We conducted a retrospective cohort study to compare the effects of
metoclopramide versus
hydromorphone for the initial emergency department treatment of
migraine headache at an urban teaching hospital. The primary outcome measure was the mean difference in the subjects' self-reported
pain scores before and after the administration of the initial medication treatment. We also estimated crude and adjusted relative risks (using Poisson multivariate regression modeling) to assess and control potential confounding by age, gender, race, and
pain score before initial medication. Two hundred subjects were included, with 51 (25.5%) receiving intravenous or intramuscular
hydromorphone, 95 (47.5%) receiving intravenous
metoclopramide, and 54 (27.0%) receiving 1 of several other medications. Using a standardized
pain scale of 0 to 10, mean
pain score reductions were 2.3 points for
hydromorphone, 3.7 points for
metoclopramide, and 2.8 points for all other medications combined (P < .001). When comparing
metoclopramide versus
hydromorphone, the crude relative risk for
pain reduction of 3 or more points was 1.76 (95% CI, 1.12-2.75, P = .01), and the adjusted relative risk was 1.60 (95% CI, 0.84-3.03, P = .15).
Metoclopramide also resulted in less use of rescue medications, faster times to discharge, and no difference in the frequency of adverse reactions.
PERSPECTIVE:
Metoclopramide appears to be an effective initial medical treatment for
migraine headaches in the emergency department setting, but its pharmacologic mechanism remains incompletely understood. A double-blinded, randomized, controlled trial comparing standard dosages of
hydromorphone versus
metoclopramide will be needed to definitively determine which medication is more effective.