Related Experiment Video
Updated: Aug 8, 2026

3D-Neuronavigation In Vivo Through a Patient's Brain During a Spontaneous Migraine Headache
Published on: June 2, 2014
Inter-observer agreement in the diagnosis of childhood headache
J R Wolstein1, S S Seshia, P Haese
1Department of Pediatrics and Child Health, University of Manitoba, Winnipeg.
Insights
Diagnosing multiple headache types in children may decrease agreement among pediatric neurologists. This study highlights challenges in consistent recurrent headache diagnosis for pediatric patients.
Area of Science:
- Pediatric Neurology
- Clinical Diagnostics
Background:
- Recurrent headaches are common in children.
- Accurate diagnosis is crucial for effective management.
- Inter-observer agreement in diagnosing pediatric headaches requires evaluation.
Purpose of the Study:
- To assess inter-observer agreement among pediatric neurologists in diagnosing recurrent headaches in children.
- To investigate the impact of diagnosing multiple headache types on diagnostic agreement.
Main Methods:
- Prospective assessment of diagnostic agreement.
- Four pediatric neurologists reviewed clinical letters of 40 children with headaches.
- Diagnoses were based on International Headache Society classifications, with an option for combined types.
Main Results:
- Percentage agreement ranged from 45% to 78% (Kappa values 0.20-0.59).
- Agreement was higher (76%) when single headache types were assigned.
- Agreement dropped significantly (4%) when multiple headache types were diagnosed.
Conclusions:
- Allowing multiple headache diagnoses may reduce inter-observer agreement in clinical practice.
- Standardizing diagnostic criteria or interpretation may be necessary for consistent pediatric headache diagnosis.
Abstract:
We prospectively assessed inter-observer agreement in the diagnosis of recurrent headaches in children. Clinical letters containing information on 40 children with headaches (age 4.3 to 17.8 years, median 10.4 years) were given to four Pediatric Neurologists. One or more headache types could be checked off on a data sheet that listed the main types recognized by the International Headache Society and an additional one, "combined migraine and tension-type headache". There were six combinational pairs of neurologists. The six pairs yielded 240 sets of diagnoses. Percentage agreement ranged from 45% to 78%, Kappa values from 0.20 to 0.59, and weighted Kappa from 0.19 to 0.52 within the six pairs. Agreement was 76% when both neurologists in a pair assigned single headache types and 4% when one or both neurologists diagnosed multiple types. The International Headache Society suggests that patients may have multiple types of headache and recommends that all types be classified. We suggest that the option of diagnosing more than one headache type from data in clinical letters may reduce inter-observer agreement.

