Table 1. Characteristics of included randomized controlled trials.

Study, location & sample size (n randomized)	Centre, design & treatment arms (n analysed)*	Diagnostic criteria	Treatment history	Patients with medication overuse	Trial stimulation and/or nerve block required	Follow-up	Sponsorship & comments
Lipton et al. 2009 (PRISM study),[16] USA, n = 140	Multicentre, parallel-group, ONS (n = 63) vs. sham (n = 62)	ICHD-2 (migraine with or without aura, and/or chronic migraine)	Refractory to ≥ 2 acute and ≥2 prophylactic medications	Included (pre-specified subgroup)	Trial stimulation done but success was not an inclusion criteria	Double-blind 12 weeks, uncontrolled open label 1 year	Industry sponsored; published only as a conference abstract
Saper et al. 2011 (ONSTIM study),[17] USA, Canada & UK, n = 67	Multicentre, parallel-group, ONS (n = 28) vs. sham (n = 16) vs. medication management (n = 17)	ICHD-2 (chronic migraine)	Refractory to ≥ 2 classes of prophylactic medications	Excluded	Successful temporary nerve block (≥50% reduction in pain) required	Single-blind 12 weeks, uncontrolled open-label 3 years	Industry sponsored; also included an non-randomized ancillary arm (n = 8), in which patients who did not respond to occipital nerve block received active ONS
Silberstein et al. 2012,[28, 30] USA, n = 157	Multicentre, parallel-group, ONS (n = 105) vs. sham (n = 52)	ICHD-2 (chronic migraine) with modification using the Silberstein-Lipton diagnostic criteria for transformed migraine	Refractory to ≥ 2 acute and ≥2 classes of prophylactic medications	Possibly included (through the criteria for transformed migraine)	Successful trial stimulation (≥50% reduction in pain or adequate paresthesia) required	Double-blind 12 weeks, uncontrolled open-label 1 year	Industry sponsored
Serra, Marchioretto, 2012,[18] Italy, n = 30	Single centre, crossover, ‘ONS ‘on’ vs ‘off’ (n = 29)	Chronic migraine or medication overuse headache	Refractory to ≥ 2 prophylactic medications	Included	Successful trial stimulation (≥50% in the number or severity of attacks) required	Controlled open-label 2 x 1 month (no washout period), uncontrolled open-label 1 year	Hospital-based, no external funding
Slotty et al. 2014, [31] Germany, n = 8	Single centre, crossover, Suprathreshold vs subthreshold vs no stimulation (n = 8)	IHS criteria for chronic migraine	Treated with ONS & reported >30% pain relief for ≥3 months, on stable medication	Not described	All patients already had good response to ONS—see ‘Treatment history’	Double-blind (except suprathreshold stimulation), 3 x 1 week (no washout period)	No external funding

*The numbers of patients actually included in the analyses by study authors

ICHD-2: International Classification of Headache Disorder 2nd edition, IHS: International Headache Society, ONS: occipital nerve stimulation.