Prospective attack reports are more reliable indicators of clinical severity in cluster headache


  • Retrospective reports of untreated cluster headache (CH) attacks tend to overestimate their duration and severity; these descriptions are influenced by recall bias.

Why this matters

  • As people with CH rarely present during attacks, most diagnoses rely on self-reporting; a better understanding of the between-patient variability of attacks will help to establish whether specific CH phenotypes can be used to predict treatment responses.