NPC Archive Item: A comparison of the effectiveness of antimanic drugs in acute mania

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

25 November 2011

A systematic review and multiple-treatment meta-analysis suggests that antipsychotic drugs are significantly more effective than mood stabilisers in the acute treatment of manic episodes in adults. Gabapentin, lamotrigine▼ and topiramate were not significantly more effective than placebo.

Level of evidence:
Level 2 (limited quality patient-oriented evidence) according to the SORT criteria.

While adding to the evidence base, this study should not by itself change practice. Health professionals should continue to follow the NICE clinical guideline on the management of bipolar disorder. If a patient develops acute mania when not taking antimanic medication, treatment options include starting an antipsychotic, valproate or lithium, taking into account patient preferences for future prophylactic use and the side-effect profile.

  • Antipsychotics (normally olanzapine, quetiapine▼* or risperidone▼* taking into account individual risk factors for side effects) are recommended if there are severe manic symptoms or marked behavioural disturbances as part of the syndrome of mania.

Valproate or lithium are recommended if symptoms have responded to these drugs before, and the person shows good compliance; however, valproate should be avoided in women of child-bearing potential. Lithium has a slower onset of action than valproate or antipsychotics, and NICE recommends lithium only if symptoms are not severe. See the NICE guideline for additional recommendations for the treatment of acute mania, including the use of drugs for people who are already taking antimanic medication and suffer a manic episode.

What is the background to this?
Mania usually occurs in association with episodes of depression, and defines the diagnosis of bipolar disorder. ‘Mood stabilisers’ (e.g. valproate, lithium and carbamazepine) and antipsychotics are the mainstay of treatment of acute mania. This study systematically reviewed the results of randomised controlled trials (RCTs) that evaluated the effectiveness of drugs for the treatment of acute mania in adults by meta-analysis of direct and indirect comparisons between active treatments or placebo.

What does this study claim?
Using data from 68 RCTs (total n=16,073), the study found that antipsychotics were significantly more effective than mood stabilisers overall for the treatment of acute manic episodes over a three-week period. This was judged by the mean change in mania rating scales (efficacy) and the number of patients who dropped out of the allocated treatment (acceptability) at three weeks.

In terms of efficacy, in order of highest to lowest effect size, haloperidol, risperidone, olanzapine, lithium, quetiapine, aripiprazole▼*, carbamazepine, asenapine, valproate and ziprasidone were all statistically significantly more effective than placebo (see study details) for individual effect sizes compared with placebo). No significant differences in efficacy were found between gabapentin, lamotrigine or topiramate and placebo. Haloperidol was significantly more effective than lithium, quetiapine, aripiprazole, carbamazepine, asenapine, valproate, ziprasidone, lamotrigine, topiramate and gabapentin. Olanzapine, risperidone and quetiapine led to significantly fewer discontinuations than did lithium, lamotrigine, topiramate, gabapentin and placebo. Olanzapine and risperidone were the two treatments most likely to be ranked as superior for both efficacy and acceptability.

How does this relate to other studies?
A systematic review of RCTs was carried out to inform the NICE clinical guideline for bipolar disorder (2006). Although identifying significant differences for antipsychotics and mood stabilisers over placebo, it identified few head-to-head trials of single agents in mania, and no convincing evidence for differential efficacy or tolerability (with the exception of a single trial comparing valproate and lithium). In view of this, NICE recommendations took into account non-RCT data and practical issues.

Recently, using a traditional meta-analytical approach, Yildiz et al (2011), compared the changes in mean mania scores for antimanic drugs with placebo in the treatment of acute mania in manic or mixed manic states of bipolar disorder. Results were generally consistent with results of the present study; aripiprazole, asenapine, carbamazepine, haloperidol, lithium, olanzapine, paliperidone, quetiapine, risperidone, valproate and ziprasidone were all significantly more effective than placebo (pooled effect sizes [Hedges’s g] ranged from 0.26 [aripiprazole] to 0.66 [risperidone]), whereas lamotrigine and topiramate were not. Improvement in mania rating scales were greater with antipsychotics than with mood stabilisers, and there was no significant difference between haloperidol and second generation antipsychotics.

So what?
The present systematic review adds to the evidence base regarding the effectiveness of antimanic drugs in acute mania. It compared and ranked the efficacy and tolerability of antimanic treatments using both pair-wise (conventional) meta-analysis and multiple-treatment meta-analysis. In multiple treatment meta-analysis, sometimes called network analysis, treatments are compared indirectly to provide a level of their effect, for example, the effect of A vs C is inferred statistically from the results of trials of A vs B and B vs C.

The findings that antipsychotics outperform other drugs in reducing manic symptoms over three weeks is consistent with NICE guideline recommendation for treatment of severe manic symptoms or marked behavioural disturbances in people with bipolar disorder. The present study found that haloperidol ranked highest in terms of efficacy, although risperidone and olanzapine were the best treatments when both efficacy and tolerability were considered together.

