In Brief: Controlling control groups
Controlling control groups
Controlled trials are the standard for proving the efficacy of drugs. Volunteers are randomly assigned to take a drug or a placebo, and the outcome is recorded. During the trial, no one is supposed to know who is taking the drug and who is taking the placebo. But finding and maintaining appropriate controls for any kind of study is no simple matter, as two recent studies show. One concerns dropout rates and the other, recruitment of volunteers.
In a review of 31 clinical trials of antipsychotic drugs, researchers distinguished placebo-controlled trials from those in which two drugs were compared head-to-head. They found that many more patients in the placebo-controlled trials dropped out, even if they weren't taking a placebo. Among patients taking an active drug, 49% dropped out in placebo-controlled trials and only 30% when two drugs were being compared. The difference of about 19% was the same whether they quit because the drug was not working or because of side effects. Patients taking a placebo dropped out at a rate of about 60% (a difference of only 11%), so the type of trial had more effect on dropouts than the difference between a drug and a placebo.