ABSTRACT
BACKGROUND: Dependence-like behaviour may complicate withdrawal and increase risk of relapse of medication overuse headache (MOH). The most effective treatment for reducing dependence-like behaviour is unknown. OBJECTIVES: To compare patient-reported outcomes among three treatment strategies for MOH. The primary outcome was change in Severity of Dependence Scale (SDS) score from baseline to 6 months. METHODS: Patients with MOH were randomized to (1) withdrawal combined with preventive medication from start (W+P), (2) preventive medication without withdrawal (P), or (3) withdrawal with optional preventive medication 2 months after withdrawal (W). At baseline, 2, and 6 months, patients filled out SDS (used for measurements of dependence-like behaviour and treatment feasibility), Headache Under-Response of Treatment (HURT) and WHO Quality of Life BREF questionnaires. RESULTS: Out of 120 patients with MOH, 100 completed the 6-month follow-up and filled out questionnaires. The W+P arm was the most effective in treating MOH. After 6 months, the SDS score was reduced by 3.69 (95% CI 3.23-4.49) in the W+P arm, by 3.19 (95% CI 2.43-3.96) in the W arm, and by 1.65 (95% CI 0.96-2.33) in the P arm (p = 0.04). At baseline and after 2 months, the P arm was considered the most feasible treatment, but at 6-month follow-up, there was no difference in feasibility score, change in HURT score, or quality of life. CONCLUSIONS: Dependence-like behaviour was reduced most in the two withdrawal arms. Withdrawal combined with preventive medication is recommended for the treatment of MOH. SIGNIFICANCE: Withdrawal combined with preventive medication from start is the treatment strategy that reduces dependence-like behaviour the most in MOH patients. Patients initially considered preventive treatment without withdrawal as the most feasible treatment. However, no difference in feasibility between the three arms was found at 6-month follow-up. Withdrawal combined with preventive medication is recommended for treatment of MOH.
