How consistent are international treatment guidelines for bipolar disorder?


Bipolar disorder is a common mental health condition affecting about 1% (in its most severe form) of the worldwide population (Merikangas et al., 2011). Bipolar disorder has an impact on functioning, employment and physical health, and also increases mortality.

Implementing evidence-based treatment for bipolar disorder can improve outcomes (Kessing et al., 2013). Thus, evidence based guidelines are important for mental health professionals and patients/carers as they represent one way of formalising which interventions a person should be receiving. If evidence is “evidence” then its interpretation should surely lead to the same set of guidelines throughout the world. If not, should we be worried?

Gordon Parker and colleagues from the Black Dog Institute in Sydney, Australia (Parker et al., 2017) have published a paper in Acta Psychiatrica Scandinavica, which reviews whether there is international consensus in evidence-based guidelines for the management of bipolar disorder. They sought to determine the extent to which guidelines were consistent, at a “high” and more fine grained level.

This new literature review compared a number of evidence-based bipolar disorder guidelines, to see how they differed.

This new literature review compared a number of evidence-based bipolar disorder guidelines, to see how they differed.


A literature review of the National Guideline Clearinghouse, the Cochrane Database of Systematic Reviews, PsycInfo and PubMed from 2005 to 2015, with retrieved guidelines analysed qualitatively.


Eleven guidelines, published in the English language, and described as evidence-based (as opposed to consensus statements) were found. They covered many well-known organisations including:

  1. National Institute of Clinical Excellence (NICE)
  2. British Association of Psychopharmacology (BAP)
  3. Royal Australian and New Zealand College of Psychiatrists (RANZCP)
  4. Canadian Network for Mood and Anxiety Disorder Treatments (CANMAT)
  5. International Society for Bipolar Disorders (ISBD)
  6. Japanese Society of Mood Disorders (JSMD)
  7. American Psychiatric Association (APA)
  8. World Health Organization (WHO)
  9. World Federation of Societies for Biological Psychiatry (WFSBP)
  10. Department of Veteran Affairs (VA).

The authors considered levels of agreement in the following areas:

Coverage of bipolar I and II disorders

  • The guidelines offered differing definitional strategies, e.g. the BAP, RANZCP and NICE used DSM (which uses the terms bipolar I, bipolar II etc), whilst the WHO use the ICD-10 (which doesn’t differentiate bipolar disorder into different types)
  • Only 3 guidelines offer disorder specific recommendations for bipolar I and II.

The management of acute mania

  • NICE recommends antipsychotics with the addition of lithium or valproate if ineffective
  • RANZCP recommends the reverse sequence
  • 4 guidelines adopt a severity linked strategy
  • All but 2 (NICE, VA) guidelines recommend adding benzodiazepines in this phase of treatment.

The management of acute hypomania

  • Few guidelines provide recommendations.

The management of bipolar depression

  • Guidelines variably recommend mood stabiliser medications and atypical antipsychotics.

Maintenance therapy in bipolar I

  • All guidelines recommend a mood stabiliser as first line for bipolar I, but with significant variation in specific agents, including if first line treatment fails
  • Several base these on specific clinical features (e.g. dominant polarity) and differences extend to what combinations might be best in different situations.

Maintenance therapy in bipolar II

  • Lamotrigine and quetiapine were most commonly recommended in the small number of guidelines that provide specific guidance.

Monitoring of lithium

  • Guidelines varied in the acceptable upper limit of plasma Lithium level ranging from 1.0 to 1.3 mmol/L.

Safety during pregnancy and breastfeeding

  • Most note the potential teratogenic effect of Lithium, Valproate and Carbamazepine
  • The WHO suggest the evidence is not clear for Valproate or Carbamazepine
  • WHO and APA make no suggestion about Lamotrigine whilst others state it is teratogenic
  • Lithium generally considered unsafe in breastfeeding in most guidelines
  • Safety recommendations vary in relation to carbamazepine in breastfeeding. APA note that it is generally considered safe, VA that it is usually without adverse risks in infants, whilst NICE say it is to be avoided.

Use of antidepressants in bipolar disorder

  • Not recommended in bipolar depression due to the risk of switching to mania in most guidelines
  • A number of guidelines suggest adding antidepressants to mood stabilisers if depression is severe or avoiding it if there are specific symptom profiles
  • The WFSBP suggest it is impossible to give a recommendation because the efficacy data are inconclusive
  • Some guidelines suggest it may be a suitable maintenance treatment for bipolar disorder II specifically.

Adjunctive psychosocial treatments

  • Psychoeducation whether in groups or individually are commonly recommended in the maintenance phase and for relapse prevention
  • Interpersonal and social rhythm therapy, CBT and family focused therapy are also commonly recommended.


