Device sensitivity and false alarms can reshape regression-to-the-mean in simulated epilepsy trials

The study examined how the sensitivity of automated seizure detection devices and the rate of false alarms affect regression-to-the-mean and placebo effect in simulated clinical trials for epilepsy. Researchers used the CHOCOLATES seizure simulator and tested 100,000 potential participants during a two-month baseline period and a three-month test period. When device sensitivity decreased from 100% to 10%, the proportion of participants exhibiting regression-to-the-mean increased from 38.2% to 64.8%, and the placebo effect increased from 14.7% to 48.1%. When the false alarm rate increased from 0 to 1 alarm per day, regression-to-the-mean increased from 38.2% to 53.2% and the placebo effect from 14.7% to 31.3%. The results demonstrate that imperfect seizure detection can significantly alter the apparent placebo response expected in clinical trials.