Twenty-seven RCTs (reported as 26 in the flow chart) (n=1,415 participants, range eight to 168) were included in the review. Twelve RCTs met all quality criteria (Class I), seven trials did not provide data on allocation concealment (Class II) and nine trials did provide data on allocation concealment and failed on at least one other criterion (Class III). Dropouts occurred in 17 RCTs (range zero to 33 patients)
Sumatriptan (three RCTs) 6mg and 12mg given subcutaneously was statistically significantly more effective than placebo for improved headache response at 15 minutes (OR 6.22, 95% CI 3.61 to 10.72; two RCTs). A third RCT showed that sumatriptan 20mg nasal spray was significantly more effective than placebo for headache response at 30 minutes (p=0.002).
Zolmitriptan (three RCTs) 5mg and 10mg was significantly more effective than placebo for improved headache response at 30 minutes when taken as a nasal spray (OR 5.03, 95% CI 2.81 to 9.01; two RCTs) or orally (one RCT).
Compared to regular air, 100% oxygen (two RCTs) was significantly more effective in relieving headache (p<0.01, one RCT) and resulting in pain-free response (p<0.001, one RCT).
One RCT each assessed cocaine/lidocaine, octreotide, dihydroergotamine, ergotamine, somatostatin and prednisone. Results were reported in the review.
Lithium (two RCTs) 900mg showed significant improvement in the headache index compared to 360mg verapamil (p<0.01, one RCT). No significant differences were found in cessation of attacks between lithium 800mg and placebo (one RCT).
Verapamil (two RCTs) in one RCT showed that 360mg verapamil was significantly more effective than placebo in reducing headache attacks per day (p<0.001). The other RCT was listed above (under lithium).
Cimetidine/chlorpheniramine (two RCTs) showed no significant differences to placebo in the prevention of cluster attacks.
One RCT each assessed civamide, suboccipital steroid injection, sumatriptan, sodium valproate, melatonin, misoprostol, 100% hyperbaric oxygen, capsaicin, nitrate tolerance and prednisone. Results were reported in the review.
The most common adverse events were reported in the review.