Management of Amphetamine and Methamphetamine Use Disorders: A Systematic Review and Network Meta-analysis of Randomized Trials

Abstract

The current evidence regarding effective management of methamphetamine and amphetamine (MA/A) use disorders is inconclusive. Therefore, we assessed the comparative benefits and tolerability of pharmacological, psychosocial, and harm reduction interventions for management of MA/A use disorders. We searched six electronic databases for randomized controlled trials of any pharmacological, psychosocial, or harm reduction interventions in adults with MA/A use disorders. We performed a random-effects frequentist network meta-analysis and used the GRADE approach to assess the certainty of evidence. We included 72 randomized trials (6836 participants). Low certainty evidence suggests quetiapine may extend abstinence compared to placebo [risk ratio (RR) 3.17 (95% confidence intervals (CI), 1.24 to 8.07)] and weekly average proportion of patients with negative urine samples [mean difference (MD) 32.17 (95% CI, 14.08 to 50.26)]. Low certainty evidence also suggested that riluzole may be associated with a higher weekly average proportion of patients with negative urine samples [MD 24.10 (95% CI, 5.54 to 42.66)]. Very low certainty evidence suggests methylphenidate alone [compared to placebo MD 10.24 (95% CI, 3.49 to 16.99)] or in combination with matrix model [MD 23.55 (95% CI, 7.64 to 39.46)] may be associated with an increased weekly average proportion of patients with negative urine samples. Compared to placebo, contingency management alone [MD 21.20 (95% CI, 11.39 to 31.00), very low certainty] or in combination with cognitive behavioural therapy [MD 34.85 (95% CI, 19.63 to 50.08), very low certainty] may be associated with longer duration of abstinence. Compared to placebo, venlafaxine [RR 0.27 (95% CI 0.08 to 0.90), low certainty] and citicoline [RR 0.69 (95% CI 0.49 to 0.99), lowcertainty] may be among the most tolerable interventions. Very few interventions may be associated with slight improvement in certain outcomes, but no intervention showed moderate- to high-certainty evidence for important changes across any patient-important outcomes.