The authors suggest the main conclusions are:

  • There are lower levels of agreement than might be expected of “evidence” based guidelines and the guidelines are distinguished more by their differences than consistency.
  • Positioning of management options is largely influenced by the opinions of the committee members as they offer their own evaluation of the literature and seek to reach consensus.
  • Minimal emphasis on any differential management of the bipolar I and II disorders is striking given that the 2 sub-types have been listed for over 2 decades in the DSM manuals. The authors suggest that this may be due to an “extrapolation model” but contrast this with specific guidelines for type I and II diabetes.
This review of international bipolar disorder guidance suggests that evidence-based guidelines are distinguished more by their differences than consistency.

This review of international bipolar disorder guidance suggests that evidence-based guidelines are distinguished more by their differences than consistency.

Strengths, limitations and implications

This study is a good quality review of bipolar disorder treatment guidelines and their international consistency. Its clinical (and scientific) importance is clear. There has been a wholehearted sign up to evidence based medicine (and the guidelines derived from this) in the previous decades. Some of the inconsistency across different guidelines is quite fine-grained with differences being omissions rather than contradictions in tackling particular mood phases. However, the divergence was significant in other areas such as prescribing for pregnant women or those breastfeeding.

Some inconsistency across guidelines is likely to be a result of when they were written (e.g. older vs more recent). Also, as the authors point out, medication recommendations will be restricted to drugs that are licensed in the country where the guidance has been developed. However, this does raise the question of whether guideline producers should take an altogether more comprehensive approach and review data for drugs that are also licensed in other countries. This might be particularly relevant in bipolar disorder where various aspects such as management of bipolar depression and treatment resistance are generally underserved areas, and in which off-label prescribing may be more widespread than we would wish.

One issue that would be interesting to explore further is the processes and paradigms that each body use to appraise evidence and produce their guidance. It would be informative to know whether or not groups hierarchically structured the quality of evidence to aid transparency. In terms of paradigms, the WFSBP guidelines use cost-benefit analysis considering both efficacy and safety and tolerability. NICE use cost-effectiveness information, whereas Parker et al. state all other guidelines appear to use efficacy as the primary outcome of interest on which recommendations are made. We could also benefit from knowing how each of the guidelines obtain and review data, including any requests to pharmaceutical companies for unpublished findings which may of course dramatically change conclusions from meta-analysis. These differing paradigms hint at one explanation of the paper’s findings. Guidelines appear not to serve a solely scientific clinical function. They have important social and policy resonance, being framed by society’s needs, wants and ability (or willingness) to pay for healthcare and can be used as the mechanism of rationing. How this framework is set between countries could be one important influence of resultant guidelines, even given the same evidence base.

The issues raised by Parker et al. chime with the recent UK public discourse around the validity and recommendations of the NICE Schizophrenia and Bipolar Disorder guidelines and the emphasis or otherwise placed on individual treatments and the make-up of decision making committees. Underlying this there is the often discrepant results of consecutive meta-analysis in specific therapeutic areas, in which apparently largely the same papers have been reviewed. This review indicates clinicians and other stakeholders might well be forgiven for wondering who to believe or what guidelines (or which bits) they should follow.

Should guidelines be produced by scientists without subject expertise reviewing the evidence? The advantages might be greater objectivity in reviewing results, less basis for criticism of panel members having a perceived vested interest (both intellectual and financial) and perhaps more consistency in evidence appraisal between differing therapeutic areas. Problems might be that clinical judgement could be invaluable when making decisions over small differences or where there is in fact little good quality evidence to review, and it is difficult to see how the patients voice could be incorporated.

Guidelines are necessary but overall this paper highlights that despite best efforts it may be difficult to take opinion out of guidelines for bipolar disorder (and perhaps other disorders). The “evidence” to some extent is dependent on who appraises it and how.

Evidence-based or not, it's difficult to take opinion out of guidelines.

Evidence-based or not, it’s difficult to take opinion out of guidelines.


Primary paper

Parker GB, Graham RK, Tavella G. (2017) Is there consensus across international evidence-based guidelines for the management of bipolar disorder? Acta Psychiatrica Scandinavica Volume 135, Issue 6 June 2017 Pages 515–526 DOI: 10.1111/acps.12717 [Abstract]

Other references

Kessing LV, Hansen HV, Hvenegaard A, Christensen EM, Dam H, Gluud C, Wetterslev J. (2013) Treatment in a specialised out-patient mood disorder clinic v. standard out-patient treatment in the early course of bipolar disorder: randomised clinical trial. The British Journal of Psychiatry, 202, 212-219.

Merikangas KR, Jin R, He J-P, Kessler RC, Lee S, Sampson NA, Viana MC, Andrade LH, Hu C, Karam EG. (2011) Prevalence and correlates of bipolar spectrum disorder in the world mental health survey initiative. Archives of General Psychiatry, 68, 241.

Photo credits

Photo by on Unsplash


Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+