Agreement Between a Daily Electronic Headache Diary and Self-Report Questionnaire from a Dose-Response Randomized Controlled Trial of Spinal Manipulation Therapy for Cervicogenic Headache

Abstract

Objective: To determine if a minimally burdensome, easy to administer self-report questionnaire based on recall can be used interchangeably with a gold-standard headache diary to measure headache outcomes in an adult cervicogenic headache (CGH) population. Design: Secondary analysis of outcomes collected in a prospective, parallel group, observer blinded dose-response randomized controlled trial of spinal manipulation for CGH (R01AT006330). Setting: General population in Minneapolis/St. Paul metropolitan region (MN, USA) Participants: 18 years of age and older with a history of chronic CGH (5 headache days per month for three months) and a pain intensity of ≥ 3 (0-10) who are otherwise in good health at baseline. Measurements: A daily, electronic headache diary (gold standard) and self-report questionnaire based on recall are used to ascertain headache frequency, measured in days, and intensity, measured using the 11 point Likert scale (0-10), collected over four weeks at baseline. Analysis: Baseline characteristics are summarized using descriptive statistics. The Bland Altman method is used to assess agreement, including limits of agreement (mean difference ± 2SDs). Linear regression is used to evaluate the presence of proportional bias. A two-tailed t-test is used as a measure of inference for mean differences between measures. Results: 87 participants are included in this analysis. The mean difference (SD) and limits of agreement (LOA) for CGH frequency and intensity are 0.77 (4.3) days (LOA: -7.6 - 9.1) and 0.14 (0.8) points (LOA: -1.43 - 1.70), respectively. Linear regression shows evidence of proportional bias for headache intensity (β=0.286, 95% CI 0.01-0.27, p=0.000). Group differences between the questionnaire and diary were not statistically significant: frequency t(86) = 1.69, p=0.09 and intensity t(86)=1.6, p=0.11. Conclusions: There is a lack of agreement between the questionnaire and electronic diary for measuring headache frequency and intensity. It is not recommended to use the questionnaire in lieu of the gold-standard daily headache diary for measuring headache outcomes in clinical research. A self-report questionnaire based on recall may be appropriate however to inform CGH management in clinical settings.