Psychostimulants may reduce cocaine and heroin use in people with cocaine dependence, but evidence is of very low certainty and trial durations are short (6 to 24 weeks).

Compared with placebo in adults dependent on or abusing cocaine, very low‐certainty evidence suggests that psychostimulants (bupropion, dexamphetamine, lisdexamfetamine, methylphenidate, modafinil, mazindol, methamphetamine, mixed amphetamine salts, or selegiline) may increase the number of people with sustained cocaine abstinence (on average, 224 vs 164 per 1000 people) or sustained heroin abstinence (on average, 644 vs 375 per 1000 people), defined by most RCTs as three weeks without use of the substance. Overall cocaine or heroin use may show a small reduction (mean proportion of cocaine‐free urinalyses across the study per participant), but analyses did not quite reach statistical significance. Fewer people seemed to experience a serious adverse event with psychostimulants than with placebo (7 vs 12 per 1000 people), but event rates were very low in both groups and rates of withdrawal due to adverse events were similar across groups.

Subgroup analyses were conducted by type of drug (bupropion, dexamphetamine, methylphenidate, modafinil), definition of cocaine use (abuse or dependence), and presence of comorbid opioid dependence or attention deficit hyperactivity disorder (ADHD). Most often, these analyses showed similar results to the main analysis or were too small to provide clinically meaningful results. Summaries are provided in the table below.

It is worth noting that trial duration ranged from 6 to 24 weeks, and all participants received psychotherapy (most received cognitive‐behavioral therapy). Generalizability of these results may be limited by the demographics of participants: 75% were male and 50% African American. In addition, when reported, lifetime cocaine use ranged from 8 to 22 years, and route of cocaine use was inhalation for 61% of participants.