Randomised trials vs observational studies
The review discusses the strengths and limitations of different types of studies. Randomised controlled trials are a robust and well recognised way of determining the effect of treatments. Whereas observational studies based on databases can generate hypotheses about associations between the use of drugs and health outcomes, randomised trials can determine cause and effect.
Ad Statistics
Times Displayed: 112999
Times Visited: 6736 MIT labs, experts in Multi-Vendor component level repair of: MRI Coils, RF amplifiers, Gradient Amplifiers Contrast Media Injectors. System repairs, sub-assembly repairs, component level repairs, refurbish/calibrate. info@mitlabsusa.com/+1 (305) 470-8013
In a randomised trial, patients are randomly divided into two or more groups - eg, one group is given a drug and the other group is given a dummy treatment (ie, a "placebo"). Researchers then compare the rates of health outcomes between these groups. Randomly allocating patients means that any difference in these rates can generally be attributed to the treatment itself (see panel 1). Additionally, by "blinding" the patients and their doctors as to whether they are taking the drug being studied or a placebo, bias in the assessment of health outcomes between the different treatment groups can be avoided.
Meta-analyses bring together evidence from randomised trials that have tested the same treatment and, by including data from a larger and more diverse set of patients, can increase the reliability and generalisability of the results.
By contrast, observational studies compare the health outcomes of people who have been given a particular treatment by their doctors as part of routine care, and people who have not been given the treatment. Doctors give treatments to selected patients for good reasons, so the health of the patients given a treatment may be very different to the health of people who are not given it. These differences in the underlying risks of the different groups of patients are hard to predict, and it is not possible to know that they have been allowed for completely in the analyses.
In addition, patients prescribed a drug in routine care know that they are getting it, and their doctors may have warned them that the drug may cause problems. Consequently, the patients may be more likely to attribute health outcomes (particularly symptoms that are subjective, eg muscle aching) to the drug, whereas people who are not taking the drug would not do so.
As a result, although observational studies may be able to detect large increases in health outcomes that would usually be rare, they are not able to produce reliable evidence about the effects of drug treatments when the health outcomes are common or the effects are not large (see panel 2).
