Massage for arthritis: WHAT IS AN RCT?

The crucial elements of a RCT are the presence of one or more control or comparison groups and random assignment to one of these groups. Blinding—which reduces experimentalbias by keeping the subjects and/or investigators ignorant of group assignments and research hypotheses—may or may not be employed. A comparison group allows the rese archer to compare the results of the treatment with those from a different intervention, such as another type of massage, to a medication, or to no treatment at all. Because many usculoskeletal conditions are self-limiting—i.e., they get better anyway—a no treatment group can be important when evaluating the effectiveness of massage. An important type of control group is the placebo, where participants receive a sham intervention that mimics the active intervention, and should have no specific effect. A placebo group helps to control for the effects of participants’ positive expectations that the intervention will help. Placebo effects can be considerable. To be evaluated as effective, researchers need to demonstrate that an intervention outperforms the placebo. In this study, a waitlist/usual care group was chosen as the control group because the investigators determined that a placebo massage wasn’t possible with the massage protocol used. While double-blinding, where neither the participants nor the investigators know which participants are receiving the active treatment, is often not possible in studies that investigate massage, study personnel responsible for collecting and analyzing the data should always be blinded as to group assignment. Random assignment to a group is a strong design feature that allows the researcher to link cause and effect—that the treatment given is responsible for the observed results. Random assignment is especially important in clinical research, where participants are out in the real world, rather than staying in a controlled laboratory setting. While researchers may try to control for factors that are known to affect the outcomes being measured, there may be unknown factors—what if someone’s eye color affects how they rate pain on a VAS? This is not as farfetched as you might think. For example, people with red hair have been demonstrated to be more sensitive to pain. The beauty of random assignment is that with a reasonably large sample, every characteristic will be evenly distributed among the study groups. With random assignment, there should be equal numbers of redheads in each group, thus eliminating a potential source of study bias. The use of this greatly increases a study’s credibility.