Nine RCTs were used to assess home-based behavioural treatments versus control.
Five RCTs were used to assess home-based behavioural treatments versus control at follow-up.
Five RCTs were used to assess clinic-based behavioural treatment versus control.
Thirteen RCTs were used to assess home-based behavioural treatments versus clinic behavioural treatment.
Four RCTs were used to assess home-based behavioural treatments versus clinic behavioural treatment at follow-up.
Seven RCTs were used to assess cost-effectiveness of home-based behavioural treatments versus clinic behavioural treatment.
Home-based behavioural treatments (HBT) vs control: d = 0.51 (95%CI: 0.27, 0.76) for all headache types and outcomes. At follow-up HBT reduced contact treatments d = 0.70 (95%CI: 0.35, 1.06).
Clinic-based behavioural treatment vs control: d = 0.52 (95%CI: 0.19, 0.85). At follow-up clinic-based treatment versus control d = 0.60 (0.13, 1.07).
Home-based behavioural treatments vs clinic behavioural treatment: overall d = 0.09 (95%CI: - 0.11, 0.29). Adult d = 0.09 (95%CI: -0.15, 0.33).
Paediatric d = 0.07 (95%CI: - 0.30, 0.44).
Results from separate analysis for individual outcome measures are given. Home-based behavioural treatments vs clinic behavioural treatment at follow-up: d = 0.11 (95%CI: - 0.28, 0.50).
Sensitivity analysis: studies providing exact effect sizes d = 0.09 (95%CI: - 0.15, 0.33) compared to studies not providing exact effect sizes d = 0.07 (95%CI: - 0.28, 0.43); studies in which allocation to groups was randomised d = 0.08 (95%CI: - 0.14, 0.29) compared to studies with suspect randomisation d = 0.13 (95%CI: - 0.35, 0.62).