Objective: Migraine treatment varies widely in the pediatric emergency department (ED). Factors associated with discharge after only initial emergency treatment were examined.
Methods: A retrospective chart analysis was conducted on patients 6 to 18 years old who presented to the St. Louis Children's Hospital ED between January 1, 2008, and December 31, 2011, with a discharge diagnosis of migraine (n = 700 visits). Associations between patient characteristics, initial treatments, and rates of discharge after only initial treatment were examined using a generalized linear model and receiver operating characteristic curves.
Results: If exclusively oral or intranasal (PO/IN) medications were given initially (n = 285), ibuprofen alone was associated with lower discharge rates compared with other PO/IN medication regimens (P < 0.05). The only other variable associated with discharge was arrival pain score (P < 0.05). When ibuprofen alone was administered, pain scores equal to or lower than 5/10 were associated with the greatest sensitivity and specificity for discharge. With administration of other PO/IN regimens, pain scores equal to or lower than 8/10 were associated with the greatest sensitivity and specificity for discharge. If intravenous (IV) medications were given initially (n = 415), ketorolac given with an antinausea medication was associated with higher discharge rates compared with independent administration of these medications (P < 0.05). Intravenous medications were associated with higher discharge rates compared with PO/IN medications (P < 0.001).
Conclusions: Arrival pain score may be used to help select initial migraine treatment in the pediatric ED. Initial use of PO/IN regimens including triptans or an antiemetic and concurrent administration of IV ketorolac with an antiemetic may be associated with higher rates of discharge after initial treatment alone.