This will be a single-center, open-label clinical trial comparing sphenopalatine ganglion blocks to standard intravenous therapy for patients who come to the emergency department for a headache.

Having outperformed other agents in head-to-head trials, intravenous (IV) dopamine antagonists, such as prochlorperazine or metoclopramide, are generally considered the 1st line treatment for headaches in the ED. However, despite the relative effectiveness of prochlorperazine, a substantial number of patients who present to the ED with a headache still have a moderate to severe headache 24 hours after discharge. Moreover, some patients may have difficult IV access, so it would be useful to employ a technique that could effectively treat headaches without the use of an IV line.

One therapy that has been suggested but not well-studied that might help treat headaches in the ED without the use of an IV line is the sphenopalatine ganglion (SPG) block via intranasal lidocaine. The sphenopalatine ganglion may play a role in the development of pain in primary headaches through the release of neuropeptides that activate or sensitize intracranial nociceptors. Several prior randomized trials have evaluated the use of intranasal lidocaine or bupivacaine vs placebo for patients with migraine headaches, and the results have been mixed.

Thus, the investigators propose a single-center, open-label, clinical trial to compare the efficacy of SPG blocks to standard IV therapy for headaches in the ED.

Adult patients who come to the ED for a suspected primary headache may be enrolled. Patients who are assigned to the standard IV therapy group will receive prochlorperazine 10 mg and diphenhydramine 50 mg.

Patients who are assigned to the SPG block group will undergo the following procedure. A cotton-tipped applicator soaked in 1% lidocaine will be placed in the nostril on the side of the headache. If the headache is bilateral, a cotton-tipped applicator will be inserted into each nostril. The cotton-tipped applicator(s) will be left in place for 15 minutes. If the patient has not had significant improvement upon removal of the cotton-tipped applicator the physician may order IV medication for the patient's headache ("rescue analgesia").

Just before the initiation of treatment, an initial visual analog scale (VAS) pain score will be obtained from the patient. The VAS pain score will be repeated 15 and 30 minutes after treatment. Additional data will be determined through chart review and telephone follow up with the patient 24-72 hours after discharge.

The primary outcome of the study will be the difference between groups in the fraction of patients with a 50% reduction in VAS pain score at 15 minutes. Secondarily, the investigators will compare groups with regards to their mean VAS pain scores as 15 and 30 minutes after the initiation of the first treatment, hospital length of stay, the need for rescue analgesic medications, complications (nose bleed, akathisia, etc.), and presence of persistent headache 24-72 hours after discharge on telephone follow up.

Inclusion Criteria:

• Any patient at least 18 years old who presented to the ED with a suspected primary headache.

Exclusion Criteria:

• pregnant women
• prisoners
• meningeal signs
• fever
• any acute abnormality on neurologic exam
• allergy to one of study drugs
• on oral anticoagulant
• already received analgesic treatment in the ED

After completion of data analysis, the data will be available indefinitely.