Antipsychotic medications for cocaine dependence


Cocaine dependence is often associated with medical, psychological and social problems for individual and public health, generating problems for the community. Users play a role in the spread of infectious diseases such as AIDS, hepatitis and tuberculosis, as well as in crime, violence and neonatal drug exposure. Use of drugs such as antidepressants, anticonvulsants and dopamine agonists to treat cocaine abuse or dependence is not supported by evidence from Cochrane reviews. The use of antipsychotic agents has also been considered, particularly because cocaine can induce hallucinations and paranoia that mimic psychosis.

Study characteristics

The review authors identified 14 randomised controlled trials involving 719 adults. One study was conducted in Italy, and the rest in the USA. They involve both inpatient and outpatient settings and had a duration of 14 to 168 days (mean 80 days). Eleven trials randomised participants to receive an antipsychotic drug or placebo using the following antipsychotic medications: risperidone (three studies, 1 to 4 mg/day and one study with injections of long-acting risperidone at a dose of 25 mg/14 days); olanzapine (three studies, 2.5 to 20 mg/day); quetiapine (two studies, 400 and 800 mg/day); lamotrigine (one study, 400 mg/day); reserpine (one study, 50 mg/day). Three trials compared two drugs; olanzapine (10 mg/day) versus haloperidol (10 mg/day), olanzapine (20 mg/day) versus risperidone (9 mg/day) and aripiprazol (10 mg/day) versus ropirinol (4.5 mg/day).

Key results

The studies used different instruments or ways to assess the outcomes of interest, limiting the possibility for us to combine the data. When we grouped together all trial results comparing any antipsychotic drug to placebo, we found that antipsychotics slightly increase those who stayed in treatment but they were not effective in reducing cocaine use during treatment (two studies), in sustained abstinence (three studies), or in reducing the urge to consume cocaine (four studies). The single comparisons of each drug versus placebo or versus another drug were made in few trials with small sample sizes, limiting the reliability of the results. However, among these comparisons, only quetiapine seemed to perform better than placebo in reducing cocaine use and craving, but results came only from one study with 60 participants. Information was limited on the acceptability of treatment in terms of side effects, abstinence from cocaine use and withdrawal symptoms. Overall we found no evidence supporting the clinical use of antipsychotic medications in the treatment of cocaine dependence.

Quality of the evidence

The major limitations of the studies were the high number of people who withdrew from them and the lack of clear reporting of the methods used to conduct the studies. Moreover, the number of participants was small, and different ways of measuring and reporting results were used, limiting the possibility for us to combine the data. Overall we judged the quality of the evidence to be moderate for dropouts and low for all the other outcomes considered. The evidence is current up to 15 of July 2015.

Funding and conflict of interest reported by the studies

The majority of trials included in this review had funding from industrial sources or declared conflict of interests for some of the researchers due to different contractual collaborations with the pharmaceutical industry. Only five of the 14 included trials reported being funded exclusively by non-industry sources, and of these just one (Grabowski 2004) disclosed no conflict of interest for the authors. Another study (Brown 2012) reported conflicts of interest for several authors and three studies (Levin 1999, Reid 2005 and Winhusen 2007) did not disclose conflict of interest of the authors. One included trial (Meini 2010) did not report information about funding sources, but disclosed no conflict of interest for the authors. The other eight studies included in this review were either funded by industry (Brown 2010; Hamilton 2009; Kampman 2003), or by a combination of industry and non-industry grants (Akerele 2007; Loebl 2008; Smelson 2004; Smelson 2006; Tapp 2015), with three (Brown 2010; Hamilton 2009; Kampman 2003) disclosing conflicts of interests for the authors, and the rest without declaration on this issue.

Authors' conclusions: 

At present, there is no evidence supporting the clinical use of antipsychotic medications in the treatment of cocaine dependence, although results come from only 14 trials, with small sample sizes and moderate to low quality of evidence.

Read the full abstract...

Cocaine dependence is a public health problem characterised by recidivism and a host of medical and psychosocial complications. Cocaine dependence remains a disorder for which no pharmacological treatment of proven efficacy exists.


To evaluate the efficacy and the acceptability of antipsychotic medications for cocaine dependence.

Search strategy: 

This review is an update of a previous Cochrane review published in 2007. We searched up to 15 July 2015 in Cochrane Drugs and Alcohol Group Specialised Register (searched in CRSLive); the Cochrane Library (including the Cochrane Central Register of Controlled Trials (CENTRAL); the Database of Abstracts of Reviews of Effects (DARE)); PubMed; EMBASE; CINAHL and Web of Science. All searches included non-English language literature.

Selection criteria: 

All randomised controlled trials and controlled clinical trials with focus on the use of any antipsychotic medication for the treatment of cocaine dependence.

Data collection and analysis: 

We used standard methodological procedures expected by Cochrane.

Main results: 

We included 14 studies (719 participants). The antipsychotic drugs studied were risperidone, olanzapine, quetiapine, lamotrigine, aripiprazol, haloperidol and reserpine. Comparing any antipsychotic drugs versus placebo, we found that antipsychotics reduced dropout: eight studies, 397 participants, risk ratio (RR) 0.75 (95% confidence interval (CI) 0.57 to 0.97), moderate quality of evidence. We found no significant differences for any of the other primary outcomes considered: number of participants using cocaine during the treatment, two studies, 91 participants: RR 1.02 (95% CI 0.65 to 1.62); continuous abstinence, three studies, 139 participants: RR 1.30 (95% CI 0.73 to 2.32); side effects, six studies, 291 participants: RR 1.01 (95% CI 0.93 to 1.10); and craving, four studies, 240 participants: RR 0.13 (-1.08 to 1.35). For all of these comparisons we rated the quality of evidence as low.

Comparisons of single drug versus placebo or versus another drug are conducted in few trials with small sample sizes, limiting the reliability of the results. Among these comparisons, only quetiapine seemed to outperform placebo in reducing cocaine use, measured by grams per week: mean difference (MD) -0.54 (95% CI -0.92 to -0.16), by US dollars spent per week: MD -53.80 (95% CI -97.85 to -9.75), and by craving: MD -1.23 (95% CI -2.19 to -0.27), but results came from one study with 60 participants.

The major limitations of the studies were the high risk of attrition bias (40% of the included studies) and low quality of reporting, mainly for the risk of selection bias, performance and detection bias, that we rated as being at unclear risk for 75% to 80% of the studies. Furthermore, most of the included studies did not report results on important outcomes such as side effects, or use of cocaine during treatment and craving, which prevented the possibility of including them in statistical synthesis.

Health topics: