OBJECTIVES: The study aimed to explore if the distribution of migraine headache days (MHD), as observed in the EVOLVE-1 (NCT02614183) and EVOLVE-2 (NCT02614196) studies in episodic migraine, can be estimated with a negative binomial probability distribution, using the approach suggested by Porter et al. (2016).

METHODS: Patient level data from two Phase 3, randomized, double-blind, placebo-controlled studies in episodic migraine (EVOLVE-1/2) were used at every month over a 12-month period. For every month and for each treatment group the raw MHD data was plotted as bar charts with the fitted Poisson, negative binomial and binomial distributions. The mean MHD per month and per treatment group was used to simulate a negative binomial distribution using a function to estimate the dispersion parameter that was suggested by Porter et al. (2016) (dispersion=6.15943-2.59754*mean+0.34453*mean RESULTS: The fitted negative binomial distribution was consistently better than the fitted Poisson and the binomial distributions at each month and for each treatment group. The fitted negative binomial was comparable with the simulated negative binomial using the raw mean and the dispersion parameter as a function of the mean.

CONCLUSIONS: The simulated negative binomial using the mean of MHDs from EVOLVE-1/2 and an estimated dispersion parameter as a function of the mean as described by Porter et al. (2016) is comparable to the raw data of EVOLVE-1/2 and the fitted negative binomial distribution. This confirms that the approach proposed by Porter et al. (2016) provides a solution to model treatment comparisons when only information on the mean MHDs are available in episodic migraine.