Researchers have devised a statistical approach that can take into account the variability in seizure frequency associated with a placebo response. In an article published in Epilepsy Research, Goldenholz et al explain their approach in more detail:
- The researchers compared the traditional 50% responder rate to a new measure called the variability-corrected score or Z v.
- The analysis revealed that the new approach can predict the expected frequency of seizures several months in advance.
- When they compared the 50% responder rates to the variability-corrected score, Z v was more useful in telling the difference between patients who had a placebo response and those who had a genuine response to a therapeutic intervention.
- Because Z v scores were able to generate higher statistical power, the authors theorize that they may allow researchers to conduct randomized clinical trials that use fewer patients and cost less than traditional randomized controlled trials.
Goldenholz DM, Goldenholz SR, Moss R, et al. Does accounting for seizure frequency variability increase clinical trial power? [Published online ahead of pint July 25, 2017]. Epilepsy Res. doi.org/10.1016/j.eplepsyres.2017.07.013