Schizophrenia is a chronic mental health condition characterised by altered perceptions, thoughts, mood and behaviour. In many cases antipsychotic medication is necessary for the management of the condition, however treatment is not always effective and some people may require the addition of other interventions to experience an improvement in symptoms and quality of life. Such interventions include the addition of other antipsychotic medication and/or benzodiazepines as well as other interventions such as psychological therapies, electroconvulsive therapy (ECT) and transcranial magnetic stimulation (TMS).
In 2009, Gallego and colleagues reported that nearly 20% of people with schizophrenia spectrum disorders across the world receive a combination of antipsychotics to manage their condition. This is despite any clear evidence to suggest they provide any additional benefit over monotherapy (a single antipsychotic drug). As the combination of antipsychotic medication may lead to an increased risk of side-effects, non-adherence and even mortality, it is very important to investigate the safety of this practice.
To investigate whether the combination of antipsychotic medication is both safe and associated with improved outcomes in schizophrenia, Oritz-Orendain and colleagues (2017) conducted a Cochrane review comparing antipsychotic monotherapy with combination treatment.
To identify relevant evidence the authors searched the Cochrane Schizophrenia Group’s Study-Based Register of Trials in January 2016. This register is compiled by systematic searches of major electronic databases (including AMED, BIOSIS, CINAHL, Embase, MEDLINE, PsycINFO, PubMed), registries of clinical trials, hand-searches, grey literature, and conference abstracts. Evidence identified through previous searches of this register in September 2010 and August 2012 was also included. Once included studies were established, reference lists were searched for additional studies and first authors contacted for unpublished trials.
|Study design||Randomised controlled trials (RCTs) and quasi-RCTs|
|Participants||Adults with schizophrenia and related disorders, including schizophreniform disorder, schizoaffective disorder and delusional disorder|
|Any combination of antipsychotic medication
|Setting||Inpatient services, outpatient programmes and community volunteer services
1. Clinical response: (a) no response or (b) relapse (any definition)
2. Leaving the study early
1. Clinical response: changes in global state, mental state, positive symptoms, negative symptoms and aggression/agitation symptoms
2. Service utilisation (e.g. hospital admission)
3. Adverse events
4. Quality of life
5. Economic burden (cost of care)
6. Behaviour (e.g. social functioning, employment, substance abuse)
Where data were available the authors carried out random effects meta-analyses for the primary and secondary outcomes, calculating relative risks (RR) for dichotomous outcomes and mean differences (MD) for continuous outcomes. Heterogeneity and publication bias were also assessed.
Outcomes were arranged into four groups according to the type of antipsychotics used in each study arm:
- Atypical antipsychotics other than clozapine
- Typical antipsychotics
- Any antipsychotic.
Where possible further analyses of the following subgroups were conducted:
- Clozapine combinations compared with non-clozapine combinations
- Enrolment of acutely exacerbated patients compared with chronically ill patients
- Treatment duration <12 weeks compared with ≥12 weeks
- Type of drug added to clozapine treatment
Sensitivity analyses were also conducted for the following factors:
- Randomisation described compared with randomisation not described
- Intention to treat analysis (ITT) compared with completer analysis
- High risk of bias
- Fixed effects meta-analysis compared with random effects analysis (primary outcomes only)
The quality of individual studies was assessed using the Cochrane risk of bias tool (Higgins et al., 2011). For individual outcomes, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework was used to assess confidence in estimates of the effect which can range from very low quality (very little confidence in estimate of effect) to high quality (confident the true effect lies close to the estimate) (Schünemann et al., 2011).
In total, 62 studies met the inclusion criteria for this review.
Most studies were less than 12 weeks in duration (k=47) and conducted in inpatient settings (k=30). Other settings were outpatient services (k=16), a mix of outpatients and inpatients (k=7) or community settings (k=2). Only seven studies reported outcomes at more than six months follow-up. On average 78 participants per study were included (4,833 people in total).
Drug combinations included the following:
|Analysis||Combination therapy||Monotherapy||Number of studies|
|1. Clozapine||Clozapine combined with a typical or atypical antipsychotic||Clozapine||k=31|
|2. Atypical||Any atypical antipsychotic combination†||Any atypical antipsychotic||k=18|
|3. Typical||Any typical antipsychotic combined with a typical or atypical antipsychotic†||Any typical antipsychotic||k=9|
|4. Any combination||Study participants were allowed to take any combination of antipsychotics||Any monotherapy||k=4|
1a. Clinical response
A meta-analysis of 29 studies indicated a reduced risk of no clinical response with combination therapy compared with monotherapy (RR=0.73, 95% CI, 0.63 to 0.84) at follow-ups ranging from six weeks to three years. There was however significant heterogeneity (I2=54%) and the quality of the evidence was rated as very low. Removing the shorter trials from this meta-analysis did however reduce the heterogeneity and maintain the effect.
When looking individually at different types of combination therapy, the effect held for the clozapine (RR=0.66, 95% CI, 0.53 to 0.83) and typical (RR=0.64, 95% CI, 0.49 to 0.84) analyses, however there was still significant heterogeneity. There was no evidence of benefit for the atypical analysis (RR=0.95, 95% CI, 0.83 to 1.09). No studies including any combination of antipsychotics reported this outcome.
Only three studies reported relapse at 8 weeks, 1 year and 3 years follow-up. As there was significant heterogeneity (I2=82%) results were not pooled. Overall, one study showed a reduced risk of relapse with combination therapy, whereas two studies showed no significant difference between combination therapy and monotherapy.
