Migraine day frequency in migraine prevention: longitudinal modelling approaches

Abstract

BackgroundHealth economic models are critical tools to inform reimbursement agencies on health care interventions. Many clinical trials report outcomes using the frequency of an event over a set period of time, for example, the primary efficacy outcome in most clinical trials of migraine prevention is mean change in the frequency of migraine days (MDs) per 28 days (monthly MDs [MMD]) relative to baseline for active treatment versus placebo. Using these cohort-level endpoints in economic models, accounting for variation among patients is challenging. In this analysis, parametric models of change in MMD for migraine preventives were assessed using data from erenumab clinical studies.MethodsMMD observations from the double-blind phases of two studies of erenumab were used: one in episodic migraine (EM) (NCT02456740) and one in chronic migraine (CM) (NCT02066415). For each trial, two longitudinal regression models were fitted: negative binomial and beta binomial. For a thorough comparison we also present the fitting from the standard multilevel Poisson and the zero inflated negative binomial.ResultsUsing the erenumab study data, both the negative binomial and beta-binomial models provided unbiased estimates relative to observed trial data with well-fitting distribution at various time points.ConclusionsThis proposed methodology, which has not been previously applied in migraine, has shown that these models may be suitable for estimating MMD frequency. Modelling MMD using negative binomial and beta-binomial distributions can be advantageous because these models can capture intra- and inter-patient variability so that trial observations can be modelled parametrically for the purposes of economic evaluation of migraine prevention. Such models have implications for use in a wide range of disease areas when assessing repeated measured utility values.Electronic supplementary materialThe online version of this article (10.1186/s12874-019-0664-5) contains supplementary material, which is available to authorized users.

Health technology assessment for the acute and preventive treatment of migraine: A position statement of the International Headache Society

“…The characteristics of the patient sample should include a distribution of MMDs and MHDs at baseline. Since models depend on assumptions about the nature of the distribution of MMDs or MHDs, and it is unrealistic to assume that migraine attacks are normally distributed, acceptable choices (with appropriate testing of model fit) include Poisson, zero-inflated negative binomial, and Beta distributions ( 56 ). Estimates of the effect of technology on that distribution, as well as the direct and indirect costs associated with an MMD or an MHD, should be made.…”

“…Models depend on assumptions about the nature of the distribution of MMDs or MHDs. Since it is improbable that migraine attacks are normally distributed, appropriate choices for use in modelling MMDs include Poisson, zero-inflated negative binomial, and Beta distributions ( 56 ). With appropriate testing of model fit, all are acceptable.…”

Economic evaluations in migraine: systematic literature review and a novel approach

“…Modelling MMD frequency as a continuous outcome, rather than as a series of categories, has several advantages over the use of arbitrary health states which have been previously reported. 16,17 The use of continuous MMD frequency distribution retains all of the observed information to accurately estimate the difference in outcomes between groups. This approach also more readily allows for indirect comparisons based on study primary endpoints (e.g.…”

Secondary Trigger Point Deactivation Surgery for Nerve Compression Headaches: A Scoping Review

“…18 Another variable that has been used to track the headache burden is the frequency of the monthly migraine days. 19 A positive response to botulinum toxin injections or nerve blocks has been shown to be a positive predictor of successful surgery, [20][21][22] with the positivepredictive value of diagnostic peripheral nerve blocks being 0.89 (95% CI, 1-0.74) and the positive-predictive value of diagnostic botulinum toxin type A injections being 89.5%. 20,22 Nevertheless, a negative response to injections does not definitively preclude surgical treatment or predict failure of trigger point deactivation surgery.…”