Opioid Prescriptions for Migraine in ED? More Future Health Resource Utilization

prescription medicine, pills
The purpose of this study was to evaluate the subsequent health resource utilization (HRU) of patients with migraine who received opioid medications at ED visits.

Among patients who present to the emergency department with migraine, those who receive opioid prescriptions are more likely to have increased future health resource utilization, suggesting that migraine management in emergency settings requires optimization. This is according to study results published in Headache.

Using data from the Baylor Scott & White Health electronic health record database, researchers conducted a retrospective cohort study of health resource utilization among people with migraines who did and did not receive opioids at emergency department visits between 2013 and 2017.

In total, 788 patients (mean age, 44.5±14.6 years; 85.9% women; 76.1% White) met eligibility criteria and were included in the study. Within this cohort, 35.9% (n=283) received 1 or more opioid medications during their index emergency department visit and 34.8% of the total cohort were opioid naïve at baseline. During the index visit, 22.6% of these opioid naïve patients received an opioid medication.

Compared with nonrecipients, patients who received opioid medications had significantly more all-cause and migraine-related opioid prescriptions during the follow-up period. This difference remained significant even when assessing opioid use as the number of prescriptions, or when grouping these prescriptions into different levels.

People who received opioids also had significantly more migraine-related and all-cause emergency department visits during the follow-up period; this remained consistent when evaluating emergency department use as the number of visits or when categorized into multiple levels.

Investigators tested time-dependent covariates for pre-index opioid use, emergency department visits, ambulatory visits, hospitalizations, preventive medication use, and index opioid use at one time. No one time-dependent covariate was significant (x2=10.17; df=6; P =.118).

In an unadjusted analysis, researchers found that people who received opioids had a significantly shorter time period before their next migraine-related emergency department visit (x2=11.36; df=1; P <.001). At the end of the 6-month follow-up period, 25.5% of patients returned to the emergency department with a migraine diagnosis; 25% of patients had a subsequent migraine-related emergency department visit within 136.5 days.

Using a Cox proportional hazard model, overall adjusted analyses were significant (x2=131.52; df=34; P <.0001). Expected hazard ratio (HR) for emergency department return visit was 1.49 times higher for people who received opioids compared with those who did not, following adjustment for all covariates (95% CI, 1.09-2.03; P =.013).

Women were also more likely to return to the emergency department (HR, 1.82; 95% CI, 1.12-2.86; P =.015), as well as people who had 10 or more nonmigraine-related opioid prescriptions prior to the index period (HR, 2.12; 95% CI, 1.24-3.65; P =.007). Other factors included 4 or more all-cause emergency department visits in the pre-index period, a hypertension diagnosis, a tobacco use disorder diagnosis, and a diagnosis of long-term medication use (HR, 2.38, 1.46, 1.45, and 4.28, respectively).

When compared with non-naïve opioid recipients, patients who were opioid naïve at baseline had significantly more migraine-related emergency department visits during the follow-up period (P <.05 for all).

Results of a Log-rank test indicated that opioid naïve opioid recipients had a significantly shorter time before their next migraine-related emergency department visit (x2=8.87; df=1; P =.003). Within this group, 16.8% of patients returned to the emergency department with a migraine diagnosis by the end of the 6-month follow-up period. Risk for emergency department return was higher among opioid-naïve opioid recipients compared with nonrecipients after controlling for covariates (HR, 2.90; 95% CI, 1.54-5.46; P =.001).

Study limitations include those inherent to observational studies; an inability to capture care received outside of this specific health system, leading to potential resource use underestimation; and possible issues with electronic health record data accuracy, completeness, and comparability due to missing variables. These results may not be generalizable to patients outside of the study area of north and central Texas.

“This real-world database analysis suggests that opioid use among patients with migraine presenting to the [emergency department] is associated with increased risk of future migraine-related [emergency department] visits, which highlights the need for optimizing migraine management at emergency settings,” the researchers concluded. “Future studies may evaluate the robustness of these findings in a larger and more diverse population.”


Shao Q, Rascati KL, Lawson KA, Wilson JP, Shah S, Garrett JS. Impact of emergency department opioid use on future health resource utilization among patients with migraine. Headache. 2021;61(2):287-299. doi:10.1111/head.14071