methodological critique of randomised trials for oral treatments
The International Headache Society (IHS) provides guidance on the conduct of trials for acute treatment of episodic tension-type headache (TTH), a common disorder with considerable disability. Electronic and other searches identified randomised, double-blind trials of oral drugs treating episodic TTH with moderate or severe pain at baseline, or that tested drugs at first pain onset. The aims were to review methods, quality, and outcomes reported (in particular the IHS-recommended primary efficacy parameter pain-free after 2 hours), and to assess efficacy by meta-analysis. We identified 58 reports: 55 from previous reviews and searches, 2 unpublished reports, and 1 clinical trial report with results. We included 40 reports of 55 randomised trials involving 12,143 patients. Reporting quality was generally good, with potential risk of bias from incomplete outcome reporting and small size; the 23 largest trials involved 82% of patients. Few trials reported IHS outcomes. The number needed to treat values for being pain-free at 2 hours compared with placebo were 8.7 (95% confidence interval [CI] 6.2 to 15) for paracetamol 1000 mg, 8.9 (95% CI 5.9 to 18) for ibuprofen 400mg, and 9.8 (95% CI 5.1 to 146) for ketoprofen 25mg. Lower (better) number needed to treat values were calculated for outcomes of mild or no pain at 2 hours, and patient global assessment. These were similar to values for these drugs in migraine. No other drugs had evaluable results for these patient-centred outcomes. There was no evidence that any one outcome was better than others. The evidence available for treatment efficacy is small in comparison to the size of the clinical problem.