The indirect nature of the multiple-treatment meta-analyses means that the findings of this analysis should be interpreted with caution. Individual studies were moderately heterogeneous and their findings could have varied as a result of differences in their patient populations (e.g. possible differences in severity of symptoms) and previous experience with the trial treatment.

The authors suggest that the results should be considered in the development of clinical practice guidelines, and it is likely that the results of this comprehensive analysis will be considered alongside other evidence when NICE next reviews its clinical guideline for bipolar disorder. However, to base choice of treatment for acute mania on the basis of these results alone does not take account that treatment of acute mania is only a small part of the overall treatment for bipolar disorder. The successful management of bipolar disorder requires the implementation of therapeutic strategies that balance long-term tolerability and effectiveness. As an accompanying editorial states ”….the short term administration of antipsychotics is often necessary, but the maintenance of a long-term perspective is essential”. The editorial also notes that clinicians rarely modify an acute drug that initially seems to be the most useful.

* Notes re intensive monitoring (black triangle scheme)

Intensive monitoring is requested only when:

  • quetiapine is used for the recently licensed indications of bipolar depression and preventing recurrence in bipolar disorder
  • risperidone is used for the recently licensed indications of short-term treatment of persistent aggression in Alzheimer’s dementia and conduct disorder in children
  • aripiprazole is used by injection (7.5 mg/ml solution for injection)

Study details
Cipriani A, et al. Comparative efficacy and acceptability of antimanic drugs in acute mania: a multiple-treatments meta-analysis. Lancet 2011;378:1306–15

Pair-wise and multiple-treatments meta-analysis of 68 double-blind RCTs (n=16,073) conducted between January 1980 and November 2010.

Adults with diagnosis of bipolar I disorder (manic or mixed episode) with acute mania.

Intervention and comparison
Comparison of one antimanic drug with another antimanic drug or with placebo as oral therapy. Combination and augmentation studies were included.

Primary outcomes

  • Efficacy: mean change scores on the Young Mania Rating Scale (YMRS) over three weeks
  • Acceptability: patients who left the study early (for any reason) during the three-week treatment period


Table 1. Pair-wise meta-analysis (direct comparison) results compared with placebo

Mean change score YMRS

Dropout rate

Medication SMD (95% confidence interval) Odds ratio (95% confidence interval)

–0.58 (–0.77 to –0.39)

0.72 (0.50 to 1.06) NS


–0.50 (–0.67 to –0.33)

0.54 (0.40 to 0.72)


–0.44 (–0.56 to –0.32)

0.60 (0.44 to 0.82)


–0.40 (–0.54 to –0.26)

0.91 (0.61 to 1.35) NS


–0.37 (–0.51 to –0.24)

0.60 (0.41 to 0.88)


–0.31 (–0.42 to –0.20)

0.86 (0.62 to 1.19) NS


–0.50 (–0.69 to –0.30)

0.71 (0.49 to 1.04) NS


–0.42 (–0.59 to –0.24)

0.80 (0.56 to 1.14) NS


–0.16 (–0.30 to –0.03)

0.79 (0.61 to 1.03) NS


–0.24 (–0.49 to –0.01)

0.93 (0.61 to 1.41) NS


0.01 (–0.21 to 0.22) NS

1.25 (0.81 to 1.96) NS


0.06 (–0.06 to 0.17) NS

1.61 (1.22 to 2.13)


–0.32 (–0.08 to 0.72) NS

1.75 (0.83 to 3.70) NS

Table 2. Multiple-treatments meta-analysis results compared with placebo

Mean change score YMRS

Dropout rate


SMD (95% credibility intervalb)

Odds ratio (95% confidence interval)


–0.56 (–0.68 to –0.43)

0.85 (0.62 to 1.15) NS


–0.50 (–0.63 to –0.38)

0.61 (0.44 to 0.83)


–0.43 (–0.54 to –0.32)

0.57 (0.44 to 0.74)


–0.37 (–0.50 to –0.25)

1.05 (0.78 to 1.43) NS


–0.37 (–0.51 to –0.23)

0.64 (0.45 to 0.91)


–0.37 (–0.51 to –0.23)

0.76 (0.55 to 1.06) NS


–0.36 (–0.60 to –0.11)

0.73 (0.42 to 1.28) NS


–0.30 (–0.53 to –0.07)

0.98 (0.57 to 1.71) NS


–0.20 (–0.37 to –0.04)

0.73 (0.51 to 1.05) NS


–0.19 (–0.37 to –0.03)

0.91 (0.61 to 1.34) NS


–0.08 (–0.34 to 0.18) NS

1.22 (0.67 to 2.21) NS


0.07 (–0.09 to 0.24) NS

1.51 (1.00 to 2.27) NS


0.32 (–0.18 to 0.82) NS

1.76 (0.62 to 5.06) NS

Note: For mean change scores values less than 0 favour the medication. For dropout rate values less than 1 favour the medication. NS – not statistically significant. a Data on risperidone and paliperidone▼ were merged. b A credible interval is the Bayesian equivalent of a confidence interval.

There was no funding source for this study.

Further information on bipolar disorder can be found on NHS Evidence and in the e-learning section of the NPC website.

Please comment on this rapid review using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you