ABSTRACT
INTRODUCTION: We determine the interobserver reliability among neurology resident physicians when diagnosing primary headaches using the 2004 International Headache Classification (IHC) compared to that obtained when compared with these observers and those of a consensus group (CG) made up of investigators and observers. PATIENTS AND METHODS: The study was performed in an adult and pediatric population from a 4th level Colombian hospital out-patient neurology clinic. After informed consent was obtained from each patient, a video was recorded, eliminating the facial images in order to prevent patient identification. Two same level residents (observer 1 and observer 2) with previous training in the international classification of headache disorders (IHD-2) examined the video. Each video was examined separately with the classification in hand, making the diagnoses. The diagnoses obtained were compared between the observers and with the CG. RESULTS: We obtained a kappa index: 0.64 comparing residents, k: 0.66 comparing resident 1 to CG and kappa index 0.70 comparing resident 2 to CG in the level of groups. The kappa value obtained for the headache type level was 0.51 interobservers, 0.61 on comparing observer one with CG and 0.65 on comparing observer two with CG. CONCLUSION: Our results indicate that there is adequate reliability in the application of the 2004 IHC classification in the diagnosis of primary headaches among neurology resident physicians.
