Abstracts

Rationale: In recent years, several important statistical features of seizure diaries have been characterized. These are: (1) the heterogeneity of seizure frequency, (2) a relation between average seizure rate and standard deviation, (3) multiple coexisting cycles, (4) seizure clusters, (5) limitations on inter-seizure intervals. To date, no simulator has incorporated all these features into a unified model.

Methods: Our approach, Cyclic Heterogeneous Overdispersed Clustered Open-source L-relationship Adjustable Temporally-limited E-diary (CHOCOLATE) simulator was based on a hierarchical model centered on a Gamma Poisson generator with several modifiers. This model accounts for each of the aforementioned statistical properties. The model was validated by simulating 10,000 randomized clinical trials (RCTs) of medication to compare with historical RCTs. The “drug” was assumed to have 30% effect, and “placebo” assumed to have 0% effect. There were 100 synthetic patients each in “drug” and “placebo” arms._x000D_
Results: The model was able to recapitulate typical 50% responder rate (RR50) and findings in historical RCTs without the necessity of introducing an additional psychological “placebo effect”. Simulated drug: 38.9 +/- 4.9%, historical drug: 43.2 +/- 13.1%, Simulated placebo 18.0 +/- 3.9%, historical placebo 21.1 +/- 9.9%. _x000D_
Conclusions: CHOCOLATE represents the most realistic seizure occurrence simulator to date, based on observations from thousands of patients in different contexts. The RCT results are an important external validation because the tool was not in any way optimized for RCTs. Given that this tool is open-source and flexible, it can be used for many applications.

Funding: This research was funded in part by NIH KL2 5KL2TR002542.