Abstract text (incl. figure legends and references)

Question: Cluster headache is a disorder often referred to as "suicide headache" and is associated with significantly disabling attacks[Newman et al 2015]. Although it is as prevalent as neurological conditions such as multiple sclerosis[Wei et al.2018, Leray et al. 2016] these patients often go over 5 years before diagnosis[Rosen et al. 2012]. This study aims to determine whether a screening tool can be retrospectively validated to reliably identify patients with a high likelihood of having cluster headache to prioritize rapid consultation with a headache specialist. Methods: A screening clinical questionnaire was developed utilizing criteria specific to cluster and to best differentiate it from other headache disorders. The tool was then retrospectively applied to all patients seen at a tertiary care specialty headache clinic between January 2021 and December 2021. Eligibility criteria include adult patients (>18 years), any reason for referral, any referring headache diagnosis and an initial consultation note including all data required to complete the scoring tool. A total score was calculated for each patient and the area under the receiver operating curve (ROC) used to identify the score with the highest sensitivity and specificity for a diagnosis of cluster headache. Results: A total of 415 patients were ultimately included in the study, of which there were 25 cluster headache patients. identified with an average screening score of 12.28. In comparison, the 255 migraine patients included had an average score of 1.45 on the screening tool. A ROC analysis indicated that a cutoff score of 8 had a 100% sensitivity and 85-87% specificity for a diagnosis of cluster headache. Conclusions: Our study was able to develop and retrospectively validate a cluster headache screening tool that can reliably differentiate cluster headache from other disorders. Figure 1. Cluster headache screening toolTable 1. Average score by headache disorderTable 2. Coordinates of the ROC.