Both epidemiological and clinical studies consistently report high rates of substance use among individuals with schizophrenia and bipolar disorder Reference Regier, Farmer and Rae(1,Reference Lubman and Sundram2). Up to 90% smoke cigarettes, while between 40 and 60% misuse other substances Reference Lubman and Sundram(2,Reference Green, Noordsy, Brunette and O’Keefe3). Co-occurring substance use disorders (SUDs) (excluding tobacco smoking) appear to be more prevalent (up to 75%) among young people with first-episode psychosis Reference Lambert, Conus and Lubman(4), as well as those who are homeless or involved with the criminal justice system. The most frequently misused substances are cannabis, alcohol and psychostimulants, mirroring patterns of use evident within the general population, although misuse of more than one substance is relatively common Reference Lubman and Sundram(2,Reference Green, Noordsy, Brunette and O’Keefe3).

Co-occurring SUDs affect medication adherence and service engagement Reference Lubman and Sundram(2,Reference Green, Noordsy, Brunette and O’Keefe3), and can precipitate mixed states and rapid cycling in bipolar disorder Reference Kemp, Gao, Ganocy, Elhaj, Bilali, Conroy, Findling and Calabrese(5). SUDs substantially increase the burden on patients, their families and the health care system, and those with co-occurring disorders generally do poorly in standard treatment settings Reference Lubman and Sundram(2). Indeed, such patients have lower rates of remission and increased rates of relapse, hospitalisation, housing instability, blood-borne virus infection [e.g. human immunodeficiency virus (HIV)], suicide, violent behaviour, incarceration and early mortality Reference Lubman and Sundram(2,Reference Green, Noordsy, Brunette and O’Keefe3). It is therefore not surprising that many clinicians feel overwhelmed or not sufficiently skilled to manage such patients Reference Lubman and Sundram(2).

Despite such clear consequences, SUDs are frequently underdiagnosed or ignored within psychiatric treatment settings, with many clinicians considering that SUD treatment should be assessed and managed independently of the mental disorder. In younger populations, psychosis or mood instability in the context of co-occurring substance misuse can present a particularly difficult diagnostic challenge, especially if the substance use is longstanding Reference Mathias, Lubman and Hides(6). As such, it is not uncommon for these patients to be excluded from psychiatric services if their substance use is deemed to be the primary problem, although there is good evidence that individuals with substance-induced psychosis are a particularly high-risk group for later transition to psychotic disorders, and have high levels of psychological morbidity Reference Mathias, Lubman and Hides(6).

Such stigma is also evident within the clinical treatment literature, with SUDs being a common exclusion criterion from the bulk of existing pharmaceutical industry trials. Thus, while we have good evidence for the efficacy of psychotropic medications in the treatment of schizophrenia and bipolar disorder uncomplicated by substance misuse, we have limited evidence for their utility in the treatment of patients with co-occurring SUDs. Whether they are as effective and whether recommended dosages are applicable to this population remain legitimate clinical questions. Despite such concerns, most clinicians prescribe similar pharmacotherapeutic regimes for patients with co-occurring SUD as those without substance use issues, with limited outcome-related research available to inform practice.

What then is the current state of the literature in terms of the pharmacotherapy of co-occurring disorders? Despite the wealth of large randomised controlled trial data available for bipolar disorder and schizophrenia, most studies examining the pharmacotherapy of co-occurring disorders have been case studies, case series, open-label trials or retrospective surveys Reference Green, Noordsy, Brunette and O’Keefe(3,Reference San, Arranz and Martinez-Raga7). While a few randomised prospective trials have been conducted, they have typically been small in size with limited long-term follow-up Reference San, Arranz and Martinez-Raga(7,Reference Baker, Hides and Lubman8). They have also tended to examine the efficacy of existing psychotropic agents (i.e. antipsychotics or mood stabilisers) rather than the role of adjunctive pharmacotherapies that specifically target each disorder. Nonetheless, despite such limitations, there is a growing body of evidence that directly informs clinical treatment.

In terms of schizophrenia, few studies with typical antipsychotics have been conducted. While there is no clear evidence that oral neuroleptics decrease substance misuse in patients with co-occurring SUDs, there are concerns that they may actually exacerbate use, particularly as extrapyramidal side-effects appear to be more commonly reported by patients with co-occurring disorders Reference Green, Noordsy, Brunette and O’Keefe(3,Reference San, Arranz and Martinez-Raga7). A number of small open-label studies have reported positive outcomes for flupenthixol decanoate over 6–12 months Reference San, Arranz and Martinez-Raga(7); however, it is unclear how much this reflects adequate antipsychotic treatment, rather than specific characteristics of the agent itself. Few other studies of depot preparations have been conducted, although a recent open-label randomised trial reported better substance use and psychosis outcomes over 6 months for patients on long-acting injectable risperidone when compared with zuclopenthixol Reference Rubio, Martinez, Ponce, Jimenez-Arriero, Lopez-Munoz and Alamo(9).