2. Leaving the study early
A meta-analysis of 43 studies indicated no difference in the number of people leaving the study early at follow-ups ranging from six weeks to one-year (RR=0.90, 95% CI, 0.76 to 1.07). The quality of the evidence was rated low.
Overall, there were mixed findings across secondary outcomes, with some trials showing some evidence of benefit and others indicating no difference between groups. Significant heterogeneity was common meaning that for many comparisons data were not pooled.
For global state and mental state there was evidence of benefit on some measures but no evidence of improvement on others. For hospital admission, very low quality evidence from three trials indicated no evidence of benefit for combination therapy compared with monotherapy (RR=0.96, 95% CI, 0.36 to 2.55). In a meta-analysis of 30 studies, very low quality evidence showed no difference between groups in the number of serious adverse events or study discontinuations due to adverse events (RR=1.05, 95% CI, 0.65 to 1.69). Quality of life was assessed in four trials but could not be combined as three different tools were used. All individual trials reported no significant difference between groups for quality of life at follow-ups ranging from 6-16 weeks.
The authors concluded that:
Currently, most evidence regarding the use of antipsychotic combinations comes from short-term trials, limiting the assessment of long-term efﬁcacy and safety. We found very low-quality evidence that a combination of antipsychotics may improve the clinical response. We also found very low-quality evidence that a combination of antipsychotics may make no difference at preventing participants from leaving the study early, preventing relapse and/or causing more serious adverse events than monotherapy.
Strengths and limitations
This is a well conducted systematic review: evidence searches were thorough, data were extracted independently, authors were contacted for missing data, risk of bias was investigated, and the quality of individual outcomes was assessed using the GRADE approach. As a reader, it was great to see all the authors’ methodological and statistical decisions reported so transparently, with challenges and limitations clearly acknowledged.
One of the main limitations of this review lies in the quality of included studies, with outcomes rated as low or very low quality which really questions our confidence in the reported estimates. There was a particularly high or unclear risk of reporting bias and attrition bias across studies and as authors included quasi-randomised trials there was a risk of selection bias in ~70% of studies. Whilst five studies were clearly quasi-randomised, for 39 studies there was not enough information to judge whether sequence generation was truly random, so the risk of bias here was unclear rather than high.
A further limitation is that the clinical history of study participants varied quite a lot from study to study, which may explain some of the heterogeneity found across outcomes. In some cases, participants were already receiving combination therapy and were randomised to either continue with the regimen, discontinue one of the drugs, or switch to a new regimen. In other cases, participants had only been receiving one antipsychotic and were randomised to either continue the same treatment, add on a further antipsychotic or switch to a new regimen. Reasons for changing treatment were due both to lack of efficacy and problems with side effects, so there is a risk here that the authors were not comparing like with like. Finally, as the authors included a large number of secondary outcomes, it was quite hard to make sense of the data reported.
As expected with a Cochrane publication, this review was well conducted and clearly reported. Findings suggested a 27% reduced risk of no clinical response with combination therapy, but no clear evidence of benefit between groups for risk of relapse and study discontinuation. Serious adverse events did not appear more likely in either group. For other secondary outcomes, findings were generally mixed, so it was difficult to draw any clear conclusions.
In a systematic review of recovery rates in schizophrenia, Jaaskelainen and colleagues (2012) found that only 13.5% (median proportion) of participants met their recovery criteria, so there is clearly a great need to explore which interventions may help support people with schizophrenia on the pathway to recovery. Sadly, I don’t think this review provides any clear answers on this. As suggested in NICE clinical guidelines for Psychosis and Schizophrenia, should someone experience a lack of response to antipsychotic medication, other aspects of care such as treatment adherence, receipt of a family intervention or cognitive behavioural therapy, and other causes of non-response (substance misuse or concurrent use of prescribed medication) should be considered before adding a second antipsychotic to augment treatment.
Ortiz-Orendain J, Castiello-de Obeso S, Colunga-Lozano LE, Hu Y, Maayan N, Adams CE. (2017) Antipsychotic combinations for schizophrenia. Cochrane Database of Systematic Reviews 2017, Issue 6. Art. No.: CD009005. DOI: 10.1002/14651858.CD009005.pub2.
Gallego JA, Bonetti J, Zhang J, Kane JM, Correll CU. (2012) Prevalence and correlates of antipsychotic polypharmacy: A systematic review and meta-regression of global and regional trends from the 1970s to 2009. Schizophrenia Research. 2012;138(1):18–28.
Hasan A, Falkai P, Wobrock T, Lieberman J, Glenthoj B, Gattaz WF, et al. (2012) World Federation of Societies of Biological Psychiatry (WFSBP) Guidelines for Biological Treatment of Schizophrenia, part 1: update 2012 on the acute treatment of schizophrenia and the management of treatment resistance. World Journal of Biological Psychiatry. 2012;13(5):318–78.
Higgins JPT, Altman DG, Sterne JAC, editor(s). Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from www.handbook.cochrane.org.
Schünemann HJ, Oxman AD, Vist GE, Higgins JPT, Deeks JJ, Glasziou P, et al. Chapter 12: Interpreting results and drawing conclusions. In Higgins JPT, Green S, editor(s), Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org.
Jääskeläinen E, Juola P, Hirvonen N, McGrath J, Saha S, Isohanni M, et al. (2012) A systematic review and meta-analysis of recovery in schizophrenia. Schizophrenia Bulletin. 2012;39(6):1296–306.
NICE (2014) Psychosis and Schizophrenia in Adults: The NICE Guideline on Treatment and Management (Updated Edition) (Clinical Guideline CG178). National Institute for Health and Care Excellence, 2014.