Massage & Bodywork

January | February 2014

Issue link: https://www.massageandbodyworkdigital.com/i/230373

Contents of this Issue

Navigation

Page 57 of 141

RANDOMIZATION Aside from the strong dependence on structure within the clinical trial, the other main factor in this study design is randomization—making sure that subjects are assigned to their treatment group randomly, or "by chance." Most researchers use a random number generator, in the form of a computer program or written table, to assign the treatment groups. Some still use the old-fashioned method of rolling dice, where even numbers mean one treatment group and odd numbers mean the other treatment group. The method of randomization doesn't much matter, as long as the treatment assignment of each patient occurs by chance, meaning neither the subject nor the clinician has any control over which group is assigned. The main benefit of randomization is that the groups become inherently even in terms of demographics and clinical characteristics. We expect that with random assignment, we would have an even number of males and females in each group, the average age would be approximately the same, and the clinical conditions would be similar between the groups. This is similar to flipping a coin. If flipped enough times, we will get approximately the same number of heads as we do tails. Without randomization, subjects would choose their own group. This may lead to significant differences between the groups. For example, if a clinical trial was designed to compare a massage treatment with surgery, subjects with more severe pain might be more likely to choose surgery and subjects with less severe pain might be more likely to choose massage. However, in research, we would like to know if massage benefits people with all levels of pain, just as we would like to know if surgery benefits all people, too. We would want to have a balance of people with mild, moderate, and severe pain in both groups, and randomization helps make this happen. Another benefit of randomization is that it decreases expectation bias. When people get to choose which treatment they will receive, they typically pick the one from which they expect the best results. This affects the study outcomes because if the subject expects to get better, they are more likely to report that they did get better. However, with randomization, subjects are much less likely to have expectations, hopefully leading to more accurate results. In research, we would like to know if massage benefits people with all levels of pain, just as we would like to know if surgery benefits all people, too. PLACEBO GROUP Clinical trials are the only study design that can have a placebo group. A placebo (or sham) is like a sugar pill. It is a treatment provided to a subject in which no benefits are expected by the clinician. There are many ethical issues surrounding placebos. For example, is it ethical to pretend that a subject will benefit from something the researchers know to be fake? Is it ethical to withhold treatment from subjects in order to provide them with a placebo? Finally, is it ethical not to tell the subjects during the informed consent process that they might get a placebo, or should they be told up front so they can truly make an informed decision as to whether they want to participate in the study? Placebo treatments began in pharmaceutical studies as a way to determine if subjects were getting better just because they were being cared for, or if the actual treatment was helping. There is an effect, called the placebo effect, in which the subject improves even though no actual treatment was rendered. In these cases, we do not know why the subject improved. It may be natural history, meaning they would have improved on their own anyway. However, many people think that the subjects get better because they believe they will get better, or because they feel cared for by health professionals. It is an interesting phenomenon that should not be dismissed. If a subject being It pays to be ABMP Certified: www.abmp.com/go/certifiedcentral 55

Articles in this issue

Links on this page

Archives of this issue

view archives of Massage & Bodywork - January | February 2014