Characterization of the adverse events profile of placebo-treated patients in randomized controlled trials on drug-resistant focal epilepsies.

In epilepsy trials a substantial proportion of patients receiving placebo has some improvement or experience adverse events (AEs) which match those related to active drug. The characterization of factors influencing these responses is crucial for a better comprehension of study results and to improve design of new trials. Seventy-one placebo-controlled, double-blind trials in drug-resistant focal epilepsies has been selected. The effect of multiple factors on some outcome measures were explored using a meta-regression model. For subjective and objective AEs, risk difference (RD) was calculated and entered in an inverse variance-weighted linear meta-regression model as independent variable to evaluate the relationship with data reported in placebo-treated patients. The number of study arms influence the percentage of patients withdrawing because of AEs and the highest dose of the experimental drug used in each RCT correlates with withdrawal because of AEs and with subjective AEs. Higher titration speed is associated with lower percentages of responders and higher reporting of both objective and subjective AEs. The correlation between proportions of placebo-treated patients with subjective and objective neurological AEs and relative RD, was significant (P = 0.002 r = 0.364 and P < 0.001 r = 0.650, respectively). Efficacy and tolerability outcomes of the placebo groups are intrinsically tied to the trial methodology and to the outcomes observed in patients treated with the active drug. The correlation for objective and subjective AEs between RD and the placebo-treated patients suggest that investigators are influenced by factors which operate within each specific trial.