Documentation of International Classification of Headache Disorders Criteria in Patient Medical Records: A Retrospective Cohort Analysis

Abstract

Objective: To determine headache diagnosis and treatment patterns in the outpatient setting, focusing on documentation of the International Classification of Headache Disorders (ICHD) criteria. Design, setting, and participants: Retrospective cohort data were collected from electronic medical records of adults aged 18-35 who presented to resident-staffed family medicine outpatient clinics in the Midwest, USA, for a new or worsening headache between 2015 and 2016. Diagnosis codes were used to summarize the overall nature and prevalence of headaches. A random subset of 30 patients each for migraine headache (MGH) with and without aura and tension-type headache (TTH) were reviewed to determine how many of the five possible ICHD criteria were documented. Demographics/clinical characteristics, ICHD criteria, number and type of medications, and healthcare utilization (imaging, primary and emergency department care) through one year following the initial visit were summarized and compared across headache types. Results: There were 716 unique patients during the study period (414 MGH, 227 unspecified headaches, 75 TTH, or others). Complete ICHD criteria were documented for two patients in total. There was partial documentation (e.g., one to four of the possible five) for 30% of TTH, 63% of MGH without aura, and 77% of MGH with aura (p<0.05). Across headache types, patients were prescribed an average of 2.3 to 3.3 medications over one year, with MGH patients generally trying more medications (up to eight for those with aura and up to 12 for those without). Abortive or rescue medications were prescribed to nearly all patients; prophylactics were prescribed for 50% of MGH with aura, 66.7% of MGH without aura, and 53.3%. Non-pharmacologic interventions were less prescribed: 33.3% of TTH patients and 3.3% of MGH types combined (p<0.05). Healthcare utilization was highest for MGH with aura (ED visits) and without aura (clinic visits) patients compared to TTH (p<0.001). Conclusion: Headache-related documentation is often incomplete, which may limit interpretation and associations between diagnoses, prescribing patterns, and healthcare utilization. Future studies should evaluate the use of electronic medical records (EMR)-based templates to improve documentation, and additional detailed studies are needed in the local setting to determine whether treatment, including the use of non-pharmacologic and prophylactic methods of treatment, is optimal.