Diphenhydramine as Adjuvant Therapy for Acute Migraine: An Emergency Department-Based Randomized Clinical Trial

Abstract

Study objective: More than 1 million patients present to US emergency departments (EDs) annually seeking care for acute migraine. Parenteral antihistamines have long been used in combination with antidopaminergics such as metoclopramide to treat acute migraine in the ED. High-quality data supporting this practice do not exist. We determine whether administration of diphenhydramine 50 mg intravenously+metoclopramide 10 mg intravenously results in greater rates of sustained headache relief than placebo+metoclopramide 10 mg intravenously.

Methods: This was a randomized, double-blind, clinical trial comparing 2 active treatments for acute migraine in an ED. Eligible patients were adults younger than 65 years presenting with an acute moderate or severe headache meeting International Classification of Headache Disorders-2 migraine criteria. Patients were stratified according to presence or absence of allergic symptoms. The primary outcome was sustained headache relief, defined as achieving a headache level of mild or none within 2 hours of medication administration and maintaining this level of relief without use of any additional headache medication for 48 hours. Secondary efficacy outcomes included mean improvement on a 0 to 10 verbal scale between baseline and 1 hour, the frequency with which subjects indicated they would want the same medication the next time they present to the ED with migraine, and the ED throughput time. Sample size calculation using a 2-sided α of .05, a β of .20, and a 15% difference between study arms determined the need for 374 patients. An interim analysis was conducted when data were available for 200 subjects.

Results: Four hundred twenty patients were approached for participation. Two hundred eight eligible patients consented to participate and were randomized. At the planned interim analysis, the data and safety monitoring board recommended that the study be halted for futility. Baseline characteristics were comparable between the groups. Fourteen percent (29/208) of the sample reported allergic symptoms. Of patients randomized to diphenhydramine, 40% (40/100) reported sustained relief at 48 hours, as did 37% (38/103) of patients randomized to placebo (95% confidence interval [CI] for difference of 3%: -10% to 16%). One hour after medication administration, patients randomized to diphenhydramine improved by a mean of 5.1 on the 0 to 10 scale versus 4.8 for those randomized to placebo (95% CI for difference of 0.3: -0.6 to 1.1). Eighty-five percent (84/99) of the patients in the diphenhydramine arm reported they would want the same medication combination during a subsequent ED visit, as did 76% (77/102) of those who received placebo (95% CI for difference of 9%: -2% to 20%). Median ED length of stay was 122 minutes (interquartile range 84 to 180 minutes) in the diphenhydramine group and 139 minutes (interquartile range 90 to 235 minutes) in the placebo arm. Rates of adverse effects, including akathisia, were comparable between the groups.

Conclusion: Intravenous diphenhydramine, when administered as adjuvant therapy with metoclopramide, does not improve migraine outcomes.

Trial registration: ClinicalTrials.gov NCT01825941.

Conflict of interest statement

We have no conflicts of interest to report.

Copyright © 2016 American College of Emergency Physicians. Published by Elsevier Inc. All rights reserved.

Figures

Figure. Group and individual change in pain scores following treatment with metoclopramide plus either placebo or diphenhydramine

The box plots show the median, 25th and 75th percentiles for the baseline pain score and the pain score measured one hour after treatment for each of the groups. The waterfall plots show the individual change in pain score for subjects in each group from baseline to one hour grouped by baseline pain score, (dots indicate no change, lines going above baseline indicate a worsening). The numbers above the waterfall plot are given to indicate every 10 patients.

Appendix Figure

Box and whiskers plot of the percent improvement in 0 to 10 pain score between baseline and one hour (Improvement in 0–10 score/Baseline score). 1.0 signifies complete improvement. 0 signifies no improvement. Negative score indicate worsening.

Appendix Figure

Line graph representing each participant’s experience at baseline and one hour later. The origin of the line depicts the 0 to 10 pain score at baseline. The terminus of the line depicts the 0 to 10 pain score at one hour. The graphs are sorted by baseline pain score. Thus, lines that rise unexpectedly depict participants whose pain worsened during the study period.

Figure 1

CONSORT flow diagram * One patient lost-to-follow-up received rescue medication in the ED. Therefore, we were able to count this participant as an outcome failure despite being unable to contact her at 48 hours.