Studies examining the effectiveness of oral atypical agents have generally been positive, although most of this research has been conducted with clozapine Reference Green, Noordsy, Brunette and O’Keefe(3,Reference San, Arranz and Martinez-Raga7). While case reports, case series and open-label trials all show improvements in both psychotic symptoms and substance use following treatment with clozapine, no randomised prospective studies have been conducted to date. Nevertheless, the consistency of these findings suggests that clozapine is an effective pharmacotherapy for patients with co-occurring disorders, although the risk of significant side-effects limits its usage in everyday practice. A number of studies have examined treatment with risperidone or olanzapine, with reports that both atypicals are well tolerated by patients with co-occurring SUDs Reference Green, Noordsy, Brunette and O’Keefe(3,Reference San, Arranz and Martinez-Raga7). Although some beneficial effects have been documented, including decreased levels of subjective craving, the few small randomised trials that have been conducted have produced conflicting findings Reference San, Arranz and Martinez-Raga(7). Similarly, several small open-label trials with quetiapine and aripiprazole have reported some positive outcomes, however, randomised prospective studies are currently unavailable Reference Green, Noordsy, Brunette and O’Keefe(3,Reference San, Arranz and Martinez-Raga7). While fewer studies have been conducted among patients with bipolar disorder and co-occurring SUDs, clinical trials have also tended to focus on examining the efficacy of standard treatments. A number of double-blind randomised trials of traditional mood stabilisers (e.g. lithium and valproate), alone or in combination, and atypical antipsychotics have reported beneficial mood outcomes in patients with co-occurring disorders, although limited improvements in substance use were observed (Reference Kemp, Gao, Ganocy, Elhaj, Bilali, Conroy, Findling and Calabrese5,Reference Brown, Garza and Carmody10–Reference Salloum, Cornelius, Daley, Kirisci, Himmelhoch and Thase12).

Several randomised, placebo-controlled, double-blind trials have reported that naltrexone may be helpful in reducing alcohol use among patients with schizophrenia and co-occurring alcohol dependence Reference Petrakis, Nich and Ralevski(13). Disulfiram has also been shown to be effective in this regard Reference Green, Noordsy, Brunette and O’Keefe(3), although concerns have been raised regarding its potential to induce psychosis. However, such events do not seem to be a significant issue in treated samples of patients with serious mental illness and alcohol dependence. While a few studies have suggested that sodium valproate may reduce drinking in patients with bipolar disorder and alcohol dependence Reference Salloum, Cornelius, Daley, Kirisci, Himmelhoch and Thase(12), no such studies have been conducted in schizophrenia.

So what does all this mean for those working at the clinical coalface? The existing evidence, although limited, highlights the importance of optimising the pharmacological management of patients with schizophrenia, with evidence that atypical antipsychotics should be considered as first-line agents. Similarly in bipolar disorder, optimising standard treatment is recommended, with some evidence for both mood stabilisers and atypical antipsychotics. For those patients with alcohol dependence, a number of adjunctive agents appear to be efficacious, although no such evidence currently exists regarding adjunctive pharmacotherapy for other drug disorders Reference Baker, Hides and Lubman(8). While inpatient detoxification should be offered to dependent patients who are not able to reduce their substance use in the community, or those who are particularly chaotic or disorganised, no empirical treatment studies of detoxification have been conducted among patients with co-occurring SUDs, with few evidence-based withdrawal guidelines currently available for this population.

Optimising the pharmacological management of patients with co-occurring SUDs should be organised in concert with integrated psychosocial approaches that are evidence based (Reference Lubman and Sundram2,Reference Baker, Hides and Lubman8,Reference Drake, O’Neal and Wallach14). In fact, more rigorous research has been conducted examining the delivery of effective psychosocial interventions and programs among patients with co-occurring disorders than pharmacotherapy trials. Findings from these studies recommend the flexible provision of integrated interventions that target both disorders (e.g. motivational interviewing, cognitive behaviour therapy, social skills training), matched to the patient's stage of change. Ideally, treatment should be offered on a long-term basis within a multidisciplinary team, with access to assertive case management (available 24 h), as well as personnel with specialty training in psychiatry and addiction medicine. Involvement with self-help or 12-step programs is also encouraged.

Together, this brief overview highlights the dearth of research evidence examining effective pharmacotherapeutic approaches to patients with co-occurring disorders. This must be considered a significant oversight, particularly given the prevalence of such presentations within routine clinical practice. Relevant large-scale prospective trials that examine the effectiveness of both psychotropic and adjunctive agents in patients with co-occurring disorders are essential to provide a strong and clinically relevant evidence base to inform practice.