Introduction of a classification interview for post-traumatic headache after concussion

Background: Mild traumatic brain injury (TBI) can cause temporary disruption to brain function, with up to half of those affected complaining of functional limitations up to 12 months after the initial injury. Mild TBI can cause a range of sequelae, most commonly post-traumatic headache (PTH). The incidence of PTH varies post mild TBI, with up to 80% affected by three months and 60% by one year, with heterogenous phenotypes reported. We aimed to introduce a standardised interview to identify and characterise PTH. The primary outcome was to identify PTH, and the secondary outcome to characterise the phenotype of PTH. Methods: Participants were prospectively recruited from a tertiary centre hospital and a military rehabilitation centre in the United Kingdom. Inclusion criteria included a diagnosis of mild TBI or concussion; normal brain imaging; aged 17 years and older; and head injury within the last 12 months at time of screening. We have excluded those with serious underlying pathology; secondary causes of headache (excluding PTH), and non-English speakers. The non-headache specialist phoned the patient primarily to run through the structured headache interview. Following this, the headache specialist conducted a telephone clinical consultation as ‘gold-standard’. Both interviewers defined PTH as headache developing ≤ 7 days as definite, probable (8–30 days) and unlikely (> 30 days). Cohen’s Kappa estimates the inter-rater reliability across categorical variables. We calculated prevalence-adjusted bias-adjusted kappa (PABAK), which adjusts the kappa value for differences in prevalence and bias across variables. Results: A total of 194 people were screened and 63 completed paired interviews. The mean age was 26.2 (SD 8.5) years and 19% were female. Most participants displayed a migraine-like phenotype (94%), followed by tension-type-like headache (13%). A very good agreement was demonstrated between the non-specialist and specialist in diagnosing PTH (PABAK 0.90) and differentiating migraine-like versus tension-type-like headache (PABAK 0.83). There was a good agreement for migraine-like versus other headache sub-phenotypes (PABAK 0.75). Conclusions: There is currently no standardised interview that aids healthcare professionals with identifying PTH and its sub-phenotype. This structured interview can be used to diagnose PTH and its sub-phenotype.