 20% of the studies used incorrect methods to conceal the principles of patient distribution.

To date, the gold standard of medical research is considered to be a double-blind randomized controlled trial that excludes actual or potential bias. The results of such a study, published in a serious medical journal, affect the practice of using medicines and treating patients. Incorrect, biased or paid results may lead to erroneous conclusions about the effectiveness of the drug or procedure. Scientists warn readers of medical journals to be critical of any information published even in a reputable respected publication.

They urge pediatricians to ask themselves two main questions when reading a new article: "How did the researchers come to this conclusion?" and "Was their analysis objective?"

"Thousands of tests in the field of pediatrics are conducted in the world, which means that the risk of using distorted results is growing. We have to be vigilant and guarantee the correctness of the test and its evaluation," says Michael Crocetti, a pediatrician at Johns Hopkins Children's Center.

Portal "Eternal youth" http://vechnayamolodost.ru13.07